Ffmpeg-devel-irc

Download

ffmpeg-devel-irc@ffmpeg.org

April 2012

1 participants
60 discussions

ffmpeg-devel.log.20120414
by burek 15 Apr '12

15 Apr '12

[00:29] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rc41ac87047 10ffmpeg/libavformat/avidec.c: [00:29] <CIA-17> ffmpeg: avidec: update size when packet is shrunk [00:29] <CIA-17> ffmpeg: Fixes out of array read [00:29] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [00:29] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [00:29] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rcca9528524 10ffmpeg/libavcodec/mpegaudiodec.c: [00:29] <CIA-17> ffmpeg: mp3on4: allocate a large enough frame. [00:29] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [00:29] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [12:36] <Nagy> btw how is the life-time of AVFilterPool handled? Is possible to still have buffer references after the filter graph have been destroyed? [12:43] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rb28c678ba8 10ffmpeg/libavcodec/wmaprodec.c: [12:43] <CIA-17> ffmpeg: wmaprodec; fix get_bits(0) case. [12:43] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [12:43] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [12:43] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r97a5addfcf 10ffmpeg/libavcodec/wmaprodec.c: [12:43] <CIA-17> ffmpeg: wmaprodec: check num_vec_coeffs for validity [12:43] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [12:43] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [12:43] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r52ec269573 10ffmpeg/libavformat/utils.c: [12:44] <CIA-17> ffmpeg: lavf: print with debug level the byte positions before and after find_stream_info [12:44] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [13:57] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r9166f483c5 10ffmpeg/libavcodec/wmaprodec.c: [13:57] <CIA-17> ffmpeg: wmaprodec: check min_samples_per_subframe [13:57] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [13:57] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [13:57] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rd442c4462a 10ffmpeg/libavcodec/wmalosslessdec.c: [13:57] <CIA-17> ffmpeg: wmalosslessdec: Make arrays WMALL_BLOCK_MAX_SIZE big and check samples_per_frame. [13:57] <CIA-17> ffmpeg: The samples_per_frame check is ported from wmaprodec.c [13:57] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [13:57] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [15:19] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * ra0abefb0af 10ffmpeg/libavcodec/wmalosslessdec.c: [15:19] <CIA-17> ffmpeg: wmalosslessdec: make mclms arrays big enough for whats written into them. [15:19] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [15:19] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [15:19] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r2a7063de54 10ffmpeg/libavcodec/wmalosslessdec.c: [15:19] <CIA-17> ffmpeg: wmalosslessdec: fix a get_bits(0) in decode_ac_filter [15:19] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [15:19] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [15:19] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rb3a4351582 10ffmpeg/libavcodec/wmalosslessdec.c: [15:19] <CIA-17> ffmpeg: wmalosslessdec: Fix reading too many bits in decode_channel_residues() [15:19] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [15:19] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [15:26] <Compnn> michaelni : btw , i mailed hte author of the openexr patch, he said he might work on exr compression next :) [15:27] <Compnn> so thanks for working on that [15:29] <michaelni> :) [16:54] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r64bd7f8e4d 10ffmpeg/libavcodec/wmalosslessdec.c: [16:54] <CIA-17> ffmpeg: wmalosslessdec: Reset put bit buffer when num_saved_bits is reset. [16:54] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [16:54] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [17:27] <Compnn> intel bought realvideo huh [17:31] <Daemon404> yeah [17:31] <Daemon404> lol [17:31] <Daemon404> perhaps they are trying to get a stronger market in china [17:31] <Daemon404> <_< [17:35] <Compnn> well [17:35] <Compnn> rmvb hwdecode on phones/netbooks [17:35] <Compnn> to win the chipwar in china :) [17:49] <Daemon404> lol [17:54] <Compn> so how long until these companies band together to destroy software patents [17:54] <Compn> or software patents destroys all of these companies ? [17:54] <Compn> i think it could go either way [18:00] <Daemon404> no [18:00] <Daemon404> they band together [18:00] <Daemon404> to form one giant patent troll [18:03] <Compn> and we watch rmvb since theres no reason to have new codecs ? [18:12] <Daemon404> indeed [18:22] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * r4c53c66a21 10ffmpeg/libavformat/mp3enc.c: [18:22] <CIA-17> ffmpeg: mp3enc: clarify meaning of "Unsupported" messages. [18:22] <CIA-17> ffmpeg: They will only cause us to skip writing the Xing header, [18:22] <CIA-17> ffmpeg: not cause any serious breakage. [18:22] <CIA-17> ffmpeg: Related to trac issue #1027. [18:22] <CIA-17> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger(a)gmx.de> [18:22] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * rd05625bd04 10ffmpeg/libavformat/avidec.c: (log message trimmed) [18:22] <CIA-17> ffmpeg: avidec: handle 0-size packets that exist only in index. [18:22] <CIA-17> ffmpeg: 0-sized packets are used to implement variable fps. [18:22] <CIA-17> ffmpeg: However there seems to be a variation where these are not [18:22] <CIA-17> ffmpeg: even stored in the main file but as 0-size index entries [18:22] <CIA-17> ffmpeg: only. [18:22] <CIA-17> ffmpeg: This fixes the sample in trac issue #957, it now plays both [18:22] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * rb44a5ec76e 10ffmpeg/libavformat/latmenc.c: [18:58] <CIA-17> ffmpeg: 03Jakub Stachowski 07master * r56dcfe87ef 10ffmpeg/libavcodec/wmalosslessdec.c: [18:58] <CIA-17> ffmpeg: Clear residues only in not coded channel. This fixes overwriting coded residues with zeros if first channel is coded and second is not. [18:58] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [18:58] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r83c7803f55 10ffmpeg/libavcodec/lagarith.c: [18:58] <CIA-17> ffmpeg: lagarith: check count before writing zeros. [18:58] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [18:58] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [18:58] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rc42efad3c3 10ffmpeg/libavformat/wtvdec.c: [18:58] <CIA-17> ffmpeg: wtvdec: fix name_size check to consider integer overflows. [18:58] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [18:58] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [18:59] <Nagy|2> how is it that avfilter_get_video_buffer is considered thread-safe in ffplay (i.e. thread_safe_callbacks == 1)? Looking at the code, the entire pool based stuff doesn't seem thread-safe? [20:17] <michaelni> Nagy|2, the calls to get_buffer() should be under a mutex [20:17] <michaelni> see pthread.c [20:18] <Nagy|2> why set thread_safe_callbacks = 1? [20:19] <michaelni> its faster [20:19] <michaelni> it says that get_buffer() can be called from a different thread [20:19] <michaelni> not from multiple at the same time [20:19] <Nagy|2> ah! [20:20] <Nagy|2> thread_safe_callbacks seems rather misleading to me [20:21] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r0b27136c65 10ffmpeg/libavformat/matroskadec.c: [20:21] <CIA-17> ffmpeg: matroskadec: fix null pointer dereference [20:21] <CIA-17> ffmpeg: Fixes: Ticket1206 [20:21] <CIA-17> ffmpeg: Found-by: John Villamil [20:21] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [20:21] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r01bf2ad735 10ffmpeg/libavcodec/indeo4.c: [20:21] <CIA-17> ffmpeg: indeo4: setup width/height properly. [20:21] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [20:21] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [20:22] <michaelni> Nagy|2, yes the name isnt very well choosen but its public API so not easy to fix and even less so when we try to stay compatible with forks [20:24] <Nagy|2> michaelni: understood, I would like to suggest to add "can be called synchronously from another thread" or something to the doc. [20:25] <Nagy|2> right now it just says "an be called from another thread", which is a bit open to misinterpretion [20:26] <michaelni> iam happy to apply such a patch if you post one ... [20:26] <michaelni> one that adds such text ... [20:27] <Nagy|2> will do [22:52] <CIA-17> ffmpeg: 03Mans Rullgard 07master * rf5be7958e3 10ffmpeg/libavcodec/qdm2.c: [22:52] <CIA-17> ffmpeg: qdm2: fix a dubious pointer cast [22:52] <CIA-17> ffmpeg: This reworks a loop to get rid of an ugly pointer cast, [22:52] <CIA-17> ffmpeg: fixing errors seen with the PathScale ENZO compiler. [22:52] <CIA-17> ffmpeg: Signed-off-by: Mans Rullgard <mans(a)mansr.com> [22:52] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r9206ac80a5 10ffmpeg/libswscale/output.c: [22:52] <CIA-17> ffmpeg: swscale: handle last pixel if lines have an odd width. [22:52] <CIA-17> ffmpeg: Fixes bug 270. [22:52] <CIA-17> ffmpeg: 03Anton Khirnov 07master * rd7bcc71dad 10ffmpeg/libavfilter/ (avfiltergraph.h graphparser.c): [22:52] <CIA-17> ffmpeg: graphparser: add avfilter_graph_parse2(). [22:52] <CIA-17> ffmpeg: Unlike avfilter_graph_parse(), it returns unlinked inputs and outputs [22:52] <CIA-17> ffmpeg: to the caller, which allows parsing of graphs where inputs/outputs are [22:52] <CIA-17> ffmpeg: not known in advance. [22:52] <CIA-17> ffmpeg: 03Anton Khirnov 07master * r7ae7c41413 10ffmpeg/libavfilter/ (buffersrc.h vsrc_buffer.c): vsrc_buffer: allow using a NULL buffer to signal EOF. [22:52] <CIA-17> ffmpeg: 03Anton Khirnov 07master * r12e7e1d03e 10ffmpeg/ (doc/filters.texi libavfilter/graphparser.c): graphparser: allow specifying sws flags in the graph description. [22:52] <CIA-17> ffmpeg: 03Anton Khirnov 07master * r4e781c25b7 10ffmpeg/libavfilter/graphparser.c: (log message trimmed) [22:52] <CIA-17> ffmpeg: 03Mashiat Sarker Shakkhar 07master * r9900dd2d9c 10ffmpeg/libavcodec/wmalosslessdec.c: [22:52] <CIA-17> ffmpeg: wmal: Skip seekable_frame_in_packet. [22:52] <CIA-17> ffmpeg: There is no point in storing the value in a variable, since it is not [22:52] <CIA-17> ffmpeg: used anywhere else in the decoder. [22:52] <CIA-17> ffmpeg: Signed-off-by: Diego Biurrun <diego(a)biurrun.de> [22:52] <CIA-17> ffmpeg: 03Stefano Sabatini 07master * r91d3cbe0fd 10ffmpeg/ (6 files in 3 dirs): [22:52] <CIA-17> ffmpeg: avfiltergraph: make the AVFilterInOut alloc/free API public [22:52] <CIA-17> ffmpeg: This is required for letting applications to create and destroy [22:52] <CIA-17> ffmpeg: AVFilterInOut structs in a convenient way. [22:52] <CIA-17> ffmpeg: Signed-off-by: Anton Khirnov <anton(a)khirnov.net> [22:52] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r7432bcfe5a 10ffmpeg/: (log message trimmed) [22:52] <CIA-17> ffmpeg: Merge remote-tracking branch 'qatar/master' [22:52] <CIA-17> ffmpeg: * qatar/master: [22:52] <CIA-17> (27 lines omitted) [00:00] --- Sun Apr 15 2012

1 0

ffmpeg.log.20120414
by burek 15 Apr '12

15 Apr '12

[01:03] <randal`> is it possible to use a named pipe for live data to ffmpeg, and have it timestamp each frame read from the pipe with the current time, rather than just incrementally adding the file and using the framerate for timing? [01:03] <randal`> i'm piping rawvideo rgb32 data [01:04] <randal`> encoding with x264 [01:11] <psycorpse_> Hey guys. I am having some issues and not really sure what to search for on the internets... I am on ubuntu 12.4 right now and I have pulled down the git package and compiled [01:12] <psycorpse_> trying to follow the tutorial from danger.com/ffmpeg and I only have 5 lines right now try to compile [01:12] <psycorpse_> http://pastebin.com/b9jXC6tM [01:13] <psycorpse_> really I have the main function and av_register_all() and trying to compile and I am getting a ton of output with mainly undefined reference to 'insert_function_here' [01:14] <psycorpse_> Any ideas on what I might be doing wrong? [01:15] <Forsaken> psycorpse_: if its not a copy-paste mistake, have a look at your first include [01:15] <psycorpse_> sorry that is a copy paste issue [01:16] <psycorpse_> I will include sample output [01:16] <psycorpse_> from the errors [01:16] <Forsaken> sometimes mistakes really are that simple ;D [01:16] <psycorpse_> I wish in this case [01:16] <Forsaken> hehe [01:16] <psycorpse_> it was kinda painful to figure out what I had to do since the split between libav and ffmpeg [01:16] <psycorpse_> and ffmpeg uses libav [01:17] <psycorpse_> very confusing [01:18] <psycorpse_> http://pastebin.com/xBaaFLn1 [01:19] <psycorpse_> has some of the error output too [01:19] <Forsaken> btw. I have a question too, is it possible with ffserver/ffmpeg to stream some video-file as it is (without converting using rtps/http)? [01:20] <Forsaken> asked to everyone ;) [01:21] <Forsaken> for me it looks like the avcodec.h requires some other header files to be included [01:23] <psycorpse_> I had to add /usr/local/include to my path, so I wonder if avcodec.h isn't able to find them then [01:23] <Forsaken> whats your compile statement? [01:24] <psycorpse_> gcc -o tutorial01 tutorial01.c -lavutil -lavformat -lavcodec -lz -lm | more [01:24] <psycorpse_> forget the | more [01:25] <Forsaken> hmm looks okay [01:25] <psycorpse_> I was trying to contain my output byt that didnt' wok [01:25] <psycorpse_> s/wok/work/ [01:26] <psycorpse_> it just sucks because I am running into issues right at the beginning [01:28] <Forsaken> you should try to change the order of the libarys, the order does matter for gcc [01:28] <Forsaken> that the only thing that comes to my mind right now [01:29] <psycorpse_> ok. Just tried it and same issue. I will continue to scratch my head and try some other things. [01:29] <psycorpse_> Appreciate it [01:30] <Forsaken> good luck :< [01:32] <Forsaken> hmm on the tutorial page (bottom of tutorial 1) the commandline for compiling looks like this: [01:32] <Forsaken> gcc -o tutorial01 tutorial01.c -lavutil -lavformat -lavcodec -lz -lavutil -lm [01:37] <psycorpse_> if I add the 2nd -lavutil I get some different output [03:48] <tsou> I'm trying to create a movie out of some jpgs. Can I specify a different speed for each jpeg? [04:06] <relaxed> tsou: it's possibel but you have to handle each seperately [04:08] <relaxed> I once wrote a script to make a slideshow; I had it take the images and output a small stream of rawvideo for each image. The duration of the stream depended on how long I wanted the image to be displayed. [04:09] <relaxed> So my script grepped a list with the images and times to process them all. [04:11] <relaxed> Then it used cat to combined them all into one stream for use as input to ffmpeg. [04:14] <relaxed> tl;dr anything is possbile with ffmpeg and shell scripting. [04:16] <tsou> relaxed: won't this result in a much much larger file though? [04:17] <tsou> i see that with imagemagick i can easily do this for an animated gif, but from what i understand, if i use ffmpeg to create a movie out of an animated gif, [04:17] <tsou> it will just extract all images from it and still require a unique framerate for the resulting output video.. [04:17] <relaxed> How many images are you using? [04:18] <tsou> around 40 [04:18] <tsou> i need the same 'delay' for most of them [04:18] <relaxed> what type of output do you need? [04:18] <relaxed> video format, etc.. [04:19] <tsou> i don't mind.. as long as it's not obscure :P [04:19] <relaxed> how long do you want the video to be? [04:20] <tsou> under a minute i guess.. most pics should delay around 1 sec, with some exceptions that delay around .5 secs each [04:21] <relaxed> Your options are a really low framerate (that not all container or codecs support) or duplicating frames with a faster frame rate. [04:22] <relaxed> say 15 fps [04:22] <relaxed> compressed with libx264 it wouldn't be that large. [04:22] <tsou> what i initially did, was duplicate all files that i needed to be around 1 second.. [04:23] <tsou> and set a -r of .5 [04:23] <tsou> err 2 [04:24] <tsou> didn't try compressing it afterwards with x264 though [04:24] <tsou> thanks ;) [04:25] <relaxed> you're welcome [04:26] <taqattack> I finally figured out how to cross compile to windows [04:26] <taqattack> feels great [05:34] <psycorpse_> would someone download this file and try to compile it? http://dranger.com/ffmpeg/tutorial01.c [05:34] <psycorpse_> gcc -o tutorial01 tutorial01.c -lavutil -lavcodec -lavformat -lz -lm [05:34] <psycorpse_> I am getting some compile errors and I am wanting to know if this is my environment [07:50] <sosman> I would like to extend a video to the same length as an audio track by reusing the last frame as a "still". Any tips please? [07:54] <sosman> I have converted the original MOV to AVI raw. [14:04] <burek> is there any way to make audio stream mono out of stereo (in which one channel was accidentally off), without reencoding the audio? [14:04] <burek> by off I mean silent, without any audio signal [14:14] <ubitux> -map_channel [14:14] <ubitux> or af pan [14:14] <burek> does it imply reencoding? [14:18] <perrth> saste: how to use ffmpeg from https://gitorious.org/~saste/ffmpeg/sastes-ffmpeg ? the audio-filter branch [14:19] <perrth> for example applying ladspa/sox filter [14:45] <taqattack> Hello there [14:46] <taqattack> I'm having some memory leak issues with ffmpeg when I'm streaming. [14:48] <cbreak> interesting [14:48] <cbreak> I found ffmpeg consuming an increasing amount of memory as well when live streaming/encoding [14:48] <taqattack> Yes that's the issue [14:48] <cbreak> but I was never able to isolate its cause or find out if it was a memory leak [14:49] <cbreak> so we just setled for rebooting the streaming box every day :) [14:49] <cbreak> kind of a hack, but what ever [14:49] <taqattack> I think it might be related to the -rtbufsize [14:50] <taqattack> SO what I think happens is, if the server doesn't accept the RTMP packets. The software keeps dumping the directshow frames into memory. [14:51] <taqattack> Because I noticed if I stream from file to server or directshow capture devices to file, the memory leak issue isn't there [14:52] <cbreak> I never used direct show [14:52] <cbreak> nor rtmp [14:54] <taqattack> I see [14:55] <taqattack> I also noticed that if I set my bitrate very high (more than my upload speed). The frames are dropped more often and there's also increase in memory [17:41] <taqattack> . [18:33] <Nagy|2> how is it that avfilter_get_video_buffer is thread-safe? [18:34] <Nagy|2> ffplay sets thread_safe_callbacks = 1 [18:34] <Nagy|2> however looking at avfilter_get_video_buffer I can't see how it can be thread-safe [19:56] <qfr> Is GPGPU video encoding being used by any high profile software? [19:57] <qfr> I don't even know how viable it is [20:16] <MrElectrooo> hi, i have a question, i've heard the linux ffmpeg encode in a batter quality than windows build of ffmpeg, i would say that is a lie, but i want to hear it from you, anyone? [20:17] <sacarasc> It should be the same. [20:17] <MrElectrooo> yeah, thats what i though [22:04] <obstinate> hi could someone direct me to the latest document on the installation process [22:05] <sacarasc> Which OS? [22:05] <obstinate> i have been trying with a couple of documents dated 2010 sep but they seem to have some version problem [22:05] <obstinate> ubuntu [22:05] <sacarasc> https://ffmpeg.org/trac/ffmpeg/wiki/UbuntuCompilationGuide [22:05] <obstinate> cool thanks will try with this [23:30] <obstinate> Hi i am trying to use pyffmpeg [23:30] <obstinate> but i am facing some compilation issues [23:30] <obstinate> is there anyone familiar with pyffmpeg ? [23:31] <Mavrik_> it's obsolete as hell and unstable [00:00] --- Sun Apr 15 2012

1 0

ffmpeg-devel.log.20120413
by burek 14 Apr '12

14 Apr '12

[00:16] <CIA-17> ffmpeg: 03Stefano Sabatini 07master * r162e400464 10ffmpeg/ (doc/APIchanges libavfilter/Makefile libavfilter/version.h): [00:16] <CIA-17> ffmpeg: lavfi: install asrc_abuffer.h header [00:16] <CIA-17> ffmpeg: The header is part of the public API, in particular is used by [00:16] <CIA-17> ffmpeg: doc/examples/filtering_audio.c. [00:16] <CIA-17> ffmpeg: 03Stefano Sabatini 07master * re21dac3a4f 10ffmpeg/doc/examples/decoding_encoding.c: [00:16] <CIA-17> ffmpeg: examples/decoding_encoding: remove unused variable "size" [00:16] <CIA-17> ffmpeg: Remove unused variable size from video_encode_example() function. [00:16] <CIA-17> ffmpeg: Fix GCC warning: [00:16] <CIA-17> ffmpeg: decoding_encoding.c:214:22: warning: unused variable size [-Wunused-variable] [00:16] <CIA-17> ffmpeg: 03Stefano Sabatini 07master * rb7a7830dd4 10ffmpeg/ (7 files in 3 dirs): (log message trimmed) [00:16] <CIA-17> ffmpeg: lavfi: port 2xsai filter from libmpcodecs [00:16] <CIA-17> ffmpeg: Partially based on the port by Niel van der Westhuizen [00:16] <CIA-17> ffmpeg: <nielkie(a)gmail.com>, done for GCI 2010. Same output as the original [00:16] <CIA-17> ffmpeg: filter and as fast. [00:16] <CIA-17> ffmpeg: See thread: [00:16] <CIA-17> ffmpeg: Subject: [FFmpeg-devel] [PATCH] Port MPlayer 2xSaI filter to libavfilter [00:16] <CIA-17> ffmpeg: 03Stefano Sabatini 07master * r5eeecde859 10ffmpeg/ (5 files in 3 dirs): [00:17] <CIA-17> ffmpeg: lavfi/mp: remove 2xsai wrapped filter [00:17] <CIA-17> ffmpeg: It was ported to a native libavfilter filter. [00:55] <CIA-17> ffmpeg: 03Diego Biurrun 07master * r2ef15b46e4 10ffmpeg/ (4 files in 2 dirs): avpacket, bfi, bgmc, rawenc: K&R prettyprinting cosmetics [00:56] <CIA-17> ffmpeg: 03Henrik Gramner 07master * r729f90e268 10ffmpeg/ (13 files in 3 dirs): (log message trimmed) [00:56] <CIA-17> ffmpeg: x86inc improvements for 64-bit [00:56] <CIA-17> ffmpeg: Add support for all x86-64 registers [00:56] <CIA-17> ffmpeg: Prefer caller-saved register over callee-saved on WIN64 [00:56] <CIA-17> ffmpeg: Support up to 15 function arguments [00:56] <CIA-17> ffmpeg: Also (by Ronald S. Bultje) [00:56] <CIA-17> ffmpeg: Fix up our asm to work with new x86inc.asm. [00:56] <CIA-17> ffmpeg: 03Diego Biurrun 07master * r7bb3a302fe 10ffmpeg/ (9 files in 7 dirs): build: Consistently handle conditional compilation for all optimization OBJS. [00:56] <CIA-17> ffmpeg: 03Ronald Bultje 07master * r992f71e95d 10ffmpeg/libavcodec/ (aasc.c bmp.c msrle.c msrledec.c msrledec.h tscc.c): [00:56] <CIA-17> ffmpeg: msrle: convert MS RLE decoding function to bytestream2. [00:56] <CIA-17> ffmpeg: Signed-off-by: Justin Ruggles <justin.ruggles(a)gmail.com> [00:56] <CIA-17> ffmpeg: 03Mans Rullgard 07master * ra31787ee3b 10ffmpeg/libavcodec/qdm2.c: [00:56] <CIA-17> ffmpeg: qdm2: use get_bits_left() [00:56] <CIA-17> ffmpeg: qdm2: simplify bitstream reader setup for some subpacket types [00:56] <CIA-17> ffmpeg: Signed-off-by: Mans Rullgard <mans(a)mansr.com> [00:56] <CIA-17> ffmpeg: 03Mans Rullgard 07master * r9f3e781eab 10ffmpeg/libavcodec/qdm2.c: [00:56] <CIA-17> ffmpeg: qdm2: make a table static const [00:56] <CIA-17> ffmpeg: Signed-off-by: Mans Rullgard <mans(a)mansr.com> [00:56] <CIA-17> ffmpeg: 03Diego Biurrun 07master * r18e8fef7c9 10ffmpeg/libavcodec/libxvid_rc.c: libxvid: remove disabled code [00:56] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rca19862d38 10ffmpeg/: (log message trimmed) [00:56] <CIA-17> ffmpeg: Merge remote-tracking branch 'qatar/master' [00:56] <CIA-17> ffmpeg: * qatar/master: [00:56] <CIA-17> ffmpeg: libxvid: remove disabled code [00:56] <CIA-17> ffmpeg: qdm2: make a table static const [00:56] <CIA-17> ffmpeg: qdm2: simplify bitstream reader setup for some subpacket types [00:57] <CIA-17> ffmpeg: qdm2: use get_bits_left() [02:57] <CIA-17> ffmpeg: 03Dale Curtis 07master * re9dc616472 10ffmpeg/libavcodec/vp3.c: [02:57] <CIA-17> ffmpeg: vp3: Assert on invalid filter_limit values. [02:57] <CIA-17> ffmpeg: Signed-off-by: Dale Curtis <dalecurtis(a)chromium.org> [02:57] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [02:57] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rbd3ea31797 10ffmpeg/ffmpeg.c: (log message trimmed) [02:57] <CIA-17> ffmpeg: ffmpeg: Compensate non monotonicity errors in the audio TSs. [02:57] <CIA-17> ffmpeg: Audio timestamps are passed through by default and when the input [02:57] <CIA-17> ffmpeg: doesnt contain clean timestamps this can lead to non monotonicity [02:57] <CIA-17> ffmpeg: errors. (rounding to a course timebase can cause this too) [02:57] <CIA-17> ffmpeg: Print a warning when the errors in the timestamps are large [02:57] <CIA-17> ffmpeg: Fixes Ticket1167 (regression since timestamps are passed through) [11:53] <kriegerod> i have a question about ffmpeg.c print_stats variable. Is it possible to turn it off? by OptionDef options[] it seems to not take argument (so cant supply with 'false/1/no'), and is initialized as true statically. [12:01] <kriegerod> figured out -nostats [12:27] <CIA-17> ffmpeg: 03Dale Curtis 07master * ra9d114dc8a 10ffmpeg/libavcodec/pthread.c: (log message trimmed) [12:27] <CIA-17> ffmpeg: pthread: Fix crash due to fctx->delaying not being cleared. [12:27] <CIA-17> ffmpeg: corrects invalid return when avpkt->size == 0. [12:27] <CIA-17> ffmpeg: Reproducible with test case and ffplay -threads 2. Stack trace: [12:27] <CIA-17> ffmpeg: http://pastebin.com/PexZ4Uc0 [12:27] <CIA-17> ffmpeg: Test case: [12:27] <CIA-17> ffmpeg: http://commondatastorage.googleapis.com/dalecurtis-shared/crash.ogm [12:27] <CIA-17> ffmpeg: 03Dale Curtis 07master * ra6b3e6d0b4 10ffmpeg/libavformat/mov.c: [12:27] <CIA-17> ffmpeg: mov: Fix memory leaks on aborted header parsing. [12:27] <CIA-17> ffmpeg: If mov_read_header exits under error, the memory allocated is [12:27] <CIA-17> ffmpeg: not freed. [12:27] <CIA-17> ffmpeg: Signed-off-by: Dale Curtis <dalecurtis(a)chromium.org> [12:27] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [12:27] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r84d74610f9 10ffmpeg/libavformat/mov.c: [12:27] <CIA-17> ffmpeg: mov: zero sc->pb in mov_read_close() [12:27] <CIA-17> ffmpeg: Its bad to free things without zeroing them. [12:27] <CIA-17> ffmpeg: This fixes a potential issue when mov_read_close() would be called twice. [12:27] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [12:27] <CIA-17> ffmpeg: 03Dale Curtis 07master * rb5ada68ead 10ffmpeg/libavformat/matroskadec.c: [12:27] <CIA-17> ffmpeg: matroska: Fix leaking memory allocated for laces. [12:27] <CIA-17> ffmpeg: During error conditions matroska_parse_block may exit without [12:27] <CIA-17> ffmpeg: freeing the memory allocated for laces. [12:27] <CIA-17> ffmpeg: Found via valgrind: http://pastebin.com/E54k8QFU [12:27] <CIA-17> ffmpeg: Signed-off-by: Dale Curtis <dalecurtis(a)chromium.org> [12:27] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [12:27] <CIA-17> ffmpeg: 03Dale Curtis 07master * rfd165a0b31 10ffmpeg/libavformat/mov.c: [12:27] <CIA-17> ffmpeg: mov: Fix harmless OOB read. [12:27] <CIA-17> ffmpeg: Convert key_off initialize to use the same sc->keyframe_count as [12:27] <CIA-17> ffmpeg: used elsewhere in the function. [12:27] <CIA-17> ffmpeg: Signed-off-by: Dale Curtis <dalecurtis(a)chromium.org> [12:27] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [12:27] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rab3fcd93fd 10ffmpeg/libavformat/mov.c: [12:27] <CIA-17> ffmpeg: mov: zero extradata_size on deallocation of extradata. [12:27] <CIA-17> ffmpeg: This prevents the possibility of extradata and its size becoming desynced. [14:57] <mrec> http://85.10.198.106/support/test.mp4 [14:57] <mrec> does anyone know why this file is not playable? [14:57] <mrec> it was created using libavcodec it's mp4/h264/aac [14:58] <mrec> I wonder if the header files are okay [15:03] <JEEB> that file barely contains anything, and I hope you didn't just cut off a part of a file from the beginning >_> the index of an mp4 file is written to the end of the file [15:04] <JEEB> fed that file to L-SMASH's boxdumper and got http://pastebin.com/PxYQ8HUv [15:04] <av500> and is needed to play it [15:06] <mrec> JEEB: another file [15:06] <mrec> just download it again now 700k [15:08] <JEEB> exactly the same result from boxdumper [15:08] <JEEB> that file barely contains any contaner-related information whatsoever [15:08] <Compn> mrec : but its libavformat that puts it into .mp4 ... [15:08] <Compn> libavcodec is audio/video :P [15:08] <Compn> was was this created ? [15:10] <mrec> Compn: I used the api libavformat yes.. [15:10] <mrec> I'm sure I'm doing something wrong [15:15] <mrec> could this be a bug in the ffmpeg code? [15:15] <mrec> I very much wonder that mp4 does not have any header [15:28] <JEEB> mrec, I'm pretty sure it's a problem on your side, but without code nothing can be said one way or another :P [15:33] <mrec> I'm digging through movenc first .. the code is nothing special [15:34] <JEEB> also if you want to try another library specific to mp4, there's always L-SMASH http://code.google.com/p/l-smash/ [15:35] <JEEB> and I recommend their boxdumper to take a look at files anyways :P [15:35] <mrec> JEEB: did you look at my uploaded file? [15:35] <mrec> thanks [15:35] <mrec> I'll check it [15:36] <mrec> I'm checking the file with "file" it does not seem to contain the typical mp4 header like isom [15:36] <mrec> at least not in a way my other sample file contains it which is recognized by "file" [15:37] <JEEB> mrec, I pushed both of those test.mp4s through boxdumper [15:37] <JEEB> that's how I got the pastebin, which pretty much says that the file contains nothing :P [15:37] <JEEB> (nothing as in nothing that could be actively taken as something mp4'ish) [15:38] <mrec> hmm maybe I should cut out audio for testing [15:43] <mrec> I'm just worried that those MP4 headers are not okay [15:43] <mrec> 00 00 00 20 66 74 79 70 69 73 6f 6d 00 00 02 00 69 73 6f 6d 69 73 6f 32 61 76 63 31 6d 70 34 31 00 00 00 08 66 72 65 65 00 00 00 00 6d 64 61 74 [15:43] <mrec> can anyone confirm that? [15:44] <JEEB> your problem is more thorough than damn headers [15:44] <JEEB> take your output through boxdumper and you'll see that you are missing a LOT of stuff [15:44] <nevcairiel> mp4 headers are fun to read in ascii [15:44] <nevcairiel> should start with a moov atom somewhere [15:44] <nevcairiel> sadly i dont know hex for moov. :P [15:44] <av500> cmov ftw! [17:30] <mrec> no cmov inside those mp4 files [17:54] <mrec> I wonder is it allowed to use av_guess_format with mp4 and set the codec to H264? [17:54] <mrec> it only creates unplayable video [18:04] <mrec> encoding a video with H264 produces a video which ffplay detects as mp3 [18:17] <mrec> did anyone ever try to use h264 via API recently? I really wonder if this is workable. The point where I'm unsure is if av_guess_format("mp4"... and afterwards setting the codec to H264 is correct? [18:39] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r2ec8789eb8 10ffmpeg/libavcodec/imgconvert.c: [18:39] <CIA-17> ffmpeg: avpicture_layout: use a endian independant format for the palette. [18:39] <CIA-17> ffmpeg: This should fix the failing fate PAL8 tests. [18:39] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [19:09] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r41a052a6ba 10ffmpeg/cmdutils.c: [19:09] <CIA-17> ffmpeg: cmdutils: fix "oswr may be used uninitialized" warning [19:09] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [19:09] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r62e5ef95ca 10ffmpeg/libavutil/imgutils.c: [19:09] <CIA-17> ffmpeg: av_image_alloc: Align the width to 8 when align >= 8. [19:09] <CIA-17> ffmpeg: This allows simd optimized routines to work in steps of 8 pixels [19:09] <CIA-17> ffmpeg: without going over the linesize. (this matters for yuv->rgb24 for example) [19:09] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [22:12] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r76538d7a78 10ffmpeg/libavcodec/x86/h264_weight_10bit.asm: [22:12] <CIA-17> ffmpeg: h264: fix 10bit biweight functions after recent x86inc.asm fixes. [22:12] <CIA-17> ffmpeg: This should have been updated in the x86inc.asm update, but was [22:12] <CIA-17> ffmpeg: accidently forgotten. [22:12] <CIA-17> ffmpeg: 03Alex Converse 07master * r2b8374e920 10ffmpeg/libavcodec/ffv1.c: ffv1: Fix size mismatch in encode_line. [22:12] <CIA-17> ffmpeg: 03Dale Curtis 07master * r27c2fb0f72 10ffmpeg/libavcodec/vp3.c: [22:12] <CIA-17> ffmpeg: vp3: Assert on invalid filter_limit values. [22:12] <CIA-17> ffmpeg: Signed-off-by: Dale Curtis <dalecurtis(a)chromium.org> [22:12] <CIA-17> ffmpeg: Signed-off-by: Ronald S. Bultje <rsbultje(a)gmail.com> [22:12] <CIA-17> ffmpeg: 03Diego Biurrun 07master * rbaaab6069a 10ffmpeg/ (8 files in 8 dirs): build: Move all arch OBJS declarations into arch subdirectory Makefiles. [22:12] <CIA-17> ffmpeg: 03Dale Curtis 07master * r97ae370078 10ffmpeg/libavcodec/pthread.c: (log message trimmed) [22:12] <CIA-17> ffmpeg: pthread: Fix crash due to fctx->delaying not being cleared. [22:12] <CIA-17> ffmpeg: Reproducible with test case and ffplay -threads 2. Stack trace: [22:12] <CIA-17> ffmpeg: http://pastebin.com/PexZ4Uc0 [22:12] <CIA-17> ffmpeg: The fields in config.fate are colon-separated so any colons [22:12] <CIA-17> (10 lines omitted) [23:15] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rd1c95d2ce3 10ffmpeg/libavcodec/dfa.c: [23:15] <CIA-17> ffmpeg: dfa: fix out of array write in decode_wdlt() [23:15] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [23:15] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [23:15] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r72b9537d88 10ffmpeg/libavcodec/dfa.c: [23:15] <CIA-17> ffmpeg: dfa: Fix out of array write in decode_dds1() [23:15] <CIA-17> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind [23:15] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [23:16] <Daemon404> ! new j00ru bugs! [23:16] <CIA-17> ffmpeg: 03Clément BSsch 07master * r3f80546d75 10ffmpeg/libavfilter/vf_drawbox.c: lavfi/drawbox: fix width displayed in place of x. [23:20] <gnafu> Daemon404: U jelly? [23:20] Action: gnafu ducks. [23:20] Action: gnafu is only joking, just read the knowyourmeme.com page for that meme. [00:00] --- Sat Apr 14 2012

1 0

ffmpeg.log.20120413
by burek 14 Apr '12

14 Apr '12

[00:08] <csholmq> Hey. Trying to compile with --enable-librtmp on Mac OSX Lion. Can't seem to find librtmp though. Despite med compiling it just earlier and installing it. [00:08] <burek> did you type ldconfig after make install [00:09] <csholmq> Nope. That might be it [00:09] <csholmq> I don't have ldconfig under OSX. Lemme check for the alternative [00:09] <DTH0> that ldconfig thing is definitly mandatory it seems... [00:10] <burek> DTH0, it should be called in make install process but.. [00:11] <DTH0> mmh... [00:12] <DTH0> and now vlc won't install without "libx264-116"... and with my luck I'm sure it'll break everything... U_U' [00:13] <burek> if you install it from repos, then no need to compile anything [00:13] <burek> it will pull all of its dependencies itself [00:13] <burek> also from the repos [00:13] <burek> but that's bad.. [00:13] <csholmq> It seems as if I should not have to use ldconfig on OS X. [00:13] <burek> since it always ends up with an old version of things [00:14] <burek> imho all of multimedia projects should be compiled, to keep up with the latest bug fixes and features and stuff (or to download static builds or something similar) [00:14] <DTH0> yes but I just succeed installing ffmpeg with latest libx264... and now vlc won't install without it's old libx264-116... really annoying... [00:14] <Mista_D> Can I detect a freeze frame wit libafilter or a chain of filter? [00:14] <burek> well, it's logical [00:15] <burek> when you install something from the repos, it always searches its dependencies in that repos [00:15] <burek> to have all the correct versions of libraries [00:16] <csholmq> Why is it so hard to compile ffmpeg with librtmp :( I just want to probe streams... [00:16] <DTH0> csholmq, can't help you... I just had to apt-get install librtmp to have ./configure detect it properly... [00:17] <DTH0> don't know how they work in mac [00:18] <DTH0> well anyhow, since it's the big compil day, I'll compil vlc as well... [00:24] <pasteeater> you don't ever want to use libx264-dev from ubuntu repo for ffmpeg. often times it can be too old. [00:25] <pasteeater> DTH0: i made the guide you linked to. i just moved it to the ffmpeg wiki from ubuntuforums a few days ago. let me double check it. [00:27] <burek> pasteeater, add ldconfig after make install :) [00:28] <pasteeater> that should not be a requirement [00:28] <DTH0> pasteeater, I don't remember using ldconfig the first time I use the guide on Ubuntu 11.04. But looks like it was necessary for having libx264 detected by ffmpeg configure script. Also, I didn't need it for vpx codec [00:28] <DTH0> there was no problem with it. [00:28] <pasteeater> looks like you added --enable-visualize? [00:28] <DTH0> yes [00:29] <pasteeater> do you know what that does? [00:29] <DTH0> it enables visualization [00:30] <pasteeater> some people added without knowing what it does [00:30] <DTH0> (I may have mistaken with visualizationS for media players) [00:30] <DTH0> that was very un-linux like from me indeed, I added it without properly checking what it was doing :3 [00:31] <pasteeater> i haven't tested that option in years, so i'm unsure if it even works [00:31] <DTH0> but anyway I removed it when I saw something was wrong and even without it, I got ffmpeg libx264 error [00:32] <DTH0> so it doesn't seem to be that option that is triggering the malfunction [00:32] <DTH0> and thanks very much for your guide. It made me discover checkinstall utility. [00:33] <DTH0> I'm very glad to be able to thank the person that introduced me to that wonderful little program. [00:37] <pasteeater> DTH0: the issue was introduced by including --enable-visualize [00:38] <DTH0> you were able to reproduce it ? [00:38] <pasteeater> yes [00:38] <DTH0> ok, my bad then [00:38] <DTH0> sorry for the trouble [00:39] <pasteeater> that's ok, but you should have mentioned how you deviated from the guide [00:40] <DTH0> I was certain to have tried to run the ffmpeg configure script without x264 having that option enabled but I guess I didn't [00:41] <pasteeater> i uninstalled x264 (as compiled with visualize); cd x264; make distclean; and then tried the configure line as shown in the guide [00:41] <pasteeater> you mentioned compiling VLC: http://www.andrews-corner.org/vlc.html [00:42] <pasteeater> another ubuntuforums refugee. [00:43] <DTH0> :) [00:43] <DTH0> I may consider since I have already problems generating the configure script with that "bootstrap" thing [00:44] <DTH0> ( was missing libtool anyway...) [00:44] <pasteeater> it's a good guide. he also made a mplayer one. [00:45] <DTH0> you mean... He's the father of mplayer ? [00:45] <pasteeater> no. the author also make a compilation guide for mplayer [00:46] <pasteeater> http://www.andrews-corner.org/mplayer.html [00:47] <DTH0> ah ok, should I check the date for codecs ? [00:48] <anxt> can ffmpeg mux an aac with SBR an PS into an mp4 container? [00:49] <DTH0> nevermind [00:53] <anxt> something like "ffmpeg -i 602272.aac -acodec aac -b 24 -f ismv test.mp4" ? [00:59] <juanmabc> nice, amv made it in upstream! [02:50] <undercash> hi [02:50] <undercash> this cmd doesn't seem to work anymore [02:50] <undercash> scale="640:trunc(ow/a/2)*2" [02:50] <undercash> what should i change [02:52] <undercash> :D [02:56] <undercash> none of examples in the doc works [02:56] <burek> can you please use pastebin.com, to show your command line and its output? [02:59] <undercash> Unable to find a suitable output format for 'scale=2*iw:2*ih:interl=1' [02:59] <burek> can you please use pastebin.com, to show your command line and its output? [02:59] <undercash> just trying out some random examples of the doc [02:59] <undercash> for "scale" [03:01] <undercash> http://pastebin.com/UNWqzTff [03:02] <burek> and what happened with -vf [03:02] <burek> replace scale=2*iw:2*ih:interl=1 [03:02] <burek> with this: -vf "scale=2*iw:2*ih:interl=1" [03:03] <burek> and remove interl [03:03] <burek> -vf "scale=2*iw:2*ih" [03:04] <burek> scale is a Video Filter, so (like any other VF) it is passed as a parameter to the -vf option and can be stacked with other VFs [03:05] <burek> for example: ffmpeg -i in.avi -vf "thumbnail,scale=300:200" [03:05] <undercash> right! [03:05] <undercash> why the hell i dont have this in my notes.. [03:06] <undercash> thx a lot [03:06] <burek> :beer: :) [03:06] <undercash> :D [09:59] <robertzaccour> I'm trying to record my webcam mic for audio but can't figure out what to input for it. I did ffmpeg -f alsa -ac 2 -i pulse -r 30 -s 1366x768 -f x11grab -i :0.0 -vcodec msmpeg4v2 -qscale 2 screencast.avi but it only records the internal mic. How do I find out what to input for the webcam mic and where do I input my finding? [10:41] <Pacefist> Can I make AVFormatContext* from list of 188 byte packages without writing a file? [10:42] <Pacefist> video format is MPEG2 [11:05] <mrec> hmm is there anything someone needs to take care when adding libx264 support? [11:05] <mrec> avcodec_find_encoder permanently fails when putting H264 in [11:06] <mrec> configure says x264 is available, even lsof shows that the application is accessing the libx264 lib through ffmpeg [11:15] <mrec> ah that one is solved next issue [11:33] <hjt> LOL [11:34] <hjt> (}) [12:36] <StaRetji1> howdy folks [12:37] <StaRetji1> can someone help me with example of how to play video playlist with ffmpeg? [12:37] <StaRetji1> I need to play n videos [12:37] <StaRetji1> thx [12:44] <robertzaccour> I'm trying to record my webcam mic for audio but can't figure out what to input for it. I did ffmpeg -f alsa -ac 2 -i pulse -r 30 -s 1366x768 -f x11grab -i :0.0 -vcodec msmpeg4v2 -qscale 2 screencast.avi but it only records the internal mic. How do I find out what to input for the webcam mic and where do I input my finding? [12:52] <kriegerod> StaRetji1: for x in video1.mp4 video2.mp4 video3.mp4; do ffplay $x; done [12:54] <kriegerod> robertzaccour: try different -i argument values [12:54] <robertzaccour> kriegerod, what should I do differently? [12:55] <kriegerod> change -i argument value [12:55] <kriegerod> i don't know what value should be [12:55] <robertzaccour> to what? [12:55] <kriegerod> now you have 'pulse' [12:55] <robertzaccour> me neither :( [12:55] <StaRetji1> hello kriegerod, thx for reply, sadly, it is windows server :/ [12:56] <kriegerod> StaRetji1: so what? [12:56] <kriegerod> you can't install bash on it? then do the same with bat script, syntax will be slightly different [12:57] <StaRetji1> oh, will try, thx [12:57] <shroomM> for %i in (video1.mp4 video2.mp4 video3.mp4) do ffplay %i [12:57] <shroomM> :P [12:57] <StaRetji1> I have 72 videos :/ [12:57] <StaRetji1> lol [12:58] <shroomM> for /f %i in ('dir /b *.mp4') do ffplay %i [12:58] <StaRetji1> I thought it can play reverse from directory [12:58] <robertzaccour> kriegerod, how do I find out what values I need to change or what the value is? [12:58] <StaRetji1> sorry, recurse [12:58] <StaRetji1> there are many dirs inside dirs [12:58] <shroomM> maybe [12:58] <shroomM> for /f %i in ('dir /s /b *.mp4') do ffplay %~dpni.mp4 [12:59] <shroomM> beware of spaces though :) [13:00] <StaRetji1> thx shroomM :) [13:00] <shroomM> np [13:01] <kriegerod> robertzaccour: try sth like -i hw1:0 [13:02] <kriegerod> there's shell command alsa-info, that generates huge file with sound devices info. Let it inspire you [13:02] <robertzaccour> what is the command to generate the sound devices? [13:03] <kriegerod> if no success - ask on maillist [13:03] <kriegerod> alsa-info [13:03] <robertzaccour> kriegerod, oh ok thanks [13:07] <robertzaccour> kriegerod, alsa-info command not found [14:10] <mrec> hmm, did anyone ever encode video to h264 for ipad? [14:12] <burek> http://stackoverflow.com/questions/3966331/ffmpeg-conversion-command-to-out… [14:13] <burek> and if you are using the latest ffmpeg (which doesn't support -vpre anymore), then this is the content of that preset file, so use those flags manually: http://code.google.com/p/kgui/source/browse/trunk/ffmpeg/ffpresets/libx264-… [14:13] <JEEB> ... [14:14] <JEEB> ipads can support level 4.1 high profile just fine btw [14:14] <JEEB> so you really don't have to set much to support them [14:14] <JEEB> just that itunes might like stuff that goes over their own encoder's capabilities :P [14:14] <burek> well, I assumed he tried already and failed and then came here for help :) [14:15] <JEEB> and you gave him something for baseline/level 3.0? (0 bframes, 1 ref, no weighted prediction or dct8x8) >_> [14:16] <JEEB> that's pre-3GS limitations IIRC [14:16] <burek> well, if it works for him, it's a strart :D [14:16] <burek> start* [14:16] <JEEB> -_- [14:17] <JEEB> mrec, ffmpeg -i derp.herp -c:v libx264 -crf 22 -preset:v fast -level:v 41 -c:a libfaac -ab 192k out.mp4 [14:19] <mrec> JEEB: I'm trying to do that with some raw input however it doesn't really work [14:19] <mrec> seems like the header are not nice or something like that [14:21] <mrec> if I encode some file with ffmpeg the file header seems to be "ISO Media, MPEG v4 system, version 1" [14:22] <burek> why do people even write things for such devices [14:22] <mrec> from application level I don't see how to set that.. [14:22] <burek> why do you even support it? [14:22] <burek> if they have idiotic policy [14:22] <burek> why do you support them end encourage them to continue existing? [14:22] <mrec> doesn't matter about ipad I'm curious why on API level I cannot set those parameters [14:23] <mrec> MPEG v4 system I'd guess it's mp4 container [14:23] <mrec> I'm stuck .. [14:23] <JEEB> <mrec> JEEB: I'm trying to do that with some raw input however it doesn't really work <- wat [14:23] <JEEB> explain yourself kthx [14:24] <mrec> I'm trying to compress YUV420P frames + PCM to h264/AAC [14:25] <JEEB> have you set the input colorspace,frame rate, frame size and so on correctly :P [14:25] <mrec> it works perfectly fine for mpeg2 mpeg2video/audio, but doing that for ipad seems to be impossible [14:25] Action: JEEB sighs [14:25] <burek> robertzaccour, try -f alsa -i hw:0 [14:25] <burek> hw:1 or hw:2 [14:26] <JEEB> mrec, I have no idea how far you're doing it right to begin with [14:26] <mrec> I'm doing it on API level, the input is raw USB packets [14:26] <JEEB> are you expecting people to be espers here? [14:26] <burek> also, do arecord -l (small L) and see the list of hw:X that you can use [14:26] <mrec> JEEB: maybe someone who did it in the past :-) [14:27] <JEEB> I would just recommend trying to encode stuff via ffmpeg first, then when you've reached your desired goal, apply it to your own crap [14:28] <JEEB> because what you just were trying to say was implying that ffmpeg-muxed mp4 files wouldn't work on iDevices [14:28] <JEEB> which is a blatant lie [14:29] <burek> buy an android :) [14:30] <mrec> JEEB: that's my current result using the API [14:30] <JEEB> then you are Doing It Wrong or blaming the wrong tree [14:30] <mrec> but I guess something's wrong with the header.. [14:30] <JEEB> ... [14:30] <JEEB> that's called "blaming the wrong tree" [14:30] <mrec> I'm surely doing something wrong ^^ [14:30] <JEEB> Yes [14:30] <JEEB> I'm very much sure of that [14:30] <mrec> I won't deny that [14:31] <JEEB> You should only specify yourself on a certain part of things only and only if you are absolutely sure that the problem's there [14:31] <JEEB> and I'm pretty sure you are just looking for random differences on a "working file" and a "non-working one" [14:31] <JEEB> herp a la derp [14:31] <JEEB> burek, shut the fuck up if you have nothing productive to say. Androids are actually a much more wider range of capabilities. You stil have Android devices that need baseline profile. [14:32] <JEEB> (thankfully those are no more in the same numbers as they used to be) [14:32] <burek> you're going on the ignore list my friend :) [14:32] <burek> bon voyage :) [14:32] <JEEB> lol [14:33] <JEEB> mrec, anyways back to the drawing board you :P [14:34] <JEEB> also, I hope you're checking actual playback capabilities VS "what iTunes lets me transfer over to the device" [14:34] <JEEB> if you need to support iTunes then it's herp derp tho :P [14:37] <mrec> the problem is if I set mp4 as format the file is not even recognized by mplayer [14:37] <mrec> ffmpeg happily writes into that file however even after it's closed by the encoder unplayable [14:37] <mrec> ffmpeg the libavcodecs.. [14:37] <mrec> while everything works fine when doing so with mpeg2 [14:38] <JEEB> then you are definitely doing it wrong [14:38] <mrec> I just wonder what :-) [14:39] <JEEB> can you output the raw H.264/AAC streams and see if those are OK by then muxing them into mp4 or something with ffmpeg/L-SMASH? [14:39] <JEEB> also, I hope your mplayer is relatively new -- libavformat has had quite a few updates into the mov/mp4 parser during the last X years [14:40] <JEEB> anyways, checking the raw streams would at least make it sure that you're encoding the streams correctly [14:40] <JEEB> after that you can move into your muxing code [14:44] <mrec> http://pastebin.com/N5mcNfVT [14:49] <mrec> is it a problem to output the file to NULL? [14:50] <JEEB> you won't be able to check it out :P [14:52] <mrec> h264 HW is coming soon at least that one is proven to work in the end :-) [15:08] <Mavrik> hmm, what do you use to encode bitrate-constrained MPEG-2 nowdays? [15:09] <taqattack> Hey guys [15:10] <burek> Mavrik, -b ? [15:10] <burek> or -b:v [15:11] <Mavrik> that makes bitrate go all over the place [15:11] <JEEB> Mavrik, check if -maxrate or -bufsize work with MPEG-2? [15:11] <burek> you can use then -minrat and -maxrate [15:11] <burek> if you want the range to be more narrow [15:11] <Mavrik> JEEB: I did, built-in MPEG2 encoder just destroys the picture quality, that's why I'm asking if anyone found anything better [15:11] <burek> I guess -b makes an average bitrate around your desired value [15:12] <JEEB> Mavrik, but did it at least keep to the bitrate limitations you set? [15:13] <Mavrik> JEEB: yes it did, but the picture just started flickering when it neared the constrained :\ [15:13] <JEEB> hmm, and I guess you're not setting the maxrate or bufsize wrong either? [15:14] <JEEB> (as in forgetting the fact that you need to add k/m if you're not in for long numbers) [15:15] <Mavrik> nope [15:15] <Mavrik> Will try to experiment a little more, but sadly I can't change the bitrate constraint :\ [15:16] <JEEB> also, I guess the settings for the mpeg2 encoder are pretty cryptic [15:16] <JEEB> so you might want to check those out, too [15:16] <taqattack> Can anyone help me please? I'm getting a memory leak issue I believe. [15:17] <JEEB> it might just be that you're setting the equivalent of profile baseline/preset veryfast in libx264 [15:17] <JEEB> as I remember the ffmpeg's defaults for !libx264 now would still be rather lulzy [15:17] <Mavrik> I might be, documentation around good defaults for MPEG2 is kinda scarce [15:17] <burek> taqattack, [15:17] <burek> can you please use pastebin.com, to show your command line and its output? [15:17] <Mavrik> can do perfect videos with x264 but mpeg2 just boggles my mind [15:17] <taqattack> one second [15:17] <JEEB> and then there's always HCEnc from that one fortran developer [15:18] <JEEB> or was it cobol? don't remember [15:18] <JEEB> Mavrik, check D_S's animation encoding test for some --preset placebo'ish examples [15:18] <JEEB> http://x264dev.multimedia.cx/archives/102 [15:19] <taqattack> http://pastebin.com/quPnhDsX [15:20] <Mavrik> ah nice [15:21] <taqattack> So when I try to stream to rtmp from device. ffmpeg uses up 1gb of memory when I'm inside the game. [15:21] <taqattack> However, when I try to save to a file, it runs perfectly fine without any memory problems. [15:22] <taqattack> If I try to stream from a file, it doesn't have any memory issues as well. [15:24] <kriegerod> taqattack: IMO librtmp can be the reason. try recompiling with --disable-librtmp --enable-protocol=rtmp [15:25] <taqattack> One second [15:25] <kriegerod> there's another, internal rtmp client implementation in ffmpeg [15:26] <taqattack> I actually used one of the static builds from ffmpeg windows site [15:26] <taqattack> Can I compile it in Visual Studio? [15:26] <kriegerod> no, install mingw on your windows [15:27] <kriegerod> thus you get bash, and can compile it [15:27] <kriegerod> ah, also you'd need yasm, and to compile libx264... [15:28] <kriegerod> ...and libmp3lame [15:28] <taqattack> do I need any more libaries? [15:28] <taqattack> dshow? [15:30] <taqattack> or perhaps pthreads? [15:31] <JEEB> both ffmpeg and (lib)x264 can use native win32 threads now [16:12] <steeve> hey folks [16:12] <steeve> anyone familiar with dvbsub ? [16:12] <steeve> dvbsub pixel format to be exact ? [17:00] <iive> steeve: i've heard it is called vobsub [17:01] <rainmaker1> Hi, does anyone knows why ffmpeg sets default udp packet size to 1472 bytes when transcoding stream to h264 in mpegts via udp? [17:01] <ckb> Is a 3 file encoding output of a 2 hour movie supposed to take 2 hours? :\ [17:06] <relaxed> ckb: not if you stream copy [17:06] <ckb> relaxed: more details? [17:07] <relaxed> show me your command [17:08] <ckb> Well, I don't have it handy right now [17:08] <ckb> can I just show you the options for each output file? [17:09] <ckb> -r 30000/1001 -b:v {$format['bitrate']}k -vcodec libx264 -vf scale=\"$resolution\" -acodec libfaac -ac 2 -pass 1 -an $file [17:09] <ckb> I'm looking to move to a single pass encoding [17:09] <ckb> I feel like dual pass would just take way too long [17:10] <relaxed> what format is the source videos? [17:11] <relaxed> Use -crf 16 and omit -b:v for one pass encoding. [17:11] <relaxed> two pass is if you targeting a specific size. [17:11] <relaxed> you're* [17:11] <ckb> well okay [17:11] <ckb> hold on [17:12] <ckb> http://pastebin.com/9qptrA9U [17:12] <ckb> This is my source file [17:12] <ckb> I'm aiming at getting 3 different formats.. let me add those to the paste too [17:13] <ckb> relaxed: http://pastebin.com/VSZdTfvt formats are at the bottom [17:14] <teegee543> hi i'm having trouble taking a single screenshot using ffmpeg: ffmpeg -ss 20 -i file.mkv -vframes 1 screen.jpg [17:14] <teegee543> i keep ending up with rainbow colored static [17:15] <teegee543> what am i doing wrong? [17:16] <relaxed> try, ffmpeg -i file.mkv -ss 20 -vframes 1 screen.jpg [17:17] <teegee543> relaxed: now i get Output file is empty, nothing was encoded (check -ss / -t / -frames parameters if used) [17:17] <teegee543> i tried various options to fix that, but didn't have any luck [17:17] <relaxed> what is the duration of the input? Shorter than 20 seconds? [17:19] <ckb> relaxed: so I shouldn't even specify a bitrate? [17:20] <ckb> it will adjust automatically to the resolution? [17:21] <ckb> blah [17:32] <ckb> marco? [17:44] <steeve> iive: figured it out :) [17:44] <steeve> iive: it's actually similar to vobsub, as it's a bitmap based format [17:45] <steeve> iive: but the decoded data isn't rgb or yuv, it's palette based, and the palette is in data[1] of the rectangle picture [17:46] <steeve> a simple call to sws_convertPalette8ToPacked32 gets your RGBA :) [18:07] <taqattack> Hello there [18:08] <taqattack> I'm having a bit of issues trying to build FFmpeg on Windows [18:08] <taqattack> Is there any links I can follow to build one the nightly source? [18:13] <andrew_> Does anyone here know anything about viewing content from FFserver on Android? [18:14] <andrew_> H.264 perferably, but not neccessary [18:19] <ckb> I'm trying to figure out why my encoded nitrate is larger than the source nitrate when using -cbf 16 [18:19] <ckb> -crf** [18:24] <ckb> relaxed: would you recommend single pass or 2 pass for 2hr+ movies? [18:32] <JEEB> ckb, because you are using a low enough crf value for your type of content that will end up like that >_> [18:32] <JEEB> also, with libx264 the "one or two passes" thing depends on what you are trying to achieve [18:32] <JEEB> either it is "exactly file size X" -> 2pass bitrate-based [18:32] <ckb> JEEB: I can't decide whether a 2 pass or single pass encoding would be best for what I'm attempting [18:32] <JEEB> or "a certain quality level" -> 1pass crf-based [18:33] <ckb> JEEB: I have a 2 hour movie we're going to deliver (plus other movies but this is my test case) [18:33] <ckb> I want to encode the movie in 3 formats& (one for iPhone, one for iPad, and one for desktop) [18:34] <ckb> it's pretty basic as of now, and I may add more formats later [18:34] <JEEB> well, those two lines I said should make it rather simple to decide :P [18:34] <JEEB> either you are hitting a super-specific size, or you are hitting a certain "quality level" (constant rate factor) [18:34] <ckb> you would say single pass in this case? [18:35] <ckb> not too concerned with size& as we're using a CDN [18:35] <ckb> I mean, I'm sure bandwidth would have something to do with that [18:35] <JEEB> if it's just about quality, then crf [18:35] <JEEB> if you are letting people watch stuff over the internet as-they-download, then you'd also want to set maxrate and bufsize [18:36] <JEEB> if it's just for-download, then maxrate and bufsize don't really matter [18:36] <ckb> JEEB: strictly RTMP [18:36] <JEEB> so streaming [18:36] <JEEB> thus, you will need a maxrate and bufsize [18:37] <ckb> see this is my first time attempting to make an encoder. gone from 2 pass to single pass, etc [18:37] <JEEB> bufsize is the size of the amount of data that the player will buffer, maxrate is the maximum average bitrate over the bufsize [18:37] <ckb> so single pass with bufsize and maxrate? [18:37] <JEEB> yes [18:37] <ckb> how should I decide what to set these to? [18:38] <JEEB> depending on your bandwidth and player's buffering [18:38] <ckb> amazon s3 [18:38] <mrec> does anyone know how to encode h264 video? is this allowed into an mpeg4 container? [18:39] <mrec> this api is pissing me off a little bit ^^ [18:40] <ckb> so you set the buffer size in the video? seems a little absurd. [18:41] <JEEB> ckb, "your bandwidth" = the minimum bandwidth to watch that clip [18:42] <JEEB> and no, it's not absurd -- the encoder has to know how you handle the reading [18:42] <JEEB> for example, if you have bufsize of 1200k and a maxrate of 450k [18:43] <JEEB> it means that if you at the beginning buffer 1200k, and the bandwidth of the client is at least 450k, the client will not have to buffer ever again while watching the video [18:43] <JEEB> as long as the user's bandwidth does not go under 450k [18:44] <JEEB> (in other words, within a buffer of 1200k the average bit rate does not go over 450k) [18:44] <ckb> now I understand the concept of a buffer and nitrates, but this is over my head [18:44] <ckb> bitrates** [18:44] <JEEB> :P [18:45] <JEEB> why do I, who hasn't gotten paid a penny, know more about this than you do? [18:45] <JEEB> mrec, yes it is -- you are failing badly that's all. [18:45] <ckb> I'm a developer.. not a video guy [18:45] <ckb> I mean [18:45] <ckb> a 1200k buffer means [18:46] <ckb> that if the video is paused a 00:00:00 [18:46] <ckb> it will only download up to 1200k of the video [18:46] <ckb> right? [18:46] <JEEB> you will have to match up your player to the stream, no-one's going to do that automatically :) [18:47] <mrec> JEEB: I don't get why the API just returns garbage.. [18:47] <mrec> http://ffmpeg.org/trac/ffmpeg/attachment/ticket/521/qthcamera.cpp [18:47] <mrec> you could look at that one it's nothing different [18:47] <mrec> so I guess there's some magic configuration required with it... [18:47] <JEEB> mrec, also I hope you really aren't setting the 'h264' video format [18:47] <JEEB> because that only has a decoder [18:48] <JEEB> libx264 is the library you can use for H.264 encoding [18:48] <ckb> JEEB: is the buffer size absolutely required? [18:48] <ckb> or just maxrate [18:48] <JEEB> yes [18:48] <JEEB> both are needed [18:48] <JEEB> how else would be the maxrate calculated? [18:48] <mrec> JEEB: [libx264 @ 0x1e1e6c0] frame= 395 QP=51.00 NAL=2 Slice:P Poc:10 I:1336 P:14 SKIP:0 size=28954 bytes [18:48] Action: ckb sighs. [18:48] <JEEB> ok, so at least you aren't doing that mistake :P [18:48] <mrec> but I can't play it [18:49] <mrec> ffplay even detects it as mp2 hah [18:49] <mrec> headers are messed up I guess [18:49] <mrec> but I can't find the switch to turn off the messing up feature ^^ [18:49] <JEEB> try outputting a raw H.264 stream [18:49] <JEEB> try manually muxing that [18:49] <JEEB> check if it works [18:50] <JEEB> if yes, then it's not your encoding [18:50] <mrec> h264-es already worked [18:50] <JEEB> if no, it's your encoding [18:50] <mrec> the input is raw YUV420P [18:50] <mrec> something like webcam [18:51] <ckb> JEEB: so if I'm encoding 3 formats for desktop/ipad/iphone [18:51] <ckb> should the bufsize/maxrates all be different [18:51] <ckb> for these devices [18:51] <ckb> you clearly don't want a large buffer for ipad/iphone, correct? [18:52] <ckb> hmmm [18:53] <JEEB> if you have control over the player, you can set the bufsize (and maxrate) as you want, based on the usual bandwidths of possible clients etc. etc., if you have no control over the player, you will have to check their documentation on what exactly is the POC going to do [18:54] <ckb> I'm going to be streaming with a SMIL file and flowplayer [18:54] <ckb> with bandwidth detection.. [18:54] <JEEB> whatever that means [18:54] <ckb> which is why I was aiming at constant bitrates [18:54] <JEEB> no, you most probably don't want constant bitrates >_> [18:54] <JEEB> you want vbv [18:55] <mrec> JEEB: is there any ffmpeg version which for sure is known to work with encoding? I'm somewhat using the latest git maybe there's some issue with it [18:55] <JEEB> which is maxrate/bufsize [18:55] <JEEB> mrec, I'm pretty dang sure it's not the case with that [18:55] <ckb> SMIL files just list the different renditions of a video (with bitrate and width) [18:55] <JEEB> mrec, so in other words, if elementary stream works, your muxing code is derped [18:55] <mrec> just last year almost the same time I figured out there was an issue with the mpeg2 decoder which came in 2 months before it was discovered [18:56] <JEEB> and you actually tested your elementary streams, right? [18:56] <mrec> maybe the mp4 muxer is bad [18:56] <mrec> yes [18:56] <JEEB> that they play back fine [18:56] <JEEB> k [18:56] <JEEB> mrec, if your command line ffmpeg can do mp4 muxing, it should be fine [18:56] <JEEB> it uses the same backend [18:57] <ckb> JEEB: it might help to know this is going to be web delivery [18:57] <JEEB> you already said rtmp >_> [18:57] <ckb> er I already told you it was RTMP yeah [18:58] <ckb> haha [18:58] <ckb> ffmpeg y u no make sense? [18:58] <JEEB> maxrate and bufsize make perfect sense [18:58] <ckb> to you -__- [18:59] <ckb> an iPhone will have slower bandwidth than a desktop on a t3 connection [18:59] <JEEB> "I want to have the player buffer Xkb and the minimum bandwidth that is needed to watch this clip without further buffering is Ykbps" [18:59] <JEEB> yes, of course [18:59] <ckb> so.. my original question.. would I need to change the buffer/maxrate for each of these? [19:00] <JEEB> it all depends on what kind of people you're aiming, what you are trying to do >_> [19:00] <JEEB> *aiming at [19:00] <JEEB> I mean, I don't know if YOU want to have separate bandwidth profiles for iPad and iPhone f.ex. [19:00] <ckb> the EASIEST example is netflix [19:00] <ckb> lets just assume I have 1 video right now that I want to be able to view on any device.. anywhere [19:01] <ckb> and have quality reflect depending on bandwidth/screen size/etc [19:01] <JEEB> the thing is, YOU are the person making the service. Thus either you are in the position to decide on these matters, or you are given a list of limitations which you have to adhere to [19:01] <JEEB> f.ex. "iPhone users on average have at least 300kbps of bandwidth" [19:01] <JEEB> and you go off that [19:01] <JEEB> (note: that was a random pseudo-fact) [19:02] <ckb> I mean, what's the difference between -bufsize 1500k and -bufsize 750k for iphones? [19:02] <mrec> JEEB: do you know if that avio interface is allowed to use mp4? [19:02] <ckb> what would happen if I set the buffer size to 400m? [19:02] <JEEB> mrec, no idea [19:02] <JEEB> ckb, it would become funny if you don't set the player to match of course [19:03] <ckb> so the player AND the video have to be set with the same options? [19:03] <JEEB> (and depending on how much bitrate your crf value would use, it would either work or not) [19:03] <JEEB> ckb, yes -- naturally [19:03] <JEEB> which I mentioned quite a few moments ago IIRC [19:03] <JEEB> <JEEB> if you have control over the player, you can set the bufsize (and maxrate) as you want, based on the usual bandwidths of possible clients etc. etc., if you have no control over the player, you will have to check their documentation on what exactly is the POC going to do [19:04] <ckb> if a video has a buffer size of 1500k and maxrate of 300k, then at any given moment in the video, the buffer is ~5s ahead? [19:04] <JEEB> yeah [19:04] <ckb> interesting [19:05] <ckb> what would happen if I set the player's maxrate to 400? [19:05] <ckb> while the video's is 300? [19:05] <ckb> I would just end up buffering again? [19:05] <JEEB> nothing really, you're just trying to get more data than you need [19:05] <JEEB> not sure if you can even do that with RTMP [19:05] <ckb> that's what I'm saying [19:05] <JEEB> since it's supposed to be "streaming" [19:05] <ckb> I didn't think RTMP used buffering [19:05] <JEEB> it does [19:06] <JEEB> every kind of streaming uses buffering [19:06] <ckb> but not like HTTP [19:06] <JEEB> if you are just reading a random stream off a protocol, it's the same [19:07] <JEEB> it's no different if I push a transport stream off my TV receiver to an http port and try to watch that, than if I would do push the same stuff to some rtmp server, and then trying to watch that [19:07] <JEEB> in both cases the player would have to buffer some amount [19:08] <ckb> so [19:08] <ckb> say maxrate was 400 [19:08] <ckb> but the client can only push 200 [19:08] <JEEB> s/push/pull/ [19:08] <ckb> thanks [19:08] <ckb> too many things going on [19:08] <ckb> that's when the video will pause to catch up the buffer back to size? [19:08] <ckb> and it would happen pretty frequently? [19:09] <JEEB> yeah [19:09] <ckb> ah [19:09] <ckb> I see [19:09] <ckb> that's why my mom's adsl connection doesn't like youtube ;( [19:09] <JEEB> of course, it also depends on your CRF value [19:09] <JEEB> you're not necessarily always filling up the buffer up to maxrate :D [19:10] <JEEB> you're just telling the encoder not to go over maxrate [19:10] <JEEB> (which in other words does indeed become the "minimum bandwidth requirement" for your video) [19:10] <ckb> ahhh [19:10] <ckb> so if the video is 9800kbps [19:11] <ckb> that's the AVERAGE? [19:11] <JEEB> I have no idea what you're talking about, but yes -- in most cases [19:11] <ckb> if I set maxrate& the average BR of the video will be less than maxrate [19:11] <ckb> ffmpeg -i gives a nitrate for a video [19:12] <JEEB> it will give an estimate [19:12] <JEEB> a rough one [19:12] <ckb> takes a sample from 1-10s or something? [19:12] <JEEB> not sure how it calculates it without going through the whole file :) [19:12] <ckb> not stored in the header? [19:13] <JEEB> I don't think many containers store the average bit rate of the clip in their headers :) [19:13] <ckb> ah [19:13] <ckb> I get this whole thing now [19:13] <ckb> so if I set maxrate to 10000 [19:13] <JEEB> but yes, in most cases if you see a number given for some bit rate on some file, it is usually the average [19:13] <ckb> that's a little high [19:13] <ckb> for standard bandwidth speeds of cable/dsl [19:14] <JEEB> 10000 is 10k, this is not x264 :) [19:14] <JEEB> ffmpeg needs k or m [19:14] <JEEB> (x264 handles kilos by default) [19:14] <ckb> libx264? [19:15] <ckb> -__- [19:15] <JEEB> well, if you are using libx264 via ffmpeg, then ffmpeg's numbering scheme is applied :) [19:15] <JEEB> as in, ffmpeg gives the values [19:15] <ckb> ya [19:15] <ckb> ok [19:15] <ckb> so 10k is a little high for maxrate? [19:15] <JEEB> and, well, no -- because you are setting a maximum average bit rate over some buffer size [19:16] <JEEB> within that clip that ended up being 9800kbps, you might have a very bitrate-intensive scene [19:16] <JEEB> which might go over that limitation [19:16] <JEEB> that would be now controlled by that limitation [19:16] <JEEB> also, you probably meant either 10000k :) [19:16] <JEEB> (or 10m/M) [19:17] <ckb> haha right [19:17] <ckb> 10k is fairly slow [19:17] <ckb> wait a second.. [19:18] <ckb> yeah.. all this bitrate talk is confusing me [19:19] <ckb> so I think I'm going to set the bufsize to 2000k and maxrate 250k [19:19] <ckb> seems reasonable [19:19] <JEEB> so you have the user buffer ~8+ seconds of content or so? [19:20] <ckb> sure? [19:20] <JEEB> if you can set that, all's fine :) [19:21] <ckb> if I can set that? [19:21] <ckb> :o [19:21] <JEEB> on the player's side [19:22] <JEEB> (I'd think that things like iphones and so on would be rather herp derp to deal with) [19:22] <JEEB> (at least people usually don't know how the heck things should be set for them) [19:22] <ckb> http://releases.flowplayer.org/apidoc-latest/org/flowplayer/model/Clip.html [19:22] <ckb> I see a buffer length [19:22] <ckb> but no maxrate [19:23] <JEEB> well, not like you could get the video faster than you can play it, right? [19:23] <JEEB> you don't really need to set the maxrate in the player [19:23] <JEEB> since it comes from the fact of whether or not the user can keep up or not [19:23] <JEEB> (to the encoder's maxrate) [19:24] <ckb> ah I see [19:24] <ckb> I think for the iPhone [19:24] <ckb> I'll probably have to go with m3u8s [19:24] <ckb> gotta be a html5 player [19:25] <eviljames> ckb: Apple has published documentation on encoding for iDevices. [19:25] <ckb> eviljames: link? [19:28] <ckb> http://flowplayer.org/plugins/javascript/ipad.html [19:28] <ckb> Ironically [19:28] <eviljames> ckb: It's somewhere on developer.apple.com - I don't have a link handy. The long & the short of it is for mobile aplications, they want HLS. [19:28] <eviljames> m3u8 + ts files that are no longer than 8-10 seconds. [19:28] <ckb> hmm [19:29] <eviljames> They also provide a tool in OS [19:29] <ckb> I know this [19:29] <eviljames> OSX rather, that should do the chunking for you from a properly formatted mp4 [19:29] <eviljames> s/formatted/transcoded/ [19:29] <ckb> the problem is [19:29] <ckb> I'm trying to automate t [19:29] <eviljames> heh, that's not my problem! ;) [19:30] <ckb> :P [19:30] <ckb> so you can't use RTMP with idevice? [19:30] <ckb> must be delivered via http? [19:31] <eviljames> AFAIK, iOS Safari will not support RTMP - because of its Adobe-ness. [19:31] <eviljames> Apple apparently has a real vendetta against Adobe. No flash, no RTMP... [19:31] <ckb> so [19:31] <ckb> I just have to host the m3u8's on the website [19:31] <ckb> rather than my CDN [19:32] <ckb> which shouldn't be a problem.. [19:32] <eviljames> Who says they can't be on the CDN? [19:32] <ckb> keyword: shouldn't [19:32] <ckb> well [19:32] <ckb> I don't think we want to enable HTTP via CDN [19:32] <eviljames> The m3u8 is a playlist of TS chunks with a bunch of meta-information (ie: chunk length, encryption used, etc) [19:32] <ckb> but that is another option that I will pass along [19:32] <eviljames> Oh, you can't feed RTMP inside the m3u8 (afaik, anyway, someone could correct me on this) [19:33] <ckb> right [19:33] <eviljames> So, you'll want/need HTTP(s) on your CDN anyway [19:33] <ckb> so thus& the m3u8's would go on the webserver [19:33] <ckb> r [19:33] <eviljames> Right, but the chunked TS files have to live somewhere, too. [19:33] <ckb> on the web server as well? [19:33] <ckb> eh [19:33] <ckb> more than likely [19:33] <eviljames> If you want your web server to incur that traffic, I suppose so. But odd are you want your CDN to incur that cost. [19:33] <ckb> we'll enable HTTP [19:34] <ckb> quick non-related question [19:34] <ckb> how to prevent rtmpdumps? [19:34] <eviljames> RTMPe is one option [19:34] <ckb> DRM? [19:34] <eviljames> DRM is the other. [19:34] <ckb> cool [19:34] <ckb> thanks eviljames [19:34] <ckb> loads of help. [19:34] <eviljames> RTMPe won't guarantee that a client doesn't dump - just that it's encrypted from CDN to client [19:34] <ckb> you too JEEB [19:35] <ckb> DRMs will though? [19:35] <eviljames> ckb: Honestly, IMHO, DRM is a waste of time. [19:36] <eviljames> I can't think of a DRM system that exists today that hasn't been broken [19:36] <ckb> video width of 320 maintaining AR, with a buff size of 1500k and a maxrate of 100k? [19:36] <ckb> eviljames: [19:36] <eviljames> If a dedicated enough person wants your content - they'll get it. [19:36] <ckb> I know this [19:36] <ckb> that's the case with anything though. [19:37] <ckb> but AFAIK the copyrights to these movies must be upheld that we'll be delivering [19:37] <ckb> I don't think people are going to offer movies to us, unless we are using DRMs [19:37] <ckb> it's not 100% (nothing is) [19:37] <ckb> but it's better than 0% [19:37] <eviljames> That's a business decision, not a technical one ;) [19:37] <ckb> that's like saying the TSA doesn't work [19:38] <ckb> so lets just take all the metal detectors out of airports [19:38] <ckb> casual effect [19:38] <eviljames> There are DRM systems available, usually for a fee, that you can implement. The real answer is to make your service more convenient and useful than just copying the content. [19:38] <ckb> which is what we're working on. :) [19:39] <ckb> and why I'm here! [19:39] <eviljames> For example, Apple doesn't use DRM on iTunes music purchases, haven't for years. [19:39] <ckb> well [19:39] <ckb> I'm sure that has something to do with Apple having more money than gold in fort knox [19:39] <ckb> (how is that even possible) [19:40] <ckb> (that was rhetorical).. [19:40] <eviljames> ckb: They weren't *quite* this rich in 2009 when they stopped offering DRM downloads. [19:40] <eviljames> and, Amazon AFAIK has never used DRM on their music downloads. [19:40] <ckb> we're offering ondemand [19:41] <ckb> not downloads :D [19:45] <steeve> hey, anyone familiar with dvbsub pts ? [19:45] <steeve> i mean, pts for dvbsub ? [19:46] <steeve> actually [19:46] <steeve> pts in dvb-t ts stream ? [19:57] <mrec> ah found my issue [19:57] <mrec> this avio stuff just sucks [19:57] <JEEB> congrats [19:58] <mrec> avio_open(..) writes the correct files including all headers [19:58] <mrec> while doing that with avio_alloc_context it doesn't seem to work here [20:00] <mrec> http://ffmpeg.org/trac/ffmpeg/ticket/984 [20:01] <mrec> so in the end still my fault hah [20:06] <JEEB> well, at least you found out the reason why you were outputting completely borked files :P [20:06] <JEEB> just saying "missing the headers" would be rather downplaying the problem :) [20:09] <mrec> the h264 stream itself should have been fine [20:09] <mrec> headerissue ^^ [20:09] <JEEB> more like packaging issue :P [20:09] <JEEB> a container is not just "headers" [20:29] <ckb> damn [20:29] <ckb> crf encodes 75% faster [20:36] <ckb> wow [20:36] <ckb> this is terrible quality [20:38] <andrew_> Can anyone explain how to tell ffmpeg to use the Baseline profile for H.264? [20:38] <JEEB> andrew_, -profile:v baseline [20:38] <JEEB> with v:c or -vcodec libx264 [20:39] <ckb> JEEB: can you give me some advice on max br and bufsize? [20:40] <ckb> for widths: 1280, 640, 320 [20:40] <ckb> I'm thinking about just keeping max br at 5000k [20:41] <andrew_> JEEB: When I try that, it errors and says that the parameters might be incorrect (bitrate, rate, width, height, etc) [20:41] <andrew_> JEEB: I am streaming this to ffserver [20:42] <JEEB> can you pastebin both your command as well as the whole output :P ? [20:42] <andrew_> sure, one moment [20:42] <JEEB> ckb, how am I supposed to give any advice when I don't know what kind of 'net speeds you're aiming at? [20:42] <ckb> average net speeds of users? :P [20:45] <JEEB> I have no idea about statistics [20:45] <JEEB> lol [20:45] <JEEB> but I guess I'd go for something a la maxrate 3500k and bufsize 9000k for 720p [20:46] <ckb> I think I'm gonna do 5/15 [20:46] <ckb> for all 3 [20:46] <ckb> since my phone is pulling 6.6 [20:46] <JEEB> lol [20:46] <ckb> I think anything slower than that [20:46] <ckb> they need to feel the wrath of the buffer [20:47] <andrew_> JEEB: http://pastebin.com/ZmJuN3e9 [20:47] <JEEB> I think I have a 512k or 1024k or so mobile connection atm [20:47] <JEEB> my interwebs is 10mbps/10mbps [20:47] <ckb> I stupidly put my first max br at 250k >.> [20:48] <ckb> quality looked like loading a iPod nano movie on a desktop [20:48] <JEEB> also, make sure your crf value is something sensible, too [20:48] <eviljames> hahah yeah, 250k ain't gonna cut it pretty much anywhere :P [20:49] <JEEB> eviljames, fine for 320x180 or so [20:49] <JEEB> depends on content of course [20:49] <eviljames> Yeah, fair enough, but modern Android/iPhone are higher res than that. [20:50] <JEEB> this also reminds me that x264 is exceptionally good at keeping to your set bit rates [20:50] <JEEB> a person I know encoded a 22min episode of an animated series with audio, into floppy size [20:50] <JEEB> lessee what the resolution was... [20:51] <JEEB> oh, he actually dropped the frame rate, too :< [20:51] <JEEB> that's not as much fun [20:51] <JEEB> (although that's not such a big problem with animated content) [20:51] <andrew_> If you are refering to me, it was 352x240 [20:51] <JEEB> nah [20:51] <ckb> JEEB [20:51] <JEEB> also, let's have a look at your pastebin [20:52] <eviljames> JEEB: 15fps ? [20:52] <ckb> how do I judge the CRF? [20:52] <ckb> what's too high? [20:52] <ckb> what's too low? [20:52] <eviljames> JEEB: I'm intrigued by getting 22minutes down to 1.44MB :D [20:53] <JEEB> ckb, "the highest crf value that looks good to you" is the rule of thumb [20:53] <ckb> well I like the number 420 [20:53] <ckb> looks good to me! [20:53] <JEEB> x264's default is 23, start from it. [20:53] <JEEB> if it looks good, push higher [20:54] <JEEB> if it looks bad, push lower [20:54] <ckb> the higher the number the less the quality? [20:54] <JEEB> less bitrate is used, yes [20:55] <JEEB> eviljames, mediainfo shows 8fps but I haven't checked the actual timestamps [20:55] <eviljames> I'd trust the metadata as far as I could throw it. [20:55] <eviljames> So, that's good enough for me. [20:55] <eviljames> (I can throw metadata a long way) [20:56] <JEEB> there really isn't "metadata" for frame rates with anything else but AVI, which actually had a frame rate field [20:56] <JEEB> everything else is an average or the delta between the first X timestamps [20:56] <eviljames> huh, I thought that it was specified in many container formats... and I learned something today! :D [20:57] <JEEB> nah, most have timestamps [20:58] <JEEB> (and really, how would you define a "frame rate" for variable frame rate content anyways :) -- other than give an average or something) [20:58] <JEEB> (easy example: a TV show that has 30fps computer graphics scenes, and 24fps live action parts) [21:00] <JEEB> andrew_, seems like libx264 isn't liking the settings you've set (or well, not set) -- I wonder why it isn't setting the medium preset by default... [21:00] <JEEB> -preset:v medium -profile:v baseline [21:01] <JEEB> (if that's not fast enough, select a faster preset) [21:01] <JEEB> http://mewiki.project357.com/wiki/X264_Settings#preset [21:05] <andrew_> JEEB: Unfortunatly, that doesn't work either [21:05] <andrew_> same errors [21:06] <JEEB> how old is your libx264 btw? [21:06] <ckb> ah [21:06] <ckb> 5000 hit the spot [21:06] <ckb> although [21:07] <ckb> I had a bad eye for quality [21:07] <ckb> have** [21:07] <Tjoppen> why not use -crf ? [21:07] <andrew_> I think only a week or so [21:07] <JEEB> Tjoppen, he is trying to decide whatever he wants for his maxrate/bufsize now it seems [21:07] <andrew_> Should I try to grab that latest one incase its only? [21:08] <JEEB> I really don't know why libx264 is finding you using the old borked ffmpeg defaults for libx264... unless ffserver has something to do with it [21:08] <JEEB> ffserver is like a black hole, no docs no love [21:09] <andrew_> I think I have gotten these errors when not using ffserver [21:10] <andrew_> aswell [21:10] <JEEB> the last time I've seen those errors was more than a year ago [21:10] <andrew_> I am using release 0.10 [21:10] <JEEB> yeh, saw that too [21:11] <andrew_> should I re-download and recompile either ffmpeg or x264? [21:11] <JEEB> unfortunately I have to say that I don't know [21:13] <andrew_> Question, what if I use an older version [21:13] <andrew_> because I haven't had these problems on the 0.8 windows build [21:13] <andrew_> (I am on Ubuntu now) [21:14] <JEEB> I think it was 0.8 or so when the new ffmpeg libx264 defaults were introduced... [21:14] <JEEB> can you try the current git HEAD for the kicks? [21:14] <andrew_> sure [21:15] <andrew_> BTW, is it ok to follow this guide to compiling x264 and ffmpeg? [21:15] <JEEB> it mostly should indeed [21:15] <JEEB> if it's the one in the topic [21:16] <andrew_> ok, Im just not super confident with linux, so I wanted to make sure this isn't doing something wrong [21:20] <andrew_> Question, so this guide has me configure x264 with --enable-static [21:21] <andrew_> Thats ok, right? [21:21] <JEEB> yes [21:31] <andrew_> Ok, build ffmpeg now, it take a little [21:47] <andrew_> Hey, so I just redownloaded and compiled x264, libvpx, and ffmpeg [21:47] <andrew_> but the same results [21:48] <andrew_> Could it possibly have something to do with the configuration of FFserver? [22:00] <Freakshow> so I'm trying to restream a feed that has no audio... however, it seems that ffmpeg doesn't like this much [22:00] <Freakshow> should I be setting something on the input side for the fact that there's no audio in the stream? [22:02] <Freakshow> basically getting this on the input [22:02] <Freakshow> Duration: N/A, start: 293.993000, bitrate: N/A [22:02] <Freakshow> Stream #0:0: Video: h264 (Baseline), yuv420p, 640x360 [SAR 1:1 DAR 16:9], 29.97 tbr, 1k tbn, 59.94 tbc [22:02] <Freakshow> Stream #0:1: Audio: none, 0 channels [22:02] <Freakshow> Unsupported codec with id 0 for input stream 1 [22:03] <andrew_> try adding -an right after your input file [22:04] <andrew_> that is working for me [22:06] <Freakshow> wouldn't that be for the output file rather than the input? [22:06] <Freakshow> not trying to be argumentative, just asking [22:07] <Freakshow> so ffmpeg -i http://path.to.whatever -an [22:07] <Freakshow> andrew_ ? [22:07] <andrew_> yea, it does mean that, as far as I know [22:07] <andrew_> But it has been working for me [22:07] <andrew_> at least with a video file with no audio [22:08] <Freakshow> sure sure... I'm trying with a static file first, but the next step will be to pull in the actual remote feed [22:08] <andrew_> well, I know that when using an RTP stream with no audio, it works [22:09] <andrew_> I don't know about other protocols [22:09] <andrew_> bbl [22:10] <andrew_> I hope that works for you [22:14] <Freakshow> hehe [22:16] <Freakshow> well, I think there's something wonky with the input in that 1. it takes too long to get the bits and 2. it would appear that ffprobe/ffmpeg thinks there's an unsupported codec with id 0 for audio input [22:17] <Freakshow> I can re-stream any of my other tests just fine, this particular input however just has no audio [22:17] <Freakshow> /sadface [22:17] <taqattack> Can someone point me the right direction to build FFmpeg on Windows [22:18] <taqattack> I'm very new to building from source [22:18] <ckb> if ffmpeg is running in the background. is there any way to tell the status of it? [22:52] <burek> taqattack, [22:52] <burek> try there [22:52] <taqattack> I have been struggling with it trying to compile it on mingw [22:52] <taqattack> it seems lot of the methods are truncated [22:53] <taqattack> is cross compiling the best method as of right now? [22:53] <burek> try the forum there [00:00] --- Sat Apr 14 2012

1 0

ffmpeg-devel.log.20120412
by burek 13 Apr '12

13 Apr '12

[00:03] <burek> does anyone have any idea what happened here: http://pastie.org/3770857 the ffmpeg used is a static build from 1 machine, that is being run on another machine :/ [00:14] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * r277f20c3c8 10ffmpeg/ (3 files in 2 dirs): (log message trimmed) [00:14] <CIA-17> ffmpeg: lavf: always call avio_flush in av_write_trailer. [00:14] <CIA-17> ffmpeg: Otherwise for muxers like e.g. latmenc that never call [00:14] <CIA-17> ffmpeg: avio_flush (and do not have a write_trailer function) [00:14] <CIA-17> ffmpeg: a part of the data will always be missing. [00:14] <CIA-17> ffmpeg: Also update references for the voc muxer, which was also [00:14] <CIA-17> ffmpeg: buggy before and did not write out all data. [02:07] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rc0af87550a 10ffmpeg/libavfilter/vf_lut.c: [02:07] <CIA-17> ffmpeg: vf_lut: fix rgba map regression [02:07] <CIA-17> ffmpeg: Fixes Ticket1184 [02:07] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [02:07] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r0cd17a6291 10ffmpeg/libavcodec/mjpegdec.c: [02:07] <CIA-17> ffmpeg: mjpeg: fix crash with lowres and flipped images [02:07] <CIA-17> ffmpeg: Fixes Ticket1155 [02:07] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [02:54] <burek> does AVI file keep index at the end of file? [02:54] <burek> the guy has got a file [02:54] <burek> that makes ffmpeg hang [02:54] <burek> but it's 2.5gb [02:54] <burek> would it be enough if only first 10mb are uploaded? [02:55] <burek> http://pastebin.com/n1FpfX0x [03:17] <burek> also, would you like me (or anyone else) to create a simple web form for submitting reports with attachments, instead of instructing people to use ftp and ML? [03:17] <burek> I guess it would be more convenient that way.. [03:18] <burek> bug reports* [03:27] <iive> burek: yes it usually keeps an index at the end. however if the file is bigger than 1gb (or 2), it would create a tree of indexes. (ODML2.0) [03:27] <burek> :S [03:27] <burek> I instructed the guy to upload the first and last 10mb of the file [03:27] <burek> he is still in the #ffmpeg [03:28] <burek> https://ffmpeg.org/trac/ffmpeg/ticket/1192 [10:07] <CIA-17> ffmpeg: 03Nicolas George 07master * r81a9d32f0d 10ffmpeg/ffmpeg.c: ffmpeg: move getutime to the start of the file. [10:07] <CIA-17> ffmpeg: 03Nicolas George 07master * r733cf0ad51 10ffmpeg/ (Changelog doc/ffmpeg.texi ffmpeg.c): ffmpeg: add -benchmark_all option. [11:41] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r75b38674b6 10ffmpeg/libavformat/utils.c: [11:41] <CIA-17> ffmpeg: lavf: compute_pkt_fields, limit the pts upcorrection to mpeg [11:41] <CIA-17> ffmpeg: The pts correction conflicts with the ogg vorbis demuxer. [11:41] <CIA-17> ffmpeg: Thanks-to: Justin Ruggles [11:41] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [11:41] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r9e938d22c5 10ffmpeg/libavcodec/msmpeg4enc.c: [11:41] <CIA-17> ffmpeg: msmpeg4enc: fix stored fps value [11:41] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [11:58] <hjt> hello [11:58] <hjt> i [11:59] <hjt> i compile ffmpeg in mingw,but ffmepg.exe didn't print any message. [11:59] <hjt> can anybody help me? [12:08] <michaelni> hjt, check http://ffmpeg.zeranoe.com/forum/ [12:09] <michaelni> but mingw compile worked fine a day or 2 ago [12:09] <michaelni> so its probably a issue on your side [12:11] <hjt> should i set loglevel while compileing ffmpeg? [12:16] <hjt> i add printf at the begin of ffmepg.c ,it didn't anything neither [13:11] <CIA-17> ffmpeg: 03Jakub Stachowski 07master * r3b00fd5985 10ffmpeg/libavcodec/wmalosslessdec.c: (log message trimmed) [13:11] <CIA-17> ffmpeg: wmalosslessdec: Reset residues to 0 if they were not decoded from bitstream. [13:11] <CIA-17> ffmpeg: If channel residues are have not been decoded from bitstream, they should be [13:11] <CIA-17> ffmpeg: initialized to 0 instead of using values from previous subframe. [13:11] <CIA-17> ffmpeg: This causes bursts of noise in silent parts of some files. [13:11] <CIA-17> ffmpeg: This patch fixes bug #1055 [13:11] <CIA-17> ffmpeg: Reviewed-by: Benjamin Larsson <benjamin(a)southpole.se> [13:11] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * re8c63981e5 10ffmpeg/libavcodec/ratecontrol.c: [13:11] <CIA-17> ffmpeg: ratecontrol: Use ticks_per_frame to calculate fps. [13:11] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [13:11] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r8274b21c09 10ffmpeg/ffmpeg.c: [13:11] <CIA-17> ffmpeg: ffmpeg: fix ticks_per_frame for the avi stream copy case [13:11] <CIA-17> ffmpeg: Should not make any difference, but its more correct. [13:11] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [13:48] <hjt> ;;) [14:06] <michaelni> hjt, you still havnt solved it ? [14:07] <michaelni> if so try maybe to compile a hello_world.c [14:08] <michaelni> if that fails its not ffmpegs fault ... [15:21] <Compn> michaelni : ask him for make output when he comes back :) [15:24] <michaelni> Compn, can you take a quick look at the 0.8.11 news entry (spelling & such) ? maybe a little later as i just pushed but better late than never [15:24] <Compn> sure [15:24] <michaelni> thx [15:25] <michaelni> s/later/late/ [15:26] <Compn> looks ok [15:26] <michaelni> :) [17:12] <ubitux> michaelni: you said yesterday ffmpeg shouldn't stop when unable to decode a frame; could you then check if it's a normal behaviour with this sample: http://lucy.pkh.me/samples/decode-error-stop.mkv [17:12] <ubitux> if so, i'll open an issue [17:13] <ubitux> but i'd like a previous check if possible :) [17:14] <ubitux> it stops decoding around frame 1090 while it shouldn't stop [17:14] <ubitux> (mplayer doesn't for instance, with the lavf/mkv and lavc/h264) [17:52] <michaelni> ubitux, looks like a bug [17:52] <ubitux> okay, will report [17:53] <michaelni> but could easily be a duplicate of 11?? [17:53] <ubitux> ticket11 ? [17:53] <michaelni> 1167 [17:53] <ubitux> :D [17:53] <michaelni> patch on ML about that one [17:54] <michaelni> didnt try with itt hough [17:54] <ubitux> why would it be related? [17:54] <burek> hi, how can I check what audio codec is this: [flv @ 0x1e83a00] Could not find codec parameters (Audio: none ([8][0][0][0] / 0x0008), 5512 Hz, 1 channels) [17:54] <ubitux> Compn: would you mind to copy this file to ffmpeg samples? (as "Matroska/decodec-error-stop-45sec.mkv" or sth like this maybe) [17:54] <burek> just to make sure if the person didn't compile needed audio encoding library [17:55] <michaelni> ubitux, how many frame is it supposed to transcode ? [17:55] <michaelni> approximately [17:55] <michaelni> with -an i get 1097 [17:57] <ubitux> mplayer goes up to ~55sec [17:57] <ubitux> so something like 10sec after the bug at 45sec [17:57] <ubitux> not sure how many frames it means [17:57] <michaelni> not a duplicate of 1167 then i suspect [17:59] <michaelni> not a regression, even ffmpeg-0.5 fails (1051 frames 45.71 sec) [18:00] <ubitux> mplayer and mencoder seems to be able to deal with it [18:00] <ubitux> but i'm not sure how :) [18:01] <michaelni> with lavf demuxer ? [18:01] <ubitux> yes afaict [19:11] <Compn> burek : in flv source/spec is 0x0008 [19:11] <Compn> burek : i think its intel audio [19:11] Action: Compn should check [19:12] <Compn> ubitux : what file ... to where ? [19:15] <Compn> arg its not in libavformat/flv.h [19:16] <silverrocker> I'm trying to load a file (that I craeted with ffmpeg -vcodec libx264 encodedfile.h264) using the example decodeing_encoding.c from doc/examples. I have changed the find codec line to avcodec_find_decoder(CODEC_ID_H264). [19:17] <silverrocker> but the example doesn't seem to work properly (giving me only a part of the frame) [19:17] <silverrocker> also giving some errors to stderr form the h264 codec itself [19:17] <silverrocker> If I open the file with ffplay, it seems to work [19:17] <silverrocker> what am I missing? [19:18] <Compn> possibly the example code isnt up-to-date [19:18] <Compn> wrt new api :\ [19:18] <silverrocker> wrt? [19:18] <Compn> with reguards to [19:18] <silverrocker> ah, so how do I go form here? [19:18] <Compn> i think the decoding api changed and no one fixed example stuff [19:18] <Compn> well uh [19:18] <Compn> have to get someone to reproduce the bug :) [19:19] <silverrocker> Compn: if I poste my code and the file that I'm trying to use, can you help me? [19:19] <Compn> silverrocker : no, i'm not good with code [19:20] <silverrocker> ah ok :s [19:29] <ubitux> Compn: this one: http://lucy.pkh.me/samples/decode-error-stop.mkv to let's say samples.ffmpeg.org; or maybe another place more appropriate? [19:29] <Compn> sounds fine [19:30] <Compn> done [19:30] <ubitux> thanks :) [19:30] <Compn> i just forgot your host addy [19:30] <Compn> full urls are nice :) [20:18] <silverrocker> ok, it seems that I can use the example decode_encode.c just for mpeg1 video. How can I extend this to be used for mpeg4 and others? [20:20] <silverrocker> I can find the video stream for any file and automatically find the codec needed but I need to do more stuff to get it working for any format (beside mpg1) it seems [20:32] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * r912a838efe 10ffmpeg/ (3 files in 3 dirs): [20:32] <CIA-17> ffmpeg: latmenc: fix muxing of byte-aligned DSE. [20:32] <CIA-17> ffmpeg: This will only work for DSEs that are first in a packet, but [20:32] <CIA-17> ffmpeg: that is enough to fix handling of the reference files in [20:32] <CIA-17> ffmpeg: fate-suite/aac (though most of them still have other issues). [20:32] <CIA-17> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger(a)gmx.de> [20:32] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * r0c1e9c2d6a 10ffmpeg/ (libavformat/latmenc.c tests/ref/lavf-fate/latm): [20:32] <CIA-17> ffmpeg: latmenc: copy pce config from proper location. [20:32] <CIA-17> ffmpeg: For the FATE test sample used, this only avoids a warning [20:32] <CIA-17> ffmpeg: message. [20:32] <CIA-17> ffmpeg: However for other samples like al05_44.mp4 the converted [20:32] <CIA-17> ffmpeg: file can be played only after this fix. [20:32] <CIA-17> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger(a)gmx.de> [20:32] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * reb24def465 10ffmpeg/libavformat/latmenc.c: [20:32] <CIA-17> ffmpeg: latmenc: Document assumptions when copying the AAC header. [20:32] <CIA-17> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger(a)gmx.de> [20:33] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * rd1a58afb95 10ffmpeg/libavformat/latmenc.c: [20:33] <CIA-17> ffmpeg: latmenc: validate extradata size. [20:33] <CIA-17> ffmpeg: Fixes potential out-of-bounds writes. [00:00] --- Fri Apr 13 2012

1 0

ffmpeg.log.20120412
by burek 13 Apr '12

13 Apr '12

[00:00] <Kingsy> burek: the latest version says Unrecognized option 'f' [00:00] <burek> o.O [00:00] <burek> can you please use pastebin.com, to show your command line and its output? [00:01] <Kingsy> sure [00:01] <Kingsy> burek: http://pastie.org/3770857 [00:02] <burek> :/ [00:02] <burek> try this one https://sites.google.com/site/linuxencoding/builds [00:02] <burek> if that also doesnt work.. f it.. [00:03] <Kingsy> 1 sec [00:03] <Kingsy> burek: ok thats a little different http://pastie.org/3770866 [00:04] <burek> hm [00:04] <burek> it doesnt support alsa, we can try with oss [00:04] <burek> let me see [00:04] <burek> ./ffmpeg -f video4linux2 -r 30 -i /dev/video0 -f oss -ac 1 -i /dev/dsp -acodec aac -strict experimental -ab 128k -ac 1 -ar 44100 -b 1000k out.mp4 [00:05] <Kingsy> ok 1 sec [00:05] <burek> or try /dev/dsp1, /dev/dsp2, ... [00:05] <Kingsy> burek: naaa just [oss @ 0x1ee6820] /dev/dsp: No such file or directory [00:05] <Kingsy> tried 1 and 2 too [00:05] <burek> just a sec [00:06] <burek> try apt-cache search oss [00:06] <burek> open sound system - something like that [00:07] <Kingsy> oss4-base ? [00:07] <burek> let me check [00:08] <burek> yes [00:08] <burek> try that [00:08] <burek> when you install it, it should create /dev/dsp [00:11] <Kingsy> naaa it didnt.. [00:11] <Kingsy> screw it [00:11] <burek> oss-compat [00:11] <burek> try that [00:11] <Kingsy> it was installed with oss-base [00:11] <Kingsy> btw [00:11] <burek> is there /dev/dsp [00:12] <Kingsy> isnt there a delay you can add to the audio? [00:12] <Kingsy> burek: no [00:12] <burek> well you can [00:12] <Kingsy> those vsync and async commands do nothing as far as I can see [00:12] <burek> -itsoffset 2 [00:12] <burek> will shift stream for 2 seconds [00:12] <burek> now, I don't remember do you put it [00:13] <burek> before -i or after [00:13] <burek> so try [00:13] <Kingsy> don't you think it would be easier to just stitch the two streams together after recording? [00:13] <burek> or that :) [00:14] <Kingsy> -itsoffset 2 doesnt do anything either [00:14] <burek> well [00:14] <Kingsy> perhaps I should just stitch them after.. [00:15] <Kingsy> do you know of a decent editor that I could use? preferably something with a GUI ? [00:15] <burek> ffmpeg -f video4linux2 -r 30 -i /dev/video0 -itsoffset 2 -f oss -ac 1 -i /dev/dsp -ab 128k -b 1000k out.mp4 [00:15] <burek> or [00:15] <burek> ffmpeg -f video4linux2 -r 30 -i /dev/video0 -f oss -ac 1 -i /dev/dsp -itsoffset 2 -ab 128k -b 1000k out.mp4 [00:15] <burek> try those two [00:15] <burek> and you'll see the difference [00:15] <Kingsy> well the first one doesnt work because I don't have oss [00:15] <burek> well, you get the idea :) [00:16] <burek> either put it after first input or after second [00:16] <burek> and then you can also use -itsoffset -2 [00:16] <burek> if you want to go to another way [00:16] <Kingsy> ah I see [00:16] <Kingsy> ok I will mess arund with that [00:16] <burek> ok :) [00:16] <burek> have fun :) [00:19] <Kingsy> burek: thanks so much for your help with that [00:19] <burek> :beer: :) [00:30] <Kingsy> damn right.. you earned a beer :) [01:39] <echelon> what causes this.. av_interleaved_write_frame(): Operation not permitted [01:40] <echelon> it doesn't happen when i use the lossless_ultrafast preset [01:40] <echelon> just when i use hq [01:40] <echelon> for libx264 [01:41] <burek> I would suggest you to ask in #libav since I was thinking it's just a library usage support [01:41] <burek> but I found out today it's a fork of ffmpeg [01:41] <echelon> oh [01:41] <burek> so I don't know.. google? :) [01:41] <echelon> i did [01:42] <burek> did you get that error by running ffmpeg cmd line tool [01:42] <echelon> yes [01:42] <burek> or using it programmatically [01:42] <echelon> the ffmpeg tool [01:42] <burek> can you please use pastebin.com, to show your command line and its output? [01:45] <echelon> http://pastebin.com/6wfivCw3 [01:46] <burek> hmh [01:46] <burek> at first look [01:47] <burek> your ffmpeg looks very old [01:47] <burek> a lot of features/bug fixes are being made each day [01:47] <burek> so 6 months old ffmpeg is really a red alert [01:48] <echelon> hmm ok [01:48] <burek> you can try [01:48] <burek> https://sites.google.com/site/linuxencoding/builds [01:48] <burek> just download and run the same command [01:48] <burek> to see if it will work [01:48] <burek> if it does, then update your ffmpeg [01:51] <echelon> Option 'croptop' has been removed, use the crop filter instead [01:51] <echelon> how do i use that [01:52] <burek> yes, you type something like -vf 'crop=x:y:z' let me find manual [01:52] <burek> http://ffmpeg.org/ffmpeg.html#crop [02:14] <echelon> i don't understand it [02:15] <burek> what exactly [02:15] <echelon> let's say i wanted to crop top 44 and crop bottom 44 on a video whose original resolution is 936x616 [02:15] <echelon> what would be the parameters for -vf crop [02:16] <burek> crop=in_w-88:in_h [02:16] <burek> no [02:16] <burek> crop=in_w:in_h-88 [02:17] <Freakshow> w and h being width and height respectively? [02:17] <burek> as written [02:17] <echelon> -vf crop=out_w:out_h:x:y [02:17] <burek> -vf 'crop=in_w:in_h-88' [02:18] <echelon> ok [02:18] <echelon> how does it tell it to from where to crop [02:18] <burek> it crops from center if you dont specify 3rd and 4th param [02:18] <burek> because it is the majority usage [02:18] <echelon> mm ok [02:19] <echelon> [crop @ 0x8085420] Invalid too big or non positive size for width '936' or height '528' [02:19] <echelon> Error opening filters! [02:19] <burek> can you please use pastebin.com, to show your command line and its output? [02:21] <echelon> http://pastebin.com/Xy45ATfr [02:21] <burek> that's not what I wrote [02:22] <echelon> w:936, h:616-88 [02:22] <burek> also, if you want to crop your input [02:22] <burek> and not output [02:22] <burek> then move -vf before first -i [02:23] <burek> you don't have to calc those values [02:23] <burek> ffmpeg -vf 'crop=in_w:in_h-88' -i WMCap\[2\].mpg -t 1284 -an -vcodec libx264 -s 800x500 -aspect 16:10 -preset ultrafast video1.mp4 [02:24] <echelon> ^_^ [02:24] <burzum> hi [02:27] <burzum> when im converting audio from a file that has no extension does ffmpeg automatically get the source format and knows how to process it right? so in other words does it matter if i use a wave file for input or a flac to convert both to mp3? [02:27] <burek> it doesn't look at extensions at all [02:28] <burek> it analyzes data [02:28] <burzum> good, because for the output i had to specify -f if i was writing an extensionless file [02:29] <burek> well yes :) how else would it know what you want :) [02:29] <burzum> i thought -acodec was enough ;) [02:29] <burek> well yes, but the same -acodec can be put into various formats (muxers) [02:29] <burek> mp4, mkv, flv, mp3, ... [02:30] <burek> http://uk.answers.yahoo.com/question/index?qid=20100808145033AADpyVX [02:32] <burzum> thanks im aware of that i just didnt know about that ffmpeg makes a difference here and was throwing errors when i have not used the -f param or whatever exactly it was doing, i dont remember [02:32] <burek> yes, because -acodec mp3 can be put into mp3 container, or avi, or flv, etc.. [02:33] <burek> and when it's about input then ffmpeg loads first part of the file in the memory and analyzes it [02:33] <burek> to see what is it actually [02:34] <burzum> but two more things i could not really figure out (i havent had to deal with audio stuff before...): how can i set a certain audiobitate for ogg vorbis? mp3 is using ab 256k for example but -aq and a value from 0 to 100 is used for ogg. how can i make it accept also 256k for example? i read that the whole thing is using a dynamic bitrate but i would like to somehow figure out whats the best [02:34] <burzum> quality/compression and preferable set it in like i do it for mp3s [02:35] <burzum> my use case is that i have wav files, convert them to flac which become the master file from which i create ogg and mp3 [02:35] <burek> -ab 256k will generally set audio quality [02:35] <burek> no matter is it mp3 or ogg [02:35] <burek> unless ogg is a codec that doesn't deal with bitrates at all, just quality indicator [02:35] <burek> but that's high unlikely [02:35] <burek> highly [02:36] <burek> burzum, did you try aac+ ? :) [02:37] <burek> it compresses several times more than mp3 with better quality [02:37] <burek> meaning you get smaller files with better quality than mp3 [02:37] <burzum> looks like you have the experience, so how would you create a mp3 file that has a very high quality? go for 256k (from what i read 192k should be already good enough) or use vbr? if vbr how do i get ffmpeg to write a vbr mp3? [02:37] <burzum> ive heard about aa+ but didnt know that [02:38] <burek> well, are you on windows [02:38] <burzum> yes [02:38] <burzum> well, yes and no [02:38] <burek> download NeroAACEnc [02:38] <burek> and try [02:38] <burzum> i develop on windows, the final app is going to run on a nix system [02:38] <burek> read the help for parameters and try to convert one file to see the result [02:38] <burzum> would that aa+ in an mp3 container play well on all hardware players to? [02:38] <burek> but make it aac+ (a.k.a. HE-AAC v2) [02:39] <burek> it depends on hardware [02:39] <burek> old hardware will probably fail [02:39] <burek> but all the new hardware is already aac+ aware [02:40] <burzum> the goal is to produce files that are playable on most hardware players but also to not have to produce 10 formats because of keeping the storage low. i do not want to have thousands of files for a single song eating up hundreds of mb [02:40] <burek> also, for vbr mp3, use quality selector in ffmpeg and not bitrate selector [02:41] <burek> mp3 is mostly supported [02:41] <burek> but is very old too [02:41] <burzum> i know, thats why ive asked about the aa+ compatibility with hardware players [02:41] <burek> so don't expect it to cover all your goals [02:41] <burek> aac+ [02:42] <burzum> so far i plan to offer flac, ogg vorbis and mp3 [02:42] <burek> well, if you offer flac [02:42] <burzum> this should satisfy all kinds of possible customers [02:42] <burek> then your chances are better with aac+ because it's older than flac [02:42] <burzum> but less known or? [02:43] <burek> AAC is from 1997 [02:43] <burek> FLAC started around 2000 or 2001 [02:44] <burzum> yes, but its not about the year but about the public perception [02:44] <burek> for that, I'm afraid I'm not competent :) [02:44] <burzum> even if aac would be from 1900 but people and hardware know it less well than flac the choice is clear [02:45] <burzum> i personally would always go for bleeding edge technology ;) [02:45] <KaosTheory-> good evening [02:45] <KaosTheory-> any devs here at all? [02:46] <burzum> the way the code is written its not a problem to reprocess the audio later into a new format [02:46] <KaosTheory-> or anyone who can help me with a problem with ffmpeg [02:46] <burek> KaosTheory-, what's the problem you have? [02:46] <burek> burzum, I agree [02:47] <burzum> i think i simply put the aac+ stuff in, 5min coding task [02:47] <KaosTheory-> I have a 1980x1080 avi file which i am trying to convert using ffmpeg i type the necessar command line (basic -i <file><File>) [02:47] <KaosTheory-> however ffmpeg just hangs and no encoding commenses [02:48] <burek> can you please use pastebin.com, to show your command line and its output? [02:48] <KaosTheory-> two secs [02:48] <burek> make it three :) [02:48] <burek> it's on the house :) [02:49] <KaosTheory-> :P [02:50] <burek> can you please paste it here [02:50] <burek> because the channel is being logged [02:50] <burek> for others to read [02:50] <KaosTheory-> ah nps [02:50] <KaosTheory-> http://pastebin.com/n1FpfX0x [02:50] <burek> hm [02:50] <burek> and then ffmpeg just hangs there? [02:50] <KaosTheory-> yeah [02:51] <KaosTheory-> does nothing [02:51] <burek> seems like a bug [02:51] <KaosTheory-> get the win7 'program not responding' popup [02:52] <KaosTheory-> so its not me doing something stupid then [02:52] <burek> KaosTheory-, do you know how to use ftp [02:52] <burek> and is that video some private video [02:52] <burek> or you can upload it as a sample [02:52] <burek> so that we can fix the bug [02:52] <KaosTheory-> nah just a Tribes Ascend SunStar capping route [02:52] <burek> is it big [02:53] <KaosTheory-> 2.5gb [02:53] <burek> hm [02:53] <KaosTheory-> wonders of using DxTory and the Lagarith Lossless Codec [02:53] <burek> would it be a problem for you to split the first 10 mb of the file [02:53] <burek> and upload it? [02:54] <KaosTheory-> i think you would get about 3 frames of it [02:54] <KaosTheory-> lol [02:54] <KaosTheory-> oh [02:54] <burek> :))) [02:54] <KaosTheory-> is it worth noting that if i run it through vDub the file works for encoding [02:55] <KaosTheory-> well it did for a 50% down scale [02:55] <burek> what did you do in vdub [02:56] <KaosTheory-> just recoded it with zero codec [02:56] <KaosTheory-> well resaved the avi [02:56] <KaosTheory-> with no codec [02:56] <KaosTheory-> with a 50% size filter on it [02:56] <KaosTheory-> and try it now with 100% scale [02:56] <burek> can you run just ffmpeg -i SunStar.avi [02:56] <burek> to see what's the output [02:56] <KaosTheory-> i can in a sec [02:57] <KaosTheory-> saving a seperate avi incase its the Lagarith Codec causing the issue [02:57] <KaosTheory-> ... for some stupid reason [02:57] <KaosTheory-> 80% [02:57] <burek> ok [02:58] <KaosTheory-> well there goes my hdd space [02:58] <KaosTheory-> 9.2gb so far [02:58] <burek> :) [02:58] <burzum> burzum do you know the acodec name for aac+ that ffmpeg is using? [02:58] <burzum> burek do you know the acodec name for aac+ that ffmpeg is using? [02:58] <KaosTheory-> ok that worked at full size [02:59] <burek> -acodec libaacplus [02:59] <KaosTheory-> perhaps codec problem? [02:59] <burzum> thank you! [02:59] <burek> burzum, http://ffmpeg.gusari.org/viewtopic.php?f=25&t=37 [03:00] <burek> KaosTheory-, just run ffmpeg -i SunStar.avi [03:00] <burzum> thanks [03:00] <burek> to see what ffmpeg detected [03:00] <KaosTheory-> burek same thing as before [03:00] <burek> hm [03:01] <burek> it would be great if you could cut first and last 10mb of that file KaosTheory- [03:01] <KaosTheory-> to note though: vdub re-coded with Uncompressed RGB/YCbCr allows the video to be encoded [03:01] <burek> so that we can investigate what's wrong [03:01] <KaosTheory-> ok let me see if i can do that with vDub [03:01] <burek> I know, as soon as you change video stream, it gets fixed [03:01] <burek> just a second [03:01] <burek> not vdub [03:02] <KaosTheory-> ok.. [03:02] <burek> I'm looking into copy dos command [03:03] <burek> it has got some param [03:03] <burek> that says how many bytes to copy [03:03] <burek> just a sec [03:03] <KaosTheory-> not sure how i will cut the first and last 10mb [03:03] <burzum> burek so -aq vs -ab is that ab will get me a fixed bitrate while -aq will create a vbr file with a given average quality (0-100)? [03:04] <burzum> if the codec supports it [03:04] <burek> KaosTheory-, http://www.virtualobjectives.com.au/utilitiesprogs/partcopy.htm :) [03:04] <burzum> and if i do not specify -ar the target file will get the same -ar as the input file? [03:05] <burek> burzum, yes [03:05] <burzum> yes for both? cool [03:05] <burek> yes :) [03:06] <burek> -aq says something like "I want constant quality" [03:06] <burzum> \o/ [03:06] <burek> where -ab on the other hand says "I want constant bit rate" [03:06] <burzum> yea, im going to validate that before i'll check the file into my storage system by parsing the ffmpeg info about that file [03:07] <KaosTheory-> ok part copy made [03:07] <KaosTheory-> well start bit anyway [03:07] <burek> KaosTheory-, http://ffmpeg.org/bugreports.html read the bottom text [03:07] <burzum> it would be a wet dream if ffmpeg would output json or xml [03:07] <burek> so just upload SunStar.avi_begin and SunStar.avi_end [03:08] <burek> with a third (text) file which will contain your pastebin text [03:08] <burek> so that we know what is what [03:08] <burek> ok ? [03:08] <burek> burzum, why do you need json or xml [03:09] <burzum> easier to parse than regexing the whole output [03:09] <burek> searching for what [03:09] <burzum> if you do not know ALL error messages you could still simply check if there are child elements in the <error> node for example and simply get its message [03:10] <burek> well, when you start ffmpeg, you check it's return value [03:10] <burek> if it is not 0 then there were errors :) [03:10] <burek> if that's what you want :) [03:11] <burzum> 2>&1 [03:11] <burek> and if there were errors, display the entire log to the user [03:11] <burek> and let him decide what to do :) [03:11] <burzum> i do not want to show an error log to a user of the web frontend :) [03:11] <burek> then log to a file, for yourself [03:11] <KaosTheory-> oh [03:11] <burzum> just a simple message "Failed because blah blah" [03:12] <KaosTheory-> do you defo want that format burek? [03:12] <burek> defo? [03:12] <KaosTheory-> i have uploaded as SunStar-Start.avi [03:12] <burek> oh, ok :) [03:12] <burek> name doesn't matter that much [03:12] <KaosTheory-> definetly = defo [03:12] <burek> just make it so that we can understand :) [03:12] <KaosTheory-> easy ;) [03:13] <KaosTheory-> where do you want the link? [03:13] <KaosTheory-> here im guessing [03:13] <burek> if you want to report a bug yourself, go here: https://ffmpeg.org/trac/ffmpeg [03:13] <burek> if not, just leave the link here [03:13] <burek> I'll create a bug report [03:13] <burek> :) [03:14] <KaosTheory-> http://www.digitallyhazardous.com/kaos/ff/ [03:14] <burek> oh :) [03:14] <burek> I thought you read the page [03:14] <burek> that I gave you [03:14] <burek> D [03:15] <burek> :D [03:15] <burek> "KaosTheory-, http://ffmpeg.org/bugreports.html read the bottom text" [03:15] <burek> it should have the effect like: "Upload the sample to the MPlayer FTP server." [03:15] <burek> :) [03:15] <KaosTheory-> oh [03:15] <KaosTheory-> lol [03:16] <burek> ^^ [03:16] <burek> do you know how to do it :) [03:18] <KaosTheory-> uploading now [03:19] <burek> thanks :) [03:19] <KaosTheory-> uploaded [03:20] <burek> great :) [03:20] <burek> now, do you want to submit a new bug report [03:20] <burek> or would you like me to do it? [03:20] <KaosTheory-> if you dont mind [03:20] <burek> not at all [03:20] <KaosTheory-> i have enough trouble remembering all the accounts and passwords i have as it is [03:20] <burek> are the names the same [03:21] <burek> file names [03:21] <KaosTheory-> SunStar-Start.avi SunStar-End.avi Error.txt [03:21] <burek> ok [03:22] <burek> https://ffmpeg.org/trac/ffmpeg/ticket/1192 [03:22] <burek> bookmark it so you know when it is resolved if you like :) [03:23] <KaosTheory-> will do [03:23] <KaosTheory-> its probably something that the codec doesnt do [03:23] <KaosTheory-> or misses out that ff isnt taking into account [03:24] <KaosTheory-> as a fully uncompressed 1980x1080 file works fine [03:24] <burek> well, as much as I can se [03:24] <burek> see [03:24] <burek> it can't even read file normally [03:24] <burek> (container) [03:25] <burek> so it didn't even come to the point to show what streams are inside the container [03:25] <KaosTheory-> hmmm [03:25] <burek> if vdub has an option [03:25] <burek> to just copy codecs (audio/video) [03:25] <burek> and just change the format (muxer) [03:25] <burek> then that could help [03:26] <KaosTheory-> you mean video compression audio compression? [03:26] <KaosTheory-> i can select Largarith codec in vDub [03:26] <burek> format is like an envelope in which are two other envelopes (audio and video) [03:26] <burek> so, if you just change the format and leave av streams intact [03:27] <burek> that might help ffmpeg recognize stuff [03:27] <KaosTheory-> ok.... [03:27] <KaosTheory-> sorry its been a while since i did any video editing and stuff [03:27] <KaosTheory-> so im a total noob to it all again [03:28] <burek> it's ok :) [03:28] <burek> noone has been born skilled :) [03:28] <KaosTheory-> unless you happen to be my m8 [03:28] <KaosTheory-> the crap he does and programs just baffles me [03:28] <burek> :) [03:29] <KaosTheory-> on a side notw [03:29] <KaosTheory-> note* [03:29] <burek> does he know how to make beer? :) [03:29] <KaosTheory-> umm [03:29] <KaosTheory-> computer wizz [03:30] <KaosTheory-> not so much alchemical wizz ;) [03:30] <burek> mh :) ok :) it was worth to try asking :)) [03:30] <KaosTheory-> side note: ffmpeg did work on a 1270x720 recording using Lagarith Lossless [03:31] <burzum> burek i know somebody who makes homebrewed beer in a good quality, need some advice? ;) [03:31] <burek> well, the thing here is that [03:31] <burek> ffmpeg didn't get to "unpack" the avi format, to read what streams are inside [03:31] <burek> (one of those streams is your Lagarith) [03:31] <burek> so I doubt the codec is a problem, it is more likely the damaged avi format or something [03:31] <burek> burzum, I need m8 that makes beer :D [03:31] <KaosTheory-> dunno [03:32] <KaosTheory-> i can only go with what i see, have experienced [03:32] <burzum> haha, who doesnt? :) [03:32] <KaosTheory-> burek i have better. i have mates that buy me beer :D [03:32] <burek> KaosTheory-, did you manage to play that file in vlc [03:32] <burek> or that :D [03:32] <KaosTheory-> the vid works fine in every player [03:32] <KaosTheory-> well Windows Media at least [03:33] <burek> can you open vlc and go to tools - messages [03:33] <KaosTheory-> it gitters [03:33] <burek> set the verbosity level to 2 [03:33] <burek> and then play that file [03:33] <KaosTheory-> but im guessin thats down to it being 2.5gb [03:33] <burek> just until it starts [03:33] <burek> then stop it and copy the log [03:33] <KaosTheory-> ok hang on [03:33] <iive> KaosTheory-: you may have already said that, but I'd ask anyway. [03:34] <KaosTheory-> downloading vlc [03:34] <iive> did you make the file on your own, or have you gotten it from somewhere else? [03:35] <KaosTheory-> iive the vid was grabbed using Dxtory [03:36] <iive> is it dv pro capture? [03:37] <KaosTheory-> burek http://pastebin.com/yyR8NDQT [03:37] <KaosTheory-> iive please elaborate [03:38] <iive> i see it is dx/opengl capture program, like fraps. [03:38] <KaosTheory-> yeah [03:38] <iive> do you know what codec does it use? [03:38] <KaosTheory-> i am using the Lagarith Lossless Codec [03:38] <KaosTheory-> for a compramise between quality and HDD space [03:39] <KaosTheory-> not to mention to retain fps within game [03:39] <iive> i think i've heard that one. wasn't it lossless? [03:39] <burzum> burek if i use -aq 15 or -aq 85 it doesnt matter, the mp3 always comes out with 96kbps? why? [03:39] <KaosTheory-> yeah as far as i understand it, and as my mate told me, its basically a huffmans codec [03:40] <iive> huffman is also lossless :) [03:40] <KaosTheory-> ;) [03:40] <burek> KaosTheory-, I updated your bug report with that VLC output :) https://ffmpeg.org/trac/ffmpeg/ticket/1192 [03:40] <iive> can you try to make few more files. [03:40] <KaosTheory-> i already have loads [03:41] <KaosTheory-> none of the 1980x1080 files work [03:41] <iive> what is the biggest file that works? [03:41] <burek> just curious, are all those that don't work, bigger than 2 gb [03:41] <KaosTheory-> been racking my brains on this for two days [03:41] <KaosTheory-> ummm [03:41] <KaosTheory-> sec [03:43] <KaosTheory-> ok resolution it worked on 1280x720 [03:44] <KaosTheory-> and no just tested a 113,453kb file and it is the same as before [03:45] <KaosTheory-> lol [03:46] <KaosTheory-> not even a 1280x720 wants to work now :/ [03:46] <burek> same error? [03:46] <KaosTheory-> same 'freezing' [03:46] <iive> KaosTheory-: the resolution is not important. Focus on the file size [03:46] <burek> with "non-interleaved AVI" message? [03:46] <KaosTheory-> yeah [03:47] <iive> can you try making even smaller files? 10mb one? [03:47] <burek> try typing ffmpeg -i filename [03:47] <KaosTheory-> same [03:47] <burek> and if you find a small one (10 mb) [03:47] <burek> can you upload it [03:48] <iive> KaosTheory-: does the program have auto split/segment functionality? [03:49] <KaosTheory-> lol i double tapped the record button. 53mb [03:49] <KaosTheory-> lol [03:49] <KaosTheory-> same error on that [03:49] <iive> good enough :) [03:49] <iive> even on that! [03:49] <KaosTheory-> yep [03:50] <iive> try smaller resolution. [03:50] <burek> omg :) [03:50] <KaosTheory-> that was using 1280x720 [03:50] <iive> computers should still be capable of 320x200 :P [03:50] <burzum> if i use -aq 15 or -aq 85 it doesnt matter, the mp3 always comes out with 96kbps. am i using it wrong? [03:50] <KaosTheory-> lolololol [03:51] <burek> burzum, [03:51] <burek> can you please use pastebin.com, to show your command line and its output? [03:51] <burzum> burek ok [03:51] <KaosTheory-> 800x600 smallest resolution. 36mb. same 'freeze' [03:52] <KaosTheory-> need drink and food brb [03:52] <iive> burzum: aq may be algorithmic quality. try -ab 128k [03:53] <burek> he wants vbr [03:53] <burzum> burek http://pastebin.com/uXce5VZ7 [03:54] <burek> burzum, try using -q instead of -aq [03:55] <burzum> 'ffmpeg: unrecognized option \'-q\'', [03:55] <burek> ok :D wait :) [03:55] <burzum> ive already looked at the documentation :) [03:56] <burek> burzum, your ffmpeg is really old.. [03:56] <burek> and i mean really old [03:56] <burzum> ok [03:56] <burek> like 3 years ago [03:56] <burek> or more [03:56] <burzum> where do i get a new windows build that is reliable and has liblame, flac and all the stuff i already use? [03:57] <burzum> ok just found that [03:57] <burzum> hmm looks like the page ive got the current one from [03:57] <KaosTheory-> right im back [03:57] <burek> http://ffmpeg.zeranoe.com/builds/ [03:58] <iive> try with "compression_level" [03:59] <iive> that one is for algorithmic quality. [03:59] <KaosTheory-> iive is that to me or burzum? [03:59] <iive> burzum: [03:59] <KaosTheory-> k [03:59] <burzum> ok [03:59] <KaosTheory-> just checking [04:00] <iive> burzum: try also "global_quality", this one must be for VBR. [04:01] <burzum> compression_level doenst work, 64k, always [04:03] <iive> burzum: add qscale to the mix. it sets the flag. [04:04] <burek> burzum, also try -aq 2 [04:04] <iive> so, you need qscale, to enable it, global_quality to set the level. [04:05] <burek> http://ffmpeg.org/pipermail/ffmpeg-devel/2011-April/110717.html [04:05] <burek> ffmpeg -i a.wav -aq 2 b.mp3 "corresponds to" lame -V2 a.wav b.m3 [04:05] <burek> and if you take a look here: http://wiki.hydrogenaudio.org/index.php?title=LAME [04:05] <burek> -V0 (~245 kbps), -V1 (~225 kbps), -V2 (~190 kbps) or -V3 (~175 kbps) [04:06] <burek> so -aq 0 will give 245 kbps, -aq 1 = 225, -aq 2 = 190 ... [04:07] <iive> KaosTheory-: btw, first I asked you about DV codec, because avi handles DV in a special way. But that's not the case. [04:07] <KaosTheory-> ok [04:08] <iive> KaosTheory-: then I asked you about size, because avi files above 1GB (or max 2GB) need a special index to work. something called ODML2.0 [04:08] <iive> but with 53MB file, this doesn't seem to be case either. [04:08] <KaosTheory-> ok [04:08] <KaosTheory-> yeah still 'froze' [04:08] <KaosTheory-> exact same log as before [04:09] <iive> so now, you just have to make the smallest files you can do, and upload it to the bug tracker, so active developers can try to reproduce the problem on their own [04:09] <KaosTheory-> tried serveral resolutions of the game, same issue [04:09] <KaosTheory-> i have uploaded a start and end of the vid [04:09] <burek> burzum, use this table for your -aq <param> http://wiki.hydrogenaudio.org/index.php?title=LAME#VBR_.28variable_bitrate.… [04:09] <KaosTheory-> 10mb in size [04:10] <KaosTheory-> with error log [04:10] <KaosTheory-> its on the ftp already [04:10] <iive> KaosTheory-: it may be better to have a whole file. [04:10] <burek> KaosTheory-, if you could upload a complete that would be the best, that's why we are asking you if you can produce ~10mb file :) [04:10] <burek> complete file* :) [04:11] <KaosTheory-> lol [04:11] <KaosTheory-> i couldnt [04:11] <burzum> burek thanks again :) already looking at it but now ive got confused by ARB vs VBR... im trying to figure out what will give me the best audio quality while keeping the file small. [04:11] <KaosTheory-> smallest i could get with a doublt tap of the record button was 53mb [04:12] <burek> burzum, those are always the opposite requirements.. you want it to be the smallest size, the best quality and the most widely supported :) [04:12] <burek> those 3 never go along :) [04:12] <iive> KaosTheory-: try the same, with even smaller resolution [04:13] <burzum> i prefer better audio quality over filesize but im trying to figure out an economic value [04:13] <iive> KaosTheory-: but if you can't 53MB is ok for an ftp. [04:13] <iive> well. [04:13] <iive> have fun. n8 ppl [04:14] <burek> o/ [04:14] <burzum> burek i guess given my requirements i would go for CBR and need to decide which one i want to use [04:14] <burek> why cbr [04:14] <burzum> constant bitrate [04:14] <burek> i just gave you the table for vbr [04:14] <burek> did you look at it? [04:15] <burzum> well but you said: you want it to be the smallest size, the best quality and the most widely supported :) [04:15] <burzum> yes [04:15] <burek> why are you talking about cbr then :) [04:16] <burzum> as I said im trying to figure out the best compromise of filesize, quality [04:16] <burzum> crap, its 04:14 here... i had the hope this would be easy to figure out [04:16] <burzum> i need to get some sleep :s the night is going to be a short one [04:16] <burek> ok :) [04:17] <burek> have a good sleep [04:17] <burek> tomorrow might give you the answers :) [04:17] <burzum> burek thanks a lot for your great support [04:17] <burek> :beer: :) [04:17] <burzum> coffee! i dont drink alcohol during the work week :) [04:17] <KaosTheory-> perfect [04:17] <KaosTheory-> managed to get a 5mb file [04:17] <burek> :) [04:18] <KaosTheory-> my finger hurts now. :/ [04:18] <KaosTheory-> thats was a quick dbl tap. lol [04:18] <burek> :D [04:19] <KaosTheory-> DryDock.avi [04:19] <KaosTheory-> uploading to ftp now [04:19] <burek> thanks :) [04:20] <KaosTheory-> done [04:20] <burek> does it also give "non-interleaved AVI" [04:22] <KaosTheory-> lol [04:22] <KaosTheory-> that worked [04:22] <burek> what [04:22] <KaosTheory-> it encoded it [04:22] <burek> well we need the smallest that still produces that err msg [04:22] <burek> :) [04:22] <KaosTheory-> urgh [04:23] <KaosTheory-> man thats gonna be an ass to find [04:23] <KaosTheory-> ok hang on [04:23] <burek> :) [04:23] <burek> just try "ffmpeg -i filename" on some small files that you already have [04:24] <KaosTheory-> LOLOLOLOL [04:24] <KaosTheory-> that would be why [04:25] <KaosTheory-> wasnt using the right codec [04:25] <KaosTheory-> hang on ;) [04:26] <KaosTheory-> nope still cant pull it lower than that 53mb [04:26] <KaosTheory-> and still wont encode it [04:27] <burek> well, just check if [04:27] <KaosTheory-> it must be something to do with the Lagarith Codec [04:27] <burek> ffmpeg -i [04:27] <burek> still gives that error [04:27] <KaosTheory-> the Cinepack worked fine [04:27] <burek> if it does, upload that 53mb file [04:27] <burek> we'll note that we couldn't produce any smaller file [04:27] <burek> :) [04:27] <KaosTheory-> nope same [04:28] <KaosTheory-> deleted the 53mb one [04:28] <KaosTheory-> smallest i got now is the 113mb one [04:28] <KaosTheory-> is that ok? [04:28] <KaosTheory-> oh sweet [04:28] <KaosTheory-> got a 59mb one [04:28] <burek> :) [04:29] <KaosTheory-> lol [04:29] <KaosTheory-> thats 2 seconds worth of video [04:29] <burek> they will look into binary data [04:29] <burek> not the images itself [04:29] <KaosTheory-> i know ;) [04:30] <KaosTheory-> ok [04:30] <KaosTheory-> TribesAscend.avi [04:30] <KaosTheory-> 59mb being uploaded [04:30] <KaosTheory-> the DryDock.avi can be scrapped [04:30] <KaosTheory-> unless you want it for evidence that its not DxTory causing the issue [04:30] <burek> ok, so that is the smallest that produces that non-interleaved AVI error msg [04:30] <KaosTheory-> yeah [04:30] <burek> ok [04:30] <KaosTheory-> smallest i can get [04:31] <KaosTheory-> want to try and keep the 1920x1080 resolution incase that has some random effect on it [04:31] <KaosTheory-> i know the likely hood is very very slim [04:31] <KaosTheory-> but just to ensure continuity [04:31] <burek> is this ok? https://ffmpeg.org/trac/ffmpeg/ticket/1192 [04:32] <KaosTheory-> yeah perfect [04:32] <burek> ok :) you have rss link at the bottom [04:32] <burek> so you can get instant info [04:32] <burek> when it is resolved :) [04:33] <KaosTheory-> i will just check daily ;) [04:33] <burek> or that :) [04:33] <KaosTheory-> many thanks to you for your help burek [04:33] <burek> :beer: :) [04:34] <KaosTheory-> i dont mind even if its the codec [04:34] <KaosTheory-> just so long as i know [04:34] <KaosTheory-> can then try and find a codec that works ok [04:34] <burek> well, I think it's a glitch in matrix, which will be fixed :) [04:34] <KaosTheory-> ok thats awesome [13:22] <JamesJJ> OK... still struggling with a flv with G711 audio. the problem is that ffmpeg is seeing the audio track but doesn't reconise the codec - I can force it to use the right codec by adding -acodec pcm_mulaw, but it is also getting the frequency wrong - Stream #0.1: Audio: pcm_mulaw, 5512 Hz, mono, s16, 44 kb/s - is there a way to change the input frequency, as I as the input data is 8000 not 5500? [13:24] <burek> can you please use pastebin.com, to show your command line and its output? [13:31] <burek> JamesJJ [13:48] <hjt> :-& [14:34] <darkb0e> hello :) [14:36] <hjt> LOL [14:36] <hjt> :-& [14:36] <hjt> LOL [14:45] <fling> hjt: :3 [14:58] <darkb0e> do you know any tutorial for streaming h264/mp3 to a flash video container? [14:58] <darkb0e> anyone/ [14:58] <darkb0e> ? [14:58] <darkb0e> i am trying with ffmpeg -> ffserver but with no luck [14:59] <burek> streaming using what protocol [15:00] <darkb0e> i am trying the 3 default that are in /etc/ffserver.conf [15:00] <burek> well what is your goal [15:00] <darkb0e> mpeg1 swf, ans asf [15:00] <darkb0e> just for testing [15:01] <burek> ffmpeg -i input.avi -vcodec libx264 -acodec libmp3lame -f mpegts udp://remoteserver:port [15:01] <darkb0e> i want to transcode with ffmpeg vcodec x264 and acodec mp3 to the ffserver, then ffserver will stream to http flv [15:02] <burek> oh [15:02] <burek> then use ffmpeg -i input.avi http://localhost:8090/feed1.ffm [15:03] <burek> and all the rest, just setup in the conf file [15:03] <darkb0e> w8 to paste bin my command and config [15:04] <burek> and also, read this thoroughly http://ffmpeg.org/ffserver.html [15:06] <darkb0e> http://pastebin.com/CfusGqRM [15:07] <darkb0e> i read the documentation [15:07] <burek> read my lines above [15:09] <burek> also, using udp input, you'll have a lot of trouble with sync-ing [15:10] <darkb0e> i was using vlc [15:10] <darkb0e> but, it crashes like hell [15:10] <darkb0e> and now i am trying to find alternative [15:15] <darkb0e> burek: you said that i need to run ffmpeg twice? or you gave me 2 examples of ffmpeg/ [15:15] <darkb0e> ? [15:16] <darkb0e> i am running first the ffserver with the config mentioned in the paste bin, and then i run ffmpeg with the arguments mentioned in the pastebin [15:28] <darkb0e> burek: are you here? [15:54] <fenrig> Hi I want to know how much I B and P frames there are in a certain video file, so I was wondering if I could do this with ffmpeg [15:57] Last message repeated 1 time(s). [16:02] <JamesJJ> Sorry for the delay - pasetbin on trying to use FLV with g711 [16:02] <JamesJJ> http://pastebin.com/3PTK5YA6 [16:03] <JamesJJ> I guess the important bit is from the flv demuxer - flv @ 0x148ac00]Unsupported audio codec (8) [16:06] <JamesJJ> G711 has been supported by flash and FLV for a while now, if there was a way to override the frequency of the input stream I would be able to get round this [16:10] <burek> darkb0e, no, just the 2nd ffmpeg line [16:11] <burek> fenrig, use ffprobe? [16:11] <darkb0e> ok, i managed to start it, but now i run into trouble with x264 presets :/ [16:12] <darkb0e> anyone with any link? with valid video sizes etc? [16:12] <fenrig> burek: THANK YOU i'll look into it [16:12] <darkb0e> thank you burek [16:12] <burek> JamesJJ, your ffmpeg is way to old [16:12] <burek> 3 years or more.. [16:12] <darkb0e> burek: thank you [16:12] <burek> :beer: :) [16:12] <burek> darkb0e, [16:12] <burek> can you please use pastebin.com, to show your command line and its output? [16:13] <JamesJJ> ah ok... more fool me for trying it on windows - I'll try a copy from the repo [16:13] <darkb0e> ofc [16:13] <burek> JamesJJ, if you use ffmpeg on windows [16:16] <darkb0e> burek: http://pastebin.com/WaSuDyTQ [16:16] <burek> where is the error msg? [16:17] <darkb0e> no error, just doesn't start to show video on vlc [16:17] <darkb0e> seems like buffering forever [16:17] <burek> what url did you used to start the stream [16:17] <darkb0e> http://<server>:8090/test.flv [16:18] <darkb0e> it shows up in the stats page [16:18] <darkb0e> but it says wait_feed [16:18] <darkb0e> although the counter up'ing [16:18] <burek> are you using latest vlc/ffmpeg [16:20] <darkb0e> yeah compiled from wiki's guide [16:20] <darkb0e> 4 times today :) [16:20] <burek> which wiki [16:20] <darkb0e> ffmpeg's wiki [16:20] <darkb0e> ubuntu compilation [16:20] <burek> do you happen to have url for that [16:20] <darkb0e> ofc [16:21] <darkb0e> https://ffmpeg.org/trac/ffmpeg/wiki/UbuntuCompilationGuide [16:21] <ckb> okay guys. 3 renditions from source: 1200Kb/s, 700Kb/s, 400Kb/s. Would it make a difference encoding in a linear route? (Encoding the 700 from the 1200 rather than src, and encoding the 400 from 700 rather than src) [16:22] <burek> darkb0e, ok, that's ok [16:22] <burek> now try opening another console [16:22] <burek> and type ffmpeg -i http://localhost:8090/test.flv [16:22] <Tjoppen> ckb: you'll get worse quality then doing it properly [16:22] <ckb> That's what I was assuming [16:22] <ckb> Tjoppen: is it significant? [16:23] <ckb> or can I get away with it? [16:23] <burek> ckb, why doing that if you can do it all in one cmd [16:24] <burek> from 1 source [16:24] <ckb> burek [16:24] <darkb0e> burek: http://pastebin.com/FzEeivbn [16:25] <Tjoppen> indeed. it'll be faster to encode everything in one go [16:25] <burek> ffmpeg -i -vcodec ... -acodec ... -b 1200k out1.avi -b 700k out2.avi -b 400k out3.avi [16:25] <ckb> seriously? [16:25] <ckb> it tells me -b is ambiguous [16:25] <burek> try -b:v [16:25] <ckb> this works with dual pass as well? [16:25] <Tjoppen> he might have to specify the stream. the syntax changed a bit recently [16:26] <burek> darkb0e, your udp input might be poor [16:26] <burek> you could use some more frequent gop's [16:26] <burek> to increase seekability [16:27] <ckb> I didn't realize you could have multiple outputs with one pass [16:27] <burek> ckb, you can set the 1st pass logfile [16:27] <burek> for each output [16:27] <ckb> burke, I'm not worried about the logfile [16:27] <darkb0e> its 3 mbits input [16:27] <ckb> burek** [16:27] <darkb0e> 720x576 [16:27] <burek> darkb0e, from what? [16:27] <ckb> how do I specify frame size for multiple outputs [16:27] <burek> ffmpeg also? [16:27] <ckb> here's my syntax: one sec [16:27] <darkb0e> gop 20? [16:27] <burek> ckb, frame size? [16:28] <ckb> yeah burke.. one sec I'll show you my cmds [16:28] <burek> darkb0e, how do you generate udp input [16:28] <ckb> ugh typos :( [16:28] <darkb0e> don't know, service provider sends me [16:30] <burek> it seems ffmpeg dropped support for -analyzeduration [16:30] <burek> too bad [16:31] <burek> now you are left behind :S [16:31] <darkb0e> burek: YEAH :) thanks man, gopsize 200 works [16:31] <burek> :) [16:32] <ckb> burek: http://pastebin.com/bpdTArvp this is a single 2 pass encoding of 700Kb/s& If I wanted to make 2 more outputs how would I specify the frame size/bitrates/etc [16:32] <ckb> right now. I'm doing this 2 other times for 1200 and 400 (all with 2 different frame sizes) [16:32] <burek> ckb what is a "frame size" ? [16:33] <burek> resolution? [16:33] <ckb> resolution? [16:33] <ckb> yes [16:33] <burek> :) [16:33] <burek> well ffmpeg's syntax is: ffmpeg -i input <options1> <output1> <options2> <output2> ... [16:33] <ckb> & lol [16:33] <ckb> that's too easy [16:34] <burek> so -s ... -b ... out1.avi -s ... -b ... out2.avi ... [16:34] <ckb> burek: thanks [16:34] <burek> :beer: :) [16:34] <ckb> :cheers: [16:44] <woshty> is -c:v or -vcodec the newer syntax? [16:44] <burek> yes [16:45] <burek> c:v [16:45] <woshty> since what version? [16:45] <burek> check out changelog [16:48] <woshty> burek: http://ffmpeg.org/changelog.html ends with 0.6 [16:49] <burek> http://ffmpeg.org/releases/ [16:51] <woshty> hmm, i am running 0.10.2 but -c:v doesnt work like -vcodec does [16:51] <ckb> burek: why is my bitrate so damn slow when having 3 output files? [16:51] <burek> woshty, that should be a bug then [16:51] <burek> ckb, ? [16:51] <ckb> like 11/Kbits/s [16:52] <burek> well your cpu encodes it at the same time (all of them) [16:52] <burek> so wait until it finishes [16:52] <ckb> so that's why it's so slow? [16:52] <burek> :) [16:52] <woshty> burek: might be debian specific, will check that, thx [16:53] <ckb> burek: http://pastebin.com/K6eRqDsG [16:53] <ckb> burek: does that look right?' [16:53] <burek> ckb, look at Stream #1:0 [16:53] <burek> 700 kb/s [16:53] <burek> it's ok [16:54] <burek> the overall kbps is 11kbps [16:54] <burek> because you have a slow cpu [16:54] <burek> but when it finishes [16:54] <burek> they'll all be ok [16:54] <ckb> ah [16:54] <burek> also you can use -t [16:54] <burek> to speed up tests :) [16:54] <ckb> tests? [16:54] <ckb> Not testing ;) [16:55] <burek> 20_test-cactus-creek_1200k.mp4 [16:55] <ckb> eh [16:55] <ckb> that's just the title of the video [16:55] <ckb> that I put in [16:55] <burek> ok then :) [16:55] <burek> wait until it finishes :) [16:55] <ckb> how can I upgrade the overall speed it encodes? [16:55] <ckb> better hardware? [16:55] <burek> usually :) [16:55] <ckb> more ram? more cores? higher speed? [16:56] <burek> more cpu :) [16:56] <ckb> ah [16:56] <burek> encoding is all about cpu :) [16:57] <ckb> now generally speaking [16:57] <ckb> if ONE output fails, will all others? [16:58] <burek> why would it fail? [17:00] <ckb> no idea [17:01] <ckb> I just want to ensure 100% that it's going to encode [17:02] <burek> then check the return value from ffmpeg process [17:03] <burek> if it is 0 everything is ok [17:03] <burek> KaosTheory-, check the trac :) [17:03] <KaosTheory-> mr. burek [17:03] <KaosTheory-> did [17:03] <KaosTheory-> same problem [17:03] <burek> I think it would be better [17:03] <burek> to register some dummy account [17:03] <burek> just to resolve that [17:03] <KaosTheory-> http://ffmpeg.zeranoe.com/builds/win64/static/ffmpeg-20120409-git-6bfb304-w… [17:04] <KaosTheory-> downloaded that revision of the encoder [17:04] <ckb> burek: the return value? [17:04] <ckb> burek: not the output? [17:04] <burek> ok [17:04] <KaosTheory-> exactly as before [17:04] <KaosTheory-> hangs [17:05] <burek> KaosTheory-, can you try downloading the latest build [17:05] <burek> (today's) [17:05] <burek> just to see if it works [17:05] <burek> ckb, when you start the ffmpeg process [17:05] <burek> you can check its return value [17:05] <KaosTheory-> umm [17:05] <ckb> well... [17:05] <ckb> I'm using PHP to execute ffmpeg [17:05] <burek> even better [17:05] <KaosTheory-> where can i find the 'todays' build? [17:05] <burek> which function [17:05] <burek> exec? [17:05] <ckb> shell_exec [17:06] <ckb> shall I use exec? [17:06] <burek> yes [17:06] <ckb> the return value of exec is the return value of the command? [17:06] <burek> and check out return_var [17:06] <burek> no [17:06] <burek> return_var [17:06] <burek> http://www.php.net/manual/en/function.exec.php [17:06] <ckb> ah [17:07] <ckb> so I can have OUTPUT and return [17:07] <burek> KaosTheory-, are there any today's build at zeranoes? [17:07] <ckb> burek: thanks for all your help, so much [17:07] <burek> ckb, output will be in $output [17:07] <burek> if you provide it [17:07] <burek> :beer: :) [17:08] <KaosTheory-> to confirm link to builds is: http://ffmpeg.zeranoe.com/builds/ [17:08] <ckb> exec($passone, $output, $encoded); -> $output to log, and $encoded == 0 if all is OK? [17:08] <burek> http://www.php.net/manual/en/function.exec.php [17:09] <ckb> right [17:09] <ckb> I'm looking at it :D [17:12] <KaosTheory-> burek if that link above is the correct one. then the lastest build is 2012-04-09 [17:12] <KaosTheory-> thats the one i have just downloaded [17:12] <KaosTheory-> and it didnt work [17:12] <ckb> burek: you said if ffmpeg's return is 0, all is ok? [17:13] <burek> KaosTheory-, hmh.. [17:13] <burek> well ok [17:13] <burek> do you have link to trac [17:14] <burek> ckb yes [17:14] <KaosTheory-> http://ffmpeg.zeranoe.com/builds/win64/static/ffmpeg-20120409-git-6bfb304-w… [17:14] <burek> KaosTheory-, trac, not build [17:14] <KaosTheory-> thats the revision i downloaded [17:14] <KaosTheory-> oh [17:14] <KaosTheory-> trac? [17:14] <burek> bug report [17:14] <KaosTheory-> oh [17:14] <KaosTheory-> hang on [17:14] <JamesJJ> OK... tried with a copy of ffmpeg from svn. Still doesn't like G711 in an flv - Stream #0:1: Audio: pcm_mulaw, 5512 Hz, 1 channels, s16, 44 kb/s - http://pastebin.com/zfSxYaRE [17:15] <JamesJJ> looks from that it is the flv demuxer at issue [17:15] <burek> JamesJJ, svn has been abandoned a long time ago [17:15] <burek> git is used [17:15] <JamesJJ> sorry 0 from git [17:15] <KaosTheory-> http://pastebin.com/SYub7kF2 [17:18] <burek> KaosTheory-, I was looking for this :) https://ffmpeg.org/trac/ffmpeg/ticket/1192 [17:18] <burek> JamesJJ, if you are using the latest git [17:18] <burek> can you please use pastebin.com, to show your command line and its output? [17:19] <KaosTheory-> oh [17:19] <KaosTheory-> sorry burek [17:19] <KaosTheory-> lol [17:20] <KaosTheory-> burek do you think its worth mentioning about the Lagarith Lossless Codec being used? [17:20] <JamesJJ> burek - http://pastebin.com/zfSxYaRE [17:21] <burek> KaosTheory-, if you've read that bug report you'd see that ffmpeg of the developer has detected that encoder [17:22] <burek> JamesJJ, can you do just ffmpeg -i camera.stream.flv [17:23] <KaosTheory-> so it has ;) [17:24] <JamesJJ> http://pastebin.com/u0hEtJsG [17:25] <JamesJJ> flv @ 0x24c4040] Could not find codec parameters (Audio: none ([8][0][0][0] / 0x0008), 5512 Hz, 1 channels <-- the audio is g711 at 8000 [17:26] <burek> hm [17:26] <burek> how did you install your ffmpeg [17:26] <JamesJJ> clean ubuntu server / got from git / make & make install [17:26] <burek> this is not from git [17:27] <burek> git versions dont look like 0.10.2.git-8274b21 [17:27] <burek> they look like N-8274b21 [17:27] <burek> something [17:27] <JamesJJ> is ffmpeg-8274b21 [17:28] <JamesJJ> oh, was snapshot... OK back in a mo [17:28] <burek> http://ffmpeg.org/download.html [17:29] <burek> just do git clone... [17:29] <burek> and make it [17:29] <JamesJJ> git://source.ffmpeg.org/ffmpeg.git? [17:29] <burek> I don't know [17:29] <burek> it says there on that page [17:29] <burek> check it out [17:30] <burek> git clone git://source.ffmpeg.org/ffmpeg.git [17:30] <JamesJJ> yup, doing that [17:33] <steinesel> I am trying to use ffmpeg to convert an avi format to divx can someone help me with the parameters for this? [17:36] <DTH0> hello, I'm trying to get a video out a big packed file which I don't know the format (.DAT, not VCD), ffplay does find a video stream though there are more than one. The thing is that I can't figure out where is the sound. ffplay is printing this : http://paste2.org/p/1976754 [17:36] <DTH0> can someone help ? [17:37] <burek> steinesel, [17:37] <burek> can you please use pastebin.com, to show your command line and its output? [17:38] <taliaraven> Are any of you mpegts experts? :) I want to modify how often PES packets are added to a stream, to increase the granularity of PTS timestamps in it. Is this possible? [17:39] <burek> DTH0, the format is Input #0, mpeg, from 'ZOE.DAT': [17:40] <burek> also, you have 5 audio streams [17:40] <burek> but your ffmpeg is too old [17:40] <burek> and it might be the reason [17:40] <burek> why it doesn't read the file correctly [17:41] <burek> try updating your ffmpeg [17:41] <JamesJJ> seems changing my ffmpeg gives the same output - http://pastebin.com/KLuXhNA5 [17:41] <burek> taliaraven, are you using ffmpeg cmd line tool or libav library [17:41] <DTH0> The thing is that the .DAT file was not meant to be read by ffmpeg [17:41] <taliaraven> burek: cli. [17:41] <DTH0> it's a file extracted from a game DVD for PS2 [17:42] <burek> JamesJJ, can you try this static build [17:42] <DTH0> so it's a bit hacking to try to read it with ffmpeg [17:42] <burek> https://sites.google.com/site/linuxencoding/builds [17:42] <burek> just download-and-run with ./ffmpeg [17:42] <DTH0> or something that's not a Playstation [17:43] <burek> DTH0, it's not hacking, ffmpeg recognized it without any problems [17:43] <burek> you can do something like ffmpeg -i input.dat -vn out.wav [17:44] <burek> to see if it extracts audio [17:44] <burek> if not, you might update your ffmpeg first [17:44] <burek> and then try again [17:44] <taliaraven> burek, in the interest of being proactive, here's the CLI output if it helps. http://pastebin.com/YdXZ19pe [17:44] <burek> taliaraven, thanks :) [17:44] <DTH0> "Output file #0 does not contain any stream" [17:44] <taliaraven> My issue is getting this ts stream to play onto an STB, and the stb seems to choke on the audio if it doesn't recieve enough PTS timestamps. [17:45] <burek> taliaraven, you might take a look at x264 options [17:45] <burek> because it's not ffmpeg specific [17:45] <burek> but rather x264 [17:45] <JamesJJ> burek - the static version gives the same - http://pastebin.com/PfMXcp3R [17:45] <taliaraven> mpeg transport stream is, burek? That doesn't make sense. [17:45] <burek> JamesJJ, you might contact zeranoe to resolve that? [17:45] <burek> taliaraven, that's what ffmpeg recognized [17:46] <JamesJJ> oh? how would I do that? [17:46] <burek> click on contact? :) [17:47] <taliaraven> burek, thank you for being both delicious and helpful. Can I ask you to clarify why an mpegts issue is resolved in x264? [17:47] <JamesJJ> I'm lost - sorry, I don't know who zeranoe is or who to contact him/her [17:48] <taliaraven> Because the PTS I need should actually be on the audio stream, which is AAC. The video plays back fine, it's the audio timing that appears to be off. THe PTS in my output stream is only delivered every 300ms, where my reference stream shows it every 20ms. I need that amount of granularity to make an STB work. [17:50] <taliaraven> sorry, I know this is a lot. Thanks for looking any way. :) [17:52] <burek> taliaraven, did you try using native aac encoder instead of faac [17:53] <burek> JamesJJ, sorry my mistake [17:53] <burek> JamesJJ, can you try: ffmpeg -i camera.stream.flv -vn out.wav [17:53] <burek> and then ffmpeg -i out.wav [17:55] <JamesJJ> http://pastebin.com/dHUTq3b8 - doesn't go far, just get a 0k file [17:59] <taliaraven> burek, I wasn't using aac native because it was listed as experimental. The issue still seems to be with my mux, rather than my encode, though. [18:00] <burek> JamesJJ, it seems your ffmpeg doesn't have linked some library (audio encoder/decoder) needed to recognize that audio stream [18:01] <burek> taliaraven, can you try changing your muxer to mp4 or flv [18:01] <burek> and see if the problem persists [18:01] <taliaraven> No, I can't, burek. MPEGTS is necessary in this case [18:01] <taliaraven> the ts container is the only one that allows me to stream across a network correctly. [18:02] <burek> because ffmpeg's mpegts implementation has usually been criticized as not stable and such [18:02] <burek> so I'm afraid you might want to try with vlc or some other ts muxer [18:03] <taliaraven> Maybe. I was hoping there's some more configuration options I wasn't aware of. [18:03] <taliaraven> brb. [18:03] <JamesJJ> Hmmm, I don't think so - the audio G711 is supported, it looks to me like the FLV demuxer needs to be updated for other than mp3 audio. It gets a 0x008 as the code for the codec and doesn't know what that codec is [18:04] <burek> JamesJJ, try typing: ffmpeg -codecs [18:04] <burek> and see is there g711 in that list [18:04] <burek> if not, then you didn't compile your ffmpeg with that lib [18:05] <DTH0> where can I fing libavutil ? [18:05] <DTH0> *d [18:05] <burek> try #libav [18:05] <DTH0> thanks [18:07] <JamesJJ> I have the codec I think becouse: [18:07] <JamesJJ> ffmpeg.exe -acodec pcm_mulaw -i camera.stream.flv -vcodec copy -ar 44100 -ac 2 -ab 192 -f mp3 test.mp3 [18:08] <JamesJJ> works, kinda... [18:08] <JamesJJ> but becouse the demuxer is getting the input frequency wrong, I get a audio track 30% longer than it should be [18:10] <JamesJJ> oops - that should be ffmpeg.exe -acodec pcm_mulaw -i camera.stream.flv -vn -ar 44100 -ac 2 -ab 192 -f mp3 test.mp3 [18:39] <fenrig> Hi I'm using ffprobe to count the I and P frames, so I'm using a python script to count the output of the "ffprobe -show_frames <file>" command output but I can't seem how to see if a frame is a B frame :s [18:42] <burek> fenrig, you can always use a video editor :) [18:42] <burek> JamesJJ, try ffmpeg -acodec ... -ar 8000 -i camera.. [18:45] <fenrig> burek: I'd rather find out how I could count all the B frames [18:46] <Durmush> fenrig: mux to mkv then use use mkvinfo -s [18:47] <Mista_D> Gents, I'm testing "lossless" 1080p `ffmpeg -i 1.mpg -sameq -vcodec mpeg4 -an 2.mpg` Getting a lot of warnings "{mpeg @ 0xd6d9940} packet too large, ignoring buffer limits to mux it...buffer underflow". Any advice? [18:47] <taliaraven> Fenrig, you didn't encode the video originally with ffmpeg, did you? [18:47] <fenrig> taliaraven: No I did not, that's not even the purpose. [18:47] <burek> Mista_D, [18:47] <burek> can you please use pastebin.com, to show your command line and its output? [18:48] <burek> taliaraven, you can try with tsmuxer tool [18:48] <taliaraven> I'm considering it burek, that's on the list of things to try next. [18:49] <burek> fenrig, try also mediainfo [18:49] <fenrig> taliaraven: I actually just wanna analyze a video file, it's for my multimedia course. We have to analyze different containers and video codecs [18:50] <fenrig> burek: Tried it, doesn't decently work on linux so I'd rather not bother :D XD [18:54] <taliaraven> what about mediainfo? [18:55] <taliaraven> no, sorry, that's just metadata and types. won't give a framecount [18:56] <fenrig> actually it's kinda funny [18:57] <taliaraven> Mmm, fenrig. Mediainfo will actually tell you if b-frames are enabled in the encoding if you pass it the -f flag [18:57] <taliaraven> Just checked [18:57] <JEEB> I think using the ffmpeg's APIs you could get the frame listing and frame types. Just don't ask me how :P [18:57] <taliaraven> but I can't get a count from it. I'll poke at probe some more cause I'm curious now :) [18:57] <fenrig> does anyone know a file with B frames in them [18:58] <fenrig> maybe i'm testing files without B frames, could be it (I know I'm not that clever -.-) [18:58] <taliaraven> sintel trailer in mp4 does. [18:59] <taliaraven> I've been using the open movie project stuff as test assets. [19:00] <fenrig> okay no my bad, I was using files without B frames, damn I'm stupid bothering you guys with this XD [19:01] <fenrig> taliaraven: I'm using them too for my analysis: big buck bunny, Elephants Dream and Sintel [19:01] <taliaraven> excellent. :) [19:02] <DTH0> hmm one think : Is x264 is warning me about some libav stuff while configuring but... do i need to compile libav in order to get x264 working so that I can compile ffmpeg ? I just found out that libav and ffmpeg aren't apparently meant to be installed together... what should I do ? (I want ffmpeg). [19:03] <ckb> burek: this is so much easier! :D [19:04] <ckb> DTH0 if you have another audio codec you plan on using (libfaac) then libav isn't required.. since you're using libx264 [19:04] <DTH0> libfaac can't be enabled in ffmpeg ? [19:04] <ckb> huh? [19:05] <DTH0> ah sorry misunderstood what you were saying [19:05] <ckb> libav is like a generic audio/video codec [19:05] <ckb> but if you know specifically what scope you're using ffmpeg for, then you're all good [19:06] <ckb> speaking of all good. Kind of pissed it's in Indiana this year [19:06] <DTH0> ok [19:39] <level09> can ffmpeg convert quick time reference files ? [19:52] <zap0> probably. [19:53] <zap0> level09, ffmpeg -formats or -codecs [20:06] <level09> thnx will check [20:31] <ckb> wow [20:32] <ckb> this 2 hour movie is taking 2 hours to encode. [20:32] <ckb> :( [20:33] <silverrocker> questions about using ffmpeg as a library in my project can be asked here? [20:37] <relaxed> silverrocker: sure [20:39] <silverrocker> I can compile the example decode_encode.c provided in doc/examples/ but this only works for mpg1 files. I would like to extend this so that it detects the file format and loads the appropeate codec for that file. I have used libavformat to find the codecid of the codec in question for the first video stream of an arbitrary file. [20:39] <silverrocker> but when I try to decode anything else then a mpg1 file, I get errors form the corresponding codecs [20:42] <fenrig> I'm having difficulties using ffprobe, I want ffprobe to report the audio bitraete [20:43] <JEEB> extract the audio track, divide size with length :P [20:44] <JEEB> and you will get the average bit rate [20:47] <fenrig> JEEB: that's a real crappy way to do it, isn't ffprobe made to analyze mediafiles. Meanwhile everyone is extracting audio to get the average bit rate :P that doens't sound too clever [20:48] <JEEB> it is, but if it isn't doing what you want :P [20:49] <JEEB> also you'd need to go thorugh the whole audio track in any case to get the whole average bit rate [20:49] <JEEB> if you can't go through it with code, it's just easier to extract it and count it like that >_> [20:50] <fenrig> JEEB: besides I need to know if it's CBR, VBR too [20:50] <JEEB> it's better to just assume VBR [20:51] <fenrig> JEEB: I have to analyze files for school [20:52] <JEEB> also, for CBR/VBR you'd have to then check the sizes of every packet, and check if the length of every packet is the same in that given format [20:52] <JEEB> if the length for every packet is the same, and the file size for every packet is the same -> CBR [20:52] <JEEB> it gets messy if the length of every packet is _not_ the same [21:00] <Durmush> Hi, i'm trying to build ffmpeg in msys-mingw, but whatever I tell it, it refuses to link to pthreads staticly. here is my configure line: --enable-gpl --enable-version3 --enable-nonfree --enable-static --extra-ldflags=-static --enable-libmp3lame --enable-libxvid --enable-pthreads --disable-doc --disable-shared [21:21] <Mista_D> Problem extracting s320m audio. http://pastebin.com/7ZYk9AtA [21:23] <Mista_D> *s302m [21:28] <bcoudurier> don't put s302m in .wav [21:29] <Mista_D> bcoudurier: it was just a test. [21:30] <bcoudurier> that test won't work [21:31] <Mista_D> bcoudurier: any way to preserve s320m in any other container? [21:36] <bcoudurier> no, ts only [21:58] <burek> Durmush, what does your config.log says at the end [21:58] <burek> say* [22:00] <burek> Mista_D, did you try ffmpeg102 -i orig.mpg -vn -acodec copy -aframes 10 out.mp4 [22:07] <iive> burek: what happened with the hanging avi? [22:08] <burek> I've updated trac pst [22:08] <burek> post [22:09] <burek> https://ffmpeg.org/trac/ffmpeg/ticket/1192 [22:10] <Durmush> burek: the problem has gone. a rename of certain file did the trick. [22:10] <burek> ok [22:10] <Mista_D> burek: tried s302m with mp4 - "[mp4 @ 0x113d1920] track 0: could not find tag, codec not currently supported in container" [22:10] <burek> I see [22:11] <burek> what is s302m anyway :) [22:29] <Mista_D> burek: some codec for high quailty audio... [22:31] <Mista_D> burek: aka BSSD "Audio data as specified in SMPTE Standard SMPTE 302M-1998 for Television - Mapping of AES3 Data into MPEG-2 Transport Stream shall use this format_identifier." so its AES3 codec in MPEG2. [22:33] <DTH0> hey, I'm following this guide : "https://ffmpeg.org/trac/ffmpeg/wiki/UbuntuCompilationGuide". The compilation and the installation of x264 seems to go well but when trying to compile ffmpeg, it's stating that there's no libx264, though I'm using last version from git [22:33] <Mista_D> burek: s302m = BSSD = 2 PCM channels wrapped in AES3, then in MPEG2-TS. [22:35] <DTH0> I'm also getting a weird version from the version.sh file ("#define X264_POINTVER "0.122.2184 5c85e0a") which doesn't look too promising... [22:36] <DTH0> I strictly followed the guide, is the mistake mine ? [22:36] <DTH0> the version.sh script is from gitting x264 [22:38] <NateW> Anyone around? [22:38] <taliaraven> plenty of people around. [22:38] <NateW> Just checking. Is it possible to use more than one audio input when using ffmpeg. [22:38] <NateW> From either pulseaudio or alse? [22:38] <NateW> *alsa [22:40] <NateW> Currently, to do a proper screen capture, I have to create a loopback of my mic and record the desktop sound, I was hoping I can record both the mic and the desktop sound without having to setup a loopback. [22:41] <NateW> It's really annoying having to listen to an echo of what I just said 100ms ago. [22:43] <NateW> I have found many forum posts about recording screencasts, but none cover recording from more than one input, so is it an ffmpeg limitation? [22:46] <NateW> taliaraven: do you have any idea? [22:48] <pasteeater> DTH0: what version of ubuntu? [22:55] <taliaraven> You should be able to combine a couple inputs; depends on how you want to do it. If you want them as seperate channels, check out this: http://ffmpeg.org/ffmpeg.html#amerge [22:56] <taliaraven> And the pan filter should let you combine two sources into stream. In either case, you're going to have to do some more complex things and build out a filter graph [22:57] <taliaraven> It explains the syntax for how to do that here: http://ffmpeg.org/ffmpeg.html#Filtergraph-description [22:57] <NateW> As I said above, I'm doing screencasts. I can set my mic level to match the audio level, I just need to be able to record from both the mic and desktop audio (They can be combined) at the same time. [22:58] <NateW> now with this: ffmpeg -f pulse -i default, could I specify multiple inputs like: ffmpeg -f pulse -i mic,desktop [22:59] <NateW> obviously I'd have to properly name the pulse inputs. [23:00] <taliaraven> Beyond what I just pointed out, I won't be much help. I've built filtergraphs in gstreamer, but not ffmpeg, and I don't have the time right now to research, sorry :( [23:00] <NateW> No probs. I was hoping someone would have the knowledge, if it is just research, I can do that. [23:01] <taliaraven> The links I posted should be a good place to start. [23:01] <NateW> Yep. [23:09] <DTH0> pasteeater, 11.10 [23:18] <DTH0> anyone have an idea ? (I was AFK, sorry) [23:26] <burek> DTH0, you are compiling ffmpeg with libx264? [23:26] <DTH0> yes [23:26] <burek> that's easy [23:27] <burek> http://ffmpeg.gusari.org/viewtopic.php?f=25&t=36 [23:27] <burek> that's for x264 [23:27] <burek> http://ffmpeg.gusari.org/viewtopic.php?f=25&t=14 [23:27] <burek> that's all together (just ignore faad and other stuff you don't need) [23:28] <DTH0> ok [23:28] <burek> or, if you prefer apt [23:28] <DTH0> I was just avoiding the make install stuff [23:28] <burek> (but that will install some old version of ffmpeg) [23:28] <burek> you can install libx264-dev [23:28] <burek> prior to compiling ffmpeg [23:28] <DTH0> I followed the ffmpeg guide [23:28] <burek> but my advice is to always compile multimedia projects [23:28] <burek> they change rapidly [23:29] <DTH0> "https://ffmpeg.org/trac/ffmpeg/wiki/UbuntuCompilationGuide" [23:29] <burek> and repos cannot follow them quickly enough [23:29] <burek> I don't have time to read that, but you get the general idea, ok? :) [23:29] <DTH0> yes [23:29] <DTH0> but [23:29] <DTH0> libx264 is properly installed [23:29] <burek> libx264-dev [23:29] <burek> not libx264 [23:30] <DTH0> I'm compiling from git [23:30] <burek> I don't know, I got to go [23:30] <DTH0> so libx264-dev files should be there as well right ? [23:30] <burek> read those links [23:30] <burek> and decide for yourself [23:31] <DTH0> It's almost the same thing that what i've done. Compilation and installation of x264 libraries go well. Problem is ffmpeg won't detect it [23:31] <DTH0> when configuring [23:31] <DTH0> before compilation [23:31] <burek> what does your config.log say [23:31] <burek> can you please use pastebin.com, to show your command line and its output? [23:37] <DTH0> http://paste2.org/p/1977085 [23:39] <burek> add --enable-shared [23:39] <burek> anyway [23:39] <burek> can you copy/paste config.log from FFmpeg's ./configure ? [23:41] <DTH0> http://paste2.org/p/1977090 [23:43] <burek> well it seems that you didn't properly install your compiled x264 [23:43] <burek> maybe checkinstall made some errors or something [23:43] <burek> or [23:43] <burek> you didn't type ldconfig [23:43] <burek> to update your libs [23:43] <burek> after checkinstall [23:43] <DTH0> i didn't entered ldconfig [23:43] <burek> try typing ldconfig [23:44] <burek> and retry ./configure for ffmpeg [23:45] <DTH0> but the output of version.sh from git is strange anyway ("#define X264_POINTVER "0.122.2184 5c85e0a") [23:46] <burek> I really don't know, I'm just trying to help you make it work.. I don't use checkinstall [23:46] <burek> I just use what I gave you on those likns [23:46] <burek> and it works everywhere and always [23:47] <DTH0> ok [23:48] <DTH0> ldconfig did the trick [23:48] <burek> cool :) [23:48] <DTH0> should be added to the guide then since it's the official one :) [23:48] <DTH0> thanks [23:49] <burek> :beer: :) [23:51] <burek> no [23:51] <burek> oops, wrong channel [23:51] <burek> :) [23:52] <DTH0> >< [23:55] <DTH0> wow ##beer exist but it's invite only ... [23:55] <burek> :D [00:00] --- Fri Apr 13 2012

1 0

ffmpeg-devel.log.20120411
by burek 12 Apr '12

12 Apr '12

[00:48] <Compn> michaelni : well i was saying that in reply to ubitux's comment. i had no actual idea that git master compile was broken ... [00:56] <Compn> btw it would be cool if we had a fate bot [00:56] <Compn> it could check fate.ffmpeg.org and report here 'fate reports compile fail!' [00:56] <Compn> or something :) [01:07] <michaelni> compile of vda was broken (i dont have the dependancies so had not noticed ...) [01:07] <michaelni> i just saw it when i looked at fate [01:08] <michaelni> a fatebot would be nice indeed .. [01:08] <michaelni> would need a bit of inteligence so as not to report known things [01:08] <michaelni> or overheating ppc boxes or intel cc running out of memory [01:09] <Daemon404> ppc just loves to overheat [01:10] Action: Daemon404 has awesome memories of that [01:21] <michaelni> btw, Daemon404, you should have OP as you are in MAINTAINERS and have a git clone of ffmpeg [01:21] <Daemon404> sure [01:21] <Daemon404> im ok with anything really [01:21] Action: Daemon404 doesnt care about epeen on irc [01:22] <Compn> its hard to tell who cares about status and having commit access [01:22] <gnafu> Daemon404: Not even just a little? [01:22] <Compn> always good to ask :) [01:22] <Daemon404> nope :P [01:23] <Daemon404> i may be young but ive been on irc for ~10 years [01:23] <Daemon404> ops are boring [01:23] <gnafu> Now's your chance to change that! [01:23] <gnafu> Be awesome. [01:23] <Daemon404> ;p [01:24] <Compn> being on irc for that long , you learn that ops only cause trouble [01:25] <Daemon404> having opsalso causes random PMs [01:25] <Compn> even amongst like-minded individuals :\ [01:25] <Daemon404> from newbies [01:25] <Daemon404> daily or so [01:25] <Compn> whitelist your pm approvals then :P [01:25] <Compn> ehe jk [01:26] Action: gnafu has gotten random PMs without being op. [01:26] <Daemon404> ive been getting PMs from gsoc applicants asking for the "results of gsoc" [01:26] <Daemon404> it's confusing. [01:26] <gnafu> Hehe. [01:28] <Compn> what channels do you hang out in ? [01:28] Action: Daemon404 looks at his irssi bar with a hilight in channel 52 [01:29] <Daemon404> on freenode? ~8 chans [01:29] <Compn> all of them are +s? hehe [01:29] <Daemon404> no [01:29] <Daemon404> i think [01:29] <Compn> cept this one [01:30] <Daemon404> [19:29] [freenode] -!- channels : #ffmbc @+#libav-devel #poky #UtVideo #x264dev @+#ffmpeg-devel [01:30] <Daemon404> (one removed that is +s) [01:30] <Compn> anyways, i think some instructions on libav gsoc say to pm an 'op' and aks for help [01:30] <Compn> thats possibly why you get gsoc questions in libavdevel [01:30] <Daemon404> yea [01:30] <Compn> well you blame being an op or the instructions to pm an op, in that case ? [01:31] <gnafu> Both, of course! [01:31] <Compn> of course! [01:31] <iive> it's Daemon404 fault of course. [01:31] <gnafu> If he wasn't an op, he wouldn't need to care about the instructions. [01:31] <Compn> well i dont actually remember the last time i was pm'd in here by newbie [01:31] <Daemon404> it's better than people thinking im an xdcc bot [01:32] <Compn> it would help if i identified more i guess :P [01:32] <Compn> hmmm [01:32] <Compn> lots of these $$ converter programs now have prores support [01:32] <Compn> wonder how they got that... [01:33] <Daemon404> through apple's kind acts of charity. [01:33] <Compn> of course [01:33] <Compn> steve jobs is a ghost in the machine [01:33] <iive> in the shell [01:33] <gnafu> s/shell/hell/ [01:33] <Daemon404> say what you want [01:33] <Daemon404> apple stuff is nice [01:33] <Daemon404> i wish i had a mac mini [01:34] <iive> Daemon404: get one second hand. If you are lucky enough to find one that didn't broke on its own. [01:35] <Daemon404> i was thinking of getting a refurbished one [03:26] <funman> main debug: TIMER : 0.000 ms - Total 2425386726513824123412325101826455280817817073886013940178100649148027127278321327603471638582279891373599947431444196332694059231614642577013085525952523027237688476255626010898605470768194704249628709211663055933138797955663315664345583084297601017174574875279360.000 ms / 8221900169169065220 intvls (Avg 0.000 ms) [03:26] <funman> oops wrong chan [03:31] <Compn> thats a lot of ms [03:52] <funman> indeed :) [05:34] <Daemon404> O_o [08:03] <ubitux> http://lists.libav.org/pipermail/libav-devel/2012-April/025688.html [08:03] <ubitux> :/ [08:04] Action: ubitux doesn't get it [10:18] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r6e9c634c36 10ffmpeg/libswresample/swresample_internal.h: [10:18] <CIA-17> ffmpeg: swr: fix copy & pasted comment to match the code. [10:18] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [10:18] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rf9a2c5bc07 10ffmpeg/libswresample/ (rematrix.c resample.c): [10:18] <CIA-17> ffmpeg: swr: simplify code by using av_get_bytes_per_sample() [10:18] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [14:30] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r6c704d9c30 10ffmpeg/ffmpeg.c: [14:30] <CIA-17> ffmpeg: ffmpeg: support changing dither parameters for swr [14:30] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [14:30] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rbe4cde226a 10ffmpeg/ (cmdutils.c cmdutils.h): [14:30] <CIA-17> ffmpeg: cmdutils: parse options for swr [14:30] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [14:30] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * ra2c92e6063 10ffmpeg/libswresample/ (dither.c swresample.c swresample_internal.h): [14:30] <CIA-17> ffmpeg: swr: pass context to swri_get_dither() [14:30] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [14:30] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r22057e8ecb 10ffmpeg/libswresample/ (swresample.c swresample.h): [14:30] <CIA-17> ffmpeg: swr: add swr_get_class() [14:30] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [14:30] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rc1d404b9e4 10ffmpeg/libswresample/ (dither.c swresample.c swresample_internal.h): [14:30] <CIA-17> ffmpeg: swr: add a dither_scale parameter to tune the amplitude of the dither. [14:31] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [16:16] <ubitux> michaelni: the av_interleaved_write_frame() dts failures are expected in quite a bunch of situations [16:16] <ubitux> i guess the source of the error can be quite large? [16:17] <michaelni> yes, 50 bugs, 1 error message, a million users who belive its the same bug [16:17] <ubitux> :) [16:17] <ubitux> reminds me the "operation non permited" because of -1 [16:18] <michaelni> :) [16:21] <Compn> ehe [16:21] Action: Compn votes to rename the non monotonically timestamps error [16:21] <ubitux> we should add a success message after a transcode btw, like "core dumped ;)" [16:21] <Compn> lol [16:22] <Compn> that only gets old users tho ubitux [16:22] <Compn> these new linux people have never seen a core dump [16:22] <Compn> so i vote 'kernel panic' [16:22] <ubitux> :) [16:23] <ubitux> i'm not sure new linux people except much "kernel panic" messages :p [16:24] <Compn> maybe we could make a BSOD joke [16:49] <ubitux> btw, can't we make ffmpeg ignore decoding error? [16:50] <ubitux> i'm transcoding a broken sample, but as soon as the problematic frame fails to decode, ffmpeg stops encoding [16:50] <ubitux> mplayer seems to be able to deal with the issue, and continue playback (output is just broken for a few frames) [16:51] <iive> there used to be error resilience, not sure what it got renamed into. [16:55] <michaelni> ffmpeg should not stop on a decoding error [17:38] -:#ffmpeg-devel- [freenode-info] channel flooding and no channel staff around to help? Please check with freenode support: http://freenode.net/faq.shtml#gettinghelp [20:40] <mbradshaw> Hello devs, I tried to ask a question about seeking in #ffmpeg but it was silent. I just want to double check I'm seeking correctly before I file a bug report. Is there anyone I can ask a few questions to make sure I'm not forgetting a function call? [20:50] <michaelni> mbradshaw, sure, ask [20:51] <mbradshaw> Thanks, Michael [20:51] <mbradshaw> So I've opened a format context and a stream, and playing the video works fine [20:51] <mbradshaw> However, if I add the call avformat_seek_file(formatContext, -1, 0, 0, 0, 0) it seeks to a pts of 12 [20:52] <mbradshaw> if I set the starting timestamp to less than zero (something like avformat_seek_file(formatContext, -1, -100000, ts, ts, 0)) it seeks to the first frame just fine [20:52] <mbradshaw> sorry, ts = 0 in the above line [20:54] <mbradshaw> the stream's starting time is zero [20:54] <mbradshaw> so I'm not sure why a negative min seek time is required [20:54] <mbradshaw> Any ideas? [20:55] <michaelni> could be a bug [20:56] <nevcairiel> dont use avformat_seek_file, it doesnt give you any advantages and just confuses you [20:56] <mbradshaw> Would you like me to file one with the full code and source sample? It's only happing with this mpeg2video I have. My h264 sample works ok. [20:56] <nevcairiel> use av_seek_frame, and specify AVSEEK_FLAG_BACKWARDS for the flags [20:58] <michaelni> that will likely fix it, still it shouldnt be needed [20:59] <nevcairiel> avformat_seek_file is just a broken API. its signature might be better, but whats the point, internally it will just go ahead and call av_seek_frame [21:00] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rca4a3f4a08 10ffmpeg/libavformat/oggparsevorbis.c: [21:00] <CIA-17> ffmpeg: oggparsevorbis: Remove code messing with cur_dts. [21:00] <CIA-17> ffmpeg: This code caused first_dts to become corrupt and in value to be [21:00] <CIA-17> ffmpeg: around relative_ts. [21:00] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [21:00] <nevcairiel> its some remnant from some attempts for better seeking that never went anywhere [21:00] <CIA-17> ffmpeg: 03Dale Curtis 07master * r309a931a38 10ffmpeg/libavcodec/h264.c: [21:00] <CIA-17> ffmpeg: Fix memory leaks on failed ff_h264_decode_init() [21:00] <CIA-17> ffmpeg: During failure conditions ff_h264_decode_init() leaks memory [21:00] <CIA-17> ffmpeg: allocated for nal units. Found via valgrind. [21:00] <CIA-17> ffmpeg: Valgrind traces: http://pastebin.com/GqTqxs8T [21:00] <CIA-17> ffmpeg: Signed-off-by: Dale Curtis <dalecurtis(a)chromium.org> [21:00] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [21:00] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r6952301f33 10ffmpeg/libavcodec/internal.h: [21:00] <CIA-17> ffmpeg: ff_samples_to_time_base: support AV_NOPTS_VALUE [21:00] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [21:00] <michaelni> nevcairiel, yes but it shouldnt need AVSEEK_FLAG_BACKWARDS [21:01] <michaelni> if theres a keyframe at ts=0 and one seeks to 0 it should find it [21:01] <mbradshaw> Yes, av_seek_frame works. [21:01] <nevcairiel> that depends on the container really, alot of formats dont even have keyframe seeking. :d [21:02] <mbradshaw> It's an mov container with a mpeg2video stream [21:03] <mbradshaw> If I don't specify AVSEEK_FLAG_BACKWARD it seeks to a pts of 12, which I'm assuming is what avformat_seek_file is doing [21:03] <nevcairiel> yes [21:04] <nevcairiel> both functions do the same, except that avformat_seek_file uses all the time values you specify to figure out the backwards flag [21:05] <mbradshaw> Yeah, I'm just curious why AVSEEK_FLAG_BACKWARD is needed, as the first frame with ts=0 is a keyframe, so I can't see why the backward flag would be needed [21:12] <nevcairiel> it looks like the binary search for the index entry is bugged [21:14] <nevcairiel> or i failed at reading [21:14] Action: nevcairiel checks again [21:16] <nevcairiel> Nop, i think i'm right. It cannot arrive at the first index [21:17] <mbradshaw> Hmmm... interesting. [21:18] <nevcairiel> hm, this is confusing, i should write it down on paper [21:18] <mbradshaw> Is this in ff_seek_frame_binary? [21:19] <nevcairiel> ff_index_search_timestamp actually [21:21] <nevcairiel> i think i was wrong, it seems to work fine, i mustve switched two variables around in my head. :d [21:21] <nevcairiel> but then, why would it not work [21:21] <nevcairiel> unless the mov forgot to mark its first frame as a keyframe [21:21] <mbradshaw> I trasncoded the mov with ffmpeg (was h264 originally) [21:23] <nevcairiel> maybe its a failure in the mpeg2 encoder or the mov muxer then [21:26] <mbradshaw> When I decode the video, the first frame is reported as a keyframe (though I'm not sure if ffmpeg is doing packet reordering magic behind the scenes that might be changing this) [21:27] <nevcairiel> that info is probably coming from the decoder then, not from the demuxer [21:27] <mbradshaw> The next keyfram is at pts=12, which is where it's seeking to without the backwarrd flag [21:27] <nevcairiel> you could check if the next av_read_frame after your seek to pts=0 (using the backwards flag) has the AV_PKT_FLAG_KEY or what its called set [21:29] <mbradshaw> Yes, it does have AV_PKT_FLAG_KEY set in the first packet after the seek with the backwards flag set [21:30] <nevcairiel> then something weird is up. :) [21:30] <mbradshaw> Ha I think so [21:31] <mbradshaw> Think I should submit it as a bug with the video samples? [21:31] <nevcairiel> sure [21:43] <Daemon404> michaelni, pingpong [21:47] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * r5735f768b0 10ffmpeg/libavcodec/ (Makefile h261data.c h261data.h h261dec.c h261enc.c): (log message trimmed) [21:47] <CIA-17> ffmpeg: h261: move tables from header to .c file. [21:47] <CIA-17> ffmpeg: Currently they end up twice in the binary, since both [21:47] <CIA-17> ffmpeg: encoder and decoder include the header and thus each gets [21:47] <CIA-17> ffmpeg: their own copy. [21:47] <CIA-17> ffmpeg: This is clearly nonsense for the const tables, but shouldn't [21:47] <CIA-17> ffmpeg: be necessary for the RLTable either. [21:48] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * re6ad943734 10ffmpeg/libavformat/latmenc.c: [21:48] <CIA-17> ffmpeg: latmenc: remove dead code. [21:48] <CIA-17> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger(a)gmx.de> [21:52] <ubitux> michaelni: i think it will be pretty hard to do a reg test on the race condition in h264 decoder [21:53] <ubitux> the issue doesn't happen often: http://fate.ffmpeg.org/history.cgi?slot=x86_64-archlinux-gcc-threads-8 [21:53] <ubitux> i've just ran a loop with h264-conformance-cama2_vtc_b test; it's been a few minutes already [21:53] <ubitux> and the race didn't get triggered [21:56] <ubitux> (i certainly need to slow down my system though) [21:58] <ubitux> to do a git bisect* [22:06] <michaelni> Daemon404, pongping [22:08] <michaelni> ubitux, how often does the failure occur ? [22:10] <Daemon404> michaelni, i have a question about a commit of yours: [22:10] <Daemon404> http://git.videolan.org/?p=ffmpeg.git;a=commitdiff;h=abe0b8e9f378a0f8781c1a… [22:10] <Daemon404> why does shiftonly only affect luma [22:11] <Daemon404> i dont find a reason for that in bt.709 [22:14] <michaelni> Date: Thu Aug 25 13:54:43 2011 +0200 [22:14] <michaelni> thats long ago [22:14] <Daemon404> yeah [22:14] <michaelni> is it causing some problem ? [22:14] <Daemon404> but someone asked me to backport it so ffms2 can use it [22:15] <Daemon404> and im trying to understand it [22:15] <Daemon404> (i.e. so ffms2 is not broken between libs) [22:20] <michaelni> it would introduce a color bias if done for chroma IIRC [22:20] <michaelni> but its long ago i might misremember [22:21] <ubitux> michaelni: well i ran ~5k times the test and it didn't happen; though on fate the failure happens sometimes (all the yellow tests on the previous link are the race) [22:22] <ubitux> i guess it's because they're run with multiple jobs at the same time [22:22] <ubitux> so some random slow down in some functions might trigger the race, dunno [22:23] <michaelni> 5k is alot [22:23] <ubitux> yes :p [22:24] <ubitux> i'm going to run a make fate-h264 with jobs instead now, and see if it triggers it [22:26] <Daemon404> michaelni, i see [22:28] <michaelni> 0x80 is the middle 0x8000 too but 0x8080 is not [22:31] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * rfa8a6385a1 10ffmpeg/libavformat/latmenc.c: [22:31] <CIA-17> ffmpeg: latmenc: remove unused return value. [22:31] <CIA-17> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger(a)gmx.de> [22:31] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * ra736eb4a60 10ffmpeg/libavformat/latmenc.c: [22:31] <CIA-17> ffmpeg: latmenc: Fix ALS in LATM. [22:31] <CIA-17> ffmpeg: "Fix" in so far as at least it will no longer overread and possibly [22:31] <CIA-17> ffmpeg: crash and makes somewhat sense, but no idea whether there is anything [22:31] <CIA-17> ffmpeg: that can play the resulting files (FFmpeg can't). [22:31] <CIA-17> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger(a)gmx.de> [22:31] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * r8e357e8e75 10ffmpeg/libavformat/latmenc.c: [22:32] <CIA-17> ffmpeg: latmenc: error out when packet size is too large. [22:32] <CIA-17> ffmpeg: Previously it would just silently write out incorrect data. [22:32] <CIA-17> ffmpeg: This also fixes a potential integer overflow in the allocation. [22:32] <CIA-17> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger(a)gmx.de> [23:07] <ubitux> michaelni: i was able to trigger the race after 2-3 minutes [23:07] <ubitux> i guess 20-30 make fate-h264 runs with jobs might be safe to assume it works (or not) [23:15] <ubitux> well, it's been 5 minutes i retried, and still no failure :p [23:18] <ubitux> anyway, i don't think a git bisect will be reliable in any way [23:21] <burek> hi, who is the owner of this page: https://sites.google.com/site/linuxencoding/builds [23:21] <CIA-17> ffmpeg: 03Diego Biurrun 07master * r679481b3b6 10ffmpeg/ (3 files in 2 dirs): [23:21] <CIA-17> ffmpeg: Drop some pointless #ifdefs. [23:21] <CIA-17> ffmpeg: The files are only compiled if the #ifdef conditions are met. [23:21] <CIA-17> ffmpeg: 03Diego Biurrun 07master * r9676d8eb67 10ffmpeg/libavcodec/interplayvideo.c: [23:21] <CIA-17> ffmpeg: interplayvideo: fix av_dlog parameter type mismatch [23:21] <CIA-17> ffmpeg: libavcodec/interplayvideo.c:909:13: warning: format %p expects argument of type void *, but argument 7 has type GetByteContext [-Wformat] [23:21] <CIA-17> ffmpeg: 03Luca Barbato 07master * r18b59956e0 10ffmpeg/libavformat/movenc.c: [23:21] <CIA-17> ffmpeg: movenc: remove redundant check [23:21] <CIA-17> ffmpeg: The proper check is already in mov_write_header. [23:21] <CIA-17> ffmpeg: Signed-off-by: Martin Storsjö <martin(a)martin.st> [23:21] <CIA-17> ffmpeg: 03Luca Barbato 07master * rebbede2265 10ffmpeg/libavformat/movenc.c: [23:21] <CIA-17> ffmpeg: movenc: small refactor mov_write_packet [23:21] <CIA-17> ffmpeg: Share the formerly internal write_packet with the hinter and move the [23:21] <CIA-17> ffmpeg: fragment flush logic to the user facing one since it is not concerned [23:21] <CIA-17> ffmpeg: Fixes bug #263 [23:21] <CIA-17> ffmpeg: Signed-off-by: Martin Storsjö <martin(a)martin.st> [23:21] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r4f6df20a00 10ffmpeg/: (log message trimmed) [23:21] <CIA-17> ffmpeg: Merge remote-tracking branch 'qatar/master' [23:21] <CIA-17> ffmpeg: * qatar/master: [23:21] <CIA-17> ffmpeg: avplay: Don't free video filters string until the end of decoding. [23:21] <CIA-17> ffmpeg: movenc: small refactor mov_write_packet [23:21] <CIA-17> ffmpeg: movenc: remove redundant check [23:21] <CIA-17> ffmpeg: interplayvideo: fix av_dlog parameter type mismatch [23:21] <CIA-17> ffmpeg: 03Alex Converse 07master * re1ce756844 10ffmpeg/avplay.c: [23:21] <CIA-17> ffmpeg: avplay: Don't free video filters string until the end of decoding. [23:21] <CIA-17> ffmpeg: av_freep()ing inside configure_video_filters() leaves a dangling [23:22] <CIA-17> ffmpeg: reference in the calling code, and the filter string is needed again when [23:22] <CIA-17> ffmpeg: reconfiguring video filters for a size change. [23:43] <cbsrobot> burek: hit the "donation" button [23:43] <burek> :) [23:43] <burek> and is john.vansickle here? :) [23:44] <cbsrobot> no idea [23:44] <cbsrobot> ping everybody and ask :) [23:45] <burek> :) [23:45] <burek> I've setup an automatic cron for static builds every day [23:45] <burek> and it works pretty good so far [23:45] <burek> so, I'd like to make it like that it contains all the popular freeware libraries included [23:46] <burek> libmp3lame, libx264, etc [23:46] <burek> and I thought to ask if anyone is interested so that we have it automated completely [23:46] <burek> and when some newbee comes and wants to play around with ffmpeg, he/she doesn't have to compile it and stuff [23:48] <burek> anyway, I thought to ask for suggestions from people that already have done it properly :) [23:48] <burek> so that I don't make some beginners mistakes :) [00:00] --- Thu Apr 12 2012

1 0

ffmpeg.log.20120411
by burek 12 Apr '12

12 Apr '12

[00:11] <ButtDog> ffmpeg -i file.avi > test.txt [00:11] <ButtDog> that just echos the output on screen and leaves an empty test file [00:13] <silverrocker> when does avcodec_find_decoder fail? [00:15] <silverrocker> to myself: when I forget to do avcodec_register_all() [00:20] <pasteeater> iive: can you update topic? link to ubuntu compile guide has changed [00:20] <pasteeater> https://ffmpeg.org/trac/ffmpeg/wiki/UbuntuCompilationGuide [00:31] <iive> pasteeater: maybe you should ask michael, also... i'm not quite sure that the longer url would fit in the topic. [00:33] <relaxed> pasteeater: http://goo.gl/ [00:34] <relaxed> Also, I vote pasteeater and I should have ops here. [00:42] <silverrocker> I have codec = avcodec_find_decoder(CODEC_ID_H264); and when I read somefile.h264 it gives cant read the frames, is there something I'm missing or is the file I have just bad [00:44] <iive> relaxed: sorry, I'm not really administering this channel. I guess the rule is the same, developers get op. [00:51] <relaxed> iive: I know. It would make sense though. [00:54] <silverrocker> so if I have a file I've created with ffmpeg like so: ffmpeg -i kate.avi -vcodec libx264 out.h264 , after that I want to load the file with avcodec_find_decoder(CODEC_ID_H264); this seems to give me errors, any ideas? [01:27] <_genuser_> does ffmpeg allow you to possibly chop / join x264 vid? [01:38] <pasteeater> iive: i meant via a short url. and thanks for the ops, but now i can't be lazy. [01:39] <iive> not as lazy as me. It is not auto op. I need to check the current chanserv commands for that. [01:40] <pasteeater> that's ok. either way is fine with me. [01:44] <_genuser_> guys, I have a video, the audio is behind the video. so I use -itsoffset 00:00:02.0 before -i video.mp4 but no matter what value is used, it doesn't change _anything_. [01:44] <_genuser_> muxing two files together. video.mp4 and audio.mp4. [01:45] <_genuser_> video has only video stream, audio has only audio stream. [02:04] <silverrocker> is it a good idea to use av_read_frame() even if the format is streambased? [06:05] <rrivera> Hey guys, scratching my head over this one -- getting "Unable to find a suitable output format for 'rtmp://[host]'" [06:05] <rrivera> Verified -protocols shows librtmp compiled in [06:07] <rrivera> This is on Win32, am currently reading in my desktop via (-f gdi -i 0,0) [06:07] <rrivera> (I added this support. Works fine writing to disk.) [06:08] <rrivera> Guessing there's something goofy going on with rtmp url parsing [07:38] <Durmush> Hi, how to troubleshoot "lib* not found" problems? i'm trying to build ffmpeg with enable-libxvid, and I installed libxvid from source to my crosstools sys-root, and yet, the configure script cannot find it. [07:49] <tdr> Durmush, its probably a header file missing if you have the lib there [07:52] <Durmush> well, the script checks first for the header, then the lib? it says in config.log: [07:52] <Durmush> cannot find -lxvidcore [07:53] <tdr> but the header check passes? [07:57] <Durmush> tdr: i'm not very sure how to know if it passed or not. do you mind seening yourself? here is the output in pastebin: http://pastebin.com/BkBTWXAJ [08:01] <tdr> Durmush, looks like the library itself, look at line 24, ld is trying to use it and fails [08:02] <tdr> or do you have config.log ? [08:05] <Durmush> tdr: the complete log? of course. here: http://pastebin.com/Mssica0Y [08:26] <Durmush> so tdr, any idea? [11:22] <rrivera> Hey guys, still scratching my head over that "Unable to find a suitable output format for 'rtmp://[host]'", any troubleshooting paths I should be going down? [11:22] <rrivera> Just to recap, verified -protocols shows librtmp compiled in, am on win32. [11:47] <StaRetji1> hello folks, just need info, is libmpeg2 compiled by default with ffmpeg? [11:47] <StaRetji1> thx [12:17] <JamesJJ> OK... still struggling with a flv with G711 audio. the problem is that ffmpeg is seeing the audio track but doesn't reconise the codec - I can force it to use the right codec by adding -acodec pcm_mulaw, but it is also getting the frequency wrong - Stream #0.1: Audio: pcm_mulaw, 5512 Hz, mono, s16, 44 kb/s - is there a way to change the input frequency, as I as the input data is 8000 not 5500? [13:11] <StaRetji1> any dev willing to help me out, I will pay for support [13:11] <StaRetji1> thx [13:26] <StaRetji1> I just need to know if ffmpeg uses libmepg2 [13:26] <StaRetji1> or I have to configure it during insta;; [13:26] <StaRetji1> install [13:26] <JEEB> it doesn't use any external libraries by default [13:27] <JEEB> and I'm not even sure if it can use libmpeg2 as it has both its own decoder and encoder for mpeg-2 in libavcodec [13:27] <StaRetji1> JEEB thx for reply [13:29] <StaRetji1> I am able to decode mpeg2, but throws many errors such as [mpeg2video @ 0x3559f00] mpeg_decode_postinit() failure [13:29] <JEEB> and as far as I can see there is no --enable-libmpeg2, thus it can be quite safely said that ffmpeg doesn't use libmpeg2 at all [13:29] <StaRetji1> I am puling mpeg2 ts stream from dreambox satellite receiver, works just fin in vlc, but I have to make it work fine in ffmpeg too [13:30] <StaRetji1> oh, that is bad news :/ [13:30] <JEEB> not necessarily [13:30] <JEEB> try current git HEAD ffmpeg [13:30] <StaRetji1> really? [13:30] <StaRetji1> I did that [13:30] <JEEB> with the files you are pulling [13:30] <StaRetji1> ehm [13:31] <JEEB> aka "try dumping the stuff into a file and see if it fails there too" [13:31] <StaRetji1> I pull live stream from satellite receiver [13:31] <JEEB> if yes, check the trac bug tracker for similar tickets. [13:31] <JEEB> and if there are none, add one with a sample attached that fails [13:32] <JEEB> (10-20MB) [13:32] <StaRetji1> I'm affraid I'm not capable doing it, it is way above my knowledge [13:32] <JEEB> not really [13:32] <JEEB> VLC has the record button afaik [13:32] <JEEB> which dumps the stream into a file :P [13:32] <JEEB> should be "easy enough", no? [13:32] <StaRetji1> ehm, yes, but let me read again what you wrote, to understand [13:33] <StaRetji1> oh, got it [13:33] <StaRetji1> dump dreambox stream to a file [13:33] <JEEB> and the ffmpeg trac is over at https://ffmpeg.org/trac/ffmpeg/ [13:33] <StaRetji1> then instead of stream try to open the file [13:33] <StaRetji1> omw [13:33] <StaRetji1> thx JEEB [13:33] <JEEB> yes [13:34] <JEEB> also if it works with a dumped stream then it might be the input protocol (whatever you're using, http/udp/whatever) [13:34] <JEEB> getting "samples" for that would be somewhat harder :D [13:35] <StaRetji1> it is http [13:35] <StaRetji1> btw, vlc doesn't offer to save it as is [13:35] <StaRetji1> only convert, if I'm not wrong [13:35] <JEEB> well, you can have it "convert" without doing anything to it, and mplayer can dump streams, too [13:36] <StaRetji1> ah, yes, DUmp raw inpu [13:36] <JEEB> http://www.mplayerhq.hu/DOCS/HTML/en/streaming.html <- mplayer's example is like this btw [13:38] <StaRetji1> i don't use mplayer [13:38] <StaRetji1> because I restream this file to rmtp [13:38] <JEEB> well, I'm just giving out options [13:38] <JEEB> :P [13:38] <JEEB> (for dumping samples) [13:38] <StaRetji1> hehe, lol, well thx [13:39] <StaRetji1> I appreciate help though [13:41] <StaRetji1> JEEB, no errors with saved file [13:41] <StaRetji1> vlc raw dump [13:42] <StaRetji1> ran with ffmpeg console, no errirs [13:42] <StaRetji1> what this means? [13:44] <StaRetji1> I mean, I understand it means it is something with http. in vlc works fine as vlc uses libmpeg2, so I assume if I could make ffmpeg to use libmpeg2 that would be solution [13:44] <JEEB> I think the problem isn't exactly in the decoder :P [13:44] <JEEB> it's in the protocol [13:44] <JEEB> or well, reading crap from http [13:45] <JEEB> can you pastebin a full log with the command line you used on what happens when you read from http? [13:45] <JEEB> also, what happens if you ghetto it with curl/wget and piping to ffmpeg? [13:46] <StaRetji1> yes, moment [13:48] <StaRetji1> http://pastebin.com/41HbMTnR [13:49] <StaRetji1> I'm not sure how to pipe it to ffmpeg, but sound like a veryu good idea, might work [13:49] <JEEB> curl http://10.0.1.18:31344 | ffmpeg -i - [13:49] <JEEB> something like this? [13:50] <JEEB> I think curl by default outputs to stdout, and thus you can pipe it to ffmpeg [13:50] <JEEB> ffmpeg takes input from stdin when input is set to "-" [13:50] <StaRetji1> got it, I will try it now [13:52] <StaRetji1> same error with pipe [13:53] <StaRetji1> curl http://10.0.1.18:31344 | ffmpeg -i "http://10.0.1.18:31344" -re -acodec libfaac -ar 22050 -vcodec libx264 -vpre ipod320 -f flv rtmp://10.0.1.123/live/livestream181792 [13:53] <JEEB> are you sure your encoding is going fast enough? [13:54] <JEEB> also what's -re ? [13:54] <StaRetji1> I don't know, I found this example on google [13:54] <JEEB> lawl [13:54] <StaRetji1> this worked well with test.ts file I created with vlc [13:54] <StaRetji1> no errors [13:54] <JEEB> VLC IIRC had some defaults for libx264 of its own [13:55] <JEEB> anyways, try -preset:v veryfast [13:55] <StaRetji1> ok [13:55] <StaRetji1> I do it now [13:56] <StaRetji1> ha, it wokrs [13:56] <StaRetji1> no errirs!!! [13:56] <StaRetji1> errors lol [13:56] <JEEB> so I was right [13:57] <JEEB> you were trying to use preset "medium" of libx264 [13:57] <StaRetji1> can you expmain please [13:57] <StaRetji1> explain* [13:57] <JEEB> and your cpu couldn't encode fast enough with that preset [13:57] <JEEB> so it had to start dropping frames [13:57] <StaRetji1> got it [13:57] <JEEB> http://mewiki.project357.com/wiki/X264_Settings#preset [13:57] <JEEB> ^ listing of presets [13:57] <JEEB> pick the slowest that still works for your realtime needs [13:58] <JEEB> also, see vbv-maxrate and vbv-bufsize [13:58] <StaRetji1> JEEB thank you so much dude!!! [13:58] <JEEB> you will need those for streaming [13:58] <StaRetji1> I really appreciate your help [13:58] <StaRetji1> thx :) [13:59] <JEEB> bufsize is basically how much you will be buffering, and maxrate is the maximum average bit rate within that buffer [14:00] <JEEB> so basically if you set a bufsize of 1200k and maxrate of 300k, it means that if you buffer 1200kb once, and your transfer speed will always be over 300kbps, you will not have to stop for buffering any more [14:00] <StaRetji1> got it [14:02] <JEEB> also, you haven't set a rate control mode, and libx264 by default sets it to crf 23, for streaming you usually want to set it to a crf value that looks "good enough" [14:02] <StaRetji1> hm [14:02] <JEEB> (lower = more bitrate) [14:02] <StaRetji1> yes [14:02] <JEEB> of course, vbv limits will be held in any case [14:02] <StaRetji1> I will play with the figures, but would gladly use your example [14:03] <StaRetji1> if you want to share it? :) [14:03] <JEEB> so if you set a too low crf for your content, it will just end up being limited by the vbv all the time :) [14:03] <JEEB> well, dunno -- 23 already is a good place to start at [14:04] <JEEB> if it looks bad, make it lower [14:04] <JEEB> if it looks good, try making it higher [14:04] <StaRetji1> Failed to set value 800k for option 'vbv-maxrate' [14:05] <StaRetji1> Unrecognized option 'vbv-maxrate' [14:05] <JEEB> heh, seems like vbv-maxrate is maxrate in ffmpeg [14:05] <StaRetji1> lol, I'm noob, I admit [14:05] Action: JEEB mostly uses x264 itself [14:05] <JEEB> and vbv-bufsize is just bufsize it seems [14:06] <StaRetji1> yes, you are right [14:06] <JEEB> anyways, that's all you need for streaming: 1) fast enough preset 2) some kind of sane rate control mode (crf or bitrate, I'd usually go with crf) 3) some kind of vbv (both maxrate and bufsize need to be set) according to your needs [14:07] <StaRetji1> thx once again, thx very much! [14:07] <JEEB> np [14:11] <JEEB> also, you can drop the vpre :P [14:11] <JEEB> you're clearly not encoding for the first ipods [14:11] <JEEB> lol [18:01] <ckb> http://pastebin.com/7NGJ7jjJ [18:01] <ckb> Can someone tell me what's wrong with this? [18:04] <ckb> It seems like I only get this problem when a certain user runs this command... [18:08] <ckb> anyone :\ [18:10] <sprzybilla> ckb: try using -minrate 1200k -maxrate 1200k -bufsize 400k [18:10] <sprzybilla> I think you have to set all of those for a CBR stream [18:10] <ckb> could it not be related to the stat file? [18:11] <ckb> and a CBR stream? [18:11] <sprzybilla> constant bitrate [18:11] <ckb> ah [18:11] <ckb> well, when I execute this encoding from my user, I don't have a problem, and it encodes fine [18:11] <ckb> but when apache runs the encoding, I get the "Error while opening encoder for output stream" [18:12] <sprzybilla> ahh - interesting [18:12] <sprzybilla> permissions for the stats file write location maybe? [18:12] <ckb> probably [18:12] <sprzybilla> try runnin it as su [18:12] <ckb> I'll change directory [18:12] <ckb> I don't have SU access [18:13] <sprzybilla> if that works you know its permissions for wherever that file is being written [18:13] <sprzybilla> ahh [18:13] <ckb> but I am running it as php /absolute/path/ as apache [18:13] <ckb> and php script.php as my user [18:13] <ckb> so that's MORE than likely the issue [18:14] <ckb> wait a second [18:14] <ckb> I'm already changing it [18:15] <ckb> & :| [18:17] <sprzybilla> maybe using the ffmpeg-php extension instead of calling it as a command would work better? http://ffmpeg-php.sourceforge.net/ [18:17] <sprzybilla> sorry, not too familliar with using it from a script [18:17] <ckb> ahhh [18:17] <ckb> I was changing it to the current script directory [18:17] <ckb> not to the tmp directory that is 777'd [18:17] <ckb> ^__^ [18:18] <sprzybilla> haha, yeah, that might help too. Best of luck! [18:20] <ckb> thanks spry! [19:52] <mcayland> hi everyone :) does anyone have a sample cmdline for converting a video suitable for viewing on an android phone? [19:53] <mcayland> i've got a sample video from the camera for comparison but can't quite work out the correct voodoo [20:32] <mbradshaw> I have a question regarding the seeking in ffmpeg that I'd like to make sure I'm doing correctly before I submit a bug report (the last time I submitted a bug report it turned out I was just doing things wrong) [20:32] <mbradshaw> Anybody familiar with seeking mpeg2video streams? [22:18] <Mista_D> any way to supress "packet too large, ignoring buffer limits to mux it" while encoding MPEG2 to MP4. I'd like to see any other errors though, so "-v quiet" won't cut it. [23:16] <Kingsy> guys, what would be a sample command to capture a webcam video stream from /dev/video0 to a file ? [23:16] <Kingsy> just something basic basic basic.. [23:19] <ubitux> http://ffmpeg.org/ffmpeg.html#video4linux2 [23:20] <Kingsy> ubitux: I didnt think -input_format was a valid flag [23:21] <Kingsy> for example ffmpeg: failed to set value 'mjpeg' for option 'input_format' [23:25] <ubitux> it works for me [23:25] <ubitux> maybe your ffmpeg is too old [23:25] <ubitux> you can just leave it out anyway, it will just not use the mjpeg source [23:25] <Kingsy> hmmm what would be a command that would work on all versions? ok [23:26] <Kingsy> ubitux: so it says [video4linux2 @ 0x136b340] Cannot open video device /dev/video0 : No such file or directory <-- but /dev/video0 is the source.. I can open it with vlc [23:26] <ubitux> the error is pretty explicit though [23:27] <ubitux> is it a weird webcam? [23:27] <Kingsy> I know.. but it doesnt really make sense.. if you can open vlc and type /dev/video0 in the source and it works.. [23:27] <Kingsy> ubitux: well I am not sure it is 100% supported.. but it seems to work in everything I have tried it in [23:28] <ubitux> mplayer tv:// works? [23:28] <Kingsy> oh it looks like the device had crashed.. [23:29] <Kingsy> weird.. [23:29] <burek> "hmmm what would be a command that would work on all versions? ok" ^^ [23:29] <burek> now, that's called optimism :D [23:29] <Kingsy> haha [23:29] <Kingsy> ok soo.. now i get Error while opening encoder for output stream #0.0 - maybe incorrect parameters such as bit_rate, rate, width or height [23:29] <burek> can you please use pastebin.com, to show your command line and its output? [23:29] <Kingsy> sure [23:30] <ubitux> i leave it to you burek ;) [23:30] Action: ubitux & [23:30] <burek> :beer: :) [23:30] <ubitux> :applejuice: [23:30] <burek> ctrl+z && bg [23:30] <burek> yes, I always forget :) [23:30] <Kingsy> burek: http://pastie.org/3770714 [23:30] <burek> MPEG1/2 does not support 15/1 fps [23:31] <burek> your webcam defaults to rawvideo (over usb I guess) and 15 fps [23:31] <Kingsy> burek: so this webcam does not do anything better than 15fps ? [23:31] <burek> try: ffmpeg -f video4linux2 -r 30 -i /dev/video0 out.mpeg [23:31] <Kingsy> or it just defaults to that? [23:31] <burek> or ffmpeg -f video4linux2 -i /dev/video0 -vcodec libx264 out.flv [23:31] <burek> it defaults to that [23:32] <burek> you can query capacity of your webcam using v4l2-ctl [23:32] <burek> it is in v4l-tools or v4l-utils package [23:32] <burek> I'm not sure [23:33] <Kingsy> burek: ok, that worked great... really smooth.. but the quality was pretty terrible.. really blocky.. can we improve that ? [23:33] <burek> yes [23:33] <burek> which one did you use [23:33] <burek> 1st or 2nd? [23:33] <Kingsy> ffmpeg -f video4linux2 -r 30 -i /dev/video0 out.mpeg [23:34] <burek> ffmpeg -f video4linux2 -r 30 -i /dev/video0 -b 1000k out.mpeg [23:34] <burek> that means you want video bitrate to be 1 mbps [23:35] <Kingsy> ok that isnt half bad.. [23:35] <Kingsy> still a bit blocky... [23:35] <burek> increase it if you like [23:35] <Kingsy> how far can you push usb 2.0 ? [23:36] <burek> I think 480 Mbps [23:36] <burek> right now you're at Stream #0.0: Video: rawvideo, yuyv422, 640x480, 73728 kb/s [23:36] <Kingsy> wow.. and I am using 1 ? [23:36] <burek> you are encoding to 1 mbps mpeg [23:36] <burek> which is pretty high [23:36] <burek> but try setting for example [23:36] <Kingsy> oh ok [23:36] <burek> 2,3,4 mbps [23:36] <burek> or use libx264 [23:36] <burek> it's way better [23:36] <Kingsy> yeah I just tried 2 [23:37] <Kingsy> so I can use -b libx264 ? [23:37] <burek> for example -f libx264 -crf 20 [23:37] <burek> without -b [23:37] <burek> ffmpeg -f video4linux2 -i /dev/video0 -vcodec libx264 -crf 20 out.mp4 [23:37] <Kingsy> Unknown encoder 'libx264' [23:37] <Kingsy> so I obviously don't have that [23:38] <burek> did you apt-get install your ffmpeg or compiled [23:38] <Kingsy> I apt-get [23:38] <burek> hmh.. what distro [23:38] <Kingsy> zubuntu [23:38] <Kingsy> xubuntu** [23:38] <burek> :D [23:38] <burek> you can try static build if you like [23:38] <burek> https://sites.google.com/site/linuxencoding/builds [23:38] <Kingsy> well I will stick with a standard -b 1000k for now [23:38] <Kingsy> burek: it looks ok, for testing [23:39] <burek> ok [23:39] <Kingsy> burek: one other thing.. [23:39] <burek> x264 is really the way to go [23:39] <burek> if you want it really good [23:39] <Kingsy> burek: and if you can get this as easy as you did that.. then I will eat my hat [23:39] <burek> what [23:39] <Kingsy> cos I have been trying to do this for 5 days [23:39] <Kingsy> heh [23:39] <burek> I hope your hat is made out of chocolate or something sweet :) [23:40] <Kingsy> I also have a condenser mic plugged into hw:2 (which I can record from using audacity so I know it works) and I want to add that to the video stream .. is that possible? [23:40] <Kingsy> I am sick of trying to do with to vlc and mencoder so I was just going to sync and stitch them together seperately [23:41] <burek> ffmpeg -f video4linux2 -i /dev/video0 -f alsa hw:2 -acodec aac -strict experimental -ab 128k -ac 2 -ar 44100 -vcodec libx264 -crf 20 out.mp4 [23:42] <Kingsy> argh I am getting that Unknown encoder 'libx264' again, can I do this with mpg ? or just avi ? [23:42] <burek> ffmpeg -f video4linux2 -i /dev/video0 -f alsa hw:2 -acodec aac -strict experimental -ab 128k -ac 2 -ar 44100 -b 1000k out.mp4 [23:42] <Kingsy> unless you know how I can install x264 on ubuntu ? [23:42] <Kingsy> ok [23:42] <Kingsy> it says Output #0, alsa, to 'hw:2': [23:42] <Kingsy> Output file #0 does not contain any stream [23:42] <burek> you can install it using apt-get install libx264 libx264-dev [23:43] <burek> can you please use pastebin.com, to show your command line and its output? [23:43] <Kingsy> sure [23:43] <Kingsy> http://pastie.org/3770776 [23:44] <burek> my bad [23:44] <burek> ffmpeg -f video4linux2 -r 30 -i /dev/video0 -f alsa -i hw:2 -acodec aac -strict experimental -ab 128k -ac 2 -ar 44100 -b 1000k out.mp4 [23:45] <burek> -i was missing :) [23:45] <Kingsy> burek: another different error --> http://pastie.org/3770791 [23:47] <burek> :) [23:47] <burek> mono mic [23:47] <burek> ffmpeg -f video4linux2 -r 30 -i /dev/video0 -f alsa -ac 1 -i hw:2 -acodec aac -strict experimental -ab 128k -ac 1 -ar 44100 -b 1000k out.mp4 [23:47] <Kingsy> oh yeah it is mono [23:48] <Kingsy> burek: it works.. but its out of sync.. [23:48] <Kingsy> this is KINDA the situation I ended up with in mencoder.. but it was ALOT worse than that.. [23:49] <burek> well, since you are not taking the input from the same hardware device [23:49] <burek> it is kinda difficult to sync [23:49] <burek> but try playing around with -async and -vsync [23:49] <Kingsy> yeah I was thinking that.. [23:49] <burek> http://ffmpeg.org/ffmpeg.html#Advanced-options [23:49] <burek> scroll down a little bit for -vsync/async [23:50] <Kingsy> burek: saying as though I am capturing from seperate devices.. would you just capture them seperatly using different programs at the same time and sync them afterwards? [23:50] <burek> your best chance is using same process to capture both if possible [23:50] <burek> syncing 2 processes is.. well.. interesting :) [23:50] <Kingsy> burek: I just don't really know what I am doing with this is all [23:51] <burek> also try switching inputs like: [23:51] <burek> ffmpeg -f alsa -ac 1 -i hw:2 -f video4linux2 -r 30 -i /dev/video0 -acodec aac -strict experimental -ab 128k -ac 1 -ar 44100 -b 1000k out.mp4 [23:52] <Kingsy> naaa its still miles out of sync [23:53] <burek> try this: ffmpeg -f alsa -ac 1 -i hw:2 -f video4linux2 -r 30 -i /dev/video0 -ab 128k -b 1000k out.mp4 [23:53] <Kingsy> burek: surely tho, if you just add -async 10 or whatever.. you could be on forever.. it could be ANY number [23:53] <Kingsy> ok [23:54] <Kingsy> burek: naaa still really far out of sync [23:54] <burek> well, I would suggest you to file a bug report [23:54] <burek> but they are going to ask you to use latest ffmpeg from git [23:54] <burek> before you submit it [23:54] <burek> so, if you can, try updating your ffmpeg [23:55] <Kingsy> is this a bug? I thought it was just a limitation of using two seperate devices ? [23:55] <ubitux> updating? you mean "using"? :) [23:55] <ubitux> he is using libav, not ffmpeg :p [23:55] <burek> isn't libav just a library [23:55] <ubitux> ;) [23:55] <ubitux> (are you serious or& ?) [23:56] <burek> no, I don't know about libav tool [23:56] <burek> i thought it's libav like lib for audio/video [23:56] <ubitux> libav is the fork [23:56] <burek> oh, so what's the diff? [23:56] <Kingsy> I think this would be easier to just record these seperatly and put them together after or something.. I am just sick of this [23:57] <burek> Kingsy, you can quickly try with latest static build if you want: http://ffmpeg.gusari.org/static/ [23:57] <burek> just download and run the same command [23:57] <ubitux> burek: sorry i'm too lazy to enter in long explanation, maybe tomorrow ;) [23:57] <burek> if it is still out of sync, well.. [23:57] <burek> :) [23:57] <burek> ubitux, ok :) I'll take a look on Google :) [23:59] <Kingsy> burek: 1 sec doing it now [00:00] --- Thu Apr 12 2012

1 0

ffmpeg-devel.log.20120410
by burek 11 Apr '12

11 Apr '12

[00:52] <meekohi> Hey devs. Is there any way to adjust saturation given the current video filters? It seems like they are all based on per-pixel-channel lookup tables? [00:53] <meekohi> lot specifically I guess [00:53] <meekohi> lut* [02:00] <michaelni> meekohi, see vf_eq / vf_eq2 [02:01] <michaelni> under vf_mp [02:01] <meekohi> michaelni: It seems like they do the same thing, just in YUV space. [02:01] <meekohi> michaelni: I need a little more control over how the channels interact. [02:02] <michaelni> elaborate, please, what are you trying to do ? [02:04] <meekohi> michaelni: I'd like to adjust pixel values by double their distance from that pixel's average value (grey). [02:04] <meekohi> doubling* [02:04] <meekohi> r = r + 2*(r - (r+g+b)/3) [02:04] <meekohi> something like this. [02:05] <michaelni> geq should be able to do it [02:05] <michaelni> it happily takes arbitrary expressions [02:06] <meekohi> michaelni: But your expression can't access the green channel when it's compute the red value and etc [02:07] <meekohi> michaelni: It basically just makes a lookup table 256 wide and applies it to each pixel. [02:07] <michaelni> no [02:07] <michaelni> it evaluates the expression for each pixel [02:08] <michaelni> you can do things like texture maping with it [02:08] <michaelni> or drawing a mandelbrot fraktal [02:08] <meekohi> okay how would one create a greyscale image using eq2 then? [02:10] <meekohi> or if you can point me at an example like what you're describing that might help too [02:11] <michaelni> -vf 'mp=geq=lum(X\,Y):128:128' [02:13] <michaelni> also if you run into some limitations of the expression evaluator, you can look into libavutil/eval* and improve it :) [02:14] <michaelni> for example functions to convert between rgb and yuv are missing and may be usefull [02:14] <meekohi> Yeah this looks very interesting, trying to wrap my head around it exactly. [02:17] <meekohi> Really I am just insanely jealous of how fast ffmpeg is. Right now I use it to write out a bunch of jpegs and then adjust color using a little C program I wrote [02:17] <meekohi> ffmpeg is done almost instantly but the C programs takes like 15 seconds [02:17] <meekohi> Makes me very sad. [02:17] <meekohi> So it'd be great to do the adjustment inside ffmpeg since I'm doing that anyway. [02:17] <michaelni> geq is not exactly fast ... [02:18] <michaelni> evaluating arbitrary expressions is not a fast operation [02:18] <michaelni> if speed matters its better to write a avfilter in C [02:19] <meekohi> Yeah makes sense. I'm hesitant to go down that road since I don't know much about writing filters or how long it will take. [02:23] <Compn> um [02:23] <Compn> meekohi : do you just want to change from a color profile to another color profile ? [02:24] <Compn> like one of those standard color profiles 702 or something ? [02:24] Action: Compn cant remember the name [02:24] <meekohi> Compn: No, trying to do some custom saturation adjustment. [02:24] <Compn> oh [02:24] <Compn> then nevermind :) [02:24] <meekohi> Compn: ;D [12:29] <j-b> why does ffv1 in 10bits outputs a YUV16bits? [12:48] <Compn> i think there was some discussion about that on the ml [12:52] <j-b> I did not see that [12:53] <av500> j-b: is there any good trolling in paris tonight? [12:54] <j-b> "is ffv1 a good codec?" at my place [12:54] <av500> cool [12:54] <av500> free beer? [12:55] <j-b> http://lacantine.org/events/digikaaniversaire booring [12:58] <Compn> what colorspace is yuv16bit ? [12:59] <Compn> oh YUV420P16 [12:59] <Compn> hmm [12:59] <av500> j-b: yeah booring [12:59] <Compn> did you specify 10bit and maybe ffmpeg defaults to 16bit ? [13:00] <Compn> ffv1: Assume encoding all bits is wanted if bits_per_raw_sample == 0. [13:00] Action: Compn hasnt a clue [13:16] <kierank> probably ffv1 never got updated to the p10 pix_fmt [13:33] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r2f0801527d 10ffmpeg/libswresample/resample.c: [13:33] <CIA-17> ffmpeg: resample: update copyright years, they where incorrect. [13:33] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [13:33] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r605bcf6101 10ffmpeg/libswresample/swresample_test.c: [13:33] <CIA-17> ffmpeg: swr-test: Add newline to test output [13:33] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [13:33] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r3d9338b1c2 10ffmpeg/libswresample/swresample.c: [13:33] <CIA-17> ffmpeg: swr: support int32 and float as internal sample formats [13:33] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [13:33] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r7f1ae79d38 10ffmpeg/libswresample/ (4 files): [13:33] <CIA-17> ffmpeg: swr: support float & int32 in the resampler [13:33] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [14:03] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r24ab1abfb6 10ffmpeg/libswresample/ (resample.c swresample_internal.h): [14:03] <CIA-17> ffmpeg: resample: support double precission resampling [14:03] <CIA-17> ffmpeg: This commit is dedicated to the audiophiles who can hear it when a [14:03] <CIA-17> ffmpeg: needle is dropped on the moon. [14:03] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [14:09] <ubitux> haha [14:43] <meekohi> lol awesome [15:01] <michaelni> j-b, just tried, when i encode yuv420p10 with ffv1 i get yuv420p10le out not 16bit [15:03] <j-b> http://www.archive.org/download/test20100622_v210_and_ffv1/DHC36.ffv1.flac.… must be 16bits then, or my user is stoopid (or both) [15:19] <ubitux> https://ffmpeg.org/trac/ffmpeg/ticket/1180 can't we solve this without reverting the commit? [15:19] <ubitux> i don't understand the reason of the block align hack [15:19] <ubitux> but indeed it seems to fix the issue [15:29] <ubitux> anyway, i should report this to the author of the patch [15:44] <ubitux> reverting it doesn't solve the remux with the -ss as input option though [15:45] <CIA-17> ffmpeg: 03Nicolas George 07master * r423047ea31 10ffmpeg/libavfilter/vf_drawtext.c: vf_drawtext: fix memory leak of glyph data. [15:45] <CIA-17> ffmpeg: 03Nicolas George 07master * rbe9b0d2c5a 10ffmpeg/libavfilter/vf_drawtext.c: vf_drawtext: fix memory leak of draw expression. [15:45] <CIA-17> ffmpeg: 03Nicolas George 07master * r40b7a27bad 10ffmpeg/ (4 files in 3 dirs): vf_drawtext: fontconfig support. [15:50] <Compn> j-b : are you taking the word of a user without reproducing it? gasp! [16:34] <silverrocker> I have installed ffmpeg but. I cant find apiexample.c as stated in the first section of http://ffmpeg.org/developer.html . I'm running archlinux and there is only one package which should have the devel files aswell but I don't seem to have the devel files [16:34] <silverrocker> I think [16:35] <ubitux> the examples are not installed [16:36] <ubitux> http://git.videolan.org/?p=ffmpeg.git;a=tree;f=doc/examples;hb=HEAD [16:36] <ubitux> maybe they should [16:36] <ubitux> and indeed the documentation is not up to date [16:37] <j-b> http://source.ffmpeg.org/?;a=tree;f=doc/examples;hb=HEAD [16:38] <ubitux> :) [16:38] <Compn> j-b : so many people ask about playing and recording at the same time. i think vlc needs to advertise that feature more :) [16:39] <Compn> and having that feature in ffmpeg would be nice too [16:40] <Compn> j-b : possibly make it easier too, just have a little checkbox 'automatically download all streams to disk' :) [16:40] <Compn> or save i should say [16:41] <silverrocker> Compn: seems like I can't clone because I'm getting a connection timeout. [16:41] <j-b> Compn: people want porn too :) [16:42] <ubitux> silverrocker: what's your clone url? [16:42] <silverrocker> "git clone git://git.videolan.org/ffmpeg.git" gives me a timeout. is the server down or so? [16:42] <silverrocker> ubitux: I have also tried git clone git://source.ffmpeg.org/ffmpeg.git [16:43] <ubitux> works here [16:43] <silverrocker> hmm, maybe campus server blocks it [16:43] <ubitux> you likely need a socks proxy [16:44] <ubitux> if you have a http proxy: http://git.videolan.org/git/ffmpeg.git [16:45] <silverrocker> so I need to git clone http://git.videolan.org/git/ffmpeg.git ? [16:45] <silverrocker> seems to work, thank you [16:50] <ubitux> here we go, gcc 4.7; let's see how much fate likes it... [16:51] <silverrocker> ubitux: it seems when I try to just make the examples, it cant find the includes. What should I add to my PATH to get this working? [16:51] <silverrocker> ubitux: I'm missing libavfilter/asrc_abuffer.h: No such file or directory [16:53] <ubitux> cd doc/examples; make doesn't work? [16:54] <silverrocker> nope [16:54] <silverrocker> but probably because I'm missing libraries [16:54] <silverrocker> and not becaues of the code [16:54] <ubitux> PKG_CONFIG_PATH needs to be set to the directory of a fairly recent enough ffmpeg installed (./configure --prefix=/tmp/ffmpeg-install && make install and then maybe PKG_CONFIG_PATH=/tmp/ffmpeg-install/lib make in the doc/examples directory) [16:54] <ubitux> or something like that [16:57] <silverrocker> ubitux: the thing is that I'm using archlinux which installs ffmpeg (which is built from the git repo) so I think I have a fairly recent ffmpeg. Also I'd rather not make install on top of what I already have, although you seem to have prefixed it to go to the /tmp/ so I should be ok? [16:59] <Compn> ubitux makes the mistake of helping user in devel channel :) [17:00] <ubitux> mmh indeed there might be a problem with asrc_abuffer.h [17:00] <ubitux> it looks like it is not installed [17:00] <silverrocker> Compn: should I go somewhere else? [17:00] <ubitux> it's ok, stay there :) [17:01] <silverrocker> :) I'm running make now [17:01] <ubitux> you can make the other examples [17:02] <ubitux> try make filtering_video for instance [17:02] <silverrocker> ok, after the make of whole ffmpeg is done [17:02] <ubitux> the makefile is independant [17:02] <ubitux> you can run make in the doc example directory [17:02] <silverrocker> ah, ok [17:03] <silverrocker> ubitux; I seem to have alot more errors sudenly [17:04] <silverrocker> give me a se [17:04] <silverrocker> c [17:04] <Compn> silverrocker : i'm just bugging ubitux , no worries [17:04] <ubitux> libavfilter/Makefile:HEADERS = avcodec.h avfilter.h avfiltergraph.h buffersink.h version.h vsrc_buffer.h [17:04] <ubitux> this line is missing asrc_abuffer.h [17:05] <ubitux> afaict [17:06] <silverrocker> ok Ill add that in a bit, give me a minute. it seems that after I have run ./configure with that prefix I have loads more errors for some reason I'm trying to figure out now [17:08] <ubitux> changing this makefile line should make the install rule deploy the file [17:09] <silverrocker> I'm now getting errors of /usr/include/bits/types.h, /usr/include/ctype.h and much more. Something is gone horribly wrong I think [17:09] <ubitux> pastebin it? [17:11] <silverrocker> http://pastebin.com/HT3vHgnW [17:11] <silverrocker> and its much more [17:13] <ubitux> what did you do? [17:13] <silverrocker> It started after ./configure in the rootdirectory of the git cloned ffmpeg [17:14] <silverrocker> I don't undestand how this affects my stdlibs [17:14] <ubitux> is this the result of a "make" in doc/examples? [17:15] <silverrocker> yes [17:15] <ubitux> just "make"? [17:15] <silverrocker> yes [17:15] <silverrocker> after a make clean [17:15] <silverrocker> I have added HEADERS = avcodec.h avfilter.h avfiltergraph.h buffersink.h version.h vsrc_buffer.h but gets no difference [17:16] <ubitux> this will only affect the make install from the root [17:16] <ubitux> can you paste the full output of the make in doc/examples? [17:16] <silverrocker> ill try [17:18] <silverrocker> http://pastebin.com/GpkGgJuj [17:18] <silverrocker> mind that I already had ffmpeg installed from the package manger of my distro, I hope it's not conflicting [17:19] <ubitux> you changed decoding_encoding.c [17:19] <ubitux> type git diff and you'll see [17:20] <ubitux> it looks like you removed the first line or something [17:20] <ubitux> (you likely saved/quit after looking in the file, without noticing you messed the file :p) [17:22] <silverrocker> thank you! that was exactly the problem [17:23] <silverrocker> ok now the makefile still gives me the same problem as before ./configuring and making in /tmp [17:23] <ubitux> seems i have issue with the decoding_encoding example too [17:24] <ubitux> (LDFLAGS=-lm make decoding_encoding seems to fix to work) [17:29] <silverrocker> so I need to replace the LDFLAGS line with that? [17:30] <ubitux> juste prefix your make [17:31] <ubitux> anyway, when everything is working for you, feel free to send patches [17:31] <silverrocker> patching the makefile you mean? [17:31] <ubitux> yup [17:32] <silverrocker> okay :) [18:27] <silverrocker> ok, now I'm looking through the example. what I want to achieve is have all the yuv data for each frame from a given file [18:27] <silverrocker> I see that you need to supply the decoder that you want to use to decode a given file with av_find_codec [18:28] <silverrocker> am I correct to assume that when I run my program using the library and doing myprogram -somecodec file.ext ; where somecodec is parsed and passed to av_find_codec I can get the yuv data? [20:11] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r5c1f312888 10ffmpeg/libswresample/rematrix.c: [20:11] <CIA-17> ffmpeg: swr: update copyright year for rematrix [20:11] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [20:11] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rdb2eadb2f5 10ffmpeg/libswresample/ (6 files): [20:11] <CIA-17> ffmpeg: swr: add dither support. [20:11] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [20:12] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r5b2be9bcc0 10ffmpeg/libswresample/swresample_test.c: [20:12] <CIA-17> ffmpeg: swr-test: include opt.h, this is needed for changing options for testing. [20:12] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [20:12] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r7c673d73ec 10ffmpeg/libswresample/rematrix.c: [20:12] <CIA-17> ffmpeg: swr: use assert in swri_sum2 to check that only supported formats are used. [20:12] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [20:25] <ubitux> Compn: you maintain the samples base on http://samples.mplayerhq.hu/ right? [20:26] <ubitux> i'd like to add a 24M sample; standard h264/aac in mkv, but which leads to yet another a/v sync issue with stream copy [20:53] <CIA-17> ffmpeg: 03Reimar Döffinger 07master * r1ac606bae6 10ffmpeg/libavcodec/aacsbr.c: [20:53] <CIA-17> ffmpeg: aacsbr: silence message for SBR extension "padding". [20:53] <CIA-17> ffmpeg: Some files contain a few additional, all-0 bits. [20:53] <CIA-17> ffmpeg: Check for that case and don't print incorrect "not supported" [20:53] <CIA-17> ffmpeg: message. [20:53] <CIA-17> ffmpeg: Fixes trac issue #836. [20:53] <CIA-17> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger(a)gmx.de> [21:31] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r0170889372 10ffmpeg/libswresample/swresample.c: [21:31] <CIA-17> ffmpeg: swr: add SWR_DITHER_RECTANGULAR to AVOptions array [21:31] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [21:31] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r0cff3e130b 10ffmpeg/libswresample/swresample.c: [21:31] <CIA-17> ffmpeg: swr: remove unused variable [21:31] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [21:31] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * rb1ef4dc406 10ffmpeg/libswresample/ (dither.c swresample.c swresample.h): [21:31] <CIA-17> ffmpeg: swr: add triangular dither support [21:31] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [22:30] <Compn> ubitux : is it on http? i can wget it easily, or you can upload it to upload.ffmpeg.org [22:31] <Compn> ubitux : btw, i maintain http://samples.ffmpeg.org , mike maintains samples.mplayerhq.hu , we've been trying to get him to rsync with us , since we cant upload/move anything around on his server :\ [22:31] <Compn> and he cant make accounts for us [22:31] <Compn> so its one big fubar [22:31] <Compn> we rsync with samples.libav.org (which iirc hasnt changed much) [22:31] <Compn> and samples.mphq [22:32] <Compn> and we could just change the dns of samples.mphq to point at our server. but uhh. mike is a good neutral party to maintain it [22:33] <Compn> so gimme a url or upload it to ftp://upload.ffmpeg.org [22:49] <ubitux> Compn: i have access to the videolan.org one; i guess they're not linked whatsoever? [22:50] <Compn> which one is the videolan.org one [22:50] <Compn> you mean streams.videolan.org ? [22:50] <ubitux> yes [22:51] <Compn> thats j-b's server, which i think i only have read access to incoming [22:51] <ubitux> mmh ok [22:51] <Compn> pretty confusing [22:51] <Compn> at least we have multiple mirrors :) [22:51] <ubitux> btw, while i'm uploading the file, do you mind syncing these in the sub directory: http://lucy.pkh.me/samples/jss/ (jss=jacosub) [22:54] <Compn> done, http://samples.ffmpeg.org/sub/jss [22:54] <ubitux> awesome, thanks a lot [22:54] <ubitux> the mkv is almost done [22:54] <ubitux> Compn: here it is: http://lucy.pkh.me/samples/ppd.mkv [22:55] <Compn> where in repo do you want me to put it ? [22:55] <Compn> (and where is txt file :P) [22:55] <ubitux> i guess in the matroska directory? [22:56] <ubitux> maybe a better name would be: paniponidash-begin-h264-aac.mkv or sth like this [22:56] <Compn> i dont think knowing its from paniponydash helps ffmpeg devels much ;P [22:56] <ubitux> not sure how to make the file is somehow "valid" [22:56] <Compn> ffmpeg-desync-bug.mkv maybe [22:56] <ubitux> well, this file is correct :) [22:56] <Compn> or bug10xxx.mkv [22:57] <ubitux> bug remuxing it with a seek fails [22:57] <ubitux> the issue is not with the file, but it's a good "pattern" file [22:57] <Compn> i'll just put it in mkv then [22:57] <ubitux> ok :) [22:58] <Compn> http://samples.ffmpeg.org/Matroska/ppd.mkv [22:58] <Compn> there it is [22:58] <ubitux> great, thanks :) [23:00] Action: Compn gets annoyed at users who wont share samples because 'its copyrighted' [23:00] Action: ubitux doesn't care a second [23:01] <ubitux> maybe i should. [23:01] Action: Compn gets more annoyed at users who are under nda and cant share samples [23:01] <JEEB> oh yes, NDAs [23:01] <JEEB> always "fun" [23:04] <nevcairiel> just tell them to sod off [23:04] <nevcairiel> if they want their bug fixed, get authorization to share a tiny sample [23:07] <ubitux> and here we go: https://ffmpeg.org/trac/ffmpeg/ticket/1180#comment:3 [23:07] <ubitux> thanks again Compn :) [23:07] <CIA-17> ffmpeg: 03Justin Ruggles 07master * rb073819bc9 10ffmpeg/avconv.c: [23:07] <CIA-17> ffmpeg: avconv: allow '-async -1' to disable timestamp sync for audio encoding [23:07] <CIA-17> ffmpeg: This will allow a workaround for cases where input timestamps are invalid or [23:07] <CIA-17> ffmpeg: when decoder delay of 1 packet or more confuses avconv into using the wrong [23:07] <CIA-17> ffmpeg: timestamps as a sync reference. [23:07] <CIA-17> ffmpeg: 03Diego Biurrun 07master * r02c39f056a 10ffmpeg/ (3 files in 2 dirs): ppc: Add/remove a number of const qualifiers to fix related warnings. [23:07] <CIA-17> ffmpeg: 03Anton Khirnov 07master * r6aadfbda09 10ffmpeg/libavcodec/avcodec.h: lavc doxy: add AVPicture functions to a doxy group. [23:07] <CIA-17> ffmpeg: 03Anton Khirnov 07master * r294b3a5074 10ffmpeg/libavcodec/avcodec.h: lavc doxy: add core functions/definitions to a doxy group. [23:07] <ubitux> these issues annoys me quite often :p [23:07] <CIA-17> ffmpeg: 03Anton Khirnov 07master * r40ca0e6a75 10ffmpeg/libavcodec/avcodec.h: [23:07] <CIA-17> ffmpeg: lavc doxy: replace \ with / [23:07] <CIA-17> ffmpeg: It's the more proper symbol to use and it prevents doxygen from thinking [23:07] <CIA-17> ffmpeg: it's a command. [23:07] <CIA-17> ffmpeg: 03Anton Khirnov 07master * rc8ef8464c2 10ffmpeg/libavcodec/avcodec.h: lavc doxy: add decoding functions to a doxy group. [23:07] <CIA-17> ffmpeg: 03Anton Khirnov 07master * rf038515f0a 10ffmpeg/libavcodec/avcodec.h: lavc doxy: add encoding functions to a doxy group. [23:07] <CIA-17> ffmpeg: 03Anton Khirnov 07master * r56512ce104 10ffmpeg/libavcodec/avcodec.h: lavc doxy: add resampling functions to a doxy group. [23:07] <CIA-17> ffmpeg: 03Anton Khirnov 07master * r30f3f62563 10ffmpeg/libavcodec/avcodec.h: [23:07] <CIA-17> ffmpeg: lavc doxy: add AVPacket-related stuff to a separate doxy group. [23:07] <CIA-17> ffmpeg: Also move AV_PKT_DATA_PARAM_CHANGE/AV_PKT_DATA_H263_MB_INFO to the [23:07] <CIA-17> ffmpeg: proper place. [23:07] <CIA-17> ffmpeg: 03Anton Khirnov 07master * re36b25d1df 10ffmpeg/libavfilter/vf_overlay.c: [23:07] <CIA-17> ffmpeg: vf_overlay: implement poll_frame() [23:07] <CIA-17> ffmpeg: Signal that it can output a frame when there are frames on the main [23:07] <CIA-17> ffmpeg: input and EOF on the overlay input, but a frame is buffered -- e.g. [23:07] <CIA-17> ffmpeg: single picture overlay. [23:07] <CIA-17> ffmpeg: 03Anton Khirnov 07master * r7c59b5c2a5 10ffmpeg/libavcodec/ (dxva2.h vaapi.h vda.h vdpau.h version.h xvmc.h): lavc doxy: add all installed headers to doxy groups. [23:07] <CIA-17> ffmpeg: 03Anton Khirnov 07master * rec57b7de74 10ffmpeg/libavcodec/avfft.h: lavc doxy: add avfft to the main lavc group. [23:07] <CIA-17> ffmpeg: 03Anton Khirnov 07master * r8a74029ea2 10ffmpeg/libavcodec/avcodec.h: lavc doxy: add remaining avcodec.h functions to a misc doxygen group. [23:07] <Compn> ubitux : ... mplayer2? :P [23:08] <CIA-17> ffmpeg: 03Martin Storsjö 07master * r14f063d294 10ffmpeg/libavutil/mem.c: (log message trimmed) [23:08] <CIA-17> ffmpeg: mem: Consistently return NULL for av_malloc(0) [23:08] <CIA-17> ffmpeg: Plain POSIX malloc(0) is allowed to return either NULL or a [23:08] <CIA-17> ffmpeg: non-NULL pointer. The calling code should be ready to handle [23:08] <CIA-17> ffmpeg: a NULL return as a correct return (instead of a failure) if the size [23:08] <CIA-17> ffmpeg: to allocate was 0 - this makes sure the condition is handled [23:08] <CIA-17> ffmpeg: in a consistent way across platforms. [23:08] <CIA-17> (37 lines omitted) [23:08] <ubitux> Compn: i was just testing playback with random players :p [23:09] <Compn> ah [23:09] Action: Compn reading whole thing slowly [23:09] <ubitux> now i need to remember all the others samples where i expected such broken behaviour [23:09] <Compn> why dont we have any sync ref tests ? [23:10] <ubitux> good question :) [23:11] <ubitux> also we don't have tests for -async afaik [23:13] <ubitux> anyway, ffmpeg seems ok with gcc 4.7 :) [23:13] <ubitux> (except the threads tests which still randomly fail) [23:14] <Compn> ubitux : i know its ffmpeg bug, but did you try mplayer -demuxer avi vs mplayer -demuxer lavf? [23:14] <Compn> on the output [23:14] <Compn> avi output of course [23:15] <Compn> mplayer uses lavf mkv demuxer by default for mkv [23:18] <ubitux> well, mplayer is using the internal by default and mplayer2 the lavf one afaik [23:18] <ubitux> i reported these results; they're both broken, in a different way :) [23:20] <Compn> ah [23:20] <ubitux> demuxer avi: no video for a few seconds, then likely ok but seek is broken ( audio gets out of sync) [23:20] <ubitux> demuxer lavf: simply out of sync [23:22] <ubitux> for the mkv, it's "just" out of sync, for both [23:22] <iive> ubitux: is that just by using gcc4.7? [23:22] <ubitux> mmh? [23:22] <ubitux> the sync issues? no, not related [23:22] <Compn> git master is broken :\ [23:43] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r8702e35b9b 10ffmpeg/libavcodec/vda.c: [23:43] <CIA-17> ffmpeg: vda: try to fix compile [23:43] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> [23:48] <michaelni> Compn, should be fixed [23:49] <michaelni> i dunno how that happened ... [23:51] <meekohi> Who knows how this thing works. [23:55] <ubitux> michaelni: + AudioData dither; ///< cached audio data (convert and resample purpose) [23:55] <ubitux> can't you propose a slightly better comment? :p [23:55] <ubitux> it looks like a copy paste you forgot to change [23:56] <michaelni> ill fix it in my next push [23:56] <ubitux> thx :) [00:00] --- Wed Apr 11 2012

1 0

ffmpeg.log.20120410
by burek 11 Apr '12

11 Apr '12

[01:53] <lidenbrock> is ffmpeg better than mencoder? [02:24] <ckb> What type of hardware is needed for an encoding server to handle a queue of videos to be encoded? [02:27] <srez> How can I get ffmpeg t =o print out information about how a wav file is encoded? (i.e., mono/stereo, 8/16kHz, etc.) [02:29] <srez> nvm [05:29] <ckb> What type of hardware is needed for an encoding server to handle a queue of videos to be encoded? [05:34] <relaxed> ckb: The fastest intel cpu you can afford. [05:34] <ckb> why specifically intel? [05:34] <ckb> and fastest? or most cores? or most ram? [05:34] <ckb> or all the above? [05:35] <relaxed> Intel is king. Yes, the more cores the better. [05:37] <lidenbrock> Hi. I need an mpeg 4.4.2 decoder plugin. [05:37] <relaxed> lidenbrock: plugin? for what? [05:40] <lidenbrock> to see I video I have created. Mplayer is accusing it doesn't exist on my system. [05:41] <relaxed> try ffplay [05:46] <relaxed> lidenbrock: I find the time it takes you to respnd distrubing. [05:49] <lidenbrock> lol [05:49] <lidenbrock> sorry [05:50] <lidenbrock> I need an plugin for microsoft 4.4.2 mpeg [05:50] <lidenbrock> that's it [05:50] <relaxed> Are you using linux? [05:51] <lidenbrock> yes [05:52] <relaxed> Do you have ffplay? [05:54] <lidenbrock> not yet [05:55] <relaxed> not yet? as in you will have it soon? [05:56] <relaxed> I've had four pints tonight and my patients for your problem is dwindling. [05:57] <lidenbrock> ;p [05:57] <lidenbrock> man [05:57] <Tthread> Hi ) [05:57] <lidenbrock> I'm aptituding [05:57] <lidenbrock> :p [05:57] <lidenbrock> I couldn't find the decoder [05:57] <relaxed> debian? [05:57] <lidenbrock> and it's driving me crazy [05:57] <lidenbrock> yeah, debian [05:57] <Tthread> did u test ffmpeg with vda support? [05:57] <relaxed> unstable? [05:58] <lidenbrock> no, stable [05:58] <relaxed> of course it's stable [05:58] Action: relaxed slaps head [05:58] <lidenbrock> LOL [05:59] <lidenbrock> man [05:59] <lidenbrock> I'll get some sleep [05:59] <lidenbrock> I'm almost falling here [05:59] <lidenbrock> thanks for your help [05:59] <lidenbrock> see you [06:02] <relaxed> Tthread: who is "u" and what is vda? [06:09] <Tthread> u means anyone :) i tell about video decoder acceleration [06:10] <relaxed> Cool, english is not you native tongue. What are you trying to do? [06:12] <relaxed> If you have a recent nvidia card, mplayer provides vdpau hardware decoding. [06:12] <relaxed> I recommend mplayer2 if that's the case. [06:12] <Tthread> I want to know is it more efficient to use vda for live video decoding. in my case videostream is about 6 mbits. and i want to decode it faster than now. [06:13] <Tthread> i use avcodec_decode_video2 function for decode for each frame [06:14] <Tthread> as i get it i should to add vds context in AvCodec context. am i right? [06:16] <relaxed> Ah, you're using the libs for your own software. I would ask in #libav [06:16] <relaxed> They're usually more helpful than here. [06:16] <variable> anyone know how to convert a .ppm file to a .mp4 file? [image2 @ 0x804c2a820] Could not find codec parameters (Video: ppm) from ffmpeg [06:17] <relaxed> even though it's the fork [06:17] <Tthread> oh, thanks a lot. you are right about tongue ) [06:17] <relaxed> variable: ppm is an image file. what is your command? [06:18] <variable> ffmpeg -i foo.ppm -vcodec libx264 -preset ultrafast -crf 15 -threads 0 -x264opts frameref=15:fast_pskip=0 foo.mp4 [06:18] <relaxed> so one frame? [06:18] <variable> honestly, ECLUE [06:18] <variable> a program called 'gource' generated a 30 GB file [06:19] <variable> with the extension .ppm [06:19] <relaxed> I believe ppm is a raw image file [06:19] <relaxed> you have one file that 30gb? [06:19] <variable> yeah [06:19] <relaxed> that's* [06:19] <relaxed> lol [06:20] <relaxed> pastebin the output of "ffmpeg -i file.ppm" [06:20] <variable> hang on [06:21] <relaxed> If it is an image it must be huge like xbox [06:22] <relaxed> DEV D ppm PPM (Portable PixelMap) image [06:22] <variable> relaxed: sample file, about 2.0 gb http://paste.pocoo.org/show/578795/ [06:23] <relaxed> clang? are you a bsd or OS X guy? [06:23] <variable> BSD [06:23] <relaxed> which? [06:23] <variable> FreeBSD [06:24] <relaxed> good for you :) [06:24] <variable> :) [06:24] <relaxed> You may want to try ffmpeg-devel [06:24] <relaxed> in ports [06:24] <relaxed> did you already try that? [06:24] <variable> no [06:25] <variable> I'll switch to -devel and try that [06:25] <variable> relaxed: I need to go to sleep nowish, but I'll report back tomorrow. Thanks! [06:25] <relaxed> you're welcome [06:49] <_genuser_> hello people. I have a question about -itsoffset. what is the offset added to? the audio or the video? [06:49] <_genuser_> I'm using -i video.mp4 -vcodec copy -i audio.mp4 -acodec copy -itsoffset 00:00:20.000 target.mp4 [06:50] <_genuser_> and it's not seeming to do anything whereas it has worked fine with different offset on other files. [07:28] <relaxed> _genuser_: I believe the inputs have to be demuxed. [07:29] <relaxed> example, ffmpeg -i blah.264 -itsoffset 00:00:20.000 -i bleh.aac .... [07:29] <relaxed> _genuser_: in that example it would affect bleh.aac [07:30] <relaxed> you dig? [07:31] <relaxed> you may be able to refernce the same input twice and use -map to do it as well. [07:31] <relaxed> reference* [11:59] <kriegerod> is there some audio frame allocation routine in API, similar to av_image_alloc()? [13:05] <silverrocker> I have an input file suzie.avi and I want to convert it to h264 but I want to use Intra PCM mode [13:05] <silverrocker> what parameters do I have to pass to ffmpeg [13:30] <jessy> hey there, i'm recording audio from my mic with ffmpeg.exe -f dshow -i audio="default input" -y \"test.mp3", how can i set sample rate frequency etc. [13:31] <ubitux> -ar? [13:34] <jessy> ubitux: is that for me ? :) i really don't have much idea about ffmpeg :S [13:35] <ubitux> -ar rate set audio sampling rate (in Hz) [13:35] <ubitux> so yes it's for you :p [13:42] <jessy> ubitux: thanks :) let me try [16:49] <JamesJJ> Hi. I'm trying to convert an flv file, which has a G711 audio track. FFmpeg lists the audio codec as 0x0008, is there a way to force the input codec? [16:55] <sacarasc> JamesJJ: -codec:a blah as an input option. [17:01] <JamesJJ> hmmm, just get unrecognized option '-codec:a' [17:17] <Diogo> hi [17:17] <Diogo> this command is correct to insert aac file in a video file? [17:18] <Diogo> sudo /servers/ffmpeg/bin/ffmpeg -i /files/HD.mp4 -i /files/music.aac -vcodec copy -acodec copy -absf aac_adtstoasc /files/VIDEO_COM_SOM.mp4 [18:03] <anoob> I didn't find nothing to GET video title or video description (metadata) ... just duration (ffmpeg -i filename.video) ... [20:06] <_r00t_> g'day [21:50] <javatexan> hey guys, can one do screen capture with ffmpeg on os x lion? I found some conflicting info online from varying times, so I thought I would ask [22:07] <cbreak> javatexan: Quicktime can do it [22:07] <cbreak> javatexan: I don't think ffmpeg has the capabilities to do any kind of screen capturing [22:09] <tdr> ffmpeg can under x11 (linux/*bsd*), not sure about under osx [22:14] <cbreak> OS X doesn't use X11 as primary output [22:15] <cbreak> it comes with an X11 server that can host individual windows though [22:15] <cbreak> maybe those could be recorded. [22:27] <javatexan> k thanks [22:47] <Mista_D> Any way to limit ffmpeg's memory usage on 64bit CENTOS? [23:31] <meekohi> If my libx264 is compiled with lavf enabled, how can I force it to not be used with ffmpeg command line args? [23:35] <silverrocker> I'm building like this: "g++ -D__STDC_CONSTANT_MACROS -lm -lavdevice -lavformat -lavfilter -lavcodec -lswscale -lavutil main.cpp" and I'm only using av_init_packet to get started. I'm including libavcodec/avcoder.h and still I have undefined reference to `av_init_packet(AVPacket*) . Any clues? [23:37] <silverrocker> If I build in c, I have no dependency errors, but when I build in c++ it gives me that linker error [23:39] <Mavrik> did you put "extern C" guards around imports? [23:48] <silverrocker> Mavrik: yes I already found that that was the problem [23:48] <silverrocker> why aren't the headers C++ safe? [23:51] <coydog> Hi folks, I think I've found a bug in ffmpeg. I'm about to try to trace it locally, but I just wanted to check in and see if it's a known issue. I couldn't find it in the ffmpeg bugzilla. [23:52] <JEEB> I think ffmpeg uses trac currently [23:52] <JEEB> https://ffmpeg.org/trac/ffmpeg/report [23:53] <coydog> If I call ffmpeg repeatedly with the same command-line options, overwriting the same output file, subsequent runs appear to duplicate some frames in the first second or so of output, which seems to make us lose audio synch. It can be worked around by always creating a new file, instead of overwriting the same file [23:53] <coydog> Has anyone else experienced this? Is it a known issue? [00:00] --- Wed Apr 11 2012

1 0