Ffmpeg-devel-irc

Download

ffmpeg-devel-irc@ffmpeg.org

December 2015

1 participants
62 discussions

ffmpeg-devel.log.20151230
by burek 31 Dec '15

31 Dec '15

[02:27:12 CET] <kierank> so I need to output a frame that's already been output (essentially a skip frame) [02:27:22 CET] <kierank> I can do that but just storing the avframe right? [03:27:26 CET] <J_Darnley> Hm. So close. Just some incorrect chroma to go. [03:39:01 CET] <atomnuker> I never thought people with so much paranoia about integer overflow existed [03:41:00 CET] <J_Darnley> Where did I see that the other day? (Some shitpost on HN no doubt.) Undefined behaviour should start nethack. [03:41:41 CET] <atomnuker> "It's undefined behavior, this is insane, anything can happen!" [03:42:00 CET] <J_Darnley> Ah the daala thread? I've been keeping it collapsed in Thunderbird. [03:49:22 CET] <BBB> atomnuker: sorry about being so persistent in that thread [03:49:32 CET] <BBB> I just really dont want us to fall in the trap of appeasing ubsan [03:49:36 CET] <BBB> its not useful [03:51:19 CET] <jamrial_> atomnuker: leaving the whole thing about overflow with invalid input on dsp code not being important aside, the decoder should definitely not crash in a normal run with invalid/fuzzed input [03:51:30 CET] <jamrial_> no idea if Andreas' crashes were all ubsan related, though [03:54:35 CET] <atomnuker> BBB: I agree with you, please keep being persistant [03:55:19 CET] <atomnuker> jamrial_: It's an RFC and the first time the decoder's been sent to the ML, let alone fuzzed [03:58:31 CET] <jamrial_> i know, just pointing it out [04:15:28 CET] <BBB> crashes should be fixed, and Im not arguing against fixing overflows in non-dsp code [04:15:59 CET] <BBB> Im just saying that dsp code overflows are probably ok, its quite likely we clip at the end of the dsp code anyway (simd code always clips when packing) [04:16:17 CET] <BBB> so although overflows occur, the output range is still defined [04:16:34 CET] <BBB> atomnuker: is it OK if I leave a full review for over the weekend? [04:16:47 CET] <BBB> (you can commit before that if you wish, but I dont think Ill have time during the week before that) [04:18:41 CET] <atomnuker> BBB: Yeah, sure, I'll have time to fix whatever I can by then and send off a v3 [09:14:59 CET] <fritsch> nevcairiel: https://github.com/mpc-hc/FFmpeg/commit/ecc953ccbc2370c8177deb4077fde1007ec… <- mind PRing that to ffmpeg master? [09:16:19 CET] <fritsch> nevcairiel: and the corresponding dxva2_hwaccel, too, please? [09:55:59 CET] <nevcairiel> there are reasons that hasn't happened yet, just unlocking 10-bit hevc dxva2 will break stuff [10:16:47 CET] <xintox> why do i have to pass \r\n manually to headers? that seems like ffmpeg should just insert those [10:25:10 CET] <fritsch> nevcairiel: can you be more concrete? does that mean the branch I linked ships broken stuff? [10:26:08 CET] <nevcairiel> that branch is generally designed to work in exactly one use-case [10:26:32 CET] <fritsch> okay - any hints to what it actually breaks? [13:06:39 CET] <J_Darnley> Is there a black and white or greyscale vsynth test in fate? [13:39:31 CET] <J_Darnley> nevermind, I'll just use testsrc and lut [13:50:53 CET] <durandal_1707> ubitux: is overlap patch for showspectrum ok? [14:02:37 CET] <J_Darnley> Can anyone point out the error in my inderstanding of the assembly code here http://pastebin.com/uUqWwSw6 [14:03:42 CET] Action: J_Darnley facepalms [14:03:50 CET] <J_Darnley> (why do I see it only now?) [14:04:25 CET] <J_Darnley> the high dqword starts two bytes further in the broken version [14:04:37 CET] <BBB> J_Darnley: try pshufb :D [14:04:59 CET] <J_Darnley> that doesn't cross lanes [14:05:09 CET] <BBB> owh crap [14:05:16 CET] <J_Darnley> indeed [14:05:22 CET] <nevcairiel> evil lane business is evil [14:05:27 CET] <J_Darnley> I think I grok them now. [14:06:51 CET] <J_Darnley> Something like conjoined independant registers [14:07:16 CET] <BBB> sorry boss, my video encoder can only encode videos up until 352x288 resolution [14:07:38 CET] <nevcairiel> 20 years ago you might have won something with that [14:07:43 CET] <BBB> oh, you want bigger video, HD you say? hm & sounds like PR bullshit to me, why dont you cut it up in many small videos and feed it to my encoder independently [14:07:55 CET] <BBB> < $intel [14:08:14 CET] <J_Darnley> Didn't ffmpeg+x264 do a BBB presentation like that at some conference? [14:08:35 CET] <J_Darnley> (that would be big buck bunny) [14:08:58 CET] <nevcairiel> now on to philosophy, did the big bunny or the BBB exist first [14:09:13 CET] <BBB> BBB stems from the 90s, I think [14:09:17 CET] <BBB> so probably me [14:09:25 CET] <BBB> (BBB as in my nickname) [14:13:38 CET] <J_Darnley> kierank: what profiling tools are available on avdev? [14:15:25 CET] <J_Darnley> ah gprof's there [14:32:59 CET] <ubitux> durandal_1707: why not use AVRational? [14:33:05 CET] <ubitux> durandal_1707: how does it affect output? [14:58:53 CET] <durandal_1707> ubitux: it gives higher framerate and more resolution [14:59:22 CET] <ubitux> can you show 2 screenshots at the same t? [15:00:18 CET] <durandal_1707> with and without it? [15:00:54 CET] <durandal_1707> overlap 0 default, gives same output [15:02:04 CET] <durandal_1707> 0.5 enlarges width by 100% [16:47:19 CET] <kierank> J_Darnley: should be able to start-timer [16:53:26 CET] Action: J_Darnley goes to look that up [16:54:13 CET] <J_Darnley> Do you mean ffmpeg's timer.h tool? [17:09:23 CET] <kierank> yes [17:29:33 CET] <Daemon404> kierank, do you still need cfhd [17:29:38 CET] <kierank> yes [17:29:42 CET] <kierank> I pinged koda dabout it [17:29:44 CET] <kierank> but he is not alive [17:30:05 CET] <Daemon404> food coma probably [17:30:32 CET] <Daemon404> let me see if we have any in the 'shitty files' archive [17:36:57 CET] <Daemon404> cant find one, might be blacklisted during ul [17:47:49 CET] <cone-234> ffmpeg 03Ganesh Ajjanagadde 07master:29bc293297a6: configure: make gcc 2 detection more robust [17:47:50 CET] <cone-234> ffmpeg 03Ganesh Ajjanagadde 07master:b272c3a5aabe: lavc/dsd_tablegen: speed up table generation [17:54:36 CET] <cone-234> ffmpeg 03Ganesh Ajjanagadde 07master:05434b0eea3f: lavc/cook: get rid of wasteful pow in init_pow2table [17:57:06 CET] <kierank> I wonder if we can convince ganesh to make thread safe table inits [17:57:14 CET] <kierank> he seems interested in working on tables [18:04:23 CET] <Daemon404> doubtful kierank [18:04:33 CET] <Daemon404> he seems mostly interested in 'interesting' 'optimizations' [18:04:39 CET] <Daemon404> table init is a slog [18:05:22 CET] <Daemon404> i started going through and marking some files here, but i got bored: https://github.com/dwbuiten/FFmpeg/commit/3d2fc03955348f49f2a6b67a66cc038d1… [18:05:48 CET] <Daemon404> i dont know how elenril and co do it [18:54:02 CET] <cone-234> ffmpeg 03Paul B Mahol 07master:2893dbfe2c31: avfilter/avf_showspectrum: add moreland color map [19:08:28 CET] <durandal_170> ubitux: 0 overlap: http://i67.tinypic.com/2q3vgx4.png [19:10:02 CET] <durandal_170> ubitux: 0.5 overlap: http://i65.tinypic.com/e1fbqa.png [20:17:52 CET] <llogan> kierank: there was an actual award to collect? [20:18:24 CET] <kierank> llogan: probably [20:18:45 CET] <durandal_1707> for what? [20:20:11 CET] <llogan> https://twitter.com/obencoder/status/682048485596737536 [20:20:15 CET] <cone-234> ffmpeg 03Michael Niedermayer 07master:389ce22b5f09: avformat/movenc: Use strict_std_compliance from the muxer layer instead of the encoder layer [20:22:14 CET] <llogan> kierank: is streamingmedia legit? [20:22:22 CET] <kierank> yes [20:22:32 CET] Action: llogan is ignorant as usual [20:23:00 CET] <kierank> well tbh they are legit enough to give us an award [20:23:07 CET] <kierank> usually these things you pay money and get an award [20:24:22 CET] <JEEB> wow [20:24:29 CET] <llogan> so, who is in California? [20:26:17 CET] <llogan> "Dear, please send monies to Dr. Elvis Guru, esteemed representative of FFmepg". [20:27:41 CET] <llogan> looks like i have to clean up after roger again [21:00:14 CET] <llogan> durandal_1707: ahistogram docs: "scroll from to bottom." looks like forgotten "top"? [21:05:45 CET] <cone-234> ffmpeg 03Vicente Olivert Riera 07master:6a7e94cb14f3: mips: add mips32r6 architecture variant [21:05:46 CET] <cone-234> ffmpeg 03Vicente Olivert Riera 07master:dde994751b16: mips: put "disable mipsfpu" in a better place for loongson [21:19:08 CET] <J_Darnley> kierank: I've got a pair of avx2 commits you can have a look at here https://gitlab.com/J_Darnley/ffmpeg/commits/v210_asm [21:19:18 CET] <J_Darnley> I'll be honest: they're pretty pathetic in terms of speedup. [21:19:58 CET] <kierank> probably still worth it [21:20:12 CET] <kierank> Might need to give you access to one of those intel boxes that downclock with avx2 [21:20:34 CET] <kierank> because the avx2 might completely negate itself [21:20:51 CET] <J_Darnley> What!? [21:21:24 CET] <J_Darnley> lol do they need an avx2slow flag? [21:21:53 CET] <jamrial> no [21:22:42 CET] <nevcairiel> "one of"? dont all do that? [21:22:48 CET] <nevcairiel> well all that have a turbo mode [21:22:50 CET] <jamrial> J_Darnley: use vpermq [21:23:09 CET] <jamrial> instead of movq/movhps + vinserti128 [21:23:48 CET] <kierank> nevcairiel: only some SKUs [21:23:53 CET] <jamrial> wait, nevermind. the mem operands you need don't allow that [21:24:05 CET] Action: llogan remembers the "turbo" button on old 486 [21:24:32 CET] <jamrial> kierank: all haswell models? [21:24:42 CET] <kierank> no afaik [21:25:05 CET] <kierank> I think e5- range only [21:25:19 CET] <kierank> it was in the x265 presentation at vdd [21:25:20 CET] <nevcairiel> and the desktops? [21:26:05 CET] <J_Darnley> jamrial: yes. I had vpermq before and struggled to see the source of my problems. Which was the +6 (basically). [21:26:15 CET] <jamrial> J_Darnley: you can however use vextracti128 with a mem operand as dest [21:26:29 CET] <J_Darnley> unaligned too? [21:26:36 CET] <jamrial> yeah [21:27:10 CET] <jamrial> every avx+ instruction except those that specify alignment (like movdqa and movaps) work with unaligned memory [21:27:32 CET] <J_Darnley> What was I looking at earlier? [21:27:47 CET] <jamrial> what do you mean? [21:28:23 CET] <J_Darnley> Oh yes. The Intel manual's large table of exception types. [21:28:46 CET] <kierank> J_Darnley: feel free to submit to the ml [21:28:58 CET] <kierank> maybe as an rfc [21:29:01 CET] <cone-234> ffmpeg 03Paul B Mahol 07master:0f387f2b6265: avfilter/avf_showspectrum: add nebulae color map [21:35:34 CET] <J_Darnley> OMFG! That makes it look much better! [22:20:16 CET] <durandal_1707> anybody likes any specific palette? [22:29:54 CET] <J_Darnley> durandal_1707: have you got a "heat map" one? [22:30:53 CET] <durandal_1707> shouldn't that be intensity, already available? [22:31:19 CET] <J_Darnley> dark blue through red onto yellow? [22:37:11 CET] <J_Darnley> Drat. I need to wait for the next nightly build. [22:40:02 CET] <J_Darnley> Anyway. Here's what I meant by heat map (an excerpt from some old code of mine) http://pastebin.com/xNA9cXWV [22:41:02 CET] <J_Darnley> Input 0..1 and it would return a nice thermal camera like image [22:41:12 CET] <J_Darnley> well colours for one [22:41:42 CET] <J_Darnley> ha ha. I do still have a rendered image from it [22:42:59 CET] <J_Darnley> http://users.telenet.be/darnley/wp-keyboard-heat.png [22:43:58 CET] <wm4> this image proves that space bar heating is overrated, and one should use the shift keys instead? [22:46:54 CET] <TD-Linux> space bar heating? [22:49:06 CET] <wm4> just an unfunny joke https://xkcd.com/1172/ [23:01:02 CET] <durandal_170> intensity option is same as heat [23:28:53 CET] <cone-234> ffmpeg 03yukari yakumo 07master:f87e29a61587: configure: support --cpu=host and Clang [00:00:00 CET] --- Thu Dec 31 2015

1 0

ffmpeg.log.20151230
by burek 31 Dec '15

31 Dec '15

[00:09:17 CET] <klaxa> nani____: so you want one output video with the audio from the first file and the video from the second file? [00:09:31 CET] <nani____> klaxa, exactly [00:09:56 CET] <klaxa> in that case: ffmpeg -i video0.mkv -i video1.mkv -c copy -map 0:a -map 1:v output.mkv [00:09:59 CET] <nani____> and the two might get out of sync by different amounts at different times [00:10:10 CET] <klaxa> well you can't fix that without manual work [00:10:17 CET] <nani____> I see [00:10:26 CET] <nani____> I'll try to hack something together [00:21:18 CET] <Disturbed1> i'm so tired.... five days and still can't get libnvenc to work.... [00:25:42 CET] <Disturbed1> In file included from /tmp/ffconf.0u565iRP.c:1:0: [00:25:42 CET] <Disturbed1> #include <uapi/linux/cuda.h> [00:25:43 CET] <Disturbed1> ^ [00:25:43 CET] <Disturbed1> compilation terminated. [00:25:43 CET] <Disturbed1> ERROR: cuda.h header not found [00:26:22 CET] <Disturbed1> but yet the file is in both /usr/include & /usr/local/include [00:27:21 CET] <c_14> Can you pastebin your whole config.log? [00:28:15 CET] <c_14> The path it's listing there isn't in your system include directory, it's part of the kernel source tree (which might be a part of an eventual -headers package on your system) [00:28:20 CET] <c_14> Not sure why it's looking for it though [00:29:22 CET] <Disturbed1> gimme a second [00:32:47 CET] <Disturbed1> http://pastebin.com/AZvHMkjf [00:32:53 CET] <Disturbed1> it's not the whole thing [00:34:17 CET] <c_14> What's your configure line? [00:34:55 CET] <Disturbed1> ../ffmpeg_libnvenc/configure --enable-nonfree --enable-gpl --enable-version3 --enable-libass --enable-libmp3lame --enable-libopencv --enable-libopenjpeg --enable-libopus --enable-libtheora --enable-libvpx --enable-libwebp --enable-opencl --enable-x11grab --enable-opengl --enable-openssl --cpu=native --enable-libnvenc [00:38:03 CET] <c_14> Where did you get these ffmpeg sources from? [00:38:13 CET] <c_14> There is no libnvenc configuration flag [00:39:11 CET] <Disturbed1> XXX://github.com/Brainiarc7/ffmpeg_libnvenc [00:39:55 CET] <Disturbed1> i've been tring so many different pdf's, guides, etc i'm loosing it lol [00:40:17 CET] <c_14> That's some random person's fork [00:40:22 CET] <c_14> That's not supported here [00:40:29 CET] <c_14> ffmpeg has nvenc support in master though [00:40:31 CET] <c_14> Just use that? [00:41:11 CET] <Disturbed1> ka gimme a sec to clear all my folders and get a fresh copy [00:54:07 CET] <Disturbed1> thank you c-14 !!! [00:54:46 CET] <Disturbed1> ~/development/ffmpeg_build$ ../ffmpeg/configure --enable-nonfree --enable-gpl --enable-nvenccut everything [00:55:10 CET] <Disturbed1> cut everything but three and it worked... [02:00:51 CET] <smolleyes> hi [02:01:50 CET] <waressearcher2> smolleyes: hallo, wie geht's ? [02:01:51 CET] <smolleyes> is it possible to pipe ffplay decoded frames to stdout like ffmpeg ? or would be hard to create a ffplay with nacl and sdl to create a browser plugin :p ? [02:02:28 CET] <smolleyes> wie geht's :) ? what does it mean :p [02:02:37 CET] <DHE> ffplay is meant to output video. if you want to output frames, output to a raw format/codec like yuv or rgb [02:02:45 CET] <The_Photographer> Hi guys [02:02:52 CET] <The_Photographer> How I could solve this problem [02:02:52 CET] <The_Photographer> "Unknown encoder 'libtheora'" [02:03:34 CET] <smolleyes> humm not clear for me DHE [02:04:07 CET] <c_14> The_Photographer: does your build support theora? [02:04:16 CET] <The_Photographer> c_14: its a binary package [02:04:49 CET] <c_14> check ffmpeg -encoders [02:06:44 CET] <The_Photographer> c_14: http://pastebin.com/piL7H1hT [02:07:32 CET] <c_14> Your build wasn't build against libtheora, try with a static build from [02:07:34 CET] <c_14> http://johnvansickle.com/ffmpeg/ [02:09:06 CET] <The_Photographer> c_14: the optinos for build? [02:09:34 CET] <c_14> --enable-libtheora [02:10:03 CET] <The_Photographer> after to build I need copy it to sbin? [02:10:33 CET] <c_14> any directory in your path _before_ the directory where your current copy is (or overwrite/delete the current one) [02:10:34 CET] <The_Photographer> c_14: I am installing right know the dependences [02:12:41 CET] <The_Photographer> c_14: ./configure --prefix="$HOME/ffmpeg_build" --enable-libtheora --disable-shared ?? [02:13:45 CET] <c_14> sure [02:16:30 CET] <The_Photographer> c_14: anyway john vasickle version work [02:34:12 CET] <The_Photographer> thanks c_14 [02:34:21 CET] <The_Photographer> c_14: work perfect [02:34:28 CET] <The_Photographer> c_14: I love ogv format [02:34:38 CET] <The_Photographer> ogv h265 [02:43:52 CET] <derekr> Is there a way to do gapless audio encoding? [02:44:43 CET] <kepstin> derekr: what codec? [02:45:00 CET] <derekr> kepstin: aac [02:45:13 CET] <kepstin> the problem with "gapless" was mostly for mp3, and iirc ffmpeg will include the xing headers for gapless playback by default [02:45:17 CET] <kepstin> dunno about aac :/ [02:46:01 CET] <derekr> I'm just trying to figure out a way around the insertion of 'silence' at the beginning and end of my transcoded segments [02:46:40 CET] <derekr> I'm starting to get the impression that this likely won't be possible on the encoding side [02:47:20 CET] <derekr> but the client side fix seems quite involved and cumbersome [02:48:33 CET] <kepstin> how does itunes do it? i think they add some metadata to the file, but I dunno what, and I dunno if other players support it [02:48:54 CET] <derekr> it seems to be largely metadata based [02:49:16 CET] <derekr> the clients parse the metadata and chop off what needs to be chopped off [03:21:17 CET] <pinPoint> so my question still stands. If a video clip has surround sound, would ffmpeg move that over to the output if an audio codec is changed? [03:21:30 CET] <pinPoint> from ac3->aac or vise versa [03:22:20 CET] <pzich> can't you just try it and see? [03:22:40 CET] <pzich> you may need to specify -ac 6 to keep the number of channels, not sure [03:22:43 CET] <pinPoint> i've got just a 2.1 [03:22:52 CET] <pinPoint> a soundbar. :/ [03:24:29 CET] <pzich> you could try making one out of mono streams: https://trac.ffmpeg.org/wiki/AudioChannelManipulation [03:24:59 CET] <pinPoint> yeah I read that page... just wasn't sure without actual hardware [04:36:27 CET] <pinPoint> soo I moved a video clip that has ac3 5.1 and ffmpeg automagically added it into aac 5.1 [04:36:49 CET] <pinPoint> it was 5.1 without me touching any thing. [04:37:16 CET] <c_14> If that didn't happen, it'd be broken. [04:37:32 CET] <pinPoint> that was the question I was asking earlier and another day [04:39:25 CET] <pinPoint> so ffmpeg does it automagically... just had to fire up a test experiment. [04:39:38 CET] <pinPoint> mediainfo and vlc confirmed the audio for me [04:41:14 CET] <pinPoint> c_14: aac 256+ can be almost lossless yeah? compared to flac? [04:41:47 CET] <kepstin> aac will never be lossless by definition [04:42:02 CET] <pinPoint> by ear yeah? [04:42:22 CET] <kepstin> a good aac encoder should be "transparent" - i.e. difficult/impossible to tell from the original - well below 256kbit [04:42:31 CET] <pinPoint> flat out aac codec [04:43:00 CET] <kepstin> a bad aac encoder (i.e. vo-aacenc) will still sound bad at 256kbit ;) [04:43:09 CET] <pinPoint> kepstin: would you say that you would save some space vs flac by ear? [04:43:24 CET] <pinPoint> i only use AAC codec only. the rest are useless to me [04:43:45 CET] <furq> why not just leave it as ac3 [04:43:50 CET] <kepstin> if you're keeping flac, it's because you have lots of spare disk space and don't want to have to worry about generation loss if you have to re-encode for some reason. [04:44:00 CET] <pinPoint> furq: that was just a test. [04:44:11 CET] <pinPoint> aac vs flac is another test for audio [04:44:23 CET] <pinPoint> ac3 is pro logic no? [04:45:23 CET] <furq> it can be [04:45:32 CET] <kepstin> any stereo codec can hold pro logic encoded audio [04:45:43 CET] <kepstin> ac3 can also do proper multi-channel instead [04:45:59 CET] <pinPoint> ok I didn't know that just unsure with audio codecs [04:46:57 CET] <furq> i don't generally think it's worth transcoding ac3 [04:47:45 CET] <pinPoint> i was just testing surround sound output. [04:47:55 CET] <pinPoint> i'm trying to write another tutorial with ffmpeg [04:47:59 CET] <furq> fair enough [04:50:01 CET] <pinPoint> you read my other one about libx265 vs libx264? [04:50:24 CET] <pinPoint> i did a comparison of encoding times @ 2M bitrate for a 10.34 clip. [04:51:08 CET] <furq> i did see that [04:51:12 CET] <furq> you might recall i had some misgivings about it [04:51:18 CET] <pinPoint> ya [04:51:47 CET] <pinPoint> one of the x265 took like 3days to complete on veryslow. :/ [04:52:41 CET] <furq> well at least it's honest [04:53:19 CET] <pinPoint> heh [05:04:43 CET] <pinPoint> furq: looks like ffmpeg also trys to match existing bitrate if you use -c:v copy or comes close [05:04:57 CET] <pinPoint> duh, it should since its a copy. nvm [05:31:12 CET] <xintox> what is yasm? i'm btrying to build ffmpeg on a mac and it says yasm not found [05:31:34 CET] <furq> an assembler [05:31:46 CET] <furq> you can build without it but you'll lose performance [05:32:23 CET] <xintox> so how do i get it for mac and debian? [05:32:47 CET] <furq> for debian it's in apt [05:43:06 CET] <xintox> ok [05:43:11 CET] <xintox> i did brew install yasm and it worked [08:31:34 CET] <MajesticFudgie> Can someone explain how to put timestamps in the filename [08:31:37 CET] <MajesticFudgie> Nothing I try works [08:31:45 CET] <MajesticFudgie> It just ends up as %H.avi [10:03:51 CET] <xintox> MajesticFudgie: hey [10:04:25 CET] <MajesticFudgie> hi [10:04:29 CET] <xintox> try video%03d.avi [10:04:35 CET] <xintox> oh [10:04:40 CET] <xintox> you wanted timestamps. not numbers [10:04:45 CET] <xintox> there's a format it supports [10:04:48 CET] <xintox> what have you tried? [10:08:32 CET] <MajesticFudgie> I managed to get a SH script to do it [10:08:39 CET] <MajesticFudgie> now I have the issue that videos appear to be double speed [10:09:30 CET] <MajesticFudgie> I've been trying to stream mjpeg url to disk and also to an RTMP stream [10:09:32 CET] <MajesticFudgie> Pain in the ass [10:17:25 CET] <xintox> i can do rtmp [10:17:31 CET] <xintox> i haven't done mjpeg though [10:17:42 CET] <xintox> oh. actually nevermind. [10:17:49 CET] <xintox> i'm can do rtmp -> hls [10:17:54 CET] <xintox> i do hls for everythig. [10:18:28 CET] <xintox> MajesticFudgie: ask on the mailing list [10:18:32 CET] <xintox> the are pretty helpful [10:18:51 CET] <MajesticFudgie> yeah [10:18:55 CET] <xintox> what are you streaming? [10:18:58 CET] <MajesticFudgie> CCTV [10:19:05 CET] <MajesticFudgie> I have it all working atm [10:19:13 CET] <xintox> oh nice [10:19:13 CET] <MajesticFudgie> But the stream stops and starts typa thing [10:19:17 CET] <MajesticFudgie> It'll stop for a minute or two [10:19:17 CET] <xintox> how you get that stream? [10:19:25 CET] <MajesticFudgie> then play and suddenly its double speed [10:19:26 CET] <xintox> hmmm. do you have a url i can test? [10:19:43 CET] <MajesticFudgie> http://stream.infermc.com/watch/CCTV [10:19:59 CET] <MajesticFudgie> It'll take ages to play [10:20:18 CET] <MajesticFudgie> You see its frozen then suddenly if a cars going past it'll go past at double speed [10:20:33 CET] <MajesticFudgie> As if ffmpeg is hanging until it has enough data to stream [10:20:50 CET] <xintox> MajesticFudgie: is that your stream? [10:20:55 CET] <MajesticFudgie> yes [10:21:02 CET] <xintox> oh. i meant what's hte source stream [10:21:08 CET] <xintox> i can try to reproduce it [10:21:12 CET] <MajesticFudgie> Sec [10:22:14 CET] <MajesticFudgie> I've noticed it to you, [10:22:32 CET] <MajesticFudgie> You'll see the FPS isn't too bad [10:22:50 CET] <MajesticFudgie> I think at some point I'm just going to gut the CCTV housing and put an RPi with a USB Webcam in [10:23:00 CET] <MajesticFudgie> Get better quality and higher FPS [10:23:12 CET] <MajesticFudgie> I'm having to do all this streaming and such on my file server [10:28:02 CET] <xintox> what is CCTV? [10:28:31 CET] <xintox> MajesticFudgie: i dind't get a pm from you [10:28:58 CET] <MajesticFudgie> Did you now? [10:29:03 CET] <xintox> yes [12:35:26 CET] <Blaxxun> Hello there! I wanted to know the reason on why the ffserver upper limit for MaxHttpConnection is 65335? [12:38:11 CET] <Blaxxun> 65535 that is [12:38:55 CET] <JodaZ> because it fits in a short xD [12:39:16 CET] <Blaxxun> yes :D but why just a short? [12:39:39 CET] <JodaZ> why care, you don't have hardware to push out to more than that anyways [12:40:40 CET] <Blaxxun> what would be wrong with the hardware? [12:43:21 CET] <JodaZ> even if you wanted to stream at just 32kbit/s 65335 users would be 2gbit/s and you can't do that with commodity hardware [12:47:28 CET] <Blaxxun> I see [12:47:51 CET] <Blaxxun> how does the ISP run iptv? they have the bandwidth? [12:48:15 CET] <Blaxxun> and do they use ffserver for iptv? (or is it a bad idea) [12:49:52 CET] <JodaZ> Blaxxun, load balancing to multiple servers or proper multicasting [12:50:18 CET] <JodaZ> and they propably don't use ffserver [12:51:14 CET] <Blaxxun> what do you think they use? [12:51:30 CET] <JodaZ> no idea [12:51:50 CET] <Blaxxun> thanks for help anyway, it's appreciated [14:55:24 CET] <Li> what is the alternative of -ss to delete specific period from end of the movie for multiple video files [14:55:28 CET] <Li> ? [14:59:43 CET] <durandal_1707> trim filter? [15:05:42 CET] <Li> durandal_1707: both of them can cut the end? [15:11:58 CET] <durandal_1707> yes [15:12:37 CET] <relaxed> -to can, or -t [15:36:57 CET] <Li> Both can not do what I need [15:37:57 CET] <Li> I've folder full of video tutorials that I need to delete last 15 seconds from all of them, but these files are not all equal in size/length [15:39:01 CET] <Li> So I need to do some cutting at the end which is behaving just like -ss 00:00:15 on the begining [15:47:47 CET] <relaxed> Li: script using ffprobe to get each duration in seconds and subtract [15:51:01 CET] <drg> salute [15:52:27 CET] <waressearcher2> drg: hallo [15:52:46 CET] <waressearcher2> drg: wie geht's es dir ? [15:53:10 CET] <drg> sorry, nope german... eng only [15:54:26 CET] <waressearcher2> so süß [15:54:57 CET] <drg> anyway... [15:55:09 CET] <drg> is there anyway to get hash codes for windows binaries? [16:09:36 CET] <c_14> drg: ask whoever provides the build to also provide the hashs [16:09:40 CET] <c_14> *hashes [16:33:05 CET] <drg> c_14: contacts page in ffmpeg.org redirects to this channel... :) [16:33:24 CET] <c_14> FFmpeg itself doesn't provide any builds for any platform [17:09:20 CET] <Franciman> Hello [17:09:46 CET] <Franciman> what is AV_SAMPLE_FMT_NB? [17:11:30 CET] <Franciman> uh ok understood [17:11:54 CET] <Franciman> it's the number of sample formats right? [17:12:03 CET] <Franciman> the number of sample formats in the enum* [18:06:58 CET] <kbs> How can I concatenate a set of mp4 files, but interleave them with a title between each join? (e.g. like the file name) [19:23:44 CET] <jonkri> Hi! It seems to me that when I'm using -vframes 1 and -ss to produce a video snapshot, a snapshot is not produced if the video is shorter than the value I provided to -ss. Is there some way to take a snapshot at a certain time in the video, and if the video is shorter than that time, at the last frame? Or perhaps take a snapshot 10% into the video or something? I'm using avconv in Ubuntu 14.04 LTS. [19:24:30 CET] <c_14> avconv isn't supported here. ask in #libav or use ffmpeg from FFmpeg (but no, not really. at least without scripting) [19:26:31 CET] <jonkri> c_14: Oh, OK, I understand. I'm somewhat constrained to using whatever is into the Ubuntu 14.04 repositories, unfortunately. [19:26:59 CET] <JEEB> be it Libav or FFmpeg you use, by now what's in 14.04 is pretty old :) [19:27:15 CET] <JEEB> so just keep that in mind [19:27:27 CET] <jonkri> JEEB: I understand that too. :) Anyway, my use case is pretty simple, or so I thought, at least. :> [19:28:33 CET] <jonkri> Does the FFmpeg project have any official information about the libav stuff, in case I want to read up about it? [19:29:12 CET] <jonkri> And by "stuff" I mean whatever happened to make Ubuntu ship libav instead. [19:29:15 CET] <JEEB> both sides have rather coloured write-ups about it, so let's just say that there's two separate forks from the FFmpeg of the olden times (up to end of 2010) [19:29:43 CET] <jonkri> And "libav" being forked. :> [19:29:47 CET] <jonkri> OK, I see. Thanks! [19:29:50 CET] <furq> the important thing is that debian and ubuntu both ship ffmpeg again now [19:30:05 CET] <jonkri> That's good. [19:30:11 CET] <furq> unfortunately not retroactively [19:30:15 CET] <JEEB> yeah, I don't agree with the discussions had but I'm fine with them picking FFmpeg :P [19:30:39 CET] <JEEB> there were some quite derpy arguments had IIRC) [19:30:48 CET] <JEEB> much laughter was had at both FFmpeg and Libav development channels [19:31:06 CET] <furq> it's the only time debian has ever dubiously picked an unpopular software package [19:31:21 CET] <jonkri> Interesting. [19:31:31 CET] <JEEB> well, Libav did had a lot of love and hopes piled onto it in the beginning [19:31:42 CET] <JEEB> it was also the reason why we got certain features finished and merged into both [19:32:01 CET] <JEEB> of course people got disillusioned or otherwise opted to not go for Libav later [19:32:02 CET] <jonkri> Good job. :) [19:32:35 CET] <JEEB> like, personally I'm not sure I like the release process @ Libav, which means that you get releases with bigger amounts of changes every year or less often than that [19:33:35 CET] <JEEB> meanwhile I dislike the FFmpeg Debian maintainer, who seems to not really have a grasp on some things other than he doesn't want to do jack shit (a bit harshly put, but that's the idea I got from his ML e-mails) [19:33:54 CET] <Franciman> Hey people, how can I transform 8 bit samples to 16 bit samples? What's the algorithm ffmpeg use? [19:33:58 CET] <Franciman> I have found this [19:34:27 CET] <JEEB> Franciman: depends on if your talking about video or audio, and all kinds of other little details [19:34:40 CET] <Franciman> //Convert 8 bits to 16 bits [19:34:40 CET] <Franciman> PeakMax := ((PeakMax8 - 128) shl 8); [19:34:43 CET] <JEEB> (like RGB or YCbCr, full range or limited range) [19:34:45 CET] <Franciman> JEEB, audio [19:34:59 CET] <Franciman> sorry [19:35:11 CET] <Franciman> shl is shift left, in C it's << operator [19:35:33 CET] <Franciman> I don't really understand how that works [19:35:56 CET] <Franciman> there PeakMax8 is a sample taken from the audio file [19:36:06 CET] <Franciman> and in uint8_t [19:37:32 CET] <jonkri> If I were to use a recent version of Ffmpeg, how would I take a screenshot 10% into the video OR after a certain time or at the last frame if the video is to short? [19:39:27 CET] <JEEB> you'd have to probe the length first, and then calculate things based on that (or use the API) [19:40:16 CET] <jonkri> JEEB: I'm invoking the binary directly. Is there some way to use the binary to have the length returned? [19:41:32 CET] <JEEB> use {ff,av}probe [19:41:48 CET] <JEEB> it can output stuff in various formats that are parse'able [19:41:52 CET] <jonkri> Thanks, JEEB. :) [19:51:26 CET] <jleclanche> https://trac.ffmpeg.org/wiki/HowToBurnSubtitlesIntoVideo is the subtitles filter lossless for mkv? [19:52:08 CET] <jleclanche> ah it just merges it with the frame, bah :/ [20:04:33 CET] <kepstin> jleclanche: i'm not sure what you mean, it's impossible to burn subtitles into video without modifying the video... what are you trying to do? [20:04:55 CET] <jleclanche> kepstin: mkv supports merged subtitle tracks [20:05:19 CET] <kepstin> not sure what you mean by "merged" there [20:05:25 CET] <jleclanche> mkvmerge -o output.mkv input.mkv input.srt [20:05:27 CET] <jleclanche> this worked [20:05:42 CET] <jleclanche> would be nice to know how to do it with ffmpeg though if its possible [20:05:57 CET] <kepstin> mkv supports having subtitle tracks in the file, so just having the subtitles as an input file in ffmpeg should work [20:06:07 CET] <furq> ffmpeg -i src.mkv -i src.srt -c copy dest.mkv [20:06:13 CET] <kepstin> ffmpeg -i input.mkv -i input.srt -c copy output.mkv ? [20:06:22 CET] <furq> if src.mkv already contains a subtitle track then you'll need to use -map [20:07:00 CET] <ChocolateArmpits> kepstin: What about bitmap subtitles? I think they run as a separate track [20:07:01 CET] <kepstin> the default output will include all subtitle tracks from all inputs, I think [20:07:32 CET] <kepstin> ChocolateArmpits: like vobsub or the bluray ones? Yeah. I dunno how ffmpeg handles those, I think it's different from text subs [20:07:50 CET] <furq> ffmpeg -i src.mkv -i src.srt -map 0 -map 1 -c copy dest.mkv [20:07:52 CET] <furq> ought to do it [20:08:15 CET] <furq> and i'm pretty sure it handles vobsub the same way as any other subs [20:08:20 CET] <jleclanche> furq: Output file #0 does not contain any stream [20:08:31 CET] <jleclanche> oops [20:08:35 CET] <jleclanche> my bad, missed an -i [20:08:46 CET] <jleclanche> thanks guys [20:08:56 CET] <c_14> ChocolateArmpits: should work exactly as with text subs (except that you can't convert them to text subs) [20:09:15 CET] <furq> yeah it's just another stream [20:26:16 CET] <DirtyCajn2> (please bare with me im very new to this, So i am grabing a m3u8 from a site and downloading a .ts from the playlist of .ts files. it somehow decides its going to be 1280x720 ... [20:26:55 CET] <DirtyCajn2> is it based on bitrate? [20:28:40 CET] <DHE> what do you mean? variant selection? [20:29:06 CET] <DirtyCajn2> the source is 2100kbps [20:29:28 CET] <DirtyCajn2> so does ffmpeg decide its 720p from that or is 720p the "default" [20:29:52 CET] <salviadud> Hello. [20:30:18 CET] <salviadud> I want to run an old version of ffmpeg with a relatively old version of the libx264 library [20:30:36 CET] <salviadud> on gentoo [20:30:41 CET] <DHE> the .ts file will contain video so it can read it directly. or sometimes the [master] playlist will have resolutions stored [20:30:55 CET] <salviadud> It shouldn't be that hard right? [20:31:59 CET] <DirtyCajn2> ok i understand so if i soley put # ffmpeg -i "address-to-.m3u8" -c copy "nameoffile.ts" then it should get all the properties from the files. [20:32:16 CET] <DirtyCajn2> ? [20:34:18 CET] <bencoh> salviadud: shouldnt be too hard, but why would you want to do that? [20:34:52 CET] <salviadud> The file I encoded didn't get the output I desired [20:35:02 CET] <salviadud> because of the x264 lib [20:35:06 CET] <salviadud> it wasn't equal [20:35:08 CET] <bencoh> hmm [20:35:14 CET] <bencoh> it might not be bitexact, but ... [20:35:20 CET] <llogan> salviadud: how can we reproduce the issue? [20:35:35 CET] <bencoh> how "different" is it? [20:35:41 CET] <salviadud> Nah, that doesn't matter, what I want is to run it the way I want [20:35:52 CET] <salviadud> I got this mp4 file [20:35:57 CET] <salviadud> ripped with handbrake [20:36:02 CET] <bencoh> well, fetch the source and build [20:36:06 CET] <salviadud> it's 720x486 [20:36:14 CET] <bencoh> it's pretty straightforward for both (x264 and ffmpeg) [20:36:26 CET] <bencoh> but I highly doubt you really need to downgrade [20:36:31 CET] <salviadud> I want to be sure which lib to use though [20:36:39 CET] <salviadud> x264 is from videolan right? [20:36:40 CET] <furq> it's best not to question gentoo users [20:36:45 CET] <salviadud> haha [20:36:54 CET] <bencoh> it's hosted on videolan servers, yes [20:36:58 CET] <salviadud> i think that ffmpeg version can be the latest [20:37:02 CET] <salviadud> but I want to have the same lib [20:37:10 CET] <bencoh> x264 api hasnt changed much (if any) in years [20:37:19 CET] <bencoh> (afaict) [20:37:40 CET] <bencoh> but how different is it? [20:38:10 CET] <salviadud> I'm doing some mad science right now [20:38:19 CET] <salviadud> so. I just want to get the formula right [20:38:32 CET] <bencoh> ? [20:38:55 CET] <salviadud> Have you ever seen those pics that have ciphered msgs in them? [20:39:03 CET] <salviadud> well, I got a video that has something similar [20:39:22 CET] <salviadud> I just need to re-encode it with x264 without any optimization flags [20:39:30 CET] <salviadud> using ffmpeg, obviously [20:39:41 CET] <salviadud> the lib has to be the same [20:40:05 CET] <salviadud> and I got to use an old machine with a single processor [20:40:09 CET] <salviadud> that's what they told me. [20:40:16 CET] <salviadud> So, hurray for science. [20:40:24 CET] <furq> "they" [20:40:36 CET] <salviadud> I tried that a couple of days ago, but I think I used the latest x264 lib [20:40:41 CET] <salviadud> WCP [20:40:58 CET] <salviadud> That's the they [20:40:59 CET] <salviadud> wcp [20:41:08 CET] <furq> the worker's communist party of canada? [20:41:14 CET] <salviadud> west coast productions [20:41:20 CET] <furq> close enough [20:43:26 CET] <waressearcher2> WCP Wasserklosettpapier [20:48:00 CET] <bencoh> salviadud: ciphered msg? do you mean you're having stride issues? [20:49:00 CET] <salviadud> Nah [20:49:20 CET] <bencoh> (something like that http://stackoverflow.com/questions/6495523/ffmpeg-video-to-opengl-texture) [20:49:27 CET] <salviadud> It's supposed to be low-level machine code that gets deciphered with a re-encode [20:50:21 CET] <bencoh> basically you want to disable every x264 SIMD optim until you get a proper result? [20:50:37 CET] <bencoh> (I still dont get how "wrong" your picture is, do you have a snapshot?) [20:50:44 CET] <salviadud> The picture is fine [20:51:12 CET] <salviadud> I'm running a basic transformation command [20:51:23 CET] <salviadud> But it takes 3 days to complete [20:51:32 CET] <salviadud> I need to move that machine to my workplace [20:51:38 CET] <salviadud> the electricity is free over here [20:52:17 CET] <llogan> none of this makes sense. i guess furq is right [20:52:48 CET] <salviadud> It won't make sense to you, because you haven't seen the damn pr0n [20:53:04 CET] <salviadud> There's like this wave of pr0n that came out all wrong on purpuse [20:53:09 CET] <bencoh> mreow. [20:53:13 CET] <salviadud> you gotta ffmpeg it to see it the right way [20:53:34 CET] <salviadud> in fact, it should be the kind of pr0n directed straight at this chan [20:53:38 CET] <salviadud> lol [20:53:59 CET] <llogan> ok. current ffmpeg requires X264_BUILD >= 118. if that is too new, then you'll need to use older ffmpeg do to whatever you are thinking that you want to do [20:54:14 CET] <salviadud> llogan, that is very good to know [20:54:41 CET] <bencoh> (and x264 118 is pretty old) [20:55:15 CET] <salviadud> I'm gonna narrow it down as soon as I compile this mediainfo software [20:55:23 CET] <salviadud> I'll know which library was used [20:55:51 CET] <llogan> then i recommend building without YASM support and instead of using the x264 encoding options you should choose contradicting and nonsense cargo-cultesque options [20:56:35 CET] <salviadud> I remember that the most important one to disable was mmx [20:56:35 CET] <llogan> adding the noise bitstream filter will also introduce some extra derpy-doo [20:56:46 CET] <furq> i recommend just finding some different porn [20:56:49 CET] <salviadud> the optimization options are for the processor [20:57:12 CET] <ChocolateArmpits> why not just choose a different coding standard? [20:57:26 CET] <salviadud> What do you mean ChocolateArmpits ? [20:57:31 CET] <ChocolateArmpits> mpeg2 :) [20:57:47 CET] <salviadud> It has to be libx264 [20:58:01 CET] <ChocolateArmpits> but that's an encoder [20:58:03 CET] <salviadud> this pr0n came out just as handbrake became gpl [20:58:46 CET] <ChocolateArmpits> so wait it's "wrongly" encoded? [20:59:18 CET] <salviadud> no, it's properly encoded, but I gotta get the right library to decode it, bit by bit [20:59:54 CET] <ChocolateArmpits> well then you shouldn't be looking at libx264 but for the "right" decoder [21:01:16 CET] <salviadud> ffmpeg -decoders shows h264 as my decoder [21:01:22 CET] <salviadud> I think it's the same one [21:01:31 CET] <ChocolateArmpits> libx264 is an encoder [21:01:33 CET] <ChocolateArmpits> it doesn't decode [21:01:39 CET] <ChocolateArmpits> h264, it encodes it [21:01:49 CET] <ChocolateArmpits> the h264 is the decoder [21:01:53 CET] <ChocolateArmpits> that's listed [21:01:58 CET] <ChocolateArmpits> no relation between the two [21:02:00 CET] <furq> i didn't know kit williams worked for west coast productions [21:02:25 CET] <salviadud> so, that decoder, where is it from? [21:02:29 CET] <salviadud> is it native for ffmpeg? [21:02:34 CET] <furq> that's the builtin ffmpeg decoder [21:03:02 CET] <salviadud> and that decoder is built into the source? [21:03:24 CET] <salviadud> I mean, if I wanted to try an earlier version of the decoder, I'd have to use an earlier version of ffmpeg [21:03:30 CET] <ChocolateArmpits> part of libavcodec I guess [21:03:52 CET] <furq> you'd have to use an earlier libavcodec [21:03:56 CET] <furq> which generally implies an earlier ffmpeg [21:04:37 CET] <ChocolateArmpits> so get a source release that was published around the time the video was encoded and see how it fares [21:04:47 CET] <ChocolateArmpits> your best bet [21:05:09 CET] <salviadud> Yeah, its from 2013 [21:05:19 CET] <salviadud> I can get a source from that time [21:05:24 CET] <salviadud> gentoo has a big attic [21:05:55 CET] <furq> i never thought i'd be helping someone have a wank in #ffmpeg [21:06:00 CET] <bencoh> :D [21:06:01 CET] <furq> i don't know how i feel about this [21:06:05 CET] <salviadud> This isn't even for a wank [21:06:25 CET] <bencoh> so, wait... you cant decode it properly anymore with ffmpeg? [21:06:25 CET] <salviadud> you can wank to it properly with stretch and aspect [21:06:35 CET] <furq> well that's a relief [21:06:37 CET] <salviadud> no, it has a hidden message inside [21:06:42 CET] <salviadud> that's the important stuff [21:06:48 CET] <salviadud> it's not even pr0n [21:06:48 CET] <ChocolateArmpits> "hidden message" [21:06:50 CET] <ChocolateArmpits> spooky [21:06:57 CET] <furq> sexy [21:07:02 CET] <ChocolateArmpits> how are you sure it's encoded in the video ? [21:07:10 CET] <salviadud> because I made it [21:07:12 CET] <salviadud> in 2007 [21:07:19 CET] <salviadud> then I did a lot of cannabis and I forgot [21:07:21 CET] <salviadud> it happens... [21:07:27 CET] <bencoh> oookay, but the decoded picture (decoder output) should be the same [21:07:30 CET] <salviadud> so, I'm brushing up on my ffmpeg [21:07:35 CET] <bencoh> (not sure about bitexact, but ... almost) [21:07:46 CET] <ChocolateArmpits> what were you watching the video with back then ? [21:08:04 CET] <salviadud> just the original crappy file [21:08:06 CET] <furq> wait [21:08:15 CET] <salviadud> no hidden message [21:08:23 CET] <bencoh> how was it "hidden"? [21:08:25 CET] <furq> so in 2007 you encoded a hidden message into some porn which came out in 2013 [21:08:28 CET] <salviadud> it's like a zip file [21:08:38 CET] <bencoh> steganography? [21:08:39 CET] <salviadud> which someone ripped in 2013 [21:08:49 CET] <salviadud> I actually never bought my own pr0n [21:08:52 CET] <salviadud> maybe I should've done that [21:09:29 CET] <ChocolateArmpits> Are you sure you didn't put the message inside the 0-16 or 235-255 levels ? [21:09:32 CET] <furq> are you the producer or the star [21:09:35 CET] <salviadud> look, if you watch the movie, there's even a part that shows a bucket [21:09:41 CET] <salviadud> it's a video message [21:09:44 CET] <furq> or that rarest of talents, both [21:09:51 CET] <salviadud> it's not some hexadecimal crap [21:10:08 CET] <salviadud> the bucket is there for u to get the ratio right [21:10:14 CET] <salviadud> it's a point of reference [21:10:29 CET] <ChocolateArmpits> is that bucket important in the case of the hidden message ? [21:10:35 CET] <salviadud> actually, yes [21:10:43 CET] <salviadud> you get a still frame from the bucket [21:10:49 CET] <salviadud> and you stretch it until it's right [21:10:55 CET] <salviadud> that's how you get the correct -s value [21:11:01 CET] <ChocolateArmpits> and then you see the message ? [21:11:09 CET] <salviadud> But, if you already know what the proper resolution is [21:11:17 CET] <salviadud> it's just kinda weird to see a bucket [21:11:24 CET] <salviadud> after a pr0n has ended [21:11:32 CET] <furq> depends what's been happening [21:11:49 CET] <bencoh> :D [21:12:01 CET] <salviadud> It's a closed bucket of paint [21:12:10 CET] <bencoh> sounds fun [21:12:25 CET] <salviadud> Now, I just found out that I probably did more [21:12:37 CET] <salviadud> Because I just re-encoded a movie that was 4:3 [21:12:46 CET] <salviadud> and the resolution was 720x480 [21:12:58 CET] <salviadud> I changed it to 848x480 [21:13:02 CET] <salviadud> It looks marvelous [21:13:10 CET] <salviadud> if you're into black girls... [21:13:19 CET] <salviadud> so, anyway. I did post-production work [21:13:32 CET] <furq> well shit [21:13:36 CET] <furq> that ruins my "into black girls" joke [21:17:02 CET] <salviadud> if you are true scientists [21:17:09 CET] <salviadud> you will see both versions of said video [21:17:11 CET] <salviadud> the original rip [21:17:17 CET] <salviadud> and my ffmpeg re-encode [21:21:03 CET] <salviadud> I got it, x264 core 130 r2273 [21:35:47 CET] <salviadud> Ok, I'm running multiple inputs [21:35:56 CET] <salviadud> like ffmpeg -i video -i audio [21:36:09 CET] <salviadud> and I'm checking the coding and it looks like my bitrate went down [21:36:31 CET] <ChocolateArmpits> salviadud: use bitrate settings [21:36:40 CET] <salviadud> I just do, b:v and some value [21:37:02 CET] <salviadud> I just want to join the files [21:37:07 CET] <salviadud> without loss [21:37:15 CET] <salviadud> and without adding quality it doesn't have [21:38:26 CET] <ChocolateArmpits> well that's impossible to do at the same original bitrate [21:38:33 CET] <salviadud> ok, that fixed it [21:38:46 CET] <salviadud> my original bitrate is 2500k [21:38:52 CET] <salviadud> average [21:39:01 CET] <salviadud> It's spiking it at 2700 [21:39:07 CET] <salviadud> I think it's fine [21:48:12 CET] <BtbN> Just use -c copy if you only want to remux stuff? [22:01:05 CET] <salviadud> that's a codec option right? [22:01:08 CET] <salviadud> not a bitrate option [22:01:34 CET] <BtbN> what? [22:01:43 CET] <salviadud> -c copy [22:01:51 CET] <BtbN> ...so? [22:02:08 CET] <ChocolateArmpits> interestingly aac is now listed as good or better than fdk_aac https://trac.ffmpeg.org/wiki/Encode/HighQualityAudio [22:02:10 CET] <BtbN> You just want to combine two files, without any change in quality or size. So it's what you are looking for. [22:02:35 CET] <salviadud> Yeah, BtbN , I'm a bit rusty with my ffmpeg [22:02:57 CET] <furq> ChocolateArmpits: yeah but that also says libvorbis is better than fdk_aac [22:03:12 CET] <squeegily> How do I set the output device with ffplay? [22:03:35 CET] <ChocolateArmpits> furq: let's say not all parts are correct but some are [22:03:38 CET] <squeegily> I'm on pulse, and I used -sinks pulse to get the output name; I just don't know what flag to feed it to [22:04:26 CET] <furq> hopefully hydrogenaudio does a new listening test with fdk and ffmpeg aac [22:06:20 CET] <squeegily> Anybody know how to set the output device? [22:09:24 CET] <BtbN> What output device? [22:09:56 CET] <squeegily> Audio [22:10:11 CET] <salviadud> damn [22:10:16 CET] <salviadud> I got an audio/video desync error [22:10:29 CET] <salviadud> after I did a remux [22:10:42 CET] <squeegily> I have a music file which crashes MPV if I try to play it through my integrated audio, but works just fine fed to the dedicated chip, and I want to see what ffplay does with it [22:17:54 CET] <durandal_1707> squeegily: it uses whatever sdl provides [22:18:43 CET] <durandal_1707> squeegily: maybe its mpv bug [22:24:51 CET] <squeegily> durandal_1707: yeah, it's absolutely an MPV bug, I just wanted to test around with ffplay to maybe have some more useful info for the mpv bug log [22:29:52 CET] <basisbit> any news on the video stuttering / frames out of order issues when using the low latency flag? [22:29:55 CET] <salviadud> Ok, aac is so crap [22:30:00 CET] <salviadud> I had to conver it to mp3 [22:30:07 CET] <salviadud> and then it ran fine, no more audio glitch [22:43:07 CET] <TD-Linux> furq, last I was aware libvorbis does generally outperform fdk_aac [22:50:58 CET] <furq> the only decent listening test i've seen only has apple aac [22:51:09 CET] <furq> which does a bit better than vorbis [22:51:59 CET] <klaxa> have you tried out opus? [22:52:03 CET] <furq> with that said there doesn't seem to be any point using vorbis over opus these days [22:52:19 CET] <furq> klaxa: opus won that test pretty handily [22:53:40 CET] <furq> http://listening-test.coresv.net/nonblocked_means_all_hd2.png [22:54:36 CET] <klaxa> oh, neat graph [22:55:50 CET] <klaxa> android also finally supports opus in ogg containers (since android 6) [22:55:59 CET] <furq> i heard that fdk was roughly on a par with appleaac, but i've not seen anything resembling evidence of that [22:56:51 CET] <furq> there's an old HA listening test from 2011 which compares them but i doubt that's of any use four years later [22:56:59 CET] <klaxa> oh? well a few months ago it was the best open-source aac encoder there was [22:57:12 CET] <klaxa> not sure how the native aac encoder of ffmpeg compares now [22:57:59 CET] <furq> i already use fdk, but it'd be nice to see a graph [22:59:15 CET] <furq> if builtin aac compares favourably then it'll save me a bit of hassle recompiling [22:59:47 CET] <klaxa> the experimental flag was removed iirc [22:59:51 CET] <klaxa> i could be wrong though [22:59:58 CET] <furq> yeah it was [23:00:16 CET] <BtbN> iirc it performs very decent for CBR. VBR still needs some work [23:19:32 CET] <TD-Linux> this is the newest of the ffmpeg aac encoder tests I've seen https://www.hydrogenaud.io/forums/index.php?showtopic=109716 [23:20:50 CET] <llogan> squeegily: see http://ffmpeg.org/ffmpeg-devices.html#toc-Output-Devices [23:23:36 CET] <llogan> TD-Linux: https://trac.ffmpeg.org/ticket/2686#comment:471 [23:23:53 CET] <llogan> unsure if he completed test yet [23:24:36 CET] <TD-Linux> ah neat he's doing another one, will be interesting to see [23:25:24 CET] <squeegily> aha, it was definitely an MPV bug. Well, I don't know that you'd call it a "bug". PA has a hardcoded 192kHz limit, evidently, and FFplay worked around it by just hardcoding in the same limit [23:25:54 CET] <squeegily> So now poor wm4 is trying to decide between having >192kHz audio broken on PA or implementing an awful workaround [23:26:05 CET] <squeegily> he's still trying to see if there's a clean solution available at all [23:27:35 CET] <BtbN> >192kHz sounds like it's broken already, where does that stuff even come from? [23:30:36 CET] <TD-Linux> whale mating call recordings [23:30:52 CET] Action: TD-Linux was already against adding >48khz to mpv [23:37:23 CET] <ChocolateArmpits> Probably orchestras with many microphones recording [23:39:48 CET] <klaxa> that would explain many channels, but why >192kHz sampling? [23:40:41 CET] <ChocolateArmpits> higher sample rate helps to "audially" preserve more high-frequency instruments instead of smugding them into one [23:41:17 CET] <klaxa> ah, makes sense, thanks for the explanation [23:41:32 CET] <fritsch> if you have the equipment to manage such high samplerates [23:41:41 CET] <fritsch> the higher those get the harder it is to work with those [23:41:55 CET] <fritsch> not talking about software here, but hardware [23:42:03 CET] <ChocolateArmpits> most likely analog [23:42:25 CET] <fritsch> yeah - at a point you need to go to the digital domain [23:42:44 CET] <fritsch> where this sampling play an important role [23:42:48 CET] <fritsch> as signals are sampled [23:42:51 CET] <fritsch> and not analog [23:42:52 CET] <fritsch> :-) [23:43:26 CET] <fritsch> https://xiph.org/~xiphmont/demo/neil-young.html <- here I like posting that one [23:43:35 CET] <fritsch> he talks also a bit about professional post processing [23:44:11 CET] <ChocolateArmpits> there's no point to high sample rate for pop music [23:44:23 CET] <ChocolateArmpits> like there's no use for it there [23:44:35 CET] <ChocolateArmpits> only monster orchestras gain anything if at all [23:46:58 CET] <TD-Linux> ChocolateArmpits, no genre of music gains anything over others [23:47:10 CET] <TD-Linux> why would inaudible sounds in an orchestra matter more than pop? [23:47:27 CET] <salviadud> Guys, want to see the fruit of my work with ffmpeg? Can I post a kat.cr link here? [23:47:47 CET] <ChocolateArmpits> salviadud: absolutely we must see it for the purposes of research [23:47:55 CET] <salviadud> https://kat.cr/let-off-in-me-5-ppralph-t11834227.html [23:48:18 CET] <salviadud> The original file was wack, this one is the real deal [23:48:56 CET] <ChocolateArmpits> >Cherokee D'Ass [23:49:00 CET] <ChocolateArmpits> these are not realy names [23:49:02 CET] <ChocolateArmpits> are they? [23:49:19 CET] <ChocolateArmpits> real* [23:49:38 CET] <salviadud> Oh come on, Cherokee is a legend. [23:49:56 CET] <salviadud> sorta... hehe [23:50:00 CET] <ChocolateArmpits> well he's not on twitter [23:50:12 CET] <salviadud> It's a she [23:50:17 CET] <salviadud> I think she's shy. [23:50:17 CET] <TD-Linux> and you're not a real person if you're not on twitter [23:50:37 CET] <salviadud> So, I worked on this proyect back in 2006 [23:50:40 CET] <ChocolateArmpits> well all real western celebrities are on twitter nowadays [23:50:42 CET] <salviadud> Well, if you could call it a proyect [23:50:59 CET] <ChocolateArmpits> pornstars included [23:51:26 CET] <ChocolateArmpits> it helps to better get in touch with the fanbase [23:51:43 CET] <salviadud> Anyway, the point is that I purposely set the resolution and aspect wrong [23:51:56 CET] <salviadud> Because I could compress it in a cheap dvd. [23:52:08 CET] <salviadud> and if you still own it, you can rip it yourself and LEARN to use ffmpeg [23:52:11 CET] <salviadud> to fix it [23:52:11 CET] <waressearcher2> Sherokee [23:52:22 CET] <salviadud> That was the whole idea... KNOWLEDGE [23:52:39 CET] <salviadud> Just check out the logo for blackice production on twitter [23:52:48 CET] <salviadud> a frozen pinguin. [23:53:00 CET] <salviadud> They know what they're doing, I taught them. [23:53:12 CET] <salviadud> and now, you guys are refreshing my memory, so much thanks. [00:00:00 CET] --- Thu Dec 31 2015

1 0

ffmpeg-devel.log.20151229
by burek 30 Dec '15

30 Dec '15

[00:01:41 CET] <cone-204> ffmpeg 03Paul B Mahol 07master:a7bf5f4117b4: avfilter/avf_showspectrum: use FF_ARRAY_ELEMS() [00:09:34 CET] <durandal_1707> atomnuker: I plan to do other things too, overlap for example, to incrrease time resolution for big number of freuqency bins [00:11:28 CET] <atomnuker> nice, I did write a pulseaudio spectrum analyzer a few months ago with libsdl and libfftw3 which had overlap [00:11:55 CET] <atomnuker> the logic is trivial, but if you want you can reuse parts of it [00:28:43 CET] <durandal_1707> atomnuker: tried baudline? There is linux binary. [00:31:14 CET] <atomnuker> I do remember looking at it back when spek didn't work [00:59:08 CET] <kierank> atomnuker: still stuck on these cfhd artefacts [00:59:10 CET] <kierank> soo close... [00:59:54 CET] <atomnuker> kierank: did you start messing around with random things to see if they disappear? [01:00:08 CET] <kierank> yes but they never disappear [01:00:31 CET] <atomnuker> anything on valgrind? [01:00:40 CET] <kierank> nop [01:01:08 CET] <kierank> it happens with only one subband [01:01:23 CET] <kierank> so it's highpass related [01:02:25 CET] <atomnuker> checked the coefficients for anything out of the ordinary? [01:02:52 CET] <kierank> can't really do that [01:02:54 CET] <kierank> there are thousands [01:02:59 CET] <kierank> millions in fact [01:04:44 CET] <atomnuker> well, print them in case they're below some ridiculious negative and some ridiculous positive number [01:21:35 CET] <kierank> https://usercontent.irccloud-cdn.com/file/B7REaavM/ [01:21:37 CET] <kierank> definitely highpass [01:30:28 CET] <atomnuker> kierank: 4*low[1*low_stride] + low[2*low_stride] [01:30:54 CET] <atomnuker> shouldn't you be skipping 2 and then 4? [01:31:02 CET] <atomnuker> and then the same for odd numbers? [01:31:42 CET] <kierank> no [01:31:44 CET] <atomnuker> oh wait the levels are not interleaved [01:37:00 CET] <kierank> it is definitely caused by low = lh3; [01:37:00 CET] <kierank> high = hh3; [01:50:33 CET] <cone-204> ffmpeg 03Ganesh Ajjanagadde 07master:63702014fa4e: lavfi/af_anequalizer: remove cabs, cexp dependencies [01:50:34 CET] <cone-204> ffmpeg 03Ganesh Ajjanagadde 07master:59ac7ce4dac5: lavfi/af_anequalizer: replace pow(10,x) by ff_exp10(x) [01:50:35 CET] <cone-204> ffmpeg 03Ganesh Ajjanagadde 07master:9f4c7b4df7f3: lavfi/af_anequalizer: replace pow(x,-2) by 1/(x*x) [02:14:42 CET] <kierank> atomnuker: found it [02:14:45 CET] <kierank> stride problem :) [02:21:14 CET] <wm4> so I keep getting files that show stuff like "aac: element type mismatch 3 != 0" on every packet or so (they appear to sound fine, and didn't print such a message in older lavc versions) [02:21:23 CET] <wm4> could this message just be silenced? [02:21:30 CET] <wm4> because it fills up terminals and log files [02:25:16 CET] <rcombs> 3 is, in fact, not equal to 0 [02:25:42 CET] <rcombs> downgrade to debug? [02:27:14 CET] <kierank> btw I tried plex for the first time [02:27:27 CET] <kierank> streaming to a tv was fine over wired [02:27:30 CET] <kierank> crap over wireless [02:27:52 CET] Action: kierank gives useless bug report [02:33:20 CET] <rcombs> thanks [02:33:27 CET] Action: rcombs files in /dev/null with the similar reports [02:33:44 CET] <rcombs> (maybe reduce maximum bitrate setting, or check if your TV's wifi is as shit as most TVs' wifi) [02:52:13 CET] <wm4> well that aac message was added with 79a98294da6cd85f8c86b34764c5e0c43b09eea3 [02:53:29 CET] <wm4> I suppose in theory it's possible that this check is too strict and triggers in valid cases too? [02:58:30 CET] <wm4> it does look a bit fishy to me, but then I have not the slightest clue about aac [12:14:35 CET] <cone-127> ffmpeg 03Paul B Mahol 07master:e30e5c83aed7: avfilter/avf_showspectrum: add horizontal orientation support [12:14:36 CET] <cone-127> ffmpeg 03Paul B Mahol 07master:efa666e0da5e: avfilter/avf_showspectrum: properly calculate w factor [13:13:56 CET] <cone-127> ffmpeg 03Mark Harris 07master:c51c08e0e70c: avcodec: Use get_ue_golomb_long() when needed [13:13:57 CET] <cone-127> ffmpeg 03Mats Peterson 07master:bd03ee1184e4: lavf/qtpalette: Treat 1-bit video as palettized [13:41:09 CET] <ubitux> durandal_1707: showspectrum has been heavily modified since my initial version [14:08:32 CET] <durandal_1707> ubitux: ? [14:08:53 CET] <ubitux> 23:37 <durandal_170> ubitux: showspectrum is missing overlap for big fft size [14:09:06 CET] <ubitux> so i'm saying i'm not really following closely the evolution of the filter [14:34:53 CET] <durandal_1707> use wavelets for an audio [14:54:37 CET] <J_Darnley> Does FFmpeg's makefiles have a target for pre-processed cc output? (i.e. gcc's -E flag)? [14:57:48 CET] <nevcairiel> i think you can ask ffmpeg to produce .i files? [14:58:01 CET] <J_Darnley> Ah .i? Let me try [14:58:10 CET] <J_Darnley> ha ha yes [14:59:08 CET] <J_Darnley> thanks [15:50:35 CET] <atomnuker> michaelni: sorry, somehow I sent the same patch twice [15:50:47 CET] <atomnuker> could you do a git pull from https://github.com/atomnuker/FFmpeg and tell me if it complies fine? [16:01:06 CET] <cone-127> ffmpeg 03Carl Eugen Hoyos 07master:e5c16e38fe15: ffserver: Cast time_t value when using it in a format string. [16:28:18 CET] <michaelni> atomnuker, builds fine [16:30:22 CET] <atomnuker> yes, just managed to test it myself [16:30:51 CET] <atomnuker> michaelni: should I sent the RFC as a responce to your last email or just make a new thread on the ML? [16:31:02 CET] <atomnuker> s/responce/response [16:31:37 CET] <michaelni> btw your tree doest pass fate [16:33:11 CET] <michaelni> about reply vs new, whatever you prefer [16:42:58 CET] <prelude2004c> hey guys.. good day.. looking for some help with ffmpeg and vdpau... i tried in regular forum but i think there is somethign wrong with the code. Basically i am using ffmpeg and vdpau ( which i know is working because i can decode mpeg2video ).. i am trying to decode h264 into the GPU instead of CPU ( using M4000 ) card.. but for some reason the h264 never uses the GPU.. is this some bug in code ? the card says it supports GPU decodin [16:50:30 CET] <J_Darnley> Does the GPU support the video profile? Is it >8 bit? Is it lossless? [16:58:14 CET] <prelude2004c> darnley that questino for me ? [16:58:26 CET] <J_Darnley> yes [18:00:52 CET] <cone-127> ffmpeg 03Ganesh Ajjanagadde 07master:4a44bc558894: ffprobe: avoid unnecessary pow and exp2 calls [18:44:33 CET] <cone-127> ffmpeg 03Andreas Cadhalpun 07master:b4466c9c1a82: oggparsedaala: check number of planes in pixel format map [19:48:26 CET] Action: kierank wonders how to store cfhd coefficients [19:51:13 CET] <durandal_1707> kierank: in struct [19:51:25 CET] <kierank> not that simple [19:51:40 CET] <kierank> I don't think I can use the layout dirac/snow uses [19:53:56 CET] <cone-127> ffmpeg 03Paul B Mahol 07master:b0e23f2a37b4: avfilter/avf_showspectrum: add rainbow colormap [19:58:30 CET] <Compn> kierank : copy from prores? [19:58:44 CET] <kierank> completely unrelated [19:59:20 CET] <Compn> ah [20:02:53 CET] <Compn> who is encoding with x264 on arm... [20:04:03 CET] <rcombs> me, on occasion [20:04:14 CET] <rcombs> usually to see if it's worth doing on the latest shiny ARM64 systems yet [20:04:15 CET] <rcombs> (no) [20:04:32 CET] <rcombs> (unless you like 480p in which case maybe) [20:11:13 CET] <atomnuker> while (topbit >>= 1) count++ << this would essentially count the number of zeroes, right? [20:12:18 CET] <llogan> is there a preferred RGB notation? making a showwavespic example [20:16:01 CET] <rcombs> atomnuker: isn't that more of a simple log2? [20:16:15 CET] <atomnuker> yes, exactly [20:21:00 CET] <durandal_1707> llogan: what? [20:22:52 CET] <llogan> users keep asking for a colored waveform so i'm adding an example. i just used RGB(66,217,150). [21:01:04 CET] <J_Darnley> Damn these avx split lanes! [21:07:31 CET] <rcombs> they're the _worst_ [21:25:04 CET] <RiCON> is gcrypt's dependancy on gpg-error optional? [21:26:24 CET] <durandal_1707> ahistogram: another multimedia A-V filter [22:36:29 CET] <J_Darnley> kierank: are you around? [22:36:36 CET] <kierank> J_Darnley: yes [22:36:49 CET] <J_Darnley> I can't connect to avdev. My key appears to be denied [22:38:38 CET] <kierank> J_Darnley: does it work now? [22:38:55 CET] <J_Darnley> No [22:41:30 CET] <J_Darnley> I wonder if I sent you the right key [22:42:24 CET] <J_Darnley> th email says 2048bit [23:28:24 CET] <J_Darnley> I need a better system for copying dotfiles around [00:00:00 CET] --- Wed Dec 30 2015

1 0

ffmpeg.log.20151229
by burek 30 Dec '15

30 Dec '15

[01:34:27 CET] <jafa> hi guys. I am looking ways to do graceful fallback when the cpu can't keep up with HD mpeg2 decode + deinterlacing [01:34:47 CET] <jafa> 720p60 -> 720p30 is quick and easy [01:35:19 CET] <jafa> looking for advice for 1080i [01:35:47 CET] <jafa> could halve the vertical resolution to output 1920x540p30 [01:36:56 CET] <jafa> is there an easy way to do that with a filter graph or should I do a simple decimate post decode pre filter graph? [02:58:56 CET] <pinPoint> does ffmpeg automagically encode out in 5.1 if it runs into an audio file or a video clip with 5.1 audio? [05:03:20 CET] <digidog> hi, is it possible to ffplay (filter>grayscale) a video file on the fly or do I always need to convert the file before? [05:08:23 CET] <venkat_330> Request some guidance in adding timestamp to to video recording that is performed by FFMPEG. When i execute the command i get "ffmpeg: unrecognized option '-vf'" as the error. http://pastebin.com/kZNTusD3 [05:23:04 CET] <furq> venkat_330: that's an ancient version of ffmpeg, i doubt you'll get much help with it [05:34:53 CET] <venkat_330> furq: I just want the timestamp to work actually [07:41:52 CET] <Logicgate> hey guys [07:42:05 CET] <Logicgate> Which pre-compiled binary would be the optimal one to get at the moment? [07:42:18 CET] <Logicgate> http://johnvansickle.com/ffmpeg/ [09:33:42 CET] <pkeuter> when i want to cut a fragmented mp4 file, the output does not start on the exact given start position. probably because it cuts in between fragments? Is there a way to force the startposition? [10:33:23 CET] <RaMcHiP-TMP> Hello all! I just installed octoprint onto a raspberry pi and it has a timelapse video feed using ffmepg. When I preview or record I get extreme distortion in the lower 2/3- 3/4 of the screen. Here is an example of a time lapse https://www.dropbox.com/s/cn5xd02loihqdvp/svapoDockingUSBFixed_201512290916… [10:35:27 CET] <RaMcHiP-TMP> Anyone know what might be causing that? [12:22:31 CET] <WongKit> Hi, does anyone know if there is an alternative AVI muxer method available in FFMPEG? [12:27:55 CET] <relaxed> WongKit: why do you need an alternative? [12:29:37 CET] <WongKit> thanks for your reply. a video file muxed with ffmpeg does not play back on a hdd recorder [12:30:05 CET] <fritsch> what does mediainfo of the resulting file say? [12:30:53 CET] <WongKit> which information of media info do you need? [12:30:58 CET] <WongKit> it is an xvid file [12:31:24 CET] <relaxed> paste it all at pastebin.com [12:31:32 CET] <WongKit> if I remuxe it with xmediarecode, the hdd recorder can play it [12:31:41 CET] <WongKit> ok, mom [12:32:48 CET] <WongKit> http://pastebin.com/42PbrJvS [12:36:30 CET] <relaxed> looks fine [12:36:32 CET] <fritsch> how does the xmediarecode look? [12:37:49 CET] <WongKit> it is format AVI with copy-mode for video and audio [12:38:10 CET] <fritsch> can you pastebin the info never the less? [12:38:21 CET] <WongKit> ok [12:38:25 CET] <relaxed> pastebin the command and ffmpeg's output as well [12:38:39 CET] <JodaZ> WongKit, did you compare the info for the file generated by ffmpeg and the one by xmediarecode? [12:39:23 CET] <WongKit> yes, they are different at few points (hex view) and result into the same mediainfo output [12:39:45 CET] <JodaZ> hmm [12:39:46 CET] <JodaZ> well [12:39:54 CET] <JodaZ> also, isn't xmediarecode just using ffmpeg, WongKit? if so, just get a good task manager and see how xmediarecode launches ffmpeg in the background [12:39:58 CET] <ethe> Logicgate: the git one, and 32/64bit depends on your system [12:42:23 CET] <WongKit> @fritsch: xmediarecode does not give me much I could copy to pastebin. i can only upload some screenshots that says "copy" [12:43:56 CET] <JodaZ> WongKit, just check how xmediarecode launches ffmpeg with process explorer [12:44:12 CET] <WongKit> it uses an dll [12:44:24 CET] <WongKit> not the .exe [12:44:29 CET] <JodaZ> :/ [12:45:16 CET] <bencoh> ffmpeg avix mux really misses a lot of avi options [12:47:00 CET] <bencoh> a/v (non-)interleaving being the most obvious [12:47:12 CET] <WongKit> output of ffmpeg http://pastebin.com/5CDKQndP [12:47:17 CET] <WongKit> note: it also contains reencoding, but it works fine after it was remuxed by xmedia [12:48:20 CET] <JodaZ> vtag xvid? you sure thats neccessary? aint those players usually DivX licensed? [12:49:07 CET] <JodaZ> ew, nvm, it uses something strange otherwise [12:49:16 CET] <WongKit> this was one of the many tries to get the file compatible with my recoder [12:49:46 CET] <WongKit> also tried DIVX as vtag, but the problem does not seem codec related [12:53:25 CET] <WongKit> btw, I used a self compiled ffmpeg binary in the upper pastebin, but also tried the "public" one from zeranoe [12:55:11 CET] <JodaZ> how many changes are there in the bindiff? [12:56:36 CET] <JodaZ> any reason you don't want mkv? player should do that too if its divx certified [12:59:12 CET] <WongKit> there are some changes in the file (estimated 0.5%) [12:59:40 CET] <WongKit> the player does not accept mkv files [13:00:04 CET] <WongKit> it is very old and only allows .avi and .divx as a file extension [13:00:33 CET] <JodaZ> .divx should be mkv [13:00:53 CET] <WongKit> let me try it [13:01:38 CET] <JodaZ> hmm, maybe really old .divx isn't mkv, recent is tho, so its worth a try [13:04:13 CET] <JodaZ> bencoh, you know how to fix this? [13:05:31 CET] <WongKit> no, the divx-mkv file is rejected with "unknown format" [13:06:46 CET] <WongKit> the avi file produces a black screen without any play back or error message [13:07:37 CET] <JodaZ> does it get the video length? [13:07:50 CET] <WongKit> no [13:12:45 CET] <bencoh> JodaZ: fix what, ffmpeg avienc.c, or WongKit's issue? [13:16:45 CET] <bencoh> WongKit: I'd try remuxing with avi-mux and/or virtualdub and playing with the parameters as their mux is more configurable than the one in ffmpeg [13:17:12 CET] <JodaZ> bencoh, both? [13:19:03 CET] <bencoh> JodaZ: regarding non-interleaved avi muxing and ffmpeg, I dont know of a simple way to do it (ie without buffering in avformat and/or seeking in resulting file) [13:19:27 CET] <bencoh> (or both, or moving data around like with the mp4 mux) [13:19:53 CET] <JodaZ> bencoh, i doubt the interleaving is much different if WongKit says the difference between a working and a non-working mux is just 0.5% [13:20:17 CET] <bencoh> indeed [13:20:25 CET] <bencoh> (in his case) [13:20:48 CET] <WongKit> @bencoh: i already remuxed it with xmediarecode and it was readable after that [13:21:30 CET] <WongKit> however, i want to prevent using more tools than i have to :) [13:21:48 CET] <WongKit> i can provide a sample file if it helps [13:21:52 CET] <bencoh> WongKit: can you parse/compare the RIFF structure? [13:23:27 CET] <WongKit> as far as i see, they are (except for the size byte) identical [13:23:45 CET] <WongKit> ffmpeg: RIFFaÈ. AVI LIST~" hdrlavih8 Â¢ ša [13:23:49 CET] <WongKit> working xmedia: RIFF É. AVI LISTî" hdrlavih8 Â¢ ž! [13:27:22 CET] <bencoh> err, that's not parsed :p [13:29:55 CET] <WongKit> it was worth a try ^^" [13:30:41 CET] <WongKit> i uploaded a video sample (big buck bunny, 30 sec) http://www.mediafire.com/download/w7uxyyysl1i4f4a/video-sample.zip [13:52:42 CET] <JodaZ> lol [14:31:54 CET] <Disturbed1_away> good morning world.... quick question... anyone else having trouble appling ffmpeg_NVIDIA_gpu_acceleration.patch? [14:35:48 CET] <WongKit> @bencoh: is there any kind of information i could offer? [14:50:41 CET] <Zucca> When I set video/audio bitrate in kbits, does it convert to 1000 times in bits or 1024 times? [14:51:43 CET] <Zucca> I haven't thought this ever in case of bitrates. [15:00:49 CET] <DHE> SI prefixes, so 1000 [15:01:21 CET] <Zucca> DHE: Ok. Thanks. :) [15:01:59 CET] <DHE> but if you want 1024 you can use ki, Mi, Gi, etc [15:02:23 CET] <Zucca> Ah. Of course. [15:06:10 CET] <Zucca> DHE: since dawn of mp3 has these bitrate prefixes been SI commonly? [15:06:35 CET] <DHE> not sure. I would assume so... [15:08:40 CET] <Zucca> I couls just test by encoding a mp3 using 128kibit and 128kbit bitrate and see if my old mp3 player accepts which file. [15:10:11 CET] <WongKit> Need to leave now. relaxed, fritsch, JodaZ, bencoh - thank you for your help! [15:13:30 CET] <pkeuter> when i want to cut a fragmented mp4 file, the output does not start on the exact given start position. probably because it cuts in between fragments? Is there a way to force the startposition? [15:27:53 CET] <pkeuter> anyone? [15:30:01 CET] <waressearcher2> pkeuter: hallo, wie geht's es dir ? [15:30:16 CET] <j105rob> WRT http://pastebin.com/PiJhkZNm does this message libx264 @ 0x7ff67d006a00] frame= 1 QP=13.00 NAL=2 Slice:P Poc:8 I:0 P:0 SKIP:1 size=12 bytes mean that the frame was skipped? If so, can you tell me why? [15:30:21 CET] <pkeuter> haha iŽm fine, but not from germany :-) [15:49:07 CET] <JodaZ> pkeuter, read up on the different seeking options in ffmpeg, theres different amounts of precission, the obvious limitation for example is that if you don't re-encode you can only cut at keyframes for example [15:49:43 CET] <pkeuter> JodaZ, thanks! I thought the keyframes would have something to do with it. [15:50:01 CET] <pkeuter> Now I am wondering what the default keyframe setting is on x264 [15:50:46 CET] <JodaZ> you can use those showinfo filter options to get an idea about how keyframe rich your file is [15:51:21 CET] <JodaZ> pkeuter, but if you do re-encode (instead of just remux) you can force it to a specific (even non-keyframe) frame [15:51:46 CET] <pkeuter> That sounds nice, but that brings me to another, more complicated question [15:51:54 CET] <pkeuter> Do you have time for that? [15:52:04 CET] <JodaZ> someone here might [15:53:25 CET] <pkeuter> iŽll just ask. it looks as if when i use afade, it wants to reprocess the entire file. which is an hour long. [15:53:38 CET] <pkeuter> well only when i use afade=t=in [15:53:43 CET] <pkeuter> afade=t=out is fine [16:13:29 CET] <jonj> Good day. I'm trying to figure out if it's possible to feed a folder of (5) image files to FFMPEG and still use filter_complex on the individual files. Command and console output is here: http://pastebin.com/y5tQ8AS3 Thank you. [16:16:30 CET] <c_14> No, the image2 demuxer turns all the images into a single stream. You can't operate on the individual images like that [16:19:52 CET] <jonj> thank you for the clarification. I guess i'll need to define each input then like I did in the first place... that works :) [17:46:43 CET] <RaMcHiP-TMP> Hello all! I just installed octoprint onto a raspberry pi and it has a timelapse video feed using ffmepg. When I preview or record I get extreme distortion in color and picture. Here is an example of a time lapse https://www.dropbox.com/s/cn5xd02loihqdvp/svapoDockingUSBFixed_201512290916… and my lsusb gives me Bus 001 Device 007: ID 0c45:62f1 Microdia [18:01:28 CET] <Disturbed1_away> grrr, four days with nvidia s870 and ffmpeg... uhg still no compile... :( [18:04:46 CET] <waressearcher2> es ist schade [18:07:06 CET] <danjanz> hi, does anybody know if it is possible to recreate ismv files from smooth streaming Manifest and all the fragments? [18:07:16 CET] <DonBastel> hi im currently trying to set up a ffmpeg listen http socket (-listen 1 http://<localip>:<port>) and it always streams as application/octet-stream. i tried to set the -mime_type to video/webm but that option only replies that it failed setting it to video/webm. is there any way to set the output of the http socket to video/webm for streaming? [18:17:54 CET] <relaxed> DonBastel: try -content_type video/webm [18:18:22 CET] <DonBastel> relaxed: already set no change [18:21:45 CET] <DonBastel> relaxed: the current cmd: ffmpeg -user-agent 42 -i http://<input url> -vcodec vp8 -vf scale=320:240 -acodec copy -f webm -content_type video/webm -listen 1 http://127.0.0.1:9000 [18:23:00 CET] <DonBastel> that works in terms of i see data flowing but the player cant play it because the mime-type is "application/octet-stream" [18:26:55 CET] <relaxed> which player [18:28:19 CET] <DonBastel> vlc [18:29:14 CET] <DonBastel> also ffmpeg on client side cant determine the streams. [18:43:18 CET] <relaxed> tcp:// work here but http:// only works with ffplay [18:43:32 CET] <furq> http works for me [18:49:30 CET] <furq> looks like the difference is on the player side though, i checked in firefox and the mime type is application/octet-stream [18:49:40 CET] <furq> but firefox and vlc play it back fine [18:49:58 CET] <wyatt8750> Anyone got advice on decoding QDMC audio? I've got a really really old .mov file that uses it [18:50:15 CET] <wyatt8750> I'm running linux but if it comes to a last resort I can boot into windows XP [18:50:33 CET] <wyatt8750> ffmpeg appears to not support QDMC decoding, that's why I ask [18:55:06 CET] <relaxed> wyatt8750: mplayer with quicktime binary codecs [18:55:43 CET] <wyatt8750> didn't know about that [18:55:48 CET] <wyatt8750> can that work in linux? [18:55:57 CET] <relaxed> yes [18:56:56 CET] <wyatt8750> where do I find/how do I install binary codecs then? [18:57:55 CET] <wyatt8750> oh, may have found them http://www.mplayerhq.hu/design7/dload.html#binary_codecs [19:04:52 CET] <DonBastel> ok thank you. [19:06:19 CET] <Logicgate> Hey guys, so I have this site which provides user a way to upload videos to the Vine app. The format of Vine's videos is either 480x480 and 720x720 (1:1 ratio). [19:06:56 CET] <wyatt8750> that worked, thanks a million. Dumped it to a .wav and now I can encode it in something slightly more modern :\ [19:07:17 CET] <relaxed> you're welcome [19:11:20 CET] <Logicgate> Which was my last message? [19:12:40 CET] <relaxed> vine takes square video [19:13:01 CET] <wyatt8750> I wonder if youtube's decoder/encoder can process QDMC. [19:13:15 CET] <wyatt8750> I feel like it is probably ffmpeg based, but I don't really know that [19:14:38 CET] <Logicgate> Now my problem is that users uploads all kind of different formats and sizes videos and I have to make a Vine compatible video for the upload sequence to be successful [19:14:43 CET] <Logicgate> One of the problems is that a lot of users complain about losing video quality / sound quality during the conversion process [19:14:55 CET] <Logicgate> here is what I'm using: ffmpeg -i <input> -ss 00:00:00 -t 00:00:06 -vf 480:480 -aspect 1:1 -c:v libx264 -preset veryslow -movflags faststart <output> [19:15:21 CET] <ChocolateArmpits> Logicgate: well you have no audio codec settings there for starters [19:15:41 CET] <ChocolateArmpits> and scaling quality can be increased by specifying scaling flags [19:15:53 CET] <llogan> you can stream copy the audio instead of re-encoding...as long as Vine will accept it [19:16:10 CET] <ChocolateArmpits> I would transcode just in case [19:16:13 CET] <Logicgate> llogan, I think that's a good idea since I don't want to preset the audio quality [19:16:17 CET] <Logicgate> More so just leave it as is [19:16:38 CET] <wyatt8750> logicgate: if you want to leave it as is then you aren't re-encoding it [19:16:54 CET] <wyatt8750> and they could use any number of impossibly obscure codecs [19:16:56 CET] <wyatt8750> (like QDMC) [19:17:18 CET] <llogan> what is the -ss 00:00:00 for? i see this often. [19:17:22 CET] <Logicgate> So would there be a way to transcode using the same bitrate as original audio track? [19:17:34 CET] <llogan> that's a fallacy [19:17:39 CET] <Logicgate> llogan, people upload videos of different length [19:17:48 CET] <Logicgate> I need to cut it down to 6 seconds, otherwise the upload fails [19:17:53 CET] <wyatt8750> logicgate: perhaps, but I would use a high standard, and 'same bitrate' means different things if you use different codecs [19:18:01 CET] <wyatt8750> some codecs handle lower bitrates better than others, for instance [19:18:07 CET] <llogan> -ss 00:00:00 does nothing [19:18:24 CET] <Logicgate> llogan, okay so -t 00:00:06 only? [19:18:31 CET] <llogan> yes [19:18:35 CET] <wyatt8750> the video starts at 0, so yeah, -t 00:00:06 is what you need. [19:18:38 CET] <Logicgate> duely noted [19:18:51 CET] <llogan> or just "-t 6" [19:18:52 CET] <ChocolateArmpits> well -t will only pass that many seconds, rather than stop at that timestamp [19:18:54 CET] <wyatt8750> true [19:19:08 CET] <Logicgate> Okay so the problem here I see is trying to handle so many different formats of video and audio codec and try to please everybody [19:19:15 CET] <ChocolateArmpits> Also you don't have bitrate settings [19:19:17 CET] <Logicgate> Am I doing the wrong thing by trying to please everyone? [19:19:17 CET] <ChocolateArmpits> specified [19:19:25 CET] <ChocolateArmpits> which results in probably low defaults [19:19:27 CET] <wyatt8750> I'd encode all videos to webm or x264 (if patents aren't an issue) [19:19:35 CET] <wyatt8750> use CRF [19:19:41 CET] <Logicgate> output needs to be MP4 [19:19:48 CET] <wyatt8750> alright then, x264 probably [19:20:03 CET] <wyatt8750> and audio in AAC (if patents aren't an issue) [19:20:16 CET] <Logicgate> ChocolateArmpits, how do you set bitrate properly? [19:20:32 CET] <kepstin> the problem with aac is that a lot of the encoders aren't very good... make sure you're not using vo-aacenc in particular [19:20:36 CET] <Logicgate> AAC should be fine wyatt8750, they're not picky about the audi track so much seems like [19:20:39 CET] <wyatt8750> I use fdk-aac [19:20:41 CET] <ChocolateArmpits> wyatt8750: both aac and h264 are free for encoding, only distributed encoders and decoders are monetized [19:21:04 CET] <llogan> the native FFmpeg AAC encoder has seen some big improvements recently [19:21:10 CET] <wyatt8750> though I don't think ffmpeg usually includes fdk-aac, right? [19:21:16 CET] <wyatt8750> I build my own :\ [19:21:16 CET] <kepstin> (the ffmpeg built-in encoder is probable the simplest option for aac encoding, just give it a bit more bitrate than you really need to workaround generation loss) [19:21:29 CET] <wyatt8750> 128kbps is my rule of thumb for aac [19:21:30 CET] <Logicgate> Okay, so I need to specify bitrate, use CRF and re-encode audio? [19:21:35 CET] <kepstin> wyatt8750: not normally, since it's a non-redistributable license combination. [19:21:46 CET] <wyatt8750> kepstin: ah, thought so. [19:22:22 CET] Action: wyatt8750 quietly pulls his ffmpeg windows/linux binaries with static linkned libfdk_aac off his server [19:22:29 CET] <Logicgate> lol [19:22:48 CET] <ChocolateArmpits> it's not like you're distributing it, are you? [19:22:56 CET] <Logicgate> Okay so could someone give me a sample of flags I should use? [19:23:03 CET] <wyatt8750> I wasn't advertising it but it was publicly available if you were to look hard enough [19:23:07 CET] <wyatt8750> my website was indexed :\ [19:23:11 CET] <wyatt8750> (google indexed) [19:23:31 CET] <Logicgate> I'd be willing to pay for a couple hours of consulting on this subject [19:23:36 CET] <Logicgate> If anyone wants to make a bit of money? [19:23:53 CET] <llogan> does vine squish the video like yours will? [19:24:04 CET] <llogan> (Ive never used vine) [19:24:08 CET] <Logicgate> llogan, well they use their own recorder on their app [19:24:18 CET] <wyatt8750> logicgate: what is 'vf 480:480'? [19:24:19 CET] <Logicgate> the output of their video format is either 720x720 or 480x480 [19:24:23 CET] <wyatt8750> do you mean to crop it? [19:24:23 CET] <ChocolateArmpits> Logicgate: -sws_flags lanczos or -sws_flags spline. former introduces some ringing when downscaling, spline is "softer" [19:24:29 CET] <Logicgate> wyatt8740, yeah I have to crop it [19:24:34 CET] <kepstin> Logicgate: the video stuff you have is mostly ok, just add a -crf flag to select a quality level (baybe around 20ish?). Then throw in "-c:a aac -b:a 140K" or so [19:24:34 CET] <wyatt8750> not resize? [19:25:04 CET] <ChocolateArmpits> kepstin: i'd suggest -crf 18, low bitrate needs higher crf most of the time [19:25:06 CET] <llogan> ChocolateArmpits: use the flags option in scale instead of sws_flags [19:25:08 CET] <ChocolateArmpits> I mean low resolution [19:25:28 CET] <Logicgate> wyatt8750 depends what the video input is right? [19:25:28 CET] <kepstin> I assume when vine gets the video they're probably gonna re-encode it on their end too, so spending some extra bits here is worth it [19:25:36 CET] <wyatt8750> logicgate: sure [19:25:40 CET] <ChocolateArmpits> llogan: does it matter really? I think it still goes through the scaling filter in the background [19:25:51 CET] <Logicgate> I have to offer the option to the user the either add black/white bars to fill the gap and keep aspect ratio [19:25:56 CET] <Logicgate> Or do full screen and resize. [19:26:03 CET] <wyatt8750> logicgate: good, I was going to suggest that [19:26:16 CET] <llogan> ChocolateArmpits: it is better because you can control which scaler flags apply to which scale if you have multiple [19:26:18 CET] <wyatt8750> if you already are doing that, then I'll not write that part of my filter chain in [19:26:33 CET] <ChocolateArmpits> llogan: well he's clearly not doing multiple concurrent outputs [19:26:33 CET] <llogan> ...and i prefer to have all filtering related stuff in the filtergraph [19:26:34 CET] <Logicgate> I have that part handled wyatt8750! Thanks [19:27:01 CET] <llogan> ChocolateArmpits: clearly, as you put it, but it makes more sense regardless [19:27:25 CET] <ChocolateArmpits> llogan: let's say the sws-flags is more straight forward in this scenario [19:27:34 CET] <llogan> how so? [19:28:28 CET] <ChocolateArmpits> llogan: less typing and generally less time spent getting to know the topic of filtering before getting results [19:28:38 CET] <wyatt8750> what AAC codecs are includable with distributable ffmpeg? [19:28:52 CET] <wyatt8750> logicgate: I'm assuming you aren't compiling your own ffmpeg? [19:29:09 CET] <Logicgate> wyatt8750 I'm using the precompiled binaries [19:29:12 CET] <Logicgate> from relaxed [19:29:26 CET] <llogan> ChocolateArmpits: it is actually more typing [19:29:43 CET] <ChocolateArmpits> llogan: well okay but the second part is true [19:30:03 CET] <wyatt8750> relaxed's? or the official ffmpeg binaries? [19:30:11 CET] <Logicgate> relaxed's [19:30:14 CET] <wyatt8750> I don't know where relaxed keeps his [19:30:16 CET] <RaMcHiP-TMP> Hello all! I just installed octoprint onto a raspberry pi and it has a timelapse video feed using ffmepg. When I preview or record I get extreme distortion in color and picture. Here is an example of a time lapse https://www.dropbox.com/s/cn5xd02loihqdvp/svapoDockingUSBFixed_201512290916… and my lsusb gives me Bus 001 Device 007: ID 0c45:62f1 Microdia [19:30:20 CET] <ChocolateArmpits> I don't think there are any "official binaries" [19:30:30 CET] <Logicgate> http://johnvansickle.com/ffmpeg/ [19:30:36 CET] <Logicgate> that's relaxed's binaries [19:30:44 CET] <relaxed> not official, I just do it to be helpful [19:31:54 CET] <Logicgate> I like your binaries, they've worked well for me :D [19:32:18 CET] <wyatt8750> I was being helpful supplying my binaries but I apparently was violating licenses :p [19:33:01 CET] <wyatt8750> so, what AAC codecs ARE included in your builds relaxed? [19:33:20 CET] <relaxed> yes, but not libfdk_aac [19:33:27 CET] <wyatt8750> which ones/ [19:33:42 CET] <relaxed> native and libvo-aacenc [19:33:46 CET] <wyatt8750> alright [19:33:54 CET] <wyatt8750> vo is the better of the two right now, right? [19:33:58 CET] <llogan> no [19:33:59 CET] <wyatt8750> I normally use fdk [19:34:01 CET] <wyatt8750> :p [19:34:12 CET] <wyatt8750> oh, native's that good now? [19:34:13 CET] <wyatt8750> sweet [19:34:13 CET] <relaxed> I should drop libvo-aacenc [19:35:08 CET] <llogan> also it doesn't support more than 2 channels [19:38:00 CET] <relaxed> bill gates said 2 channels was all anyone would ever need [19:41:16 CET] <ChocolateArmpits> hey if we have only 2 ears why would we need more than 2 channels? [19:41:52 CET] <furq> so i can spend more money on crystal-plated interconnects [19:43:28 CET] Action: llogan uses clothes hangers [19:47:25 CET] <furq> did you at least paint arrows on them [21:01:12 CET] <Franciman> Hello [21:01:33 CET] <Franciman> I am decoding a list of AVPackets [21:01:39 CET] <Franciman> with av_decode_audio4 [21:01:47 CET] <Franciman> avcodec_* [21:01:52 CET] <Franciman> now when I have to flush [21:02:12 CET] <Franciman> can I use a dummy AVPacket not allocated with av_init_packet [21:02:18 CET] <Franciman> or do I need to allocate it? [21:02:51 CET] <Franciman> (provided that I will do packet.size = 0; and packet.data = NULL; ) [21:12:12 CET] <micechal> what is the convention for ffmpeg version numbers? ie. why is n2.6.6 newer than n2.7.4 and n2.8.4? [21:12:28 CET] <fritsch> it was tagged later [21:12:36 CET] <fritsch> .6 of 2.6 was tagged later [21:13:02 CET] <micechal> so there are like three main branches? [21:13:12 CET] <micechal> 2.6, 2.7 and 2.8? [21:13:14 CET] <fritsch> three versions that are still supported [21:13:18 CET] <fritsch> and receive dot updates [21:13:31 CET] <DHE> why is windows XP up to service pack 3 but windows 7 only has service pack 1? [21:13:35 CET] <micechal> I see. thanks [21:35:56 CET] <aep> hi. any idea how i can get the latency down when using isml? [21:36:03 CET] <aep> this is over local network [21:36:58 CET] <aep> i mean, i'm testing with no network for now: [21:37:00 CET] <aep> ffmpeg -re -f v4l2 -i /dev/video0 -g 10 -c:v libx264 -crf 22 -movflags faststart -c:v libx264 -f ismv -movflags isml+frag_keyframe - | vlc - [21:37:27 CET] <aep> -g 10 is probably terrible, but without i get like 30 seconds lag [21:41:09 CET] <dystopia_> can you encode as fast as realtime? [21:41:21 CET] <aep> uuh [21:41:41 CET] <aep> i dont see why not. but i guess i can recduce the framerate to try? [21:41:58 CET] <dystopia_> your doing -re to limit yourself to encode at input frame rate [21:42:16 CET] <dystopia_> so if it was 25fps you are limiting encoding to 25fps [21:42:21 CET] <aep> yeah thats fine [21:42:27 CET] <aep> by lag i dont mean framerate [21:42:37 CET] <dystopia_> but if your hardware can't actually encode the video at 25fps you will get lag [21:42:38 CET] <dystopia_> ok [21:42:40 CET] <aep> it takes 30 seconds before the first frame is shown from 30 seconds ago [21:42:49 CET] <aep> and then the framerate is consistent [21:42:50 CET] <dystopia_> hmm [21:42:53 CET] <aep> just 30 seconds behind [21:42:59 CET] <dystopia_> yeah i see what you mean [21:43:24 CET] <aep> also mplayer stops doing anything after another 10 seconds or so [21:43:33 CET] <aep> i suspect about when another keyframe comes [21:43:55 CET] <aep> since the time it takes to do both decreases witha higher -g [21:46:42 CET] <aep> its also eating all of my 4 cores. i wonder why mjpeg is the only format that doesnt do that :/ [21:55:16 CET] <ChocolateArmpits> aep: processing requirements can be adjusted through the preset parameter [21:55:21 CET] <ChocolateArmpits> for x264 [21:55:30 CET] <ChocolateArmpits> By default it's running at medium [21:55:38 CET] <aep> yeh, its already pretty low quality by default :( [21:55:39 CET] <aep> oh well [21:55:43 CET] <aep> i'd be glad if it works at all [21:55:57 CET] <furq> preset shouldn't affect quality (too much) [21:56:00 CET] <ChocolateArmpits> As for reducing latency, use -analyzeduration as an input option [21:56:10 CET] <furq> you can try -tune zerolatency but i doubt that'll help much [21:57:17 CET] <aep> hm not sure where -analyzeduration goes [21:57:25 CET] <ChocolateArmpits> something like -analyzeduration 500000 will tell ffmpeg only to scan the input for 0.5 seconds [21:57:28 CET] <ChocolateArmpits> before the -i [21:57:31 CET] <furq> "as an input option" = before -i [21:57:31 CET] <aep> ffmpeg -r 10 -f v4l2 -analyzeduration -i /dev/video0 -g 10 -c:v libx264 -crf 22 -c:v libx264 -f mp4 -movflags frag_keyframe+faststart - [21:57:35 CET] <aep> ah duh [21:57:39 CET] <aep> with a number :D [21:57:59 CET] <furq> https://www.ffmpeg.org/ffmpeg-formats.html#mov_002c-mp4_002c-ismv [21:58:04 CET] <furq> at a guess, you want to tune some of those options [21:58:09 CET] <aep> same thing [21:58:20 CET] <furq> i assume the delay is down to the fragment size [21:58:24 CET] <aep> yeah [21:58:25 CET] <furq> i've never touched ismv before though [21:58:36 CET] <aep> pretty sure its related to whatever -g does to fragments [21:58:51 CET] <furq> oh you're already using frag_keyframe [21:59:00 CET] <ChocolateArmpits> you shouldn't need the faststart flag [21:59:13 CET] <aep> aye. was just a random try [21:59:27 CET] <ChocolateArmpits> as the gops are already fragmented [21:59:32 CET] <furq> i'd have thought faststart would make it worse [21:59:49 CET] <ChocolateArmpits> it should have no effect, or actually any use at all with smoothstream [22:00:14 CET] <ChocolateArmpits> faststart is for progressive download sources, while smoothstream is adaptive [22:00:35 CET] <aep> makes sense [22:01:00 CET] <ChocolateArmpits> aep: you should also try configuring the analyzeduration for vlc however I don't know the command for that there [22:01:04 CET] <aep> i noticed when i just let it print to stdout, it also takes several seconds before it does anything [22:01:11 CET] <ChocolateArmpits> you can try piping it to ffplay instead and using the same command there [22:01:14 CET] <aep> aye [22:01:28 CET] <furq> is it actually encoding at 10fps [22:01:35 CET] <aep> ffplay dies [22:01:39 CET] <aep> [mov,mp4,m4a,3gp,3g2,mj2 @ 0x7fb378000920] stream 0, offset 0x3e563: partial file [22:01:47 CET] <aep> prints that in a loop at 100% cpu [22:01:56 CET] <ChocolateArmpits> ok [22:04:59 CET] <aep> ocassionally it works. analyzeduration has no effect tho [22:08:02 CET] <aep> its the same with webm [22:08:08 CET] <aep> maybe there's some buffer? [22:11:10 CET] <aep> hm webm only has about 5 seconds delay [23:54:13 CET] <nani____> I want to extract from two "mostly" in sync video files the audio and video, such that video 1 has the audio from video 0. These two files might go out of sync by different amounts througout their duration. The video is identical (but out of sync), encoding quality aside, the audio is completely different. [23:54:33 CET] <nani____> Is there an easy ffmpeg one liner for this or do I have to hack something together in python? [00:00:00 CET] --- Wed Dec 30 2015

1 0

ffmpeg-devel.log.20151228
by burek 29 Dec '15

29 Dec '15

[03:12:18 CET] <kierank> why is there no function to free a VLC table? [03:13:45 CET] <kierank> oh there is I am blind [03:18:26 CET] <cone-197> ffmpeg 03Mats Peterson 07master:57631f1851ef: avformat: factor ff_get_qtpalette() out of mov.c [03:18:26 CET] <cone-197> ffmpeg 03Mats Peterson 07master:797360384373: avformat/matroskadec: palettized QuickTime video in Matroska [03:31:01 CET] <J_Darnley> Wow. Somehow gcc has done these intrinsics much better on Windows than on Linux. (It is more likely to be 4.9 vs 5.3) [03:45:30 CET] <llogan> wasn't expecting any of that "palettized" stuff would actually get pushed [04:00:24 CET] <RiCON> what an adventure [04:08:27 CET] <Daemon404> indeed [04:19:45 CET] <Compn> ancient quicktime is arcane [04:56:19 CET] <tmm1> can someone help me understand the pts handling in the w3fdif and yadif filters? [04:57:34 CET] <tmm1> the second frame emitted has out->pts = cur_pts + next_pts [05:02:05 CET] <tmm1> having a hard time grokking why that makes sense [05:53:33 CET] <tmm1> oh, it works because all the other frames have their pts doubled [10:18:20 CET] <cone-204> ffmpeg 03Michael Niedermayer 07master:ca9e3cb3ce7a: avformat/qtpalette: Move ff_get_qtpalette() doxy to header [10:18:20 CET] <cone-204> ffmpeg 03Michael Niedermayer 07master:3c6781b48acd: Revert "ffplay: Fix auto insertion point of rotation filter" [10:24:50 CET] <cone-204> ffmpeg 03Hendrik Leppkes 07master:50401f5fb7d7: avcodec: properly check pkt_timebase for validity [12:45:46 CET] <cone-204> ffmpeg 03Paul B Mahol 07master:b841fe002a2b: avfilter/af_silenceremove: lower number of operations in for loop [12:45:47 CET] <cone-204> ffmpeg 03Paul B Mahol 07master:47aaebd63e40: avfilter/af_silenceremove: make size of window user configurable [12:45:48 CET] <cone-204> ffmpeg 03Paul B Mahol 07master:397774603690: doc/filters: add one more silenceremove example [14:37:24 CET] <durandal_170> tmm1: still need help? [16:11:46 CET] <cone-204> ffmpeg 03Rodger Combs 07master:4caa3e1c6cf4: lavf: add API to apply a list of bsfs to a packet [16:11:47 CET] <cone-204> ffmpeg 03Rodger Combs 07master:a5fd3a1a2bd2: ffmpeg: use lavf API for applying bitstream filters [16:11:48 CET] <cone-204> ffmpeg 03Rodger Combs 07master:7a161b74ad13: lavf/tee: use lavf API for applying bitstream filters [16:11:49 CET] <cone-204> ffmpeg 03Rodger Combs 07master:1f9139b07b8a: lavf: add automatic bitstream filtering; bump version [16:11:50 CET] <cone-204> ffmpeg 03Rodger Combs 07master:822e80fde39f: lavf: add internal API to append a bsf to a stream's list [16:11:51 CET] <cone-204> ffmpeg 03Rodger Combs 07master:b287d7ea17f4: lavf/matroskaenc: add automatic bitstream filtering [16:11:52 CET] <cone-204> ffmpeg 03Rodger Combs 07master:1b5bd4051d1e: lavf/mpegtsenc: add automatic bitstream filtering [16:13:43 CET] <ubitux> @_@ [16:30:23 CET] <nevcairiel> cant we just ban that guy already [16:31:09 CET] <durandal_170> who? [16:32:34 CET] <JEEB> the matias guy? [16:32:40 CET] <ubitux> :D [16:35:25 CET] <rcombs> no, me [16:36:33 CET] <nevcairiel> yeah, how dare you make the life of ffmpeg cli users easier [16:36:42 CET] <nevcairiel> they deserve to suffer! [16:39:42 CET] <saste> ah cool, that got in, kudos rcombs :-) [16:39:46 CET] <rcombs> \o/ [16:40:28 CET] <atomnuker> holy shit that's a huge thread [16:41:00 CET] <ubitux> it's mostly the work of a single man derping at himself [16:41:25 CET] <ubitux> the guy sounds really crazy [16:41:48 CET] <durandal_170> atomnuker: still gonna do showspectrumpic? [16:42:20 CET] <atomnuker> that's a good idea actually, was just wondering what to do [16:48:45 CET] <BBB> rcombs: \o/ [16:49:00 CET] <BBB> rcombs: now we need to automatically pack vp9 superframes in the encoders (ivf, mkv) [16:49:18 CET] <rcombs> BBB: I've no idea how those work [16:49:45 CET] <nevcairiel> BBB just needs to write a bsf to do it [16:49:46 CET] <BBB> its the inverse of vp9_parser.c [16:49:51 CET] <BBB> bt Im lazy [16:49:59 CET] <BBB> I want someone else to do it [16:50:20 CET] <rcombs> in the encoder or in the muxer? [16:50:28 CET] <BBB> you have to think of future generations, theyll notice that any different guys contributed to vp9, so vp9 must have been really important [16:50:40 CET] <BBB> I dont think rcombs has vp9 patches yet, so Im trying to get him to write one :-p [16:50:42 CET] <BBB> rcombs: muxer [16:50:47 CET] <nevcairiel> rcombs: in between :D [16:50:57 CET] <BBB> its for the transmuxing case, where no encoder exists [16:51:09 CET] <BBB> vp9_parser will split but nothing will join them back together [16:51:12 CET] <BBB> which is kind of annoying [16:51:26 CET] <BBB> like -c:v copy [16:51:28 CET] <rcombs> if it's just "if it's VP9, put it through this BSF", ^^that can handle it [16:51:41 CET] <rcombs> there just has to actually be a BSF for it [16:51:45 CET] <nevcairiel> its just that this particular BSF does not exist yet :) [16:51:55 CET] <BBB> oh I see you guys want me to write the bsf [16:51:58 CET] <BBB> grmbl [16:52:11 CET] <rcombs> I know jack shit about VPX [16:52:18 CET] <nevcairiel> you actually know what needs to happen in there! [16:52:36 CET] <BBB> legit argument [16:52:38 CET] <BBB> but! [16:52:45 CET] <BBB> I could teach you what needs to happen in there [16:53:06 CET] <durandal_170> everything should be usable through lavfi [16:53:22 CET] <nevcairiel> i already know way more about vp9 than I want to [16:53:46 CET] <BBB> :( [16:54:33 CET] <nevcairiel> but it can be hardware decoded now, so thats a bonus [16:54:56 CET] <cone-204> ffmpeg 03Joel Holdsworth 07master:b9c46b5242bf: avformat/http: Documented http_proxy option [16:55:02 CET] <rcombs> my users still don't care about VPX [16:55:16 CET] <wm4> but they should [16:55:45 CET] <BBB> doesnt youtube use it or so? [16:56:07 CET] <rcombs> sure [16:56:30 CET] <rcombs> but I don't think I ever have to remux it [16:56:32 CET] <nevcairiel> vp9 probably is quite a substantial part of streamed internet video, just by being on youtube [16:56:46 CET] <ubitux> mmh, isn't youtube-dl remuxing the vp9 parts? [16:56:49 CET] <wm4> isn't the problem that there's still no good vp9 encoder? [16:57:04 CET] <nevcairiel> its not that terrible anymore [16:58:30 CET] <atomnuker> 1.4 is über slow and its debian maintainer has gone AWOL so we're stuck with it [16:58:44 CET] <nevcairiel> you choose to use debian [16:58:45 CET] <nevcairiel> so... [16:58:48 CET] <wm4> which debian maintainer? [16:59:24 CET] <atomnuker> wm4: Sebastian Dröge, https://packages.qa.debian.org/libv/libvpx.html [16:59:58 CET] <wm4> don't know how to read this, but doesn't look very AWOL [17:00:23 CET] <nevcairiel> apparently 1.5 was accepted into unstable just today [17:02:09 CET] <atomnuker> awesome, maybe now I can encode 5 seconds @24fps in less than 3 hours [17:40:58 CET] <kierank> michaelni: can you explain this line: https://github.com/FFmpeg/FFmpeg/blob/master/libavcodec/snow.c#L562 [17:41:08 CET] <kierank> why do you add stride >> 1 to buf [17:44:56 CET] Action: Daemon404 wonders why kierank is reading snow [17:45:08 CET] <kierank> gotta make cfhd do the same thing [17:45:28 CET] <kierank> not 100% sure how to store the coefficients [17:47:28 CET] <Daemon404> o [17:57:12 CET] <michaelni> kierank, dont remember exactly but it must have been to have buf at the rigt spot [18:03:53 CET] <atomnuker> anyone likes reading 6000+ line patches on the ML? [18:04:20 CET] <ubitux> TLDR LGTM [18:05:12 CET] <wm4> "probably ok" [18:08:36 CET] <smarter> Daala decoder incoming? [18:10:39 CET] <kierank> what's a bit odd about cfhd is that there are extra coefficients [18:13:02 CET] <Daemon404> atomnuker, the bitstream format is frozen now isnt it? [18:13:04 CET] <Daemon404> or did they give that up [18:14:23 CET] <smarter> Daala? We're pretty far from freezing the bitstream [18:14:27 CET] <atomnuker> not frozen by a longshot [18:15:02 CET] <smarter> Yeah there was a plan for freezing at end of 2015 but that was before the IETF standardization effort, AOM, etc [18:16:38 CET] <Daemon404> smarter, so the '2015 freeze' that was sung for 2 years was given up on [18:16:43 CET] Action: Daemon404 predicted correctly [18:16:58 CET] <smarter> congrats I guess :) [18:23:32 CET] <Daemon404> ubitux, do we still need arm stuff for FATE? [18:23:39 CET] <Daemon404> i have a a9 that has sat idle for 2 years [18:23:47 CET] <Daemon404> either itll be a fate cli or ill bin it [18:23:52 CET] <ubitux> no idea [18:23:59 CET] <ubitux> i think we actually need more tests [18:24:02 CET] <ubitux> but well :p [18:24:07 CET] <Daemon404> i could add a native clang/arm [18:28:23 CET] <Daemon404> oh good im 2 versions of fedora out of date on this arm box [18:28:29 CET] <Daemon404> see you in 2 weeks when it finishes updating [18:30:56 CET] <cone-204> ffmpeg 03Paul B Mahol 07master:67771ac4b894: avfilter/avf_showspectrum: add rscroll sliding mode [18:39:26 CET] <rcombs> so Comcast is switching from MPEG-2 to AVC, if anyone cares [18:39:33 CET] <rcombs> for HD channels [18:40:48 CET] <Daemon404> welcome to 10 years ago? [18:40:56 CET] <Daemon404> i wonder how bad their encoder will be [18:41:18 CET] <kierank> it's an ongoing process [18:42:24 CET] Action: TD-Linux can't wait for his AVC reencodes of MPEG-2 streams [18:43:05 CET] <Daemon404> that descriebs mostly any scene release [19:04:51 CET] <cone-204> ffmpeg 03Paul B Mahol 07master:45b3e6e04e8e: avfilter: move window function generation into separate file [19:04:52 CET] <cone-204> ffmpeg 03Paul B Mahol 07master:f88546b426af: avfilter/avf_showspectrum: use ff_generate_window_func [19:09:28 CET] <llogan> rcombs: do the docs need to be updated to mention autobsf? [19:10:03 CET] <rcombs> probably something somewhere [19:10:40 CET] <llogan> doc/bitstream_filters.texi [19:35:08 CET] <BBB> atomnuker: oo a daala decoder [19:35:51 CET] <BBB> atomnuker: so why is it blockier? does that mean you didnt implement loopfilter yet? [19:36:00 CET] <BBB> and pframe support sounds missing also? [19:36:09 CET] <nevcairiel> he did mention all that =p [19:36:15 CET] <nevcairiel> missing filter, only I frames [19:37:08 CET] <Daemon404> can the daala encoder even make p frames [19:38:00 CET] <BBB> who needs p frames anyway [19:38:10 CET] <nevcairiel> go straight to B! [19:38:39 CET] <rcombs> clearly we must invent new types of frames [19:39:20 CET] <rcombs> anyone with ideas for how T-frames should work should send their suggestions to someone else [19:39:44 CET] <atomnuker> BBB: there is a loop filter, but there is no deringing filter [19:58:15 CET] <llogan> looks like gstreamer does regression tests too https://ci.gstreamer.net/ [20:39:33 CET] <cone-204> ffmpeg 03Paul B Mahol 07master:4020787b5bbd: avfilter/avf_showspectrum: make colors for log scale more user friendly [21:07:56 CET] <smarter> Daemon404: the daala encoder has had p frames for a long time, and got b frames recently [21:14:15 CET] <BtbN> What's even the state of Daala? I thought it was still unfinished and changing a lot [21:14:42 CET] <smarter> yes [21:16:59 CET] <smarter> atomnuker: you wrote "ratser" instead of "raster" in a bunch of places [21:21:05 CET] <cone-204> ffmpeg 03James Almer 07master:6e243d17e911: x86/vf_stereo3d: optimize register usage [21:21:06 CET] <cone-204> ffmpeg 03James Almer 07master:1817643d4f50: x86/vf_stereo3d: make ff_anaglyph_sse4 work on x86_32 [21:28:03 CET] <Daemon404> smarter, o ok [21:55:43 CET] <atomnuker> smarter: fixed, thanks [22:01:07 CET] <atomnuker> libvpx 1.5.0: "This release improves upon the VP9 encoder and speeds up the encoding and decoding processes." [22:01:12 CET] <atomnuker> "frame= 195 fps=0.0 q=0.0 size= 208kB time=00:00:06.55 bitrate= 260.2kbits/s speed=0.000856x" [22:01:33 CET] <atomnuker> this has been going on for 3 hours, this is just as slow as 1.4.0 [22:01:56 CET] <nevcairiel> sounds like your system is a RPi or you are using insane settings [22:02:06 CET] <nevcairiel> i can get decent speeds out of 1.5 now [22:02:25 CET] <atomnuker> -b:a 1M -quality best sounds reasonable [22:02:36 CET] <atomnuker> s/b:a/b:v [22:03:06 CET] <nevcairiel> they dont recommend using best, fwiw [22:03:11 CET] <nevcairiel> its like placebo [22:03:19 CET] <nevcairiel> use good with some cpu-used parameter [22:04:20 CET] <smarter> never use best [22:04:22 CET] <nevcairiel> ie, good with --cpu-used 0 should be much faster than best at a minimal quality loss [22:04:25 CET] <smarter> there's no point [22:04:32 CET] <smarter> exactly [22:04:50 CET] <atomnuker> turns out I hadn't even set -b:v [22:05:06 CET] <atomnuker> I get 0.2x realtime now [22:05:10 CET] <atomnuker> pretty decent [22:05:38 CET] <atomnuker> still only 1 thread though [22:05:57 CET] <nevcairiel> didnt they introduce some sort of slice threading [22:18:59 CET] <Daemon404> i keep hoping theyll fix RC [22:19:08 CET] <Daemon404> but every member of teh vp9 team i talked to said its a non-issue [22:19:11 CET] <Daemon404> not even on their TODO [22:31:43 CET] <llogan> we could use a vp9 encoding guide on wiki [22:31:55 CET] <c_14> llogan: there is one [22:32:17 CET] <c_14> Might need updates though [22:32:27 CET] <c_14> https://trac.ffmpeg.org/wiki/Encode/VP9 [22:32:50 CET] <llogan> oh, duh [22:34:03 CET] <llogan> is that an orphaned page? [22:35:02 CET] <llogan> what use is this? https://trac.ffmpeg.org/wiki/NullRenderer [22:35:59 CET] <llogan> yet another of Roger's Ramblings" [22:36:56 CET] <J_Darnley> I was wondering whether you mean "what use is this page" or "what use is -f null" :) [22:37:13 CET] <llogan> i use null often [22:49:38 CET] <Daemon404> everyone does [22:58:38 CET] <beastd> -f null is quite useful :) [22:59:16 CET] <beastd> though i could use -f lol from time to time ;-) [23:13:20 CET] <atomnuker> hm, the mp3float decoder is 2x (500xrealtime) the speed of mp3 (~250xrealtime) [23:13:59 CET] <kierank> probably simd [23:15:28 CET] <wm4> I always wondered what was the difference between them [23:17:35 CET] <atomnuker> I think it'd be nice to have mp3float made default on x86 [23:17:46 CET] <atomnuker> since currently mp3 is the default [23:18:08 CET] <BBB> llogan: http://wiki.webmproject.org/ffmpeg/vp9-encoding-guide [23:18:17 CET] <BBB> llogan: (its all ffmpeg stuff) [23:29:22 CET] <llogan> BBB: thanks [23:29:35 CET] <llogan> "FFMpeg" [23:36:12 CET] <durandal_170> atomnuker: you really writting showspectrumpic? [23:37:06 CET] <durandal_170> ubitux: showspectrum is missing overlap for big fft size [23:41:08 CET] <atomnuker> why, do you want to do it? [23:43:38 CET] <atomnuker> if you do, then go ahead, it would probably take me much longer [00:00:00 CET] --- Tue Dec 29 2015

1 0

ffmpeg.log.20151228
by burek 29 Dec '15

29 Dec '15

[00:03:29 CET] <furq> angeldorado: are you running ubuntu 14.04 or 15.10 [00:04:24 CET] <furq> if you're going to be using ffmpeg/libavxxx then it's better to use 15.10 [00:06:18 CET] <angeldorado> ubuntu 15.10? [00:06:49 CET] <furq> whichever ubuntu derivative you want [00:07:15 CET] <furq> for various convoluted reasons ffmpeg isn't in 14.04 or 14.10 [00:07:22 CET] <angeldorado> oh I see [00:07:41 CET] <furq> if you're using 15.10 then you can install it from apt and save yourself a lot of bother [00:07:54 CET] <angeldorado> well, good news is ubuntu studio 15.10 [00:07:59 CET] <furq> well yeah [00:08:01 CET] <furq> apt-get install ffmpeg [00:08:12 CET] <furq> unless you absolutely need ffmpeg 2.8 [00:08:21 CET] <angeldorado> well I have to install it, I have the previous version. but it does exist [00:08:28 CET] <angeldorado> great stuff. what's the ppa? [00:08:45 CET] <furq> i don't know of a ppa for 2.8 [00:08:49 CET] <furq> 2.7 is in the main repo [00:09:39 CET] <furq> don't take that to mean that there isn't a ppa, i don't use ubuntu [00:09:59 CET] <angeldorado> well I've been looking for hours [00:10:11 CET] <angeldorado> and its not like any other thing ive installed in linux [00:10:13 CET] <furq> do you specifically need 2.8 [00:10:46 CET] <angeldorado> im not sure exactly which version I need. I just know that almost every forum post I find has someone mentioning ffmpeg [00:10:56 CET] <angeldorado> trying to solve some video/audio render problems im having [00:11:13 CET] <furq> you might need to add universe to your /etc/apt/sources.list [00:11:54 CET] <angeldorado> how do I do that? [00:12:02 CET] <angeldorado> ill look google it [00:22:34 CET] <rjp421> http://cdn.c3voc.de/s1_native_slides.webm [02:53:58 CET] <salviadud> Hi [02:54:04 CET] <salviadud> I want to share what I'm doing [02:54:12 CET] <salviadud> which I feel is kind of odd [02:54:25 CET] <salviadud> I got this compressed mp4 video in h264 format [02:54:45 CET] <salviadud> The ratio is wrong, it should be 1080p [02:54:54 CET] <salviadud> but it's 720x486 [02:55:04 CET] <salviadud> so, I'm stretching it and changing the aspect ratio [02:56:20 CET] <salviadud> the command goes something like this: ffmpeg -i input.mp4 -s 1920x1080 -aspect 16:9 -b:v 3000k -vcodec libx264 -an out.mp4 [02:56:31 CET] <salviadud> now, it's doing it [02:56:35 CET] <furq> you don't need to reencode to change the aspect ratio [02:56:46 CET] <salviadud> but, at certain intervals [02:56:50 CET] <salviadud> it asks me for a target [02:57:00 CET] <salviadud> it's weird [02:57:02 CET] <furq> you do if you want the output to be 1080p but that's pointless if the source is 480p [02:57:25 CET] <salviadud> oh, but this is no ordinary video [02:57:38 CET] <salviadud> It has something ciphered in it... [02:57:40 CET] <pzich> it's extraordinary [02:58:05 CET] <salviadud> It has to do with a native algorithm inside handbrake [02:58:14 CET] <furq> it probably won't have something ciphered in it once you've reencoded it [02:58:28 CET] <salviadud> why not furq? [02:58:46 CET] <furq> i can't see how that would work [02:59:09 CET] <salviadud> well, I might be wrong, and there's nothing there [02:59:18 CET] <salviadud> but it is a misterious video, nonetheless [02:59:43 CET] <salviadud> Now, my only doubt is why ffmpeg asks for a command [02:59:46 CET] <salviadud> every now and then [02:59:52 CET] <salviadud> it's version 2.3 [02:59:53 CET] <furq> anyway if you just want to change the AR then use -aspect 16:9 -c copy [03:00:23 CET] <furq> and if you really think scaling it to 1080p will help anything then at least use -crf instead of -b:v [03:00:33 CET] <salviadud> what's -crf ? [03:00:50 CET] <furq> https://trac.ffmpeg.org/wiki/Encode/H.264#crf [03:01:51 CET] <salviadud> ok, I get it. thanx for the crf [03:02:02 CET] <salviadud> But still, is it common for ffmpeg to pause [03:02:09 CET] <JodaZ> salviadud, what do you think stretching a 720x486 to 1080p will acomplish? you are upscaling... [03:02:09 CET] <furq> it's not something i've ever seen [03:02:11 CET] <salviadud> and ask for arguments, while doing a re-encode [03:02:57 CET] <salviadud> another thing... [03:04:30 CET] <salviadud> The output goes something like this: I get a constant line that repeats itself, in blue cyan, libx264@[somememoryaddress] and the usual frame by frame line [03:04:48 CET] <salviadud> Is that common when reencoding h264? [03:05:09 CET] <furq> that sounds like a warning which is being overwritten by the progressbar [03:05:17 CET] <furq> it's probably something you can safely ignore [03:05:33 CET] <furq> although maybe not if this is a mysterious james bond diagnosis murder video [03:05:41 CET] <salviadud> Its xxx [03:05:44 CET] <salviadud> haha [03:05:52 CET] <salviadud> the original file is all wrong [03:05:56 CET] <salviadud> straight from source [03:06:01 CET] <salviadud> the dvd looks awful [03:06:07 CET] <salviadud> It's like hacker xxx [03:06:14 CET] <salviadud> you need to reencode it to see the original [03:06:47 CET] <furq> i don't see how scaling would make a difference [03:07:09 CET] <salviadud> It's an experiment of mine. [03:07:22 CET] <salviadud> I lost ffmpeg on an old computer that has a particular processor [03:07:31 CET] <salviadud> I'm running a gentoo live dvd on it [03:07:43 CET] <salviadud> Had to compile ffmpeg 2.3 to get it working [03:07:57 CET] <salviadud> I disabled the mmx flag [03:08:13 CET] <salviadud> It's doing the most basic of processor optimizations. [03:08:27 CET] <salviadud> That I belive is the key to decipher the vid [03:08:51 CET] <salviadud> Soooo, I'll find out in a couple of days [03:08:56 CET] <salviadud> since it's taking forever. [03:09:07 CET] <salviadud> I will report back to you guys [03:10:32 CET] <pzich> maybe use -t # to see if the first bit is looking right? [03:33:17 CET] <Li> How to cut 1 minute from the begining of all video files in a given folder? I know for a single file it goes like ffmpeg -i input.filename -ss time output.filename [03:33:53 CET] <pzich> you just run that for each file? [03:33:53 CET] <c_14> for file in *; do ffmpeg -i <blargh>; done [03:34:00 CET] <c_14> or any other scripting language [03:52:43 CET] <Li> c_14: what is <blargh> how to replace it? [03:53:13 CET] <c_14> -i "$file" -ss time "output$file" [06:51:41 CET] <thomedy> hello [06:52:29 CET] <thomedy> im not entirely sure how to get this very specific encoding [06:52:32 CET] <thomedy> plese help [06:52:44 CET] <thomedy> h264 main level 3 avc1 and aac [06:55:18 CET] <pinPoint> ffmpeg -i input -c:v libx264 -profile:v baseline -level 3.0 -b:v 1000k output.mp4 [06:55:40 CET] <pinPoint> https://trac.ffmpeg.org/wiki/Encode/H.264 [06:58:23 CET] <rjp421> pick a width/height/fps within the level/profile youre encoding for. baseline to play on real old devices or main otherwise [07:02:14 CET] <rjp421> https://sonnati.wordpress.com/2011/08/30/ffmpeg-%E2%80%93-the-swiss-army-kn… [07:02:19 CET] <thomedy> thats funny i thought it would have been alot harder but i found it in my serach finally [07:03:57 CET] <thomedy> now ijust have to figure out how to do the width height and fps [07:04:49 CET] <pinPoint> https://trac.ffmpeg.org/wiki/Scaling%20(resizing)%20with%20ffmpeg [07:04:58 CET] <pinPoint> thomedy: ^^^ [07:05:26 CET] <pinPoint> hell, https://trac.ffmpeg.org/wiki [07:06:39 CET] <thomedy> oh my god i did it [07:07:22 CET] <thomedy> im less concerned about the ffmpeg conversion but i needed it to accomplish my goal [07:07:27 CET] <thomedy> and it just happened [07:07:39 CET] <pinPoint> which is [07:09:31 CET] <thomedy> i am building a website where i can scroll through the video in forwrd or reverse based on the window position [07:09:43 CET] <thomedy> its not all that technical honestly [07:09:49 CET] <thomedy> but i really liked the desing [07:09:54 CET] <thomedy> design [07:10:28 CET] <thomedy> now the more that i think about it ... it may have been less about the encode and more about the frame size [07:10:50 CET] <thomedy> whatever i have it now [07:10:55 CET] <thomedy> thank you [07:11:20 CET] <pinPoint> ya [07:26:32 CET] <sati> hi anybody there which can help me [07:27:08 CET] <sati> I want use ffmpeg for video recording from analog camera [07:28:00 CET] <sati> I want know for more than 8 recording (h.264) which hardware is compatible for ffmpeg [07:36:36 CET] <rjp421> sati v4l2-ctl -d /dev/video0 --list-formats-ext --list-fields [07:36:57 CET] <rjp421> or something along those lines [07:40:04 CET] <sati> rjp421 : no Actually our pc i5 intel have limitation , that processor reach 100% usages on 3 h.264 video encoding [07:40:14 CET] <sati> I want more than 8 [07:40:51 CET] <sati> so there have any another hardware solution which have low cost [07:41:10 CET] <sati> because intel processor is a high cost pc [07:41:45 CET] <rjp421> sati keep the lowest bitrate, fps and resolution you can stand [07:44:14 CET] <rjp421> which os? [07:44:39 CET] <sati> rjp421 : yes bro I do alot of R & D [07:44:49 CET] <sati> that is linux Os [07:44:52 CET] <sati> Ubuntu [07:45:51 CET] <sati> rjp421 : now I think , we can use external hardware for video encoding [07:45:57 CET] <sati> is it possible [07:46:23 CET] <rjp421> ffmpeg -f v4l2 -list_formats all -i /dev/videoX [07:46:55 CET] <rjp421> you may get lucky and have cams that can output h264 [07:47:18 CET] <rjp421> like raspberry pi cams [07:47:45 CET] <rjp421> that will save cpu transcoding [07:51:42 CET] <sati> rjp421 : No rasberry pi is very small board that device can not handle video encoding and decoding more than 1 channel [07:53:04 CET] <sati> I mean to say that there have any external h.264 chip set available which is supported by ffmpeg and that will be attached with my pc [07:53:25 CET] <rjp421> i mean h264 is one of the cameras output formats, and doesnt need to be transcoded.. i think youre going to have to skimp on some of the quality to get 8 smooth h264 streams [07:54:16 CET] <sati> no I have analog camera . these camera producing raw data [07:54:21 CET] <rjp421> especially if theyre all usb devices, yikes [07:54:44 CET] <sati> like ?? [07:55:00 CET] <sati> or any device serial no if you know please [07:55:06 CET] <rjp421> like your usb bud being overwhelmed [07:55:09 CET] <rjp421> bus [07:57:23 CET] <sati> rjp421 : bro I don't understand what you want to say [07:58:46 CET] <sati> I want know there have a dedicated hardware which can perform encoding operation under ffmpeg externally and my cpu consumption decreases [08:04:50 CET] <flux> USB2.0 should be plenty for even 10 h264-encoded video streams? USB3.x even more so. [08:06:53 CET] <flux> sati, you may be looking for some professional equipment to do hardware-based h264-encoding capture from 8 channels.. perhaps a device that talks IP as-is already, no need for ffmpeg? [08:08:31 CET] <flux> sati, just did an ebay search on h264 encoder, and on the paper this seems applicable: http://www.ebay.com/itm/150772072175 (you would need two) [08:08:51 CET] <flux> I hope though those 8 video streams don't need to be synchronized, though :) [08:09:39 CET] <sati> flux : if you know intel have quick sync technology which is paid and that is operated by Our ffmpeg , I want like that device , but make a low cost device [10:43:38 CET] <Disturbed1_away> gm, does anyone know of a decent guide for compiling ffmpeg with nvenc? [10:54:14 CET] <saste_> Disturbed1_away, you only need the nvenc header from the NVidia SDK, then you build ffmpeg the usual way [10:54:25 CET] <saste_> with --enable-nvenc [10:55:44 CET] <saste> Disturbed1_away, https://trac.ffmpeg.org/wiki/HWAccelIntro#NVENC [10:59:36 CET] <Disturbed1_away> ka, i'll give it a go over.... thought i was being smart buying a tesla s870... come to learn after purchase finding files for a old legecy device are impossible thru nvidia site.... [11:18:20 CET] <jonascj> Hi all. I am doing h264 two pass encoding and I want to scale my video at the same time (1080p to 720p). [11:18:40 CET] <jonascj> Do I need "-vf scale=-1:720" in both "-pass 1" and "-pass 2" command? [11:20:02 CET] <jonascj> if pass1 is just going through the file looking for statistics to use in pass 2 then it seems "stupid" to also scale it in that pass. So any scaling could be ignored when "-pass 1" is there, but I am not sure ... [11:53:26 CET] <relaxed> jonascj: both [11:59:37 CET] <Li> c_14: I tried this but it didn't work ... for file in *; do ffmpeg -t "$file" -ss 18 -vcodec "output$file"; done [12:00:59 CET] <jonascj> relaxed, both? [12:03:53 CET] <relaxed> jonascj: you need to scale on both passes [12:05:02 CET] <jcay> Li, -t is time AFAIK , you need -i before filename [12:18:42 CET] <Li> jcay: yes I changed it but it still not working [12:18:49 CET] <Li> for file in *; do ffmpeg -i "$file" -ss 00:00:18 -vcodec "output$file"; done [12:22:50 CET] <jcay> you didnt give the vcodec [12:24:37 CET] <rjp421> http://streaming.media.ccc.de/32c3/hallg/#schedule [12:44:51 CET] <jonascj> relaxed, alright, that is also what I am doing [12:46:12 CET] <jonascj> after some time of -pass 1 it begins with "Invalid UE golomb code=25.0" [13:01:22 CET] <jonascj> The bitrate it reports in "-pass 1" for libx264 encoding, (frame=384754 fbps=58 q=25 time=03:04:03.23 bitrate=4300 kbits/s speed=2.4x), what bitrate is that? [13:02:00 CET] <jonascj> In my two-pass I am targeting 4100kbit/s video and 250kbit/s audio, which is 4350kbit/s, is that it? [13:22:15 CET] <rjp421> http://cdn.c3voc.de/s2_native_sd.webm [13:23:10 CET] <dystopia_> bitrate will fluctuate even with a target [13:23:21 CET] <dystopia_> 2pass encoding to x264 is kinda pointless [13:23:35 CET] <dystopia_> use single pass and correct crf for content type [13:28:29 CET] <Li> How to delete 5 seconds from the end of the file without having to substraction operation to calculate duration -t [13:29:20 CET] <Li> in other words an equivalent for the switch -ss to delete from file's end [13:42:01 CET] <rjp421> damnit wheres their irc... someone should remind them of rtmfp and its ability to handle the udp packet loss [13:50:44 CET] <Li> huh [14:23:10 CET] <Disturbed1_away> does the latest build of ffmpeg not support nvenc 4.0 ? ( [14:26:11 CET] <Disturbed1_away> trying to compile ffmpeg with cuda kit 5 & nvenc 4 (and x264) so i can use legecy nvidia S870 hardware... [14:35:16 CET] <Disturbed1_away> if so, what version of ffmpeg does support nvenc4 >? [14:42:40 CET] <saste> Disturbed1_away, from the check in configure, NVENC API version 4 or older is not supported [14:43:13 CET] <saste> basically, ffmpeg only supports nvenc API >= 5 [15:00:11 CET] <Disturbed1_away> so from what i have figured out nvenc5 isn't compatible with the driver 340.xx required to run the s870... only nvenc4 is.... [15:01:00 CET] <Disturbed1_away> so i'm kinda forced to use a legecy copy of ffmpeg if i'm ever going to use s870 to encode with [15:03:27 CET] <Disturbed1_away> what i need to know is at what ffmpeg version did nvenc4 get dropped so i can compile a version before that... [16:01:57 CET] <Disturbed1_away> ugh [16:08:49 CET] <jonascj> dystopia_, two-pass is recommended when you have a target file size as far as the documentation goes. [16:27:30 CET] <LEOXD> hi [16:28:55 CET] <waressearcher2> LEOXD: hallo [16:29:14 CET] <waressearcher2> LEOXD: wie geht's es dir ? [16:30:39 CET] <LEOXD> waressearcher2: Kenn ich dich? [16:31:19 CET] <saste> Disturbed1_away, see commit b08caa87c35a768ec0abb16b1e99c3a85f1df28e [16:31:50 CET] <waressearcher2> LEOXD: nein, ich glaube nicht [16:33:08 CET] <saste> Disturbed1_away, I think it requires nvenc >= 5 since its first commit [16:33:44 CET] <LEOXD> anyways, I was wondering whether dale's info from Jan 9 about ffvp9 are still correct https://groups.google.com/a/chromium.org/forum/#!msg/chromium-dev/2Dz5Q5gCT… [16:36:51 CET] <durandal_170> lot have changed in ffvp9 [16:39:38 CET] <LEOXD> durandal_170: yeah, I thought so. But I can't find anything useful regarding what exactly changed. [16:40:15 CET] <durandal_170> see git log [16:40:50 CET] <LEOXD> link? [16:42:16 CET] <durandal_170> source.ffmpeg.org [16:42:38 CET] <durandal_170> locate vp9 in libavcodec [16:44:12 CET] <LEOXD> thanks [17:31:49 CET] <jonascj> stopping and resuming an ffmpeg encoding, is that possible? [17:32:08 CET] <c_14> As long as you don't reboot, sure [17:32:49 CET] <c_14> Use SIGSTOP and SIGCONT [17:33:13 CET] <jonascj> ah yes, how would sigstop followed by a suspend to disk work? [17:33:24 CET] <jonascj> okay I suppose [17:33:33 CET] <c_14> Should be fine afaik [17:34:23 CET] <jonascj> and of course this job is running on windows :( [17:34:36 CET] <jonascj> thanks, i'll go look at windows sigstop sigcont capabilities [17:44:06 CET] <kepstin> you should be able to just suspend to disk with ffmpeg running, the os will take care of stopping/resuming everything. [17:50:53 CET] <jonascj> kepstin, good point. [18:23:56 CET] <JodaZ> hmm, theres these hacks to suspend a single process to disk on linux [18:24:50 CET] <JodaZ> otherwise, i am kind of very interested in ways to stop/start encoding at a specific point in a file because that would make HLS (encoding and caching) much more elegant [18:25:59 CET] <JodaZ> but last time i tried there was no way to encode from a specific frame (and have the audio not somehow generate half filled chunks and stuff) [19:15:52 CET] <yongyung> So I'm trying to encode a video so that sony vegas can read it, I've used this options for different sources already and vegas could always read them no problem. But somehow it can't with this source, even though the source shouldn't even influence it right? Vegas thinks it's audio-only and the audio just sounds like random data, it's definitely not reading the actual audio stream. If I use a different container it doesn't recognize the [19:15:52 CET] <yongyung> file at all. Any ideas what I could try? The exact command line is [19:15:55 CET] <yongyung> ffmpeg -ss 00:02:44 -i "clip01.mp4" -t 30 -c:a copy -c:v libx264 -qp 16 -preset veryfast -tune fastdecode -movflags +faststart test.mp4 [19:17:20 CET] <fritsch> depends on the source [19:17:25 CET] <fritsch> post a mediainfo of the source [19:17:36 CET] <fritsch> it might be interlaced [19:18:12 CET] <yongyung> Like this? http://codepad.org/tkiiG0Oq [19:18:45 CET] <fritsch> its 1080p60 [19:18:55 CET] <fritsch> not sure the device can keep up with 60 progressive frame [19:18:57 CET] <fritsch> s [19:19:30 CET] <yongyung> I've edited 1080p180 with qp=12 before so I don't think it should be a problem [19:19:38 CET] <yongyung> Or what do you mean [19:19:41 CET] <ChocolateArmpits> transcode the audio too [19:20:02 CET] <ChocolateArmpits> maybe it expects aac and then goes crazy when there's mp3 there [19:28:23 CET] <yongyung> ChocolateArmpits: It works! But vegas only likes mp4, and mp4 doesn't support pcm, and I don't want to get another audio reencoding step... well let me try to just strip the audio [19:28:45 CET] <fritsch> yongyung: that sentence makes no sense [19:29:00 CET] <yongyung> fritsch: ? [19:29:13 CET] <fritsch> that sentence has nothing to with with what ChocolateArmpits said [19:29:14 CET] <ChocolateArmpits> yongyung: just use high bitrate if you want transparency [19:30:42 CET] <Li> what is the wrong with this syntax and how to correct it? ffmpeg -ss 00:00:10 -i file1.webm -c copy file2.webm [19:31:00 CET] <Li> I'm trying to cut 10 second intro from audo file [19:31:03 CET] <Li> audio* [19:32:08 CET] <ChocolateArmpits> Li: do you want intro removed or preserved ? [19:32:25 CET] <Li> ChocolateArmpits: nope I want to get rid of it [19:33:58 CET] <yongyung> Well I don't have libfdk_aac and the aac encoder doesn't seem that great.. but it doesn't matter, stripping audio seems to work [19:34:43 CET] <ChocolateArmpits> yongyung: aac is pretty good now, people say, at least than libvo definitely [19:35:00 CET] <yongyung> ChocolateArmpits: Any idea why I get tons of "past duration too large" yellow messages when I try to encode the file without -ss, but with -ss 00:00:02 it works just fine? Is the first frame somehow not correctly aligned in the source file or something? [19:35:46 CET] <ChocolateArmpits> yongyung: is this the same transcode? [19:35:52 CET] <yongyung> Yeah [19:35:58 CET] <yongyung> with -an [19:36:47 CET] <ChocolateArmpits> Probably some problem with initial timestamps, the warning message is very generic [19:39:50 CET] <yongyung> Hmm well good to know that I can avoid it at least. It comes up so much the i/o actually significally slows encoding lol [19:40:43 CET] <ChocolateArmpits> You can hide those messages using -loglevel error I think [19:41:14 CET] <ChocolateArmpits> or using setpts filter to completely reset timestamps for each frame [19:45:53 CET] <yongyung> Yeah the problem with loglevel is that it also hides all the information output since it's technically "less important". The other method I never tried, if -ss doesn't work some day I'll remember that [20:13:45 CET] <gabi__> hi, i have an audio file (adts mp3) and i am wondering how i could transform that to a regular mp3 file, because my car can not play adts, just regular mp3. can someone tell me how to do that with ffmpeg? [20:15:20 CET] <gabi__> even when i convert it to .wav and then reconvert to mp3 with sox/lame/ffmpeg etc. i always get a mp3 in adts [20:15:33 CET] <gabi__> Audio file with ID3 version 2.3.0, contains: MPEG ADTS, layer III, v1, 192 kbps, 44.1 kHz, Stereo [20:32:59 CET] <yongyung> gabi__: Can you post the ffprobe output of your file to pastebin and the link here? [20:43:54 CET] <gabi__> yongyung, jep, one second [20:44:27 CET] <gabi__> yongyung, http://pastebin.com/k4Urr9bA [20:44:53 CET] <BtbN> That's a perfectly normal mp3 file. [20:45:19 CET] <gabi__> hm ok, i used file before and that reports adts [20:45:43 CET] <gabi__> Adele - Hello.mp3: Audio file with ID3 version 2.3.0, contains: MPEG ADTS, layer III, v1, 192 kbps, 44.1 kHz, Stereo [20:49:05 CET] <gabi__> BtbN, so is file reporting something wrong_ [20:49:14 CET] <furq> file reports that for every mp3 i just tried [20:49:28 CET] <gabi__> ah ok :) so its a feature of file... [20:54:00 CET] <furq> yeah it looks like libmagic always refers to mp3 as "MPEG ADTS, layer III" [21:00:32 CET] <yongyung> gabi__: So does your car play the file? [21:06:46 CET] <gabi__> yongyung, no [21:06:57 CET] <gabi__> unfortunately not, i thought maybe thats the reason [21:07:13 CET] <furq> is it vbr [21:07:22 CET] <furq> some very old hardware mp3 players will only play cbr [21:07:28 CET] <gabi__> how can i check that? [21:09:56 CET] <furq> mediainfo will tell you [21:12:29 CET] <furq> http://sprunge.us/YfXI [21:13:07 CET] <yongyung> gabi__: Did you buy the music? If so, are you able to download a .flac or .wav version of the file or do you have a CD? Because in that case you could just encode the music yourself and try out what works on your car. Re-encoding usually doesn't sound as good, though I guess it's not the end of the world for car speakers [21:13:41 CET] <gabi__> no i downloaded it from youtube with a online converter [21:15:03 CET] <gabi__> http://pastebin.com/jLWaZEZg [21:15:19 CET] <gabi__> so i guess thats CBR, as it says constant bit rate [21:15:23 CET] <yongyung> gabi__: Alright if you want to go the youtube route (instead of.. ehem more shady ways to get the music), use a program or website to just download the raw video file, and convert that with ffmpeg. What you should try in your car is [21:15:52 CET] <yongyung> ffmpeg -i fromyoutube.mp4 -c:a libmp3lame -c:b 320k test1.mp3 [21:16:05 CET] <furq> use youtube-dl [21:16:08 CET] <yongyung> ffmpeg -i fromyoutube.mp4 -c:a libmp3lame -c:b 128k test2.mp3 [21:16:27 CET] <yongyung> ffmpeg -i fromyoutube.mp4 -a:a libmp3lame -aq 0 test3.mp3 [21:16:34 CET] <furq> -q:a 0 [21:16:57 CET] <yongyung> ffmpeg -i fromyoutube.mp4 -c:a copy test4.aac [21:17:00 CET] <furq> also -b:a, not -c:b [21:17:05 CET] <gabi__> i have a document which says that the radio should work with regular mp3 files with 96...256k, if i remember correctly [21:17:10 CET] <furq> and please use .m4a, not .aac [21:17:11 CET] <yongyung> uhh yeah mit b:a LOL [21:17:40 CET] <furq> with hardware players there could be a million things that it doesn't like [21:18:09 CET] <yongyung> gabi__: Well, then try ffmpeg -i fromyoutube.mp4 -c:a libmp3lame -b:a 256k test1.mp3 [21:18:16 CET] <furq> i take it there's no line-in port for you to plug a real mp3 player into [21:18:44 CET] <gabi__> no, there is no line in [21:18:52 CET] <gabi__> only usb [21:19:38 CET] <gabi__> i will generate a mp3 from the raw file and see if that works. if not, i will be back :) [21:20:02 CET] <furq> i can't help but notice "Mode extension: MS Stereo" in that mediainfo [21:20:06 CET] <yongyung> gabi__: Good and the parameters are -c:a for codec selection and -b:a for bit rate selection, I mistyped that a couple of times [21:20:18 CET] <furq> i'm not sure what that is but it sure sounds like something which could break an elderly player [21:39:21 CET] <gabi__> ok, i have the new .mp3 file without ms stereo extension. will test this tomorrow. many thanks :) [21:39:36 CET] <gabi__> bye [21:55:04 CET] <BtbN> A lot of players also play AAC, which you can download straight from youtube, without any transcoding [23:13:03 CET] <jonascj> "Invalid UE golomb code=27" while encoding x264. It started around half-way through the encoding (framewise). Anything to be worried about? Not much can be seen on the mailing list. [23:13:57 CET] <jonascj> Mostly people / developers talk about polluting the log files with it, and worry that certain checks pertaining to this UE golomb code may slow down x264 encoding which they don't want because it is a popular encoding proceduire... [00:00:00 CET] --- Tue Dec 29 2015

1 0

ffmpeg-devel.log.20151227
by burek 28 Dec '15

28 Dec '15

[01:26:02 CET] <J_Darnley> I might have to ask this again later but... [01:28:42 CET] <J_Darnley> actually ignore me, again [01:33:48 CET] <J_Darnley> I guess my linux platform doesn't use aslr [01:55:31 CET] <Compn> J_Darnley : what distro do you use ? [01:56:16 CET] <J_Darnley> I put arch on it [02:22:48 CET] <cone-230> ffmpeg 03Ganesh Ajjanagadde 07master:71af38954b18: avcodec/on2avc: fix regression on icc since 5495c7f [03:17:44 CET] <cone-230> ffmpeg 03Lou Logan 07master:0779d54a218c: doc/filters: note NetCDF dependency for sofalizer [05:31:14 CET] <cone-230> ffmpeg 03Ganesh Ajjanagadde 07master:4e7cfefa16af: lavfi/vf_hue: replace rint by lrint [05:31:15 CET] <cone-230> ffmpeg 03Ganesh Ajjanagadde 07master:3e2e303e4b0f: lavfi/avf_showspectrum: replace rint by lrint [05:31:16 CET] <cone-230> ffmpeg 03Ganesh Ajjanagadde 07master:e09edc62cdad: lavc/texturedsp: replace rint by lrint [05:31:17 CET] <cone-230> ffmpeg 03Ganesh Ajjanagadde 07master:5979c740f5a9: lavc/dds: replace rint by lrint [05:31:18 CET] <cone-230> ffmpeg 03Ganesh Ajjanagadde 07master:c5b3c4c74171: lavc/snowenc: replace rint by lrint [09:20:07 CET] <cone-962> ffmpeg 03Reynaldo H. Verdejo Pinochet 07master:0e5c1dc9a32b: ffserver: make our 404 explicitly HTML5/UTF-8 [09:20:07 CET] <cone-962> ffmpeg 03Reynaldo H. Verdejo Pinochet 07master:daaa53586727: ffserver: HTML encode msgs instead of blindly stripping chars out [09:20:07 CET] <cone-962> ffmpeg 03Reynaldo H. Verdejo Pinochet 07master:b25ac3c999b0: ffserver: add a doctype heading to our HTML pages [09:20:07 CET] <cone-962> ffmpeg 03Reynaldo H. Verdejo Pinochet 07master:baf4c489e5f4: avio: add detail to avio_printf() size warning [09:42:41 CET] <Prelude2004c2> hey guys.. anyone around that could help with with ffmpeg and vdpau ? i think there may be something wrong with the ffmpeg source code [09:43:07 CET] <Prelude2004c2> i can' get mpeg2video to decode just fine using the GPU but i can't seem to get it to work with h264 no matter what i do .. and. i am using the latest M4000 card [09:43:15 CET] <Prelude2004c2> so i am stumped on it for days now [09:43:21 CET] <Prelude2004c2> did i say days.. i mean weeks [17:06:55 CET] <cone-058> ffmpeg 03Michael Niedermayer 07master:3215342121c1: avcodec/on2avc: Fix stability issues with scale_tab generation [18:35:19 CET] <Prelude2004c2> good day everyone. Looking for assistance with vdpau and ffmpeg [18:37:33 CET] <Compn> developing with ffmpeg or developing your own project ? [18:38:05 CET] <Compn> people in this channel only like talking about ffmpeg development in here. [20:12:14 CET] <nevcairiel> that was one weird bug, and I wonder how it was never found before [20:18:35 CET] <BtbN> which one? [20:23:49 CET] <nevcairiel> the one i just sent [20:35:08 CET] <tmm1> i'm trying to write a vf filter that doubles framerate, kinda like vf_yadif in field mode [20:36:53 CET] <tmm1> can i call ff_filter_frame() twice from the filter_frame callback? [20:51:51 CET] <cone-884> ffmpeg 03Paul B Mahol 07master:1809894b2afd: avfilter/af_silenceremove: add peak detector [20:59:51 CET] <durandal_1707> tmm1: sure, see yadif and w3fdif [21:39:20 CET] <jamrial> patch v9 [21:52:19 CET] <tmm1> i was confused by the use of request_frame, but emitting twice from filter_frame worked [22:18:55 CET] <ubitux> so we still need libzvbi to decode teletext? [22:19:56 CET] <kierank> yes decoding teletext is tedious stuff [22:20:11 CET] <cone-884> ffmpeg 03Joel Holdsworth 07master:b4eb1f29ebdd: avformat/http: Added http_proxy option [22:20:12 CET] <cone-884> ffmpeg 03Joel Holdsworth 07master:9cdccd404be9: avformat/hls: Remember to free HLSContext::headers [22:20:13 CET] <cone-884> ffmpeg 03Joel Holdsworth 07master:c48122d731be: avformat/hls: Added http_proxy support [23:29:50 CET] <llogan> not very interesting zample images... [23:31:43 CET] <J_Darnley> Meh. They're a good place to start from. [00:00:00 CET] --- Mon Dec 28 2015

1 0

ffmpeg.log.20151227
by burek 28 Dec '15

28 Dec '15

[01:01:36 CET] <Hybridesque> Okay, I've cooked up my own ffmpeg for Ubuntu 15.10, and it works normally using x264 (after a bit of work of installing the necessary libaries on the transcoding server for mp3, aac etc etc) [01:01:53 CET] <Hybridesque> It works fine on the x264, but when I try to use the qsv, it's not working. [01:02:01 CET] <Hybridesque> I followed the compile instruction at https://trac.ffmpeg.org/wiki/CompilationGuide/Ubuntu [01:02:18 CET] <Hybridesque> with exception of compiling the mfx dispatcher, and using --enable-libmfx [01:02:50 CET] <Hybridesque> The logs I'm getting from ffmpeg are here: http://pastebin.com/MKUqHmyK [01:50:17 CET] <Terra> Quick question with ffmpeg and rtmp streams (input), when the stream ends or if ffmpeg does not receive any data, it does not simply die. It continues running instead of just stopping and has to be manually killed [01:50:22 CET] <Terra> Is there something I'm doing wrong? [01:51:03 CET] <Terra> This is with nginx-rtmp as the rtmp server if that's any help. [03:12:35 CET] <waressearcher2> Terra: hallo, wie geht's ? [03:13:00 CET] <Terra> ? Sorry I only speek English. [03:14:18 CET] <pepee> lol [03:14:25 CET] <Terra> Though I got that much at least, I'm fine thanks :P [03:30:58 CET] <_6U54N0_> does ffmpeg_2.0.4 has support to convert FLTP to S16? [03:33:34 CET] <_6U54N0_> cause, i am trying to convert a decode frame in format FLTP to frame in format S16, but the result is wrong. [03:44:39 CET] <Disturbed1_away> sup all.... [03:46:09 CET] <Disturbed1_away> need a little bit of help... [03:49:55 CET] <Disturbed1_away> i just spent couple days getting ibm systemx3650(7979) working with centos6.7/64bit, with a nvida tesla s870... finally got everything working perfect, ( LOL ) so my next task is to try ffmpeg to use the gpu's.... after searching web for a bit i cant seem to recall the command to tell ffmpeg to use the gpu's.... here a copy of the code i'm using... " for %%a in ("*.*") do ffmpeg -i "%%a" -s [03:49:55 CET] <Disturbed1_away> 1920x1080 -c:v libx264 -strict experimental -crf 20 -c:a aac -ab 128k -ac 2 -ar 48000 -vprofile main -g 48 -threads 0 "newfiles\%%~na_master.mp4" [03:49:55 CET] <Disturbed1_away> " where would be the best place to put the gpu command? [05:04:31 CET] <pepee> waressearcher2: hallo, wie geht's ? [05:05:44 CET] <yongyung> When concatenating a couple of clips, is there a way to tell ffmpeg to create 2 seconds of black frames between them as a sort of "transition"? [05:41:41 CET] <waressearcher2> pepee: es geht's, und selbst ? [06:15:02 CET] <pepee> waressearcher2, es geht's. I just learned a bit of german today :P [09:59:31 CET] <thomedy> hello is anyone working right now [11:33:43 CET] <neXyon> is 48kHz the default sample rate of anything digital nowadays? [15:45:34 CET] <Hybridesque> argh, this qsv thing on linux is going my nut. [15:45:39 CET] <Hybridesque> *doing my nut [15:50:24 CET] <BtbN> Just forget it exists, it's not worth it. [15:56:30 CET] <Hybridesque> sodding Intel. [19:26:46 CET] <senoo> hi, i want to blur a part on the center of my udp video input, how can i do this? [19:35:27 CET] <waressearcher2> senoo: hallo, wie geht's ? [19:36:24 CET] <microchip_> waressearcher2: english please [19:42:24 CET] <Fjorgynn> hello [19:42:32 CET] <Fjorgynn> I've got an hls mp3 audio stream [19:42:39 CET] <Fjorgynn> how do I dump it? [19:43:16 CET] <sylvain> hi [19:43:42 CET] <Guest40501> any plans to make ffply available thru nacl ? might be very cool to create a new web browser plugin [19:43:50 CET] <Guest40501> ffplay* [19:44:43 CET] <Guest40501> https://code.google.com/p/native-client-ffplay/ found this but seems inactive :/ [20:06:38 CET] <Fjorgynn> I can't download [20:38:23 CET] <Fjorgynn> [mpegts @ 004fe8c0] Non-monotonous DTS in output stream 0:0; previous: 60306610, current: 60301908; changing to 60306611. This may result in incorrect timestamp [20:38:24 CET] <rjp421> http://cdn.c3voc.de/s1_native_hd.webm [20:38:26 CET] <Fjorgynn> s in the output file. [20:38:29 CET] <Fjorgynn> what does that mean? [21:34:08 CET] <m4t> i'm struggling trying to get a gif input to loop for the duration of an audio input, when combining the two into a video [21:34:46 CET] <m4t> trying like: ffmpeg -i giphy.gif -i *.m4a -strict -2 -loop 0 protein.mp4 [21:35:14 CET] <m4t> if i move the "-loop 0" elsewhere, i.e. after the "-i giphy.gif" like "-i giphy.gif -loop 0", it says option loop not found [21:35:59 CET] <furq> -loop is an input option [21:36:03 CET] <furq> move it before the gif [21:37:03 CET] <m4t> ya tried that too, thanks: -loop: No such file or directory [21:37:16 CET] <furq> before -i [21:37:46 CET] <m4t> yeah tried that too :( : Option loop not found. [21:38:39 CET] <furq> i can't say i've tried with a gif but that should work fine for images [21:42:16 CET] <m4t> hmm, -ignore_loop 0 works [21:42:30 CET] <m4t> but it uh, seems to go on forever :-) [21:42:39 CET] <furq> -shortest [21:43:01 CET] <furq> that one's an output option [21:44:10 CET] <m4t> perfect, thanks [22:20:54 CET] <Terra> Asked yesterday but didn't get a reply. Does anyone have any idea why ffmpeg doesn't die/timeout whenever an RTMP stream stops or doesn't exist? It has to be manually killed. [22:21:05 CET] <Terra> Not sure why or what I'm doing wrong [22:24:26 CET] <JEEB> Terra: did the server die as well? [22:24:49 CET] <Terra> ffmpeg -i rtmp://127.0.0.1/live/$uid -vframes 1 -an -s 220x130 -y /var/www/lthumb/$uid.png <- This is the command I use for generating screenshots but if it hits a stream that doesn't exist it just hangs [22:24:51 CET] <JEEB> because rtmp(e) servers can basically have nothingness in the stream [22:24:55 CET] <Terra> JEEB, nope. [22:25:04 CET] <JEEB> so it's a valid stream, it just doesn't contain anything [22:25:11 CET] <Terra> Yeah basically that. [22:25:26 CET] <Terra> Is there a way to make it timeout if not actual data is recieved from it if that makes sense [22:25:39 CET] <JEEB> probably not that limited, but there could be general timeouts [22:25:49 CET] <JEEB> at least I remember finding one for UDP [22:25:56 CET] <Terra> -timeout and -stimeout doesn't seem to do diddly [22:26:18 CET] <Terra> huh udp timeout [22:26:24 CET] <JEEB> for UDP it was in the URL, also make sure options you set are in the "decoder" side of things [23:10:41 CET] <pepee> waressearcher2: hallo! [23:12:15 CET] <waressearcher2> pepee: genug eigentlich [23:12:49 CET] <pepee> lol [23:17:38 CET] <_6U54N0_> porque ffmpeg se traga toda la memoria ram? [23:18:26 CET] <_6U54N0_> me refiero, cuando se usa la liberia con C/C++ para decodificar video stream. [23:20:29 CET] <pepee> _6U54N0_, has la pregunta en ingles [23:21:24 CET] <_6U54N0_> no me gusta ese puto dialecto. [23:22:19 CET] <pepee> ooh, we got a badass over here [23:25:10 CET] <_6U54N0_> i dont know what did you say ... but ... if by flys hdp. [23:41:15 CET] <angeldorado> hello [23:42:08 CET] <waressearcher2> angeldorado: hallo [23:44:49 CET] <angeldorado> hi, trying to install ffmpeg static 32 bit on ubuntu studio [23:45:19 CET] <angeldorado> not really sure how to do it. I tried ./ffmpeg on the folder it is downloaded but that just brought up a help menu it seems [23:45:43 CET] <c_14> there's no installer [23:45:47 CET] <c_14> just put it somewhere in your path [23:46:12 CET] <angeldorado> so if I want to run ffmpeg -i <file> [23:46:17 CET] <angeldorado> how do I do it? [23:46:19 CET] <angeldorado> that does not work [23:46:26 CET] <c_14> just put it somewhere in your path [23:46:31 CET] <angeldorado> idk what you mean [23:46:38 CET] <c_14> echo $PATH [23:46:42 CET] <c_14> put it in one of those directories [23:47:13 CET] <angeldorado> so do, echo $PATH <file directory> [23:47:15 CET] <angeldorado> then run it? [23:47:27 CET] <angeldorado> im a linux amateur [23:47:52 CET] <c_14> put the ffmpeg binary in one of the directories listed when you run the command "echo $PATH" [23:48:35 CET] <angeldorado> so echo $PATH calls a process from some other folder. [23:50:22 CET] <c_14> mv ffmpeg $(echo $PATH | cut -d: -f1) [23:50:32 CET] <rjp421> JEEB, im assuming ffmpeg will never support the udp rtmfp (p2p/multicast) protocol? i feel more people would use it if they knew how useful it could be [23:50:39 CET] <angeldorado> echo $PATH /usr/local/bin/ffmpeg-2.8.4-32bit-static/ffmpeg ffmpeg -i ending.mp4 [23:50:39 CET] <angeldorado> /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games /usr/local/bin/ffmpeg-2.8.4-32bit-static/ffmpeg ffmpeg -i ending.mp4 [23:52:08 CET] <angeldorado> this is too hard. [23:59:53 CET] <angeldorado> ok, I've taken a breather now. What about software like openshot or blender that might use ffmpeg. how do I get those to access it? [00:00:00 CET] --- Mon Dec 28 2015

1 0

ffmpeg-devel.log.20151226
by burek 27 Dec '15

27 Dec '15

[00:34:30 CET] <tmm1> J_Darnley: so i've been staring at this yadif code and i noticed FFABS is double-evaluating all over the filter_line function [00:36:28 CET] <J_Darnley> Are you sure? It does many of them but I don't think any do the same two pixels. [00:36:33 CET] <nevcairiel> compilers are generally trusted to remove that [00:36:59 CET] <J_Darnley> Oh you mean when it advances to the next pixel [00:37:07 CET] <tmm1> i mean in the macro itself [00:37:08 CET] <tmm1> #define FFABS(a) ((a) >= 0 ? (a) : (-(a))) [00:37:31 CET] <J_Darnley> oh, well yes, as nevcairiel said [00:38:00 CET] <tmm1> ah i didn't realize the compiler was smart enough to calculate `a` only once there [00:38:24 CET] <J_Darnley> Well, if you use -O0 it probably won't [00:38:51 CET] <tmm1> sure, it won't do anything smart in that case [00:38:52 CET] <J_Darnley> (but ffmpeg doesn't configure itself with that) [00:39:36 CET] <J_Darnley> (someone should point Ganesh at this code) [00:43:07 CET] <J_Darnley> tmm1: ffmpeg has lots of macros which end up doing that, the MAX and MIN ones are features in yadif [00:43:15 CET] <J_Darnley> *featured [00:47:36 CET] <iive> tmm1: what do you mean by double evaluating? How would you write that macro so it doesn't do that... [00:49:36 CET] <tmm1> for a callsite like FFABS(c-e), it would do ((c-e) >= 0 ? (c-e) : (-(c-e))), so c-e is performed twice [00:50:43 CET] <nevcairiel> just need to be careful not to put anything with side-effects in there [00:50:49 CET] <nevcairiel> but otherwise trust the compiler [00:51:10 CET] <iive> i can't find it in the moment, but if you have (i++) isn't it possible that it evalutates it 3 times? [00:52:03 CET] <tmm1> yea i believe that's possible, hence the side-effects warning [00:52:34 CET] <tmm1> it does look like the compiler produces smart output w/ -O2 [00:52:48 CET] <tmm1> so much for that alleged easy win.. [00:53:46 CET] <iive> that's kind of strange... i've seen compiler optimize common algebra calculates even when explicitly written separately... [00:53:50 CET] <J_Darnley> I doubt that you can make the C much faster. [00:54:17 CET] <iive> you might try using built-in [00:54:35 CET] <J_Darnley> Does arm gain anything if you operate on 16-bit ints rather than 32-bit? [00:55:07 CET] <iive> does it even have 16 bit registers? [00:57:16 CET] <tmm1> not sure [00:58:01 CET] <tmm1> i think i might try using neon compiler intrinsics as a first pass rather than rewriting the whole thing in asm [00:58:16 CET] <tmm1> trying to figure out which parts of the algorithm can actually be optimized with SIMD though [00:59:09 CET] <J_Darnley> One thing you should definitely doublecheck is that the if(is_not_edge) branch is made constant. [00:59:46 CET] <J_Darnley> At one time that was not constant (thanks Libav) and killed performance. [01:00:56 CET] <iive> tmm1: you want to port yadif to arm? [01:01:00 CET] <tmm1> should evaluate to if(1) from what i can tell [01:01:01 CET] <J_Darnley> again, the compiler should be smart enough [01:01:13 CET] <tmm1> yea seems like an obvious optimization for the compiler [01:01:37 CET] <tmm1> iive: yea, trying to improve performance on aarch64 [01:03:02 CET] <iive> nice :) [01:14:47 CET] <J_Darnley> If you want to try simd then you'll probably have to do the whole filter_line function [01:15:19 CET] <J_Darnley> but I might suggest that you start with the code in the CHECK macros [01:16:07 CET] <J_Darnley> That's where most of the x86 code is [01:17:47 CET] <tmm1> cool, i was wondering what the mmx/sse optimized versions did differently [01:17:54 CET] <J_Darnley> Nothing [01:18:02 CET] <J_Darnley> They produce identical output [01:18:49 CET] <J_Darnley> Are you looking for the files containing the x86 code? [01:20:05 CET] <tmm1> i meant, what does vf_yadif.asm's filter_line do differently than the C version [01:20:22 CET] <J_Darnley> (they are libavfilter/x86/{vf_yadif.asm,yadif-10.asm,yadif-16.asm} [01:21:08 CET] <tmm1> yep i see them, my assembly is just really rusty [01:26:36 CET] <iive> good luck polishing it :) [01:51:53 CET] <tmm1> getting there, slowly [01:54:40 CET] <tmm1> the x86 version is using the mmx registers and doing parallel loads/ops, obviously [03:29:49 CET] <J_Darnley> it uses mm-regs for the mmx version and xmm-regs for the sse2 version [03:30:18 CET] <J_Darnley> Well, that doesn't really matter. [03:31:21 CET] <J_Darnley> (plus he left) [06:02:24 CET] <tmm1> i saw the messages i missed in the channel log [06:02:42 CET] <tmm1> can the spatial checks be skipped entirely when in nospatial mode? [07:33:47 CET] <prelude2004c> hey anyone aroun d? [14:44:12 CET] <J_Darnley> Does the fate website list the machine triplet thing (as in gcc -dumpmachine) anywhere? [14:44:53 CET] <J_Darnley> oh wait, nevermind [18:26:45 CET] <ubitux> michaelni: sorry to ask you again, but can you upload http://b.pkh.me/empty-events-2167.srt ? (MD5=6b9dd871e29faf5104be85b1c494bc47) [18:27:00 CET] <ubitux> (same directory, fate-samples/sub) [18:44:15 CET] <michaelni> ubitux, uploaded [18:44:55 CET] <ubitux> thanks [18:49:57 CET] <metRo_> Hi, I need to compile ffmpeg, should I expose my problem here or in in ffmpeg channel? [18:51:16 CET] <Daemon404> #ffmpeg [18:52:46 CET] <metRo_> ok [18:56:30 CET] <metRo_> since anyone asnwer me at ffmpeg if anyone knows about pkg-config, can you help me there? [19:18:04 CET] <J_Darnley> What's wrong with that thing now? [19:21:39 CET] <Compn> ffmpeg's arch nemesis, pkg-config! [19:22:03 CET] <JEEB> dunno if it's pkg-config's fault if your pc files contain paths from another chroot :P [19:22:13 CET] <JEEB> (I helped him on the correct channel) [19:22:41 CET] <J_Darnley> Ah [19:23:21 CET] <J_Darnley> Well, it can only spit out the config it's given (garbage in; garbage out) [20:23:17 CET] <cone-066> ffmpeg 03Michael Niedermayer 07master:e9e87822022f: avformat/img2dec: Skip checking the input files existence if it has already been opened [20:23:18 CET] <cone-066> ffmpeg 03Michael Niedermayer 07master:e70d56b8ad5d: avformat/img2dec: Reuse main IO context instead of reopening a single file [00:00:00 CET] --- Sun Dec 27 2015

1 0

ffmpeg.log.20151226
by burek 27 Dec '15

27 Dec '15

[00:57:30 CET] <DrunkOnHotCoco> Hello! looking for help with our videos getting shortened by ffmpeg when being converted to FLV, it started out happening every now and then and now happens every time. We are using Java to get this done as seen in this code http://escapekeys.com/ffmpeg-and-coldfusion/18 [00:59:45 CET] <DrunkOnHotCoco> Any help is appreciated! [02:57:03 CET] <el_juliano> what is better: OPUS 160k or AAC 256k? I guess AAC because of the higher bitrate but I am not sure [06:23:45 CET] <kepstin> if el_juliano was still here, i'd answer "at those bitrates, it's somewhere between very difficult and impossible to tell the difference" :/ [06:24:30 CET] <kepstin> (and it depends a lot on the quality of the aac encoder used) [07:50:10 CET] <rjp421> how do i make ffmpeg listen on 0.0.0.0:1234 instead of localhost:80 when encoding to mjpeg? [10:06:51 CET] <wuqiong> anybody knows ffmpeg on iOS avformat_open_input api hangs when opening microphone [13:50:02 CET] <jleclanche> hey; anyone knows the status of bink2 video support? I see bk2 audio is supported [13:53:04 CET] <durandal_1707> where is bk2 audio supported? [14:00:19 CET] <jleclanche> durandal_1707: https://ffmpeg.org/pipermail/ffmpeg-devel/2014-January/152661.html [14:02:34 CET] <durandal_1707> jleclanche: theres incomplete bink2 patch on codecs.multimedia.cx [17:41:42 CET] <rjp421> how do i make ffmpeg listen on 0.0.0.0:1234 instead of localhost:80 when encoding to mjpeg? [17:46:38 CET] <_6U54N0_> how to sort a buffer from planar FLTP to S16 interleaved? [17:46:47 CET] <_6U54N0_> my frame buffer is: [17:46:48 CET] <_6U54N0_> uint8_t *output; [17:47:21 CET] <JEEB> you use {av,sw}resample [17:47:27 CET] <JEEB> you get an optimized conversion [17:48:03 CET] <_6U54N0_> i use: [17:48:29 CET] <_6U54N0_> data_size = av_samples_get_buffer_size(); [17:48:48 CET] <_6U54N0_> struct SwrContext ... [17:49:08 CET] <_6U54N0_> swr_alloc_set_opts(); [17:49:25 CET] <_6U54N0_> swr_init(); [17:49:55 CET] <_6U54N0_> av_samples_alloc(); [17:50:17 CET] <_6U54N0_> swr_convert(); [17:50:29 CET] <_6U54N0_> in that order. [17:53:03 CET] <_6U54N0_> but the audio obtened is noise and bad sounds or un understandible sounds. [17:55:36 CET] <_6U54N0_> i would like an example to sort a buffer from planar FLTP to interleaved S16. [17:55:50 CET] <_6U54N0_> audio buffer. [17:57:44 CET] <JEEB> I think there's an example for it in examples [17:58:32 CET] <_6U54N0_> not, there is not. [18:19:55 CET] <realies> hi [18:20:15 CET] <realies> how can I transcode a video without changing anything but the bitrate? [18:20:28 CET] <realies> ffmpeg -i input.ts -c:copy -b:v 5000k output.ts [18:20:29 CET] <realies> ? [18:20:37 CET] <c_14> that won't work [18:20:52 CET] <realies> how should I approach it? [18:20:52 CET] <c_14> It depends on the input codec as well, but that is a very complicated problem. [18:21:05 CET] <c_14> It also depends on what exactly you mean by "anything" [18:21:08 CET] <realies> h264 [18:21:17 CET] <realies> in a ts [18:21:19 CET] <c_14> There are a lot of encoding options you probably don't care about, do you want those to remain the same as well? [18:21:26 CET] <c_14> i.e. exactly where the keyframes are etc [18:21:31 CET] <realies> yes [18:22:14 CET] <c_14> The "easy way out" would be ffmpeg -i input.ts -map 0 -c copy -c:v libx264 -b:v 5000k out.ts [18:22:31 CET] <c_14> That won't preserve keyframe locations, profile, preset etc though [18:22:37 CET] <realies> is -map 0 the audio channels? [18:22:47 CET] <c_14> audio, subtitle etc [18:23:35 CET] <realies> right [18:25:56 CET] <c_14> ffmpeg will automatically map one audio, one video and one subtitle stream. If you don't have/don't need/want more than that you can leave the map out. [18:27:46 CET] <realies> thanks c_14 [18:27:52 CET] <realies> you're saving my ass right now :) [18:29:42 CET] <realies> is this correct for changing resolution as well? ffmpeg -i input.ts -map 0 -c copy -c:v libx264 -b:v 5000k -vf scale=1280:720 out.ts [18:36:28 CET] <c_14> ye [18:36:55 CET] <c_14> you might want to use -2 instead of 720 to avoid stretching in the case that an input video isn't perfectly 16:9 [18:38:17 CET] <realies> c_14, you mean scale=-2? [18:38:24 CET] <c_14> scale=1280:-2 [18:38:33 CET] <realies> right [18:38:38 CET] <c_14> The -2 means scale to the nearest round number preserving aspect ratio [18:38:39 CET] <realies> good to know [18:38:50 CET] <realies> cool [18:45:02 CET] <realies> should I use -threads for making it faster? [18:48:58 CET] <metRo_> Hi, can anyone help me setting up pkg-config option on configure to cross compile pkg-config? [18:57:30 CET] <JEEB> metRo_: you don't need to generally cross-compile pkg-config [18:57:36 CET] <JEEB> you might have to set PKG_CONFIG_LIBDIR, though [18:58:11 CET] <JEEB> for example I have a cross-prefix-pkg-config that is just a shell script that calls normal pkg-config and sets PKG_CONFIG_LIBDIR to the related architecture's pkg-config libdir [18:58:24 CET] <thereh> I have image1 1920x800, I cut out piece w=1422:h=800:x=249:y=0 from it then scaled that piece to 1280x720 and got image2, then cut out piece from image2 w=925:h=520:x=177:y=200 and got image3, but how to cut out image3 right from image1 ? I can't add x1+x2 and y1+y2 because after I cut out image2 it was scaled to 1280x720 [18:58:29 CET] <thereh> I mean you can say 249+177 would be offset for image3 in image1 but as I said I scaled image2 [18:58:51 CET] <thereh> so I need w,h,x,y from the image1 to get right to image3 [18:59:09 CET] <thereh> or you can say its a video not image [18:59:41 CET] <realies> c_14, is there a way to not be doing the transcodes realtime? [18:59:46 CET] <metRo_> JEEB I know I don't need to cross compile, the problem is that the *.pc file for each package is point to /usr/lib/... and /usr/include/... however on my host the libraries are in /home/ze/raspi/sysroot/usr/include and /home/ze/raspi/sysroot/usr/lib [18:59:47 CET] <realies> i mean to do them faster than realtime [19:00:04 CET] <metRo_> so should I set PKG_CONFIG_LIBDIR to /home/ze/raspi/sysroot? [19:00:33 CET] <c_14> realies: you don't need to set threads for libx264 it's done automatically. You can set a faster preset (-preset fast faster or veryfast), but this will impact the quality at constant bitrate encoding [19:00:37 CET] <JEEB> usually things install their pkg-config pc files to ./lib/pkgconfig or so [19:00:47 CET] <JEEB> so find that directory and set PKG_CONFIG_LIBDIR there [19:01:33 CET] <JEEB> PKG_CONFIG_LIBDIR overrides the pkg-config's default search path, unlike PKG_CONFIG_PATH which appends things to it [19:01:58 CET] <JEEB> in case of cross compilation you most probably want to override, as to not have your system stuff come up :) [19:03:22 CET] <metRo_> JEEB right now pkg-config can find the right files at /home/ze/raspi/sysroot/usr/lib/arm-linux-gnueabihf/pkgconfig/ because if I go there and change the pc file to point the path to correct directory, per example ./configure is failing at smbclient because it says it can't find the library. The pc file for smbclient return the following -I/usr/include/samba-4.0 -L/usr/lib/arm-linux-gnueabihf if I manually change that so the smbclient. [19:04:02 CET] <metRo_> -I/home/ze/raspi/sysroot/usr/include/samba and -L/home/ze/raspi/sysroot/usr/lib/arm-linux-gnueabihf then ./configure can pass the smbclient test [19:04:13 CET] <metRo_> however it fails at next package test [19:04:20 CET] <JEEB> well yeah, if your sysroot is meant to be done otherwise then you'd also have to set the sysroot variable for your toolchain [19:04:38 CET] <metRo_> would you mind to look at my confiure? [19:04:39 CET] <JEEB> see your toolchain's documentation on how to set sysroots [19:04:43 CET] <JEEB> yes I would [19:04:57 CET] <JEEB> I think the configure script even has a thing for sysroots [19:05:04 CET] <JEEB> look at the output of ./configure --help [19:05:06 CET] <JEEB> and figure it out [19:05:10 CET] <metRo_> I have configure the sysroot path [19:05:14 CET] <JEEB> ok [19:05:24 CET] <metRo_> configured* [19:05:28 CET] <JEEB> I don't remember if I had to set something else as well for the NDK f.ex. [19:05:50 CET] <JEEB> been a while since I've used toolchains with their pc files having things meant for / instead of the prefix [19:06:27 CET] <JEEB> I think you should look into the general gcc/ld et al documentation [19:06:29 CET] <metRo_> JEEB aren't you confuse PKG_CONFIG_PATH with PKG_CONFIG_LIBDIR? [19:06:33 CET] <JEEB> no [19:06:38 CET] <JEEB> _PATH appends [19:06:41 CET] <JEEB> _LIBDIR overrides [19:07:01 CET] <JEEB> both actually otherwise do the same thing as far as I can see [19:07:05 CET] <metRo_> ah ok [19:07:25 CET] <JEEB> so if I have my own custom prefix for the same arch [19:07:31 CET] <JEEB> like in my home dir [19:07:33 CET] <JEEB> I use _PATH [19:07:45 CET] <JEEB> if I have a completely different arch and a prefix for it [19:07:49 CET] <JEEB> then I use _LIBDIR [19:08:15 CET] <JEEB> in your case it's just a case of that thing wanting a chroot/sysroot so that root actually becomes prefix in the compiler/linker [19:08:27 CET] <JEEB> you could try running shit in a chroot if you want to, that would be one way of poking at it [19:08:36 CET] <JEEB> but not sure if you'd have the normal tools then :P [19:10:59 CET] <metRo_> JEEB the last two sentences was above my knowlegde on this... :/ [19:21:48 CET] <metRo_> JEEB: thank you, searching for PKG_CONFIG_LIBDIR I end up find that the variable I need to set up is PKG_CONFIG_SYSROOT_DIR [19:21:48 CET] <metRo_> :D [19:21:53 CET] <metRo_> lets try! [19:22:26 CET] <JEEB> not sure what that sets [19:22:46 CET] <JEEB> but sure, if that overrides what the pc files' root means [19:22:48 CET] <metRo_> "At the time of writing, with version 0.25 of pkgconfig package, this is achieved mainly through the PKG_CONFIG_SYSROOT_DIR variable, which is set to the path of the sysroot, and is inserted in-between the -I or -L flags and the following path." [19:23:00 CET] <JEEB> ah [19:23:01 CET] <JEEB> ok [19:23:02 CET] <JEEB> yeah [19:23:22 CET] <JEEB> aka stuff to fix your shit when someone did chroot installations of things 8) [19:23:31 CET] <metRo_> I'm just installed other pakackes ffmpeg need but at least now it could test smbclient :) [19:24:10 CET] <metRo_> how should I have done it? I copied the files from rpi to my pc (host) [19:24:39 CET] <JEEB> dunno if it's strictly incorrect, but you have to take into mention in that case that your possible coded paths are gonna be wrong as they were for you [19:24:52 CET] <JEEB> you just have to make something fix that [19:25:01 CET] <JEEB> if that pkg-config variable does that, sure good [19:25:16 CET] <Hybridesque> Evening everyone [19:26:38 CET] <Hybridesque> Would I be right in thinking the static binaries (like from johnvansickle.com) don't have libmfx included in them? [20:35:03 CET] <waressearcher2> Hybridesque: hallo, wie geht's es dir ? [20:39:47 CET] <nimzter> Hi Everyone, just a simple question: How to loop invideo text. I'm fighting all day :) Thanks [20:40:49 CET] <nimzter> the filter: drawtext --> option: enable [20:44:55 CET] <Hybridesque> waresearcher2: Hallo! Nicht so gut. Entschuldigung, mein Deutsch schwach ist. [21:20:09 CET] <colinb> hello all! I am on Windows 7, been using ffmpeg for a while now (currently 20151105-git-c878082). I am doing a simple encoding of a video file, but I am not seeing any actual output file! ffmpeg thinks it has output a file, as it asks for overwrite conformation. but it is nowhere to be found. Even when I specify an absolute path. Any ideas on what I must be obviously doing wrong? [21:20:40 CET] <c_14> Hybridesque: afaik no [21:24:28 CET] <colinb> http://pastebin.com/JwB8uEhk [21:24:33 CET] <colinb> There we go! [21:28:19 CET] <c_14> It should be in C:\. try doing cd C:\ and then dir [21:28:54 CET] <colinb> one would think so .. but that is not the case [21:30:45 CET] <colinb> the baffling thing to me is that ffmpeg asks for the overwrite confirmation [21:31:58 CET] <c_14> Is this in the Windows terminal or something like cygwin? [21:32:27 CET] <c_14> Is the file maybe in the cwd? [21:32:41 CET] <Hybridesque> c_14: so I'd have to compile a build if I want to include libmfx to enable qsv for a linux static build. [21:32:55 CET] <c_14> Hybridesque: yes [21:34:18 CET] <colinb> Windows terminal [21:34:31 CET] <colinb> also, an interesting thing, the hard drive space is allocated [21:34:49 CET] <colinb> checked the cwd as well [21:34:59 CET] <colinb> as well as other semi-relevant directories [21:35:24 CET] <c_14> colinb: does ffmpeg -i C:\cut-1.mp4 find it? [21:37:06 CET] <colinb> as a matter of fact, it does! [21:38:00 CET] <colinb> perhaps it is one of those strange things that a good 'ol-fashioned reboot would take care of? [21:41:37 CET] <colinb> I'm gonna give that a go, akin perhaps to blowing on a NES cartridge but, I am out of ideas [21:45:10 CET] <Hybridesque> c_14: I've been reading the guide for compiling a build of ffmpeg, but not sure how to tie in the information about libmfx into it. bit new to linux and even newer to compiling. [21:45:46 CET] <prelude2004c> hey everyone good day :) [21:46:34 CET] <c_14> Hybridesque: you have to build the mfx dispatcher stuff, make sure it's somewhere ffmpeg will pick it up and then configure with --enable-libmfx [21:47:00 CET] <c_14> Hybridesque: then you also need to patch your kernel or something [21:47:28 CET] <colinb> Aaand that didn't do it. [21:47:30 CET] <prelude2004c> still looking for a bit of help with h264 decoding using vdpau. Took a few days off thinking that i was just tired but clearly not :) i still have not addressed the issue. Basically i am using the new nvidia M4000 card and it has h264 decoding capabilities. everytime i use the vdpau with an mpeg2video source it uses teh decoder just fine but soon as the source is h264 the encoder is used with the nvenc but the [21:47:30 CET] <prelude2004c> decoder sits at 0% usage. Anyone know why that could BE ? [21:48:01 CET] <colinb> even the windows file search doesn't find the files anywhere. [21:48:13 CET] <colinb> but the free space is diminished! [21:48:50 CET] <Hybridesque> c_14: didn't seem to be necessary from what I read though but I'll go away and have more a lookie. [21:50:25 CET] <c_14> colinb: it's a kind of magic. [21:50:42 CET] <c_14> colinb: if you do `ffmpeg -f lavfi -i testsrc:d=5 out.mp4' can you find out.mp4? [21:51:03 CET] <c_14> colinb: Have you tried checking if the file is hidden? [21:51:35 CET] <colinb> no such filter for testsrc:d [21:51:43 CET] <c_14> eh, testsrc=d=5 [21:51:46 CET] <c_14> Always mess that up [21:51:54 CET] <colinb> yes, I am displaying hidden files, as well as searching in the terminal [21:52:15 CET] <colinb> yup, I can see that out.mp4 [21:52:52 CET] <colinb> could it have anything to do with that last notice on my output which says 1 frame left in queue? so, hard drive space is being allocated but the final file isn't being created? [21:53:13 CET] <c_14> nah [21:53:56 CET] <c_14> Anyways, you can't allocate space without creating a file for it. [21:54:56 CET] <colinb> okay well it was the 'magic' reboot [21:55:36 CET] <colinb> now files are being created normally! sorry for bringing a fluke to the table [21:55:54 CET] <c_14> Windows is Magic [21:56:16 CET] <prelude2004c> so can anyone hint me why vdpaul wont' decode h264 but does mpeg2video ok ? nvidia M4000 which supports h264 so i don't get it [22:17:25 CET] <rjp421> how do i make ffmpeg listen on 0.0.0.0:1234 instead of localhost:80 when encoding to mjpeg? [22:19:31 CET] <c_14> rjp421: what's your command? [22:20:15 CET] <c_14> mjpeg itself doesn't have anything to do with ip addresses [22:22:34 CET] <rjp421> c_14, ffmpeg -loglevel info -r 10 -vcodec mjpeg -s 320×240 -f video4linux2 -i /dev/video0 "http://0.0.0.0:1234/webcam.mjpeg" [22:23:14 CET] <rjp421> i have multiple cams i want listening on their own port [22:23:39 CET] <c_14> rjp421: set -listen 1 [22:23:59 CET] <c_14> otherwise it won't start a listening server... [22:24:13 CET] <rjp421> ah ok ty [22:35:28 CET] <Hybridesque> c_14: just looks like I just need to build that mfx_dispatcher, don't need kernel patching it seems before compiling a build... still working on it so don't know the final result [23:07:57 CET] <Hybridesque> didn't work, going to try over from scratch, but feel like I'm on a good path now, I'm probably messed up in following the instructions [00:00:00 CET] --- Sun Dec 27 2015

1 0