Ffmpeg-devel-irc

Download

ffmpeg-devel-irc@ffmpeg.org

April 2013

1 participants
60 discussions

ffmpeg-devel.log.20130414
by burek 15 Apr '13

15 Apr '13

[01:18] <cone-86> ffmpeg.git 03Paul B Mahol 07master:2952ed2b6076: doc/filters: move telecine from audio to video filters [02:25] <cone-86> ffmpeg.git 03Paul B Mahol 07master:03e2ec32b88f: lavfi: add smptehdbars source [02:27] <ubitux> - * smptebars is by Paul B Mahol. [02:27] <ubitux> + * smptebars and smptehdbars is by Paul B Mahol. [02:27] <ubitux> durandal_1707: "are" ^ [02:28] <durandal_1707> ubitux: received too late [02:31] <ubitux> sent too late [02:31] <ubitux> i'm reviewing on cvslog [02:34] <durandal_1707> i dont read cvslogs [02:35] <durandal_1707> i wondering about filter that is doing something reverse of interleave [02:36] <durandal_1707> for example see: split[a][b],[a]select=not(mod(n\,2))[aa],[b]select=(mod(n\,2))[bb],[aa][bb]blend=average' [02:37] <durandal_1707> this is suboptimal, because if you change from even/odd queue gets full [02:59] Action: ubitux just did something kinda cool [03:00] <durandal_1707> sent patch to l? [03:01] <ubitux> haha no [03:01] <ubitux> i'll send the patchset soon, just need a few more adjustments :) [03:03] <durandal_1707> saste: did explored that histeq behaviour? [03:08] <ubitux> here we go. [03:08] <ubitux> cool patch is 5/5 [03:08] <ubitux> i'm not sure it's relevant in practice but well [03:10] <ubitux> i want to add some string interpolation with bprint and dict now [03:11] <durandal_1707> isn there already better/faster one with trees? [03:11] <durandal_1707> see libavutil/dict.h [03:11] <ubitux> possibly :) [03:12] <ubitux> but since our code use a lot of av dict [03:12] <ubitux> i thought it was a good idea :p [03:13] <ubitux> maybe it would be possible to use the tree inside the dict, but i'm not sure it will really simplify that code [03:16] <durandal_1707> maybe i will port virtualdub deflicker [03:17] <ubitux> yeah [03:17] <ubitux> that would be nice :) [03:17] <ubitux> a flicker filter as well btw [03:18] <durandal_1707> doesn't that just apply curves to luma [03:19] <ubitux> i was thinking of a degradation filter actually [03:19] <ubitux> with the vertical lines and random artefacts [03:27] <durandal_1707> looking at code is soo trivial.... [03:27] <cone-86> ffmpeg.git 03Clément BSsch 07master:d9be6e69cf00: lavfi/testsrc: grammar fix in comment after 03e2ec32. [10:35] <cone-349> ffmpeg.git 03Clément BSsch 07master:f359be96cacb: lavfi/smptehdbars: fix priv_class pointer. [11:36] <cehoyos> Hi! Can somebody explain the option "-to" to me? [11:36] <cehoyos> (What is the difference between -t and -to ?) [11:38] <ubitux> -to is an absolute time [11:39] <ubitux> while -t is a duration [11:39] <cehoyos> I tested the following: [11:39] <cehoyos> ffmpeg -ss 10 -i fate-suite/svq3/Vertical400kbit.sorenson3.mov -to 20 -an out.avi [11:40] <cehoyos> ffmpeg -ss 10 -i fate-suite/svq3/Vertical400kbit.sorenson3.mov -t 20 -an out2.avi [11:40] <cehoyos> out.avi and out2.avi are identical. [11:40] <ubitux> what if you move the -ss as output option? [11:41] <cehoyos> I just did and it works, but I don't find that 100% obvious. [11:41] <ubitux> maybe it's a bug [11:41] <cehoyos> It's at least not a regression [12:31] <ubitux> any idea why av_dict_copy is ignoring suffixes? [12:37] <ubitux> mmh i may be able to get this stuff a bit faster [12:46] <ubitux> michaelni: speedloss only happens with the non-default -sws_flags +full_chroma_int? [12:55] <cehoyos> Iiuc: Workaround for several reported bugs is -sws_flags +full_chroma_int which causes a speed regression, Michael's patches fix output for default, but also produces speed loss [12:56] <michaelni> cehoyos, yes [12:58] <cehoyos> Ticket 2467 looks interesting if somebody works on filters... [13:31] <cone-349> ffmpeg.git 03Michael Niedermayer 07master:662664551c80: sws/output: factor yuv2rgb_write_full() out [15:10] <ubitux> isn't it today that we can have some feedback on gsoc? [15:11] <ubitux> or i'm mistaken? [15:16] <michaelni> ubitux, dunno, this year saste & logan are admin and responsible for gsoc [15:24] <JEEB> it feels like an end of an era when you don't have ffmpeg (or its forks) nor VideoLAN and things under its umbrella in GSoC :< [15:24] <ubitux> yes but there is phpmyadmin so it's ok [15:25] <JEEB> hahaha [15:27] <michaelni> it would be interresting to know if there are other large projects that where in gsoc for long and then where no longer accepted multiple years [15:35] <michaelni> and of course whenever that "why was your project not accepted" thing is, we should have a few people at #gsoc and ask there [15:37] <michaelni> Re: [FFmpeg-devel] [RFC] Google Summer of Code 2013: We can get feedback as to [15:37] <michaelni> why we were not accepted in #gsoc (Freenode) on Friday, April 19, 2013 [15:37] <michaelni> at 16:00 UTC. [15:39] <cone-349> ffmpeg.git 03Michael Niedermayer 07master:9204a7dc8ee6: sws/output: add +yuv2rgb_full_2_c_template() [15:39] <ubitux> ah, 19, my bad, ok [16:00] <cone-349> ffmpeg.git 03Clément BSsch 07master:ab0ad6eccf38: lavfi: rename decimate to mpdecimate. [16:00] <cone-349> ffmpeg.git 03Clément BSsch 07master:7a92ec93c650: lavfi: port IVTC filters from vapoursynth. [16:00] <cone-349> ffmpeg.git 03Clément BSsch 07master:458d956b0964: doc/filters: move mpdecimate doc to a more appropriate position. [16:09] <ubitux> ok, what remaining mp filter is important? [16:12] <ubitux> btw, did anyone look at the differences between tinterlace and interlace? [17:08] <cone-349> ffmpeg.git 03Michael Niedermayer 07master:a4b5e45e2bbc: sws/output: add yuv2rgb_full_1_c_template() [17:08] <cone-349> ffmpeg.git 03Michael Niedermayer 07master:41ebb64511dc: swscale/output: fix alpha scale in the new functions [19:56] <jomp16> Hey, newbie question: how to compile FFMPEG for Android ARMv7 NEON? [20:06] <cone-349> ffmpeg.git 03Michael Niedermayer 07master:d4956b0bfbf9: sws/input: replace hardcoded rgb2yuv coefficients by table [20:08] <michaelni> jomp16, did you try google and if that fails the ffmpeg-user mailing list ? [20:09] <cehoyos> Yuvi: Do you have access to PPC64 OSX? [20:09] <michaelni> also there are android targets on http://fate.ffmpeg.org/ the parameters used to build them should be there toom these might be a startpoint as well [20:10] <cehoyos> gas-preprocessor seems to fail there. [20:10] <cehoyos> (fail: libavcodec/ppc/fft_altivec_s cannot be compiled) [20:17] <jomp16> michaelni, so, i can use the configure prefix and put a neon instruction? [20:21] <cone-349> ffmpeg.git 03Clément BSsch 07master:1e3104cd3cd1: Add avpriv_dsputil_init() and use it in lavfi where relevant. [20:33] <cone-349> ffmpeg.git 03Stefano Sabatini 07master:fba3a3bbfba3: doc: document -opencl_options option in ff* tools manuals [20:33] <cone-349> ffmpeg.git 03Stefano Sabatini 07master:f6c9a325b727: ffmpeg: show error message in case of av_buffersrc_add_frame_flags() failure [20:35] <ubitux> saste: so, is it possible to do some kind of timeline with interleave filter? [20:35] <ubitux> and how much can it does it? [20:35] <saste> ubitux, "timeline"? [20:36] <ubitux> basically can i blur from 1 minute to 5 minutes using the filter? [20:36] <ubitux> select='between(t,60,60*5):branch=1 [x], boxblur=2:2, [x] interleave' [20:37] <ubitux> would that work? [20:37] <cehoyos> Regarding "Add avpriv_dsputil_init": deshake currently crashes on sse2 for rx%16 != 0 - would it be a solution to pass the expected alignment to dsputil_init() ? [20:38] <saste> ubitux: I don't think so [20:38] <ubitux> saste: why does the random work then? [20:38] <ubitux> cehoyos: huh? a problem with the recent patch? or it just make you think of another issue? [20:39] <cehoyos> It just made me think of ticket 2443 [20:39] <ubitux> ok [20:39] <cehoyos> (There are two issues: It made me think of rx=20, there is a - probably different - issue with rx=64) [20:39] <saste> ubitux, interleave needs to have both inputs with available frames [20:40] <ubitux> saste: with your flickering blur effect it doesn't look like to be the case either, unless i'm missing something? [20:40] <saste> so it starts, ask for the first input, get the first frame [20:40] <saste> now the second queue still doesn't have frames, so it asks select [20:41] <saste> select will return frames from the first path, so it will fill the first queue and will start to drop frame [20:41] <saste> when the first frame arrives in the second output, you dropped already a lot of frames [20:41] <saste> of course we could extend the queue, but in that case we can end up by filling a lot of memory [20:42] <saste> right now we have a queue of 32 frames [20:42] <ubitux> right so that's the reason it works "most of the time" with the random 0.2 [20:43] <durandal_1707> hmm. cant interleave call request_frame when it actually needs one? [20:43] <saste> ubitux: yes, if you increase the decrease that value it starts to drop frames [20:43] <ubitux> durandal_1707: calling request_frame on select will only raise you the frames you don't want [20:43] <ubitux> if you're not in the range [20:43] <saste> durandal_1707, how do you know which frame to send if you don't have at least one frame in each queue? [20:44] <durandal_1707> i was wondering about reverse of interleave, but cant find good name (split is already taken, and was already bad pick, as split is more like clone) [20:44] <ubitux> anyway, so timeline not for today :( [20:44] <ubitux> durandal_1707: torn [20:44] <saste> ubitux, global enable function [20:44] <saste> and callback [20:44] <ubitux> hehe right [20:44] <durandal_1707> saste: i mean request frame inside interleave filter [20:45] <durandal_1707> and not using select because that fills queque [20:45] <saste> durandal_1707, what do you mean by "inverse" of interleave? [20:45] <saste> you could extend select, indeed i plan to do that [20:46] <durandal_1707> but avisynth have feature iirc that it does not produce frame that will be later dropped [20:46] <durandal_1707> saste: select takes frames and drop random ones .... [20:46] <saste> durandal_1707, what features do you want to achieve? [20:46] <durandal_1707> by inverse interleave i mean filter that take single input and produce multiple outputs [20:47] <saste> durandal_1707, see my branch option to select, and the idea of nicolas [20:47] <durandal_1707> and puting each input frame to one of outputs ... [20:47] <durandal_1707> saste: select should just select, and not do some other stuff ... [20:48] <durandal_1707> you are complicated already complicated filter [20:48] <saste> durandal_1707, you can have multiple outputs, and let select "selects" the output [20:48] <saste> for example [20:48] <durandal_1707> that would still fill up queque [20:48] <saste> select=n=3:e='mod(n,3)' [20:49] <saste> durandal_1707, we can avoid frame drop, but there is the possibility to fill memory [20:49] <ubitux> select=between(t,60,60*5) [a]; select=not(between(t,60,60*5)) [b]; [a][b] interleave [20:50] <saste> but i agree the queue size should be configurable [20:50] <ubitux> this would work, right? [20:50] <saste> yes [20:50] <ubitux> what if select would have 2 contexts? [20:50] <durandal_1707> queue size eats memory [20:50] <ubitux> (internally) [20:50] <ubitux> (to reproduce that behaviour) [20:50] <ubitux> is that possible? [20:51] <durandal_1707> what if i want to use blend filter on input and input+few frames forward/backward.... [20:52] <durandal_1707> what about those filter descriptions that have . at end and others that do not have? [20:53] <ubitux> i remember adding the periods to all of them [20:53] <ubitux> but maybe some new filters don't have them [20:54] <ubitux> ah i forgot in fieldmatch [20:54] <ubitux> otherwise only 3 filters are missing it, the filters from libav [20:54] <durandal_1707> codecs/formats do not have them [20:54] <durandal_1707> i find that dots little funny .... [20:55] <saste> durandal_1707, we discussed the convention [20:55] <saste> IIRC that was a suggestion from diego [20:56] <saste> that's because filter description is not a nominal form, but a full sentence with verb [20:56] <saste> note that we have the same framedrop problem with overlay as well [20:56] <saste> which is sometimes annoying [20:57] <saste> so I'd like at least to be able to extend the queue size with no recompilation [20:57] <durandal_1707> saste: that actually fix problem (to me its just band-aid) [20:58] <saste> durandal_1707, the problem is that we don't want to fill all the memory [20:59] <durandal_1707> so changing 32 to 64 is really fix? [20:59] <cone-349> ffmpeg.git 03Clément BSsch 07master:dfac37afd250: lavfi: add missing periods in filter descriptions. [21:01] <durandal_1707> michaelni: did you have patch that address that infinite loop when someone (like me) messes up .priv_class ? [21:01] <durandal_1707> ubitux: what happened to that flush patch? [21:01] <ubitux> durandal_1707: i felt uneasy :p [21:02] <ubitux> i could push it now though [21:02] <durandal_1707> i see no problem as default is to always flush .... [21:03] <ubitux> yes, but if you disable flushing it should not break [21:04] <ubitux> i'll push in a moment [21:06] <ubitux> so anyway [21:06] <ubitux> now that we have some ivtc filters [21:06] <ubitux> what do we drop? [21:06] <ubitux> cehoyos: you are interested in keeping pullup? [21:07] <ubitux> durandal_1707: what's left in mp except eq2 and spp? [21:07] <ubitux> maybe mcdeint? [21:07] <durandal_1707> ow (what/where/when/how it helps) [21:08] <ubitux> is it efficient? [21:09] <saste> perspective [21:09] <ubitux> it seems to fuck up the pts at least somehow [21:10] <ubitux> or maybe that's mp [21:10] <saste> sab? [21:10] <ubitux> ? [21:11] <durandal_1707> ubitux: its slow, its wavelet denoiser, so only reason to keep it is if it gives better output than other denoises in some if not all scenarios [21:11] <saste> mp=sab [21:11] <ubitux> saste: no, ow [21:12] <ubitux> ah, relevant filters. [21:12] <ubitux> ok [21:12] <ubitux> durandal_1707: ok [21:12] <ubitux> i want a fft denoiser... [21:12] <ubitux> :( [21:12] <ubitux> but no 2d fft (yeah i know i repeat myself) [21:12] <Compn> why is ivtc called fieldmatch? [21:12] <Compn> :P [21:13] <ubitux> Compn: because ivtc is fieldmatch+decimate [21:13] <durandal_1707> what is this phase filter? [21:13] <Compn> ubitux : what do they call it in vapoursynth ? fieldmatch ? [21:13] <ubitux> Compn: tfm in avisynth (tritical fieldmatch, tritical being the name of the dude), vfm in vsynth (i let you guess what it means) [21:14] <cone-349> ffmpeg.git 03Stefano Sabatini 07master:bc1847addf34: doc/filters: remove mention to telecine mp filter [21:14] <Compn> lol ok [21:14] <Compn> just making sure ... [21:14] <durandal_1707> softpulldown is another teleciner [21:15] <Compn> ubitux : i just want to make sure if someone tells a user he needs 'ivtc' that he can find out it means 'field match' [21:15] <durandal_1707> Compn: they already know that ... [21:16] <ubitux> Compn: first paragraph of fieldmatch says it all [21:16] <cone-349> ffmpeg.git 03Clément BSsch 07master:8de9bb6e5e52: lavf: remove some flushing in write_packet muxers callbacks. [21:17] <durandal_1707> the only reason to keep pullup is if it is faster than fieldmatch for 3:2 case [21:17] <ubitux> yes fieldmatch is not meant to be used for realtime [21:18] <durandal_1707> and there is no realtime flag? [21:18] <ubitux> you can make it faster by playing with the combmatch mode [21:18] <ubitux> also, decimate is slow and needs to queue 5 frames [21:20] <durandal_1707> well one dont need to use decimate too [21:20] <ubitux> well actually it does [21:20] <ubitux> the flickering is visible [21:24] <ubitux> ./ffplay ~/fate-samples/smjpeg/scenwin.mjpg -vf 'split=3[a][b][c]; [a]pad=iw*3,drawtext=text=origin:fontcolor=yellow[p]; [b]hqdn3d,drawtext=text=hqdn3d:fontcolor=yellow[h]; [c]mp=ow,drawtext=text=ow:fontcolor=yellow[o]; [p][h]overlay=w[p2]; [p2][o]overlay=w*2' [21:24] <ubitux> of course, configuration etc. [21:27] <durandal_1707> ubitux: i dont have drawtext, so tell what that does [21:27] <ubitux> http://ubitux.fr/pub/pics/_orig-hqdn3d-ow.png [21:30] <durandal_1707> it read some args that are never used [21:31] <cehoyos> ubitux: Yes, you told me it is faster than ivtc [21:31] <ubitux> cehoyos: so you're porting? ;) [21:32] <cehoyos> I think that would be exceptionally unproductive [21:32] <ubitux> :) [21:33] <cehoyos> actually inefficient, sorry [21:33] <ubitux> ask l. to do it [21:33] <ubitux> and compete with them [21:33] <durandal_1707> ubitux: porting? its just select=(3 of 5) [21:33] <ubitux> durandal_1707: pullup? [21:34] <ubitux> i believe split in multiple files and the asm all around is not just a decimate [21:36] <saste> ubitux, should I still implement per-frame eval support in volume? [21:36] <saste> the same way I did in overlay [21:36] <saste> (maybe with eval=init by default) [21:37] <ubitux> as you wish, i just want av_eval_is_const [21:37] <ubitux> and i can deal with the rest [21:37] <ubitux> :D [21:37] <michaelni> ./ffplay fate-suite/smjpeg/scenwin.mjpg -vf 'noise=1:20,split=3[a][b][c]; [a]pad=iw*3[p]; [b]hqdn3d[h]; [c]mp=ow=10:16:16[o]; [p][h]overlay=w[p2]; [p2][o]overlay=w*2' [21:38] <ubitux> is it a boxblur? [21:38] <ubitux> :D [21:41] <durandal_1707> boxblur give blur output [21:43] <ubitux> ow=10:16:16 too [21:43] <ubitux> somehow :) [21:43] <michaelni> try a fft :) [21:43] <durandal_1707> ubitux: its not that blur.... [21:45] <durandal_1707> mp=ow=20 segv here [21:45] <michaelni> yes, it doesnt like large depths [21:49] <michaelni> durandal_1707, patch for handling duplicate AVClasses posted [21:50] <durandal_1707> does it slows stuff? [21:51] <michaelni> ubitux, to compare with boxblur: ./ffplay fate-suite/smjpeg/scenwin.mjpg -vf 'noise=1:20,split=3[a][b][c]; [a]boxblur=1,pad=iw*3[p]; [b]boxblur=2[h]; [c]mp=ow=10:16:16:0[o]; [p][h]overlay=w[p2]; [p2][o]overlay=w*2' [21:51] <durandal_1707> but i wanted assert/warning and not workaround ... [21:52] <ubitux> :) [21:52] <michaelni> durandal_1707, easy to do by changing the code from the patch but should we disallow it at all ? [21:53] <ubitux> i think so for now [21:53] <durandal_1707> michaelni: i dont think allowing it is smart idea [21:55] <durandal_1707> michaelni: i still wanted to know what was meant to be used with unused args like delta & mode [21:56] <michaelni> i am fine with disallowing it too but does it make that much sense? i mean assert or infloop is about the same [21:56] <michaelni> mode was for switching between hard and soft thresholding IIRC [21:56] <michaelni> was MEANT for [22:00] <durandal_1707> so pullup is actually useful? [22:01] <michaelni> about the speed effects of the dup class patch, it is probably a 50% slowdown for the function [22:02] <durandal_1707> michaelni: i was just wanted to fix loop, so assert is better than machine hang [22:03] <durandal_1707> so if it is possible to detect loop in fast way - its optimal solution to me [22:05] <michaelni> it can be done with 50% slowdown or fast with ABI break or fast with complex code, iam not really likeing any of these options though [22:08] <ubitux> saste: i'm going to try to implement the enable thing [22:08] <ubitux> unless you want to do it [22:09] <saste> ubitux, go on [22:09] <saste> i have too many pending patches to work on [22:09] <ubitux> yes [22:14] <durandal11707> hmm, could lavfi do tehnicolor/sepia ? [22:16] <saste> durandal11707, curves? [22:22] <durandal11707> example needed [22:22] <ubitux> -vf curves=vintage [22:23] <ubitux> feel free to add more profiles [22:23] <ubitux> hint: google for "photoshop + curves + technicolor|sepia" [22:24] <ubitux> grab the settings from tutorials and use them in vf curves [22:27] <Yuvi> cehoyos: nope, all my ppc machines died [22:41] <ubitux> AVOptions are real dark magic :) [22:41] <ubitux> actually, more the classes [22:56] <cehoyos> Yuvi: Does the error message help you? [22:56] <cehoyos> The problem is that I don't even know how to debug gas-preprocessor... [22:58] <cehoyos> (ie: How do I show its output?) [22:58] <Yuvi> what error message? [22:58] <saste> cehoyos, the vid.stab dev preferred to work on an external lib so he has more development freedom [22:59] <Yuvi> set GASPP_DEBUG and it'll dump the preprocessed file to stdout [22:59] <saste> and it can be used outside the scope of ffmpeg, which is perfectly fine imo [22:59] <cehoyos> And he should have been told that this is not a good idea. [22:59] <cehoyos> Of course it is "fine", but that doesn't make it a good idea. [22:59] <cehoyos> We don't even test compilation for all external libraries. [23:00] <cehoyos> Yuvi: "./configure --cc='gcc -m64'" [23:00] <saste> cehoyos, we discussed that to death on ML [23:00] <cehoyos> (OSX gcc defaults to -m32) [23:01] <saste> it is imo a good idea [23:01] <cehoyos> The library is doomed to rot imo [23:01] <cehoyos> (Sorry, I did not start the discussion) [23:01] <Yuvi> is it llvm-gcc? [23:02] <Yuvi> because I don't think llvm's ppc backend was ever fully working [23:02] <cehoyos> Yuvi: "make V=1 libavcodec/ppc/fft_altivec_s.o" fails with "libavcodec/ppc/fft_altivec_s.S:693:Invalid mnemonic 'tocbase,'" [23:02] <saste> cehoyos, when it will rot will port it [23:03] <saste> (as know, ffmpeg preys on ill and rotting projects) [23:03] <saste> *as known [23:04] <cehoyos> I am unsure, but I suspect OSX ppc predates llvm-gcc [23:04] <cehoyos> "gcc version 4.0.1 (Apple Inc. build 5493)" <- is that sufficient to know? [23:04] <cehoyos> It fails identically with 4.2 [23:04] <Yuvi> yeah, that's pure gcc (that I don't have anymore actually) [23:04] <durandal11707> ubitux: gonna write colorbalance? [23:05] <ubitux> nope [23:05] <Yuvi> set GASPP_DEBUG and paste the preprocessed output? [23:05] <cehoyos> durandal11707: Do you want to look at the fate-random failure? [23:05] <ubitux> durandal11707: the only filter remaining in my TODO list is the vignette/lens one [23:05] <cehoyos> export GASPP_DEBUG=1 ? [23:05] <Yuvi> yeah [23:05] <durandal11707> ubitux: than you are leaving? [23:05] <ubitux> nope [23:06] <ubitux> i'm adding timeline support in lavfi [23:06] <durandal11707> cehoyos: i'm in filter land now, do not disturb me [23:06] <ubitux> i would love to write more filters, but i want a better fft api :p [23:06] <ubitux> haha [23:06] <cehoyos> LOL [23:07] <cehoyos> Yuvi: ok, I have a huge asm file now [23:07] <cehoyos> But the file numbers in the error message still don't match the file... [23:08] <cehoyos> But I found: ".quad .ff_fft_calc_altivec, .TOC.@tocbase, 0" [23:08] <Yuvi> yeah they won't, but something with tocbase should be obviously wrong [23:08] <Yuvi> hm [23:08] <Yuvi> I'm not familiar with ppc relocations [23:08] <cehoyos> tocbase is there twice, and the error is shown twice. [23:09] <cehoyos> I am not sure I know what relocations are ;-) [23:10] <Yuvi> does that work for linux ppc64? [23:10] <Yuvi> I don't think it ever worked for os x ppc64 at least [23:10] <cehoyos> Yuvi; There is a second error:" libavcodec/ppc/fft_altivec_s.S:726:Invalid mnemonic 'got(r2)'" - "ld r6, fft_data@got(r2)" [23:11] <Yuvi> yeah those are coming from extfunc/movrel in asm.S; gas-preprocessor is expanding things correctly [23:11] <cehoyos> Yuvi; Could you check fate.libav.org ? (My ip is blocked) [23:11] <Yuvi> dunno if they're valid for any ppc64 or if mach-o just doen't support them [23:12] <Yuvi> OK, looks like linux ppc64 works [23:12] <cehoyos> Ticket 56 indicates that it works [23:12] <Yuvi> err but darwin ppc64 says it works too [23:13] <cehoyos> It fails here (I tested it) [23:13] <cehoyos> How does the configure line look like? [23:13] <cehoyos> (Actually: How does configure's output look like?) [23:13] <Yuvi> --enable-gpl --arch=ppc64 --cc='ccache gcc-4.2 -m64' --disable-debug --enable-shared [23:13] <cone-349> ffmpeg.git 03Paul B Mahol 07master:6ffe9113026a: lavfi/testsrc: unbreak smptebars only build [23:14] <Yuvi> http://pastie.org/7549319 [23:14] <Yuvi> so no gas preproc [23:16] <cehoyos> Yes, it works fine without gas-preprocessor ;-) - LOL [23:16] <cehoyos> It's funny that configure requests something that does not work. [23:18] <cehoyos> Yuvi: OT: Did you never receive an email from me asking to merge gas-preprocessor changes (before it was discussed on ffmpeg-devel and here)? [23:20] <Yuvi> oh looks like I did, sorry I didn't see it [23:21] <cehoyos> np, I just wanted to konw... [23:21] <cehoyos> What would you need to fix gas / how do we find out if it can be fixed? [23:22] <Yuvi> oh wait is ARCH_PPC64 set? [23:22] <Yuvi> and CONFIG_PIC [23:23] <Yuvi> err, of course it is [23:23] <Yuvi> tocbase isn't anywhere else [23:23] <cehoyos> CONFIG_PIC is 1 [23:25] <cehoyos> CONFIG_PIC can be 0 (--disable-shared) or 1 (--enable-shared), both fail [23:25] <Yuvi> easiest way would be to check for [23:25] <Yuvi> .quad .Blah, .TOC.@tocbase, 0 [23:25] <Yuvi> or just disable asm for ppc64 / darwin [23:25] <Yuvi> since it's never worked [23:25] <cehoyos> There are a few asm optimizations outside of .S files [23:26] <Yuvi> isn't there a separate flag for those? [23:26] <cehoyos> ? [23:26] <Yuvi> inline-asm or something [23:28] <Yuvi> like just disable gnu_as [23:29] <Yuvi> I'll try to figure out how to do mach-o ppc64 relocations though, but I'll just be guessing [23:29] <cehoyos> I can test;-) [23:35] <cehoyos> Reimar could theoretically help imo, but he currently explains MPlayer code to wm4 ;-) [23:36] <ubitux> rha i can't get that passthrough working :( [23:37] <durandal11707> of what? [23:38] <ubitux> i'm adding the global filter option enable [23:38] <ubitux> so far the expression works fine [23:38] <ubitux> but i'm unable to make a passthrough filter_frame working [23:38] <ubitux> it seems to assert somewhere for some reason [23:38] <ubitux> i'm almost there! :( [23:42] <ubitux> AHAH! [23:42] <ubitux> it works! [23:42] <ubitux> timelines in lavfi \o/ [23:43] <ubitux> ./ffplay ~/samples/GoneNutty.avi -vf 'curves=vintage:enable=between(t\,2\,5)' [23:43] <ubitux> \o/ [23:44] <ubitux> 50 lines of diff [23:44] <ubitux> that was too easy [23:46] <durandal11707> hmm you added that to single filter? [23:46] <ubitux> no [23:46] <ubitux> generic option [23:46] <ubitux> works theorically with every filter [23:46] <durandal11707> i cant add enable option to my filter? [23:47] <ubitux> i can adjust the code so it is possible and take the priority [23:47] <ubitux> but you should not need it [23:47] <ubitux> i'll make a special callback for the passthrough for filters who need some local updates [00:00] --- Mon Apr 15 2013

1 0

ffmpeg.log.20130414
by burek 15 Apr '13

15 Apr '13

[01:38] <fatpony> is it possible to apply a filter between specific frames? i want to run yadif on one section of a clip but not on the rest to avoid quality loss [01:41] <fatpony> maybe using vf select? [02:59] <elkng> if I use two pass encoding ffmpeg first generates some log file with this kind of content http://sprunge.us/WdBI, that means it scan for every frame and decides what will be bitrate for every frame instead of one global bitrate ? so when I create that video how it can later say what bitrate of that video is if it was generated per frame basis ? [03:04] <hi117> elkng: in lossy audio when you do that it tries for an average bitrate [03:04] <hi117> so if you want 100 bits per second, it might say 51 bits for this second and 150 for the next, the average is ~100 [03:06] <hi117> my completely uneducated guess at the internals of video encodes say that they do the same thing though [03:06] <klaxa> depends on the codec in use though [03:06] <hi117> encoders* [03:07] <klaxa> for h264, according to the wiki at least, two-pass encoding is only really used if you want to match a certain filesize [03:08] <klaxa> >Will two-pass provide a better quality than CRF? [03:08] <klaxa> >No, though it does allow you to target a file size more accurately. [03:08] <klaxa> taken from http://ffmpeg.org/trac/ffmpeg/wiki/x264EncodingGuide [03:17] <elkng> I use "-c:v mpeg4" option for codec, but with two passes for same codec and same bitrate there is a huge difference, less noice and there are no visible "squares", but it takes about 3 times longer to encode [03:17] <elkng> s/noice/noise [03:19] <klaxa> sounds like it's what you want then, no? filesizes are similar i guess? [05:33] <elkng> http://i.imgur.com/86FakDo.jpg [06:24] <Trix> Hey there, using http://pastebin.com/6YhjhpCK I get huge amounts of blocking it seems, with any preset... [06:24] <Trix> bad colors/etc~ [06:24] <Trix> I cant figure it out [06:24] <Trix> tried a bunch [06:28] <Trix> actually [06:28] <Trix> i may be dumb [06:28] <Trix> seems to be with 10bit files [09:00] <dv--> ALSA buffer xrun... [09:00] <dv--> facepalm [09:24] <LithosLaptop> dv--: did you manage toget the colors working? [09:24] <LithosLaptop> *to get [09:25] <dv--> no, i gave up. it's not super important [09:26] <LithosLaptop> I mentioned last time that you try yuvj444p instead of yuv444p [09:26] <LithosLaptop> oh ok [09:26] <dv--> i need to use yuv420p anyway because i'm ultimately uploading youtube [09:26] <LithosLaptop> oh [09:26] <dv--> to* [09:26] <LithosLaptop> yeah, then there is nothing you can do [09:26] <dv--> well, it's weird [09:27] <dv--> i mean the color orange clearly exists in yuv420p [09:27] <dv--> but the encoder is replacing it with grey in parts [09:28] <LithosLaptop> weird [10:02] <crackerjackz> i'm trying to make this .mp4 video smaller i found something on the net that says to run ... ffmpeg -i input.mp4 -acodec mp2 output.mp4 but i'm not exactly sure that's what i want? what does that do exactly, i ran it over an hour ago and its still encoding. its a 700 mb file, does it normally take that long? [10:11] <zordiack> hi, could someone help me out with HLS streaming? [10:11] <zordiack> i'm trying get a working ffserver.conf but having no luck with it [10:15] <zordiack> http://pastebin.com/wR9pD0Gu [10:15] <zordiack> here is the current config [10:16] <zordiack> server starts and input stream is connected and feedind, but output is not working [10:16] <zordiack> i get this in ffserver output [10:16] <zordiack> Sun Apr 14 11:16:25 2013 Error writing output header [10:16] <zordiack> Sun Apr 14 11:16:25 2013 94.237.88.69 - - [GET] "/studio.m3u8 HTTP/1.1" 200 79 [10:17] <zordiack> and client says invalid data when prosessing input [10:18] <zordiack> anyone? [10:24] <LithosLaptop> crackerjackz: you need to figure out what bitrate you need to use to get the target size you want [10:25] <crackerjackz> LithosLaptop, how do i do that? isn't there like a gui i can use or something? [10:26] <crackerjackz> what if i just convert it to .flv or something,, will that reduce the file size? [10:26] <LithosLaptop> no, flv is just a container format like mp4 is [10:27] <LithosLaptop> do a ffmpeg -i input.mp4 [10:28] <LithosLaptop> past the result and let me check the bitrate of the source [10:29] <crackerjackz> LithosLaptop, http://pastebin.com/ahehZt5m [10:30] <LithosLaptop> hmm thats not what I wanted :) [10:31] <crackerjackz> LithosLaptop, today is the first time i ever used ffmpeg [10:31] <zordiack> anyone with HLS streaming help? [10:31] <crackerjackz> no idea how to use it... just found a command on a website and hope it didn't kill my computer :p [10:32] <LithosLaptop> yeah, I just wanted the output from 'ffmpeg -i input.mp4' [10:32] <LithosLaptop> without any output file specified [10:32] <crackerjackz> LithosLaptop, okay give me a sec ill pb that info to you [10:34] <crackerjackz> LithosLaptop, http://pastebin.com/setVRY4q [10:35] <LithosLaptop> as you can see the average bitrate of the file is 13744 kb/s [10:36] <LithosLaptop> you need to make that less to make a smaller file [10:36] <LithosLaptop> example: [10:37] <LithosLaptop> ffmpeg -i input.mp4 -c:v libx264 -b:v 9000k -c:a copy outfile.mp4 [10:38] <LithosLaptop> that would use a bitrate of 9000kbit/s for video and just copy the audio unaltered [10:39] <LithosLaptop> to calculate the bitrate you want to use for a specific size: total size = (bitrate of video + bitrate of audio) * duration of video in seconds [10:39] <LithosLaptop> I have to go afk [10:39] <LithosLaptop> bbl [10:45] <zordiack> please, i need to get iOS comptible stream out of ffserver [11:07] <LithosLaptop> baseline maybe? [11:07] <LithosLaptop> o it is [11:07] <LithosLaptop> nm [11:10] <zordiack> LithosLaptop: the pois is i need live streaming [11:11] <zordiack> i couldn't get mp4 to work over http [11:11] <zordiack> now i'm trying HLS but it's not working either [11:12] <zordiack> i have webm and flv streams which work just fine [11:12] <zordiack> but i need Apple support too [11:15] <LithosLaptop> I am not 100% sure what resolutions,codecs and container formats ios supports over HTTP [11:17] <zordiack> HLS container and MP4 encoding [11:17] <zordiack> also MP4 with rtmp streaming i think [11:17] <zordiack> but ffserver can't do rtmp serving [11:23] <LithosLaptop> have you tried mpeg4 instead of libx264? [11:28] <zordiack> i did, it gave me an error too [11:28] <zordiack> i'll reproduce it [11:29] <LithosLaptop> ah ok [11:29] <zordiack> [mpeg4 @ 00000000059222e0] timebase 333333/10000000 not supported by MPEG 4 stan [11:29] <zordiack> dard, the maximum admitted value for the timebase denominator is 65535 [11:29] <zordiack> Output #0, ffm, to 'http://piknik.wappuradio.fi:9001/feed2.ffm': [11:30] <zordiack> i didn't understand that at all [11:34] <LithosLaptop> yeah that is greek [11:47] <zordiack> there is also and option -f hls present but no documentation on how to use it in encoding? [12:00] <zordiack> anyone here with experience about HLS streaming? [12:07] <elkng> why ffmpeg wasn't that way it is now like 5 years ago ? [12:27] <fatpony> is it possible to apply a filter to a specific section of the video? i want to apply yadif to one section of a clip but not to the rest [12:30] <LithosLaptop> hmm [12:30] <LithosLaptop> no idea [12:30] <fatpony> :/ [12:39] <zordiack> now i got hls stream to work but it stops after ~minute [12:40] <LithosLaptop> wow, progress [12:40] <elkng> section as on some frames or region as on some place in frame ? [12:40] <zordiack> yeah but frustrating [12:40] <zordiack> i'm using /tmd/feed1.ffm as an input [12:41] <zordiack> frame= 2509 fps=313 q=34.0 q=34.0 size=N/A time=00:01:22.86 bitrate=N/A dup=11 dframe= 2570 fps=301 q=34.0 q=34.0 size=N/A time=00:01:24.90 bitrate=N/A dup=11 dframe= 2570 fps=284 q=34.0 q=34.0 size=N/A time=00:01:24.90 bitrate=N/A dup=11 d [12:41] <zordiack> it just stops like that [12:41] <zordiack> time is not progressing anymore and it stops producing segments [12:41] <elkng> I heard people invented pastebin like sites long time ago, does anyone tryed to use them ? [12:42] <LithosLaptop> usually for lots of text [12:42] <zordiack> your irc buffer overflew? [12:42] <LithosLaptop> 3 lines doesn't make sense [12:42] <elkng> no but my patience can [13:14] <LithosLaptop> does ffmpeg automatically apply dithering when for example converting from 24bit/96Khz tp 16bit/44.1Khz? [13:14] <LithosLaptop> *to [13:22] <xlinkz0> is it possible to change the metadata of a file 'in place' ? [13:23] <xlinkz0> just add the metadata not use an output file.. this takes a lot of time because it copies the input file frame by frame [13:24] <ubitux> xlinkz0: no you need to remux [13:24] <ubitux> but -c copy should make things faster. [13:24] <xlinkz0> but i just set metadata to the container not in the individual streams.. [13:26] <xlinkz0> but i lose the old metadata :( [13:37] <xlinkz0> how do i set more than one metadata key=value pair in one call? [13:38] <xlinkz0> ah nvm [13:38] <xlinkz0> but is it possible to keep the old metadata? [14:32] <zordiack> well if anyone is interested i think i nailed it at least somehow [14:32] <zordiack> it's bit of a hack but it's working [14:34] <zordiack> you create .ts stream output with ffserver and then use ffmpeg segmenter on that with -c:v copy and serve the files with some www-server [14:53] <DylanCl> Hello. I'm using a script to stream to twitch.tv that uses FFMPEG. Is it possible to make my script do something like this: http://i.imgur.com/bVw3Y1F.png. It makes it easyer to stream then :/ [15:26] <StFS> Hi. I'm trying to record a specific window on my desktop and I've pretty much figured it out except that my command now records my microphone instead of recording the actual sound that's being played in my speakers... can someone help me figure out how I can do that? [15:26] <StFS> The command I have is: ffmpeg -f alsa -ac 2 -i pulse -f x11grab -r 30 -s $WIN_GEO -i :0.0+$WIN_XY -acodec libvo_aacenc -vco dec libx264 -vpre lossless_ultrafast -threads 0 -y $1.mkv [15:41] <Stels> StFS: Open "pavucontrol", and select capture settings. [15:48] <StFS> Stels: hmmm... I don't see any Capture settings there? [15:54] <LithosLaptop> not related to the problem, but any reason why you are using libvo_aacenc? [15:55] <LithosLaptop> even the experimental built-in aac encoder gives better quality than libvo_aacenc :) [15:59] <StFS> oh? no... I just saw a warning when I tried using aac... it told me I had to add some option or use this one... [15:59] <LithosLaptop> -strict -2 [16:00] <StFS> ok... I'll use that then :) [16:00] <LithosLaptop> oh wait [16:00] <LithosLaptop> did it tell you other ones also? [16:00] <StFS> nope [16:00] <LithosLaptop> oh ok [16:02] <LithosLaptop> for the built-in aac encoder I find a bitrate of 256kbit/s and a -cutoff 15000 to sound ok [16:03] <StFS> ok so now -vpre doesn't seem to work [16:12] <StFS> LithosLaptop: nevermind... I had changed the command to avconv and that apparently doesn't support -vpre [16:12] <LithosLaptop> ah ok [16:13] <StFS> buuuut... unfortunately I'm still having trouble getting it to record the audio from the application rather than the mic [16:17] <StFS> when I record I see the application pop up in pavucontrol and I can change the "capture from" for ffmpeg but none of them work except the microphones [16:22] <StFS> ahh.. the "Monitor of Built-in Audio Analog Stereo" input device was muted. Things work fine when I unmuted that :) [16:22] <StFS> I would like to be able to do this from the CLI though... but this is better than nothing :) [18:52] <jomp16> Hey, newbie question: how to compile FFMPEG for Android ARMv7 NEON? [18:53] <atiti> hey guys, I'm trying to encode some v4l footage to mpeg-ps and have it returned in a buffer in the memory. For that I've setup a custom AVIOContext and did avformat_open_input() with an empty ("") file [18:54] <atiti> but I'm getting "Invalid data found when processing " while trying to do avformat_open_input() [19:14] <atiti> i guess noones here [19:16] <zap0> why? [19:18] <atiti> cos noones talking [19:39] <atiti> http://pastebin.com/1zbLnLA9 < anyone has any idea why it segfaults at av_set_parameters() ? [19:45] <klaxa> when in doubt, valgrind? [19:57] <atiti> not a bad idea, i thought your gona say gdb tho :p [19:59] Action: ubitux wonders what av_set_parameters is [20:01] <atiti> "Set the output parameters of the output video context (see output_example). Must be done even if no parameters in order to prepare the output format context." (http://wiki.aasimon.org/doku.php?id=ffmpeg:av_set_parameters) [20:49] <Smashcat> Hi, I'm having a weird issue with ffmpeg trying to create a video and transpose it "90 degrees clockwise") using ffmpeg -y -r 25 -i %04d.png -c:v libx264 -s 720x1280 -pix_fmt yuv420p -crf 15 -vf "transpose=1" out.mov - it rotates the content, but the video remains in portrait format. Using the exact same command in avconv, it's landscape. [20:50] <Smashcat> ... However, with some PNG's, ffmpeg does rotate it to landscape. Even though the input PNGs are exactly the ame width/height. Is there some metadata ffmpeg reads that prevents it from rotating correctly? [21:44] <Smashcat> Seems that reversing the width height works in ffmpeg. Love it when projects fork and they start changing options to break compatibility :) [00:00] --- Mon Apr 15 2013

1 0

ffmpeg-devel.log.20130413
by burek 14 Apr '13

14 Apr '13

[00:00] <cone-607> ffmpeg.git 03highgod0401 07master:8f99494c1ca5: avutil/opencl: merge opencl env to opencl context [00:01] <saste> thanks [00:28] <ubitux> durandal11707: i've just tried to de-telecined a telecined material with -vf telecine and it seemed to work fine, without the need for any settings [00:29] <ubitux> saste: do you want to have a look to the ivtc filters before i push? [00:29] <durandal11707> ubitux: change default pattern (23) [00:30] <saste> ubitux, if you see no reply from me in a day you can push it safely [00:30] <ubitux> ok [00:30] <ubitux> saste: i can wait a little more if your priority is to deal with the graph parse issue ;) [00:30] <ubitux> i'm not particularly in a hurry [00:30] <ubitux> i'd like some desperate comments on the ass/ssa stuff though [00:31] <saste> ubitux, not my area, maybe wait for nicolas? [00:31] <ubitux> yeah i know :) [00:31] <saste> which is the name of the patch anyway? [00:32] <ubitux> "[PATCH] pain in the ASS" or something like that [00:32] <ubitux> "subtitles: introduce ASS codec id and use it." [00:33] <ubitux> btw, it seems our mkv demuxer doesn't deal that well with the reference mkv files [00:42] <Compn> aurel was the ass and mkv maintainer :) [00:42] <Compn> dont let it drive you insane ubitux!!! [00:42] <durandal11707> so how that color issue in #2465 gonna be fixed? [00:42] <durandal11707> ubitux: what reference files? [00:43] <Compn> files on matroska.org ? [00:44] <ubitux> durandal11707: test7.mkv to be specific [00:44] <ubitux> from the official test suite from matroska.org indeed [00:44] <durandal11707> what it have as codec(s)? [00:44] <Compn> ubitux : got url is what we're asking for :) [00:44] <ubitux> i don't remember but that's not the issue [00:45] <durandal11707> l have it too? [00:46] <Compn> http://www.matroska.org/downloads/test_w1.html [00:46] <ubitux> durandal11707: it's not a regression or anything afaik [00:46] <Compn> damaged mkv file [00:46] <Compn> :P [00:46] <Compn> is test7 [00:46] <durandal11707> huh, is this site nonsense, or my opera got hacked? [00:47] <ubitux> https://sourceforge.net/projects/matroska/files/test_files/matroska_test_w1… [00:47] <Compn> durandal11707 : whats it look like (i'm also using opera) [00:47] <Compn> just looks like a wiki to me [00:47] <cone-607> ffmpeg.git 03Clément BSsch 07master:dcea58502c9b: lavfi: drop now unused shorthand field from AVFilter. [00:47] <durandal11707> Compn: stupid cursor when you hever over some underlined stuff [00:48] <Compn> oh the double underlined links ? [00:48] <durandal11707> this is 176.mb [00:48] <Compn> thats some ad revenue generating script on the website. but no idea if they put it there or someone added it to the wiki :P [00:48] <durandal11707> :))) [00:50] <ubitux> so, when is anton going to send us a new storm? i'm getting bored [00:51] <durandal11707> i searched for that file and found #545 bug report [00:51] <saste> ubitux, you could implement filter reconfiguration, and then fix it when anton reimplement it differently [00:51] <saste> or we can just wait [00:51] <saste> i plan to work on 0mq, but i should fix sox (and splice) before [00:52] <ubitux> it's funny how we are able to add so much filters given the constraints of lavfi, while l is doing the opposite trying to rethink the whole framework without having any testing filters yet [00:53] <durandal11707> something like tile, but takes multiple inputs (better that StackVertical/Horizontal from vapour) [00:53] <saste> worse is better [00:54] <saste> then we should tackle scripting and high level binding [00:55] <Compn> someone requested an audio filter to get rid of audio pops and glitches [00:55] <Compn> like from vinyl album recordings [00:55] <Compn> need to be de-popped [00:55] <durandal11707> but stackh/v could be done with splice+tile, no? [00:57] <durandal11707> Compn: it should be possible, just port relevant sox filters [00:57] <durandal11707> the only thing i'm not sure is possible, but is required is writing to file [00:58] <durandal11707> so filter would scan audio and output data .... [00:58] <durandal11707> A->? [00:58] <durandal11707> than another filter takes data and audio and produce de-popped audio: A?->A [00:59] <durandal11707> i don thing how good/useful/reliable would be to make it single pass: A->A [01:27] <Compn> i dont think theres any actual technical reason we cant do 2pass video encoding in one pass [01:27] <Compn> just that no one does it [01:27] <Compn> like reverse playback. its possible , but one of those things [01:33] <BBB-> Compn: do 2 pass encoding in 1 pass? [01:33] <BBB-> what does that even mean [01:34] <BBB-> you mean ffmpeg -i file -vcodec mpeg4 -pass 1 and 2 at the same time out.file ? [02:40] <Compn> BBB- : yep [02:40] <Compn> just needs more buffer or something [02:42] <BBB-> just seek back after pass1 [10:10] <cone-7> ffmpeg.git 03Janne Grunau 07release/1.2:f0dc5c04194f: vaapi: fix argument for ff_vaapi_common_end_frame call (cherry picked from commit b066d90211072c7532e17c0c54d8475f10fc97ad) [10:36] <kurosu> I'm about to add a test for aac; I can send a patch for the makefile part, but what about the 3MB worth of data files? [10:56] <kurosu> michaelni: regarding r5q = countq, I guess I need to allocate 1 more reg for UNIX64 and set NOISE_TABLE to r6q then ? [10:56] <kurosu> I can't test that [11:05] <kurosu> bah, I lost my current tree somewhere between a rebase and a cherry pick... sigh :/ [11:26] <michaelni> kurosu, if you want to add a fate test just post a link or upload to the files for fate to our ftp, also smaller files are preferred if possible [11:27] <michaelni> about unix/pic/x86_64, yes another register seems the obvious solution [11:28] <kurosu> michaelni, ok, but there are issues (license, big difference between conformant output and current output) so I mailed alex converse for clarifications [11:28] <kurosu> so I'll postpone that part [11:28] <michaelni> ok [11:29] <kurosu> sorry for the fumbling, I'm not very proficient at git, and managing reviews on every side and merging the result is not easy for me [11:29] <kurosu> regarding ac3, I prefer not being in charge of anything, my time being limited - it just happens that I have 2-3 weeks where I can have time off / enough motivation now [11:31] <michaelni> well, you can just review the ac3 stuff as long as you have interrest/motivation [11:39] <burek> we discussed some time ago something regarding yuv444p in libx264, which confuses people, since most players expect yuv420p, and we concluded that if the output is yuv444p then the logical behavior would be to just copy that to output, but this: http://pastebin.com/bWbWgurP [11:40] <burek> shows that the input is bgr0 and the output is still yuv444p (default i guess), when in fact it should (imho) be yuv420p [11:40] <nevcairiel> why [11:40] <nevcairiel> bgr is a 4:4:4 format [11:40] <nevcairiel> it should use the format as close to the original as possible [11:40] <burek> oh that's the idea.. ok [12:03] <ubitux> < kurosu> bah, I lost my current tree somewhere between a rebase and a cherry pick... sigh :/ // git reflog can save you [12:38] <durandal_1707> saste: why i cant limit fps for sdl muxer? [12:39] <saste> durandal_1707, don't know [12:58] <kurosu> ubitux, thanks I was indeed pointed to that solution [13:36] <cehoyos> durandal_1707: Do you think that there really is a xv driver that supports anyhting else than yuv420p ? [14:12] <saste> durandal_1707, why telecine is documented amongst audio filters? [14:14] <cone-86> ffmpeg.git 03Stefano Sabatini 07master:87dd62e14198: lavfi/testsrc: make use of AV_OPT_TYPE_DURATION [14:46] <burek> does anyone know where can i ask questions about strange diff output? :) [14:46] <burek> is there an official website for that tool or something [14:48] <ubitux> man diff [14:48] <ubitux> GNU diffutils home page: <http://www.gnu.org/software/diffutils/> [14:49] <ubitux> what's the "strange diff output"? [14:49] <burek> it creates such a diff where it marks all the lines with - and then marks new changes with + (for the whole file) when there are just 2 new lines added.. [14:49] <burek> http://ffmpeg.gusari.org/paste.txt [14:50] <burek> i'll try with that url, thanks [14:50] <iive> burek: whitespaces ? [14:50] <burek> -b? [14:51] <iive> more like -w [14:51] <nevcairiel> sounds like EOL differences [14:51] <iive> e.g. it could be difference in cr/lf [14:52] <burek> oh, i see [14:52] <ubitux> tab vs spaces [14:52] <ubitux> maybe :) [14:52] <burek> yes yes i got it, thanks a lot for that tip :) i'll check it :) [14:59] <burek> thanks guys :) it was windows/unix crlf issue :) [15:08] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:9b672d4017bb: vf_mp: mp buffers are not compatible with the reference count system [15:54] <nevcairiel> really, how many problsm can be in one 10 line function [16:09] <cone-86> ffmpeg.git 03Hendrik Leppkes 07master:cac309611cf3: lavf/mxf: fix parsing of the month from mxf timestamps [16:10] <mateo`> michaelni: i didn't saw your reply :( obviously the bit masks are wrong :( [16:11] <mateo`> forgot an extra F :( [16:12] <michaelni> mateo`, want to send a patch ? [16:13] <ubitux> why the 0xf are wrong? [16:14] <mateo`> it does not cover 8bits [16:14] <ubitux> oh, 8 bits, ok [16:14] <ubitux> year is on 16 bits? [16:15] <ubitux> mateo`: you could use an union btw [16:15] <michaelni> other software had a year 2000 bug, we have a 16th hour and 16th minute bug [16:17] <mateo`> ubitux: i used mxfenc as reference, i better check the specs [16:17] <mateo`> sorry about this mistake guys [16:24] <nevcairiel> i noticed the 0xf stuff only after i sent the patches, or i would've fixed it directly as well [16:24] <nevcairiel> i only looked at the month line, where it worked =p [16:26] <nevcairiel> anyhow, the other patches are more important, they actually fix msvc, this one i just found weird :p [16:32] <mateo`> nevcairiel: shouldn't we issue a warning if month and day are not valid ? [16:33] <nevcairiel> you tell me how the mxf spec describes those values [16:34] <nevcairiel> the file in fate has one date with the timestamp 0 [16:35] <nevcairiel> and month -1 or day 0 cause errors in strftime [16:40] <mateo`> from S377M: mxf timestamp is year (16), month (8), day (8), hour (8), min (8), sec (8), msec (8). A value of 0 for each field means timestamp 'unknown' [16:46] <nevcairiel> so what do you do if the month is "unknown"? :) [16:46] <nevcairiel> only print a time? [16:47] <BBB-> kurosu: hey you're online [16:50] <mateo`> nevcairiel: from what i understand, "unknown" is the value of the time if all field are set to 0 [16:51] <mateo`> in theory month/day/hour/... fields should be checked ... [16:52] <nevcairiel> zero hours/minutes is at least valid [16:53] <mateo`> i was thinking about values above 60 for these fields [16:53] <nevcairiel> oh right [16:53] <nevcairiel> not a problem right now [16:53] <nevcairiel> :P [16:54] <mateo`> sure ... [16:54] <nevcairiel> cant even reach 16 :) [16:57] <cone-86> ffmpeg.git 03Hendrik Leppkes 07master:8da2a3aba3a7: lavf/mxfdec: use more widely supported time format string [17:02] <mateo`> nevcairiel: i'll double check the ts values with other mxf tools [17:07] <kurosu> BBB-: yeah, been so for a week [17:07] <kurosu> trying to optimize my still inefficient patch submission [17:17] <mateo`> nevcairiel: maybe year should be checked as well ? [17:17] <nevcairiel> -1900 is at least valid for msvcs strftime [17:18] <nevcairiel> there is a upper limit of something like 8900 [17:18] <nevcairiel> not sure how much checking is needed there [17:19] <nevcairiel> i guess the limit means years 0 to 10000 are supported [17:19] <nevcairiel> or something [17:24] <mateo`> what is the behaviour of strftime with invalid values in msvcs ? [17:25] <nevcairiel> it looks like its trying to call some error handler, but for some reason that causes it to crash [17:25] <mateo`> doh ! [17:25] <nevcairiel> may only be crashing on my system because i enforce no-execute on all apps [17:25] <nevcairiel> which isnt the default [17:26] <nevcairiel> and the error lists that as the reason [17:30] <RobertNagy> why does video_get_buffer use pic->format to get the AVPixFmtDescriptor, and in all other places use s->pix_fmt? [17:30] <RobertNagy> seems inconsistent... [17:32] <mateo`> nevcairiel: i guess the patch is ok then [17:37] <mateo`> nevcairiel: isn't the mday check superflous since the value (timestamp >> 32 & 0xFF) is of type uint64_t ? [17:37] <cone-86> ffmpeg.git 03Matthieu Bouron 07master:c8b364449424: lavf/mxf: fix parsing of timestamps [17:37] <nevcairiel> but a day of 0 is not valid [17:37] <nevcairiel> month starts at 1 [17:38] <mateo`> nevcairiel: i confused month/day ... [18:12] <cone-86> ffmpeg.git 03Hendrik Leppkes 07master:423089e96441: lavf/mxfdec: validate month/day before date formatting [18:13] <nevcairiel> yay that should make fate happy again [18:21] <ubitux> possibly not only fate [19:58] <cone-86> ffmpeg.git 03Clément BSsch 07master:73180f5b0efd: lavfi: flag more dynamic i/o filters. [19:58] <cone-86> ffmpeg.git 03Clément BSsch 07master:2b44195fc9db: cmdutils: print if filters are sources or sinks. [20:09] <cone-86> ffmpeg.git 03Clément BSsch 07master:8137ea358c71: doc/filters: use common options introduction string. [20:19] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:ded2f187a272: avcodec: fix ignorecrop AVOption offset [20:23] <ubitux> michaelni: seems there are more in v4l2, drawtext, and fps [20:23] <ubitux> michaelni: is it fixing a bug? [20:28] <michaelni> no, not that i know [20:29] <michaelni> also having the offsets equal to the parent field of a constant isnt that wrong having the offset being equal to a random value is wrong [20:31] <ubitux> sure [20:31] <ubitux> ok [20:34] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:536baf6cc5d1: oggenc: Fix pref_duration AVOption type [20:44] <ubitux> michaelni: do we have a fate instance with assert level >= 2? [20:45] <michaelni> good question [20:47] <cone-86> ffmpeg.git 03Nicolas George 07master:6fbb21d6858b: lavfi/buffersink: factor checks for lists sizes. [20:47] <cone-86> ffmpeg.git 03Nicolas George 07master:76c8060654f0: lavu: add av_pure to av_int_list_length_for_size. [20:47] <cone-86> ffmpeg.git 03Nicolas George 07master:2a1d7ea5f8ba: lavu: add parens to macro argument. [20:47] <cone-86> ffmpeg.git 03Nicolas George 07master:9e8e03de38b3: lavu/opt: check int lists length for overflow. [20:47] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:38e66af43b06: Merge remote-tracking branch 'cigaes/master' [21:24] <cone-86> ffmpeg.git 03James Almer 07master:572e38a51313: log: Remove redundant va_list creation from missing_feature_sample() [23:53] <cone-86> ffmpeg.git 03Clément BSsch 07master:28ac8426e646: lavu/dict: cosmetic realign. [00:00] --- Sun Apr 14 2013

1 0

ffmpeg.log.20130413
by burek 14 Apr '13

14 Apr '13

[10:27] <dv--> what I'm recording: [10:28] <dv--> oops. what I'm recording: http://i.imgur.com/2w3C7RX.png, ffmpeg -q 0: http://i.imgur.com/5WJXXoQ.png [10:28] <dv--> is there any way to improve that? [10:32] <relaxed> dv--: ? png is lossless [10:34] <dv--> http://pastebin.com/q8N8Qpt0 [10:35] <relaxed> you want lossless? [10:36] <dv--> i'd like something closer to lossless than what i'm getting [10:36] <relaxed> try -crf 20 [10:37] <dv--> the orange text in particular looks really bad [10:37] <relaxed> lower the value for higher quality [10:37] <dv--> isn't -q 0 already -crf 0? [10:38] <relaxed> -qp 0 [10:40] <dv--> i tried it but with -q 0 -qp 0 it looks identical to just -q 0 [10:44] <relaxed> did you try -crf 0? [10:45] <dv--> yeah, it's the same [10:47] <relaxed> show me the pastebin [10:49] <dv--> http://pastebin.com/bWbWgurP [10:51] <dv--> if I use yuv420p it looks even worse [10:52] <LithosLaptop> and if you use libx264rgb? [10:54] <dv--> hm [10:54] <dv--> i can't play it with vlc, looks terrible in mplayer (very desaturated), also when recording i'm getting alsa buffer xruns [10:55] <LithosLaptop> hmm ok [10:57] <dv--> maybe i'll just live with slightly suboptimal font rendering [11:01] <LithosLaptop> tried huffyuv or ffv1? [11:02] <dv--> huffyuv works but the files are huge [11:02] <dv--> i'll try ffv1 [11:02] <relaxed> it's strange you getting bad quality with libx264 lossless [11:02] <relaxed> it should look exactly the same [11:03] <LithosLaptop> yeah [11:08] <LithosLaptop> dirac has a lossless mode, but I am not sure how one specifies it in ffmpeg [11:31] <cryptopsy> after i used ffmpeg to make an avi from an mp3, sound doesn't work, and /etc/init.d/alsa restart, doesn't fix it [11:33] <cryptopsy> what for? [11:34] <burek> for the balance among the stars.. [11:34] <cryptopsy> $FFMPEG -loop 1 -i $1 -i $2 -vcodec mpeg4 -threads 4 -acodec copy -y -t $TIME $3 [11:34] <burek> forget it [11:34] <cryptopsy> that's it [11:37] <cryptopsy> burek: alsa can i eat a dick, i can reboot after saving my instance instead of wasting hours debugging for the alsa team [11:37] <cryptopsy> can eat* [11:37] <burek> dv--, relaxed, does bgr0 -> yuv444p has got anything to do with that [11:37] <cryptopsy> have no idea who you're talking to [11:37] <burek> it's ok, i know who am i talking to and that's all that matters [11:38] <cryptopsy> k then i'll disregard [12:14] <dv--> burek: how would I check that... [12:15] <burek> dv--, try something like: ffmpeg ... -vcodec libx264 -pixel_format yuv420p ... [12:16] <dv--> yeah i've tried yuv420p. it looks worse [12:20] <LithosLaptop> and libx264rgb doesn't work in vlc? [12:23] <dv--> it looks solid green [12:28] <LithosLaptop> ah ok [12:28] <LithosLaptop> and yuvj444p? [12:29] <LithosLaptop> the j was intentional [12:54] <burek> if you used -crf 0 then only thing that could change the visual perception of colors is the colorspace conversion due to changing pixel formats, i guess.. [12:57] <LithosLaptop> yuvj444p is 0-255 and yuv444p is 16 to 235 I think [16:50] <dregan> hi guys, who has a link to latest ffmpeg compiled or rpm for centos 6? [18:23] <burek> dregan [18:30] <JodaZ> 2.4 ? [21:48] <elkng> http://sprunge.us/VceO those are 4 commands I use to convert same "mkv" video into "avi" but using different bitrate, and http://sprunge.us/LTDL are "avi" files, but they have bitrate not that I expected, those are "800.avi - 939 kb/s" "1000.avi 1139 kb/s" "1200.avi 1142 kb/s" "1500.avi 1143 kb/s", file that should be about 800 kb/s bitrate is quite close to that bitrate but others that should be close to 1000,1200,1500 arn't they are all about same bitrat [21:48] <elkng> codec decides to make it that way ? [21:52] <elkng> any ideas of that behaviour ? I'm using ffmpeg version 1.2 [22:24] <relaxed> The encoder probably can't use that many bits for the given source. [22:24] <relaxed> Just because you use -b:v 1000000000M doesn't mean the encoder can comply :) [22:37] <elkng> now that I think about it initial mkv video has 1333 kb/s bitrate [22:38] <elkng> so it can't be converted to 1500 kb/s ? [23:07] <cbreak> sure it can [23:07] <cbreak> just add noise [23:07] <cbreak> or something else that's hard to compress [23:07] <cbreak> then reencode [23:07] <cbreak> of course the quality will suffer, but you will get a bigger file [23:24] <elkng> you eat clowns on breakfast ? [00:00] --- Sun Apr 14 2013

1 0

ffmpeg-devel.log.20130412
by burek 13 Apr '13

13 Apr '13

[00:03] <ubitux> no one wants to deal with scale filter? :( [00:06] <michaelni> ubitux, you could just drop the single unnamed parameter case [00:07] <ubitux> well it's not that it's hard, it just takes me quite some time to do all these filters [00:07] <ubitux> i'll go for pan now [00:08] <ubitux> which is tricky as well [00:08] <saste> michaelni, i have a few patches to push [00:09] <saste> do you want me to wait, or you're already done [00:13] <ubitux> pan is actually gonna get extremely tricky [00:13] <ubitux> unless i run some escaping just like mp [00:14] <michaelni> saste, i wrote a script to rebase a merge so its less annoying now, at least for simple case [00:14] <michaelni> cases [00:21] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:a77454ec145b: lavfi/overlay: reindent constants in option list [00:21] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:ed2c827575a6: lavfi/overlay: add repeatlast option [00:21] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:aff6cebb4166: lavfi/overlay: fix crash in case of invalid expression [00:21] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:b2e22731e4ee: doc/filters: apply minor fixes to hue Commands section [00:21] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:fdedfc029b35: doc/filters: rework Commands section for overlay filter [00:21] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:9da369604ecf: lavfi/overlay: improve feedback in case of invalid expression [00:22] <ubitux> saste: thx for a77454ec :) [00:34] <ubitux> well i'm really too tired to deal with the remaining filters [00:39] <cone-607> ffmpeg.git 03Anton Khirnov 07master:bc1a985ba030: lavfi: replace avfilter_open() with avfilter_graph_alloc_filter(). [00:39] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:86070b8e5a0e: Merge commit 'bc1a985ba030e9861d24965d42792850b43a43ea' [00:44] <cone-607> ffmpeg.git 03Anton Khirnov 07master:c2c9801bc9bc: lavfi: deprecate avfilter_graph_add_filter(). [00:44] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:5c68e14e42c1: Merge commit 'c2c9801bc9bce688d51d1a96f5f3ea93933e2dee' [00:48] <durandal11707> ubitux: did you tested ivtc with our telecine? [00:48] <ubitux> no, not yet [00:51] <cone-607> ffmpeg.git 03Anton Khirnov 07master:111367263af4: lavfi: add AVFilterContext.graph. [00:51] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:4fde70539675: Merge commit '111367263af41c88a44bd763ceefc11d53a7f655' [00:59] <cone-607> ffmpeg.git 03Anton Khirnov 07master:1565cbc65cbb: lavfi: make avfilter_free() remove the filter from its graph. [00:59] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:eb0f774d4bfd: Merge commit '1565cbc65cbb9f95c11367314a080068895e0cf0' [01:02] <ubitux> any idea what anton is planing with the filter removing from a graph? [01:05] <durandal11707> what that silly thing that guy removed in tinterlace filter? [01:05] <ubitux> ? [01:08] <saste> ubitux, dynamic reconfiguration? [01:09] <ubitux> saste: that's what i thought, but why is it necessary? [01:10] <ubitux> i mean, why removing the filter from the graph actually helps [01:10] <saste> only anton knows, and they don't discuss design on list, they just say "ok" [01:10] <ubitux> :D [01:38] <cone-607> ffmpeg.git 03Anton Khirnov 07master:48a5adab62bd: lavfi: add avfilter_init_str() to replace avfilter_init_filter(). [01:38] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:710b0aa8012a: Merge commit '48a5adab62bd2a553f5069d41fa632a0701835e5' [02:40] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:9ab796127a38: avfilter: Update scale to dictionary based options [09:40] <cone-607> ffmpeg.git 03Pavel Koshevoy 07master:8f3c440a39e8: lavfi/atempo: switch to an AVOptions-based system. [09:54] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:16e5e13c7e46: buffersrc: switch to an AVOptions-based shorthand system. [10:00] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:c378ba19b5c0: abuffer: switch to an AVOptions-based shorthand system. [10:20] <cone-607> ffmpeg.git 03!5@6 8A:>@>4>2 07master:9470b541e5c4: af_pan: fix parsing arguments from the example in the documentation [10:33] <durandal_1707> anybody working on remaining filters? [10:34] <durandal_1707> this is getting ridiculous, how long it passed why I uploaded gif disposal background to fate and still same machines have not done rsync [10:34] <durandal_1707> s/why/when [10:36] <ubitux> durandal_1707: i'm not on them [10:37] <michaelni> durandal_1707, which machines miss it ? [10:39] <durandal_1707> michaelni: only one, cygwin something, http://fate.ffmpeg.org/history.cgi?slot=x86_64-win7-cygwin-gcc-4.5.3 [10:41] Action: michaelni is updating af_pan [10:42] <michaelni> michaelni, sent mail to beastd about cygwin (i think he is admin of it) [10:43] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:3c821e755086: af_pan: switch to an AVOptions-based shorthand system. [10:45] <durandal_1707> michaelni: that commit added some funny history in log output [10:47] <durandal_1707> ugh, ignore, i did not know about addititonal hacks added there [10:48] <ubitux> michaelni: will that pan commit enough to deal with the '='? [10:48] <ubitux> pan = 5 : c0=c0 : c1=c1 : ... ? [10:49] Action: ubitux cleaning the doc [10:49] <durandal_1707> but why we are updating filters if init_filter still have >1 args? [10:50] <michaelni> ubitux, do you have a failing command line ? [10:52] <ubitux> michaelni: if pan=stereo:c0=c1:c1=c0 works then it should be ok [10:54] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:80116b6e6755: af_pan: avoid spaces before % to avoid possible issues with MSVC [10:55] <michaelni> ubitux, no errors just the | vs : spam [10:56] <ubitux> ok well then i guess it's good [10:59] <cone-607> ffmpeg.git 03Clément BSsch 07master:8876b0816d39: doc/filters: update blackframe options description. [10:59] <cone-607> ffmpeg.git 03Clément BSsch 07master:fc15ad408c78: doc/filters: update boxblur options description. [10:59] <cone-607> ffmpeg.git 03Clément BSsch 07master:9811b7942db3: doc/crop: update boxblur options description. [10:59] <cone-607> ffmpeg.git 03Clément BSsch 07master:9c5f392cbcf0: doc/cropdetect: update boxblur options description. [11:00] <ubitux> looks like it's gonna take me the day [11:00] <durandal_1707> what? [11:01] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:04924bc923ef: avfilter: remove old shorthand system [11:02] <ubitux> michaelni: i think aconvert will need some update as well actually (if you want to drop the args from the init) [11:03] <ubitux> durandal_1707: most of the filters need an update of their doc; i need to check for aliases, order, non-merged doc changes, etc [11:03] <durandal_1707> am i the only one that think vf_copy code can be simplified? [11:04] <ubitux> what is copy used for? [11:04] <durandal_1707> how to handle this destruct deprecated warning in lavf? [11:04] <durandal_1707> ubitux: to copy frame unchanged [11:05] <ubitux> sounds like an ineffective passthrough [11:05] <durandal_1707> see b5670209438481f21addc8b7c4891627dd5ddac7 [11:05] <durandal_1707> avfilter_draw_slice is gone [11:06] <durandal_1707> and it was much simpler in old days [11:06] <ubitux> the filter looks completely useless nowadays [11:06] <ubitux> unless i'm missing something [11:07] <durandal_1707> leave it to l to deprecate it? [11:07] <ubitux> i guess so yes [11:07] <ubitux> if you want to play with copies, just use (a)perms [11:07] <ubitux> otherwise i don't see much point in this [11:08] <durandal_1707> i thought about replacing that code with clone_frame [11:08] <durandal_1707> or clone != copy? [11:09] <ubitux> it will copy if necessary [11:09] <ubitux> but really it makes no sense to "copy" anyway afaik [11:10] <ubitux> except eventually testing your ram [11:10] <durandal_1707> :) [11:11] <ubitux> michaelni: can we drop shorthand from AVFilter now? [11:11] <ubitux> (it's not public, we can drop without any problem, right?) [11:16] <durandal_1707> anybody fixing scale spam? [11:16] <ubitux> and we'll likely have to deprecate av_opt_set_from_string now [11:16] <ubitux> durandal_1707: 'not working on it [11:16] <durandal_1707> [auto-inserted scaler 0 @ 0x29c50680] The <w>:<h>:flags=<flags> option syntax is deprecated. Use either <w>:<h>:<flags> or w=<w>:h=<h>:flags=<flags> [11:19] <michaelni> ubitux, feel free to drop shorthand [11:20] <ubitux> ok [11:32] <durandal_1707> where is all scale filter auto inserted? [11:40] <ubitux> libavfilter/avfiltergraph.c: if (!(filter = avfilter_get_by_name("scale"))) { [11:41] <ubitux> anyone heard of vaapi crashes? [11:42] <durandal_1707> michaelni: patch to silence warnings sent [11:43] <durandal_1707> *scale warnings [11:47] <cone-607> ffmpeg.git 03Paul B Mahol 07master:ae8d435ba8bc: doc/filters: add quick emboss example [11:50] <durandal_1707> ubitux: iirc you were mentioned twitter? [11:55] <cone-607> ffmpeg.git 03Clément BSsch 07master:dee2db523d40: doc/filters: update volume options description. [11:55] <ubitux> durandal_1707: huh? [11:56] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:491d261adece: avfilter/asrc_aevalsrc: check correct variable [11:56] <cone-607> ffmpeg.git 03Anton Khirnov 07master:d69a4177b988: lavfi: remove now unused args parameter from AVFilter.init [11:56] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:fd6228e65711: lavfi: remove now unused args parameter from AVFilter.init and init_opaque [11:58] <durandal_1707> ubitux: oh it was saste [11:59] <ubitux> 491d261ad is wrong; args1 can not be set in that scope; you can just ret = AVERROR(ENOMEM) here [11:59] <ubitux> michaelni ^ [12:02] <michaelni> ubitux, hmm [12:03] <michaelni> ill fix [12:03] <ubitux> michaelni: or well, you should check for the string in the context to be set instead [12:05] <ubitux> - ret = args1 ? AVERROR(ENOMEM) : AVERROR(EINVAL); [12:05] <ubitux> + ret = eval->exprs ? AVERROR(EINVAL) : AVERROR(ENOMEM); [12:05] <ubitux> this should do the trick [12:05] <ubitux> ... [12:05] <ubitux> without my stupid switch [12:05] <durandal_1707> ubitux: what's about that s=%s comment? [12:06] <ubitux> ? [12:06] <ubitux> durandal_1707: i just prefer the s= syntax, it's simpler [12:06] <ubitux> and i just hope it's still supported [12:07] <ubitux> possibly "faster" btw, since it's not parsed as an expression [12:07] <durandal_1707> ubitux: if i use scale=s=800x600 i get again that retarded spam (which is incorrect btw) [12:08] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:3d7d819aad52: avfilter/asrc_aevalsrc: fix return code, 2nd try [12:08] <ubitux> durandal_1707: huh? [12:08] <durandal_1707> this is big mess, switch all filter users to opt based thing, snprintfs hacks is wasting by brain cells [12:08] <ubitux> there is no reason to warn in the scale=s=800x600 case [12:08] <ubitux> if it warns there is a bug [12:09] <ubitux> and i believe that bug is because of the limited-libav-based wrapper in avfilter.c [12:10] <ubitux> the reason is that libav doesn't support s/size [12:13] <ubitux> strchr(args, ':') < strchr(args, '=')) { [12:13] <ubitux> this check is wrong [12:14] <ubitux> scale=s=800x600 NULL < 0x... [12:14] <ubitux> it needs something similar to what's done for aevalsrc [12:15] <ubitux> anyway, the compat wrapper need to be reworked [12:15] <michaelni> ubitux, why does scale need this hack in avfilter.c at all ? [12:16] <ubitux> it might not be needed, except to warn [12:16] <ubitux> maybe it's not needed actually [12:17] <michaelni> ok, ill try to remove it [12:18] <ubitux> yeah right it's not needed at all [12:18] <ubitux> it's just a wrapper because libav doesn't support mixed/long [12:18] <ubitux> so they can't have w:h:flags=... [12:18] <cone-607> ffmpeg.git 03Nicolas George 07master:af0d270aac86: lavu: add helper functions for integer lists. [12:18] <cone-607> ffmpeg.git 03Nicolas George 07master:e48ded855117: lavfi/buffersink: accept parameters as options. [12:18] <cone-607> ffmpeg.git 03Nicolas George 07master:62d888ef0e3b: ffmpeg: use options for abuffersink. [12:20] <ubitux> shorthand? [12:20] <ubitux> again :( [12:22] <ubitux> durandal_1707: after the avfilter scale hack is dropped, your patch should not be necessary [12:25] <michaelni> ill remove the shorthand in a moment [12:25] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:f1e62af0e03f: avfilter: remove mixed args hack for scale. [12:26] <ubitux> yay :) [12:29] <ubitux> michaelni: libavfilter/vf_libopencv.c:100:9: error: args undeclared (first use in this function) [12:29] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:4a8898925460: buffersink: remove shorthand [12:31] <michaelni> ubitux, ERROR: opencv not found [12:32] <durandal_1707> opencv need switching to opt [12:32] <durandal_1707> (i said that yesterday i think) [12:32] <ubitux> durandal_1707: it has a compat layer [12:32] <ubitux> michael dropped too much "args" in a recent patch [12:33] <durandal_1707> ahh [12:33] <ubitux> the sub functions takes an args grabbed from the local context [12:33] <durandal_1707> could yust remove compat layer? [12:33] <ubitux> noo [12:33] <ubitux> just give me a minute i'll fix it [12:33] <durandal_1707> i mean in ocv [12:34] <ubitux> ocv already has opt anyway [12:34] <durandal_1707> ahh it insternal thing ...... confusing everyone [12:35] <durandal_1707> diracdec still touch deprecated stuff.... [12:35] <durandal_1707> is it actually still working? [12:38] <cone-607> ffmpeg.git 03Clément BSsch 07master:51e71d6e1127: lavfi/ocv: fix build after fd6228e65. [12:39] <durandal_1707> what will be done with l setpts filter? [12:42] <ubitux> wow ocv is nicely broken [12:44] <cone-607> ffmpeg.git 03Clément BSsch 07master:490786c02e85: lavfi/ocv: fix crash with no filter name specified. [12:45] <ubitux> ok, now another crash.. [12:45] <ubitux> seems that was well tested [12:45] <ubitux> there is even a warning about the problem [12:45] <ubitux> :p [12:46] <durandal_1707> ffmpeg -f lavfi -i smptebars=hd980 [12:48] <ubitux> didn't you remove some smptebars opt thing? :p [12:48] <durandal_1707> no this crash if you set invalid video size string [12:49] <durandal_1707> happens with testsrc too [12:50] <durandal_1707> for some reason life swithced to monob as default format [12:54] <ubitux> wasn't it already the case? [12:54] Action: ubitux fixing ocv [12:57] <durandal_1707> michaelni: know reason of that regression? [12:58] <durandal_1707> something broke in opt handling recently [13:00] <xlinkz0> are there any examples for 0.5 version? [13:00] <durandal_1707> that is extremly old [13:01] <durandal_1707> and unmaintained [13:01] <xlinkz0> well.. it's the one on debian :( [13:01] <xlinkz0> squeeze [13:01] <cone-607> ffmpeg.git 03Clément BSsch 07master:f16a6f667c99: lavfi/ocv: fix crash with no dilate arguments. [13:01] <durandal_1707> are you sure that is really ffmpeg and not libav? [13:01] <ubitux> xlinkz0: 0.5 please... [13:02] <xlinkz0> it's both [13:02] <michaelni> libav did not exist at 0.5 times [13:02] <xlinkz0> i just need to get some info about the file.. nothing fancy yet [13:02] <ubitux> xlinkz0: 0.5 was release beginning 2009... [13:03] <xlinkz0> that's just 4 years ago ! [13:03] <ubitux> just 30k commits ago [13:03] <xlinkz0> so no examples yet? [13:03] <xlinkz0> ok i'll try to compile latest [13:05] <xlinkz0> is yasm 0.8 old? [13:06] <ubitux> ... [13:06] <durandal_1707> michaelni: actually bug is in lavfi [13:06] <durandal_1707> *device [13:06] <ubitux> xlinkz0: you should seriously think about upgrading your system [13:06] <xlinkz0> "upgrading" a stable distro makes in unstable [13:07] <ubitux> yes but maybe usable [13:09] <ubitux> michaelni: you switched the way of deleting filters from a graph, and it seems to be what's causing the -f lavfi -i smptebars=hd980 crasg [13:09] <ubitux> crash* [13:17] <J_Darnley> <xlinkz0> "upgrading" a stable distro makes in unstable <-- That must be why I'm still using XP [13:17] Action: Compnn still on win2k. [13:17] <Compnn> ehe [13:17] <J_Darnley> :) [13:17] <xlinkz0> you really don't afford more than 4gb ram? [13:17] <J_Darnley> Of DDR? Of Course I can't afford it! [13:18] <Compn> i chose my software via how not bloated it is [13:18] <Compn> so my web browser only uses 300mb ram for 100 tabs :P [13:18] <Compn> javascript disabled [13:18] <kierank> oh [13:18] <kierank> which browser? [13:18] <kierank> can't be firefox [13:18] <kierank> ie is pretty light on memory [13:18] <Compn> old opera [13:18] <kierank> ah [13:19] <Compn> new opera sucks, chrome wont run on win2k, firefox wont even run on win2k anymore :P [13:19] <kierank> upgrade to xp [13:19] <kierank> xp is the best OS [13:19] <kierank> I'll be sad to see updates go [13:19] <Compn> everyone links against xp dlls when building [13:20] <JEEB> well, there are some actual new APIs with XP [13:20] <JEEB> which I can see useful [13:20] <Compn> all of mplayers hard fought windows video support probably wont work too well in new windows :\ [13:20] <Compn> i remember it shuts down aero if using the wrong vo [13:21] <JEEB> atomics and unaligned alloc etc. [13:21] <Compn> and people love that aero crap [13:21] <JEEB> desktop composition is nice [13:21] <JEEB> overall vsync for the whole screen [13:21] <kierank> not that useful [13:23] <JEEB> also at least for me the newer opengl-based renderer by wm4 works Just Fine on 7+, and I guess if someone /really/ wanted a well-supported NT6+ renderer they'd plug in EVR to mplayer/mplayer2/mpv in one way or another [13:36] <Compn> i thought someone was working on icod [13:38] <ubitux> durandal_1707: http://pastie.org/7463412 fixes your crash [13:40] <durandal_1707> ubitux: so you want approval or what? [13:41] <ubitux> no i'm just wondering about the code aroung [13:41] <ubitux> around [13:45] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:a1e7e02eaffd: abuffersink: switch from opaque to AVOptions for params [13:52] <durandal_1707> ubitux: so you push it or? [13:52] <ubitux> in a few minutes eventually [13:53] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:3fc7b47145da: buffersink: switch from opaque to AVOptions for params [14:00] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:0acf7e268b2f: avfilter: remove opaque passing hacks [14:13] <cone-607> ffmpeg.git 03Clément BSsch 07master:6b5ec762830d: lavfi: fix forgotten chunk in eb0f774d. [14:22] <durandal_1707> hmm lavfi does not support creating/filtering frames only on demand? [14:30] <durandal_1707> there is no filter to pass N inputs to one output? [14:30] <cone-607> ffmpeg.git 03Anton Khirnov 07master:1ba95a9cca57: lavfi: add avfilter_init_dict() for initializing a filter with a dict. [14:30] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:46de9ba59815: Merge commit '1ba95a9cca57b023b9b9de071a5671fc05b05e58' [14:38] <durandal_1707> huh buffer example mentions size but that is not in code [14:50] <cone-607> ffmpeg.git 03Anton Khirnov 07master:7cdd737ba81b: lavfi: mark filters with dynamic number of inputs or outputs with special flags [14:50] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:16fc24b240ca: Merge commit '7cdd737ba81b5c2c9521c4509edf0ac315fabc65' [14:50] <durandal_1707> ubitux: see http://www.vapoursynth.com/doc/functions/splice.html#std.Splice [14:54] <cone-607> ffmpeg.git 03Anton Khirnov 07master:7e8fe4be5fb4: lavfi: add a function for counting elements in AVFilterPad arrays. [14:54] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:f4db6bfeb86f: Merge commit '7e8fe4be5fb4c98aa3c6a4ed3cec999f4e3cc3aa' [15:14] <cone-607> ffmpeg.git 03Anton Khirnov 07master:fa2a34cd40d1: lavfi: change the filter registering system to match the other libraries [15:14] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:835cc0f2e709: Merge commit 'fa2a34cd40d124161c748bb0f430dc63c94dd0da' [15:16] <ubitux> durandal_1707: ok.. ? :p [15:16] <ubitux> michaelni: "This removes the ability to put AVFilters in read only memory and having them shareable." [15:16] <ubitux> what does that mean? [15:18] <durandal_1707> ubitux: vsynth:splice = lavfi:concat, lavfi:splice == vsynth:interleave [15:18] <cone-607> ffmpeg.git 03Anton Khirnov 07master:f160c6a18a57: lavfi: do not segfault on NULL passed to avfilter_get_by_name() [15:18] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:9a91c67478cd: Merge commit 'f160c6a18a574b4c7d063e0a748037cbbd9912ee' [15:20] <ubitux> ah, ok [15:22] <durandal_1707> ubitux: note that l added trim filter which is afaik exact in behaviour to vsynth one [15:22] <durandal_1707> perhaps they gonna clone vsynth? [15:23] <durandal_1707> or it is just coincidence [15:23] <ubitux> as long as it's not introducing compat break idc [15:24] <burek> would it be possible to create a simple voip-like thing with ffmpeg, using ffmpeg/ffplay/ffserver, using for example a simple wrapper (let's call it ffvoip or ffspeak or so) which could use all the features already present in ffmpeg/ffplay/ffserver [15:24] <burek> it would only call all those tools with proper params [15:25] <burek> it might boost popularity of ffmpeg even more and the best thing is that all the parts needed, are already implemented in various parts of ff tools [15:26] <durandal_1707> the tile example does not work any more [15:27] <burek> i'll try to draw an example and create a ticket, it will be easier that way [15:30] <durandal_1707> ubitux: do you get assertions now when using testsrc with ffplay? [15:30] <ubitux> indeed. [15:31] <cone-607> ffmpeg.git 03Anton Khirnov 07master:bee2d75b66e2: avconv: update to the new options syntax. [15:31] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:7b6c0bda9e94: Merge commit 'bee2d75b66e2f02e0877796c01202299a4b56cfb' [15:31] <durandal_1707> perhaps gmake clean is needed [15:38] <durandal_1707> seems to be ffplay only issue [15:52] <cone-607> ffmpeg.git 03Anton Khirnov 07master:dc574658d15a: avconv: print filter options in -h full output. [15:52] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:f81a8e89a1c2: Merge commit 'dc574658d15ad8f171dbdecbdce7197acf9424a0' [15:52] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:292f68154b66: avfilter: Filter options that are used for both video and audio should have both flags [15:52] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:1007de703305: asetpts: add AVOptions [15:52] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:3dedcef8b889: avfilter: add missing AV_OPT_FLAG_FILTERING_PARAM [16:01] <cone-607> ffmpeg.git 03Anton Khirnov 07master:e19e8aeeaaff: avtools: move media_type_string() from avprobe to cmdutils. [16:01] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:608d86013a86: Merge commit 'e19e8aeeaaff3d4bc8f6065764c93f0298782ac6' [16:03] <ubitux> durandal_1707: i think the example will break if you don't use the vsync opt [16:05] <durandal_1707> ubitux: does not break here, perhaps i need some special file [16:06] <ubitux> durandal_1707: look at #1644 [16:06] <durandal_1707> ubitux: the older example works with avi and mpeg4 , but not with mkv and avc [16:11] <durandal_1707> anybody found why ffplay crashes? [16:13] <durandal_1707> ubitux: using -skip_frame nokey is much faster because it decodes less... [16:13] <durandal_1707> but i don't think that can work with avc [16:18] <ubitux> http://starlogs.net/#ffmpeg/ffmpeg [16:21] <cone-607> ffmpeg.git 03Anton Khirnov 07master:921046008f53: cmdutils: allow -h filter=<name> to print information about a filter. [16:21] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:7e68491367ce: Merge commit '921046008f5335601f9d76dc3a768eb5cd91e1b9' [16:24] <durandal_1707> ubitux: what is that? [16:24] <av500> click on it [16:25] <durandal_1707> i only heard loud music [16:26] <av500> open your eyes then [16:27] <durandal_1707> av500: i see only one big image [16:28] <av500> it shows commit logs [16:29] <durandal_1707> i don't see them it opera [16:32] <ubitux> too bad [16:37] <av500> people in Belarus will suffer most [16:37] <cone-607> ffmpeg.git 03Vittorio Giovara 07master:3fce13679837: lavfi: new interlace filter [16:37] <cone-607> ffmpeg.git 03Vittorio Giovara 07master:25882a7ff642: FATE: add a test for the interlace filter [16:37] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:ebbd4fd5f06a: Merge remote-tracking branch 'qatar/master' [16:40] <durandal_1707> now we have 3 useless interlacers [16:42] Action: michaelni thought there where more [16:42] <av500> interlacers or deinterlacers? [16:44] <durandal_1707> number of useless deinterlacers is even higher [16:44] <av500> is there more than one way to interlace? [16:45] <av500> maybe vertically too? [16:45] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:227b4458fb43: vf_interlace: fix frame rate [16:46] <michaelni> av500, yes you can interlace more than 2 fields [16:46] <michaelni> for example one can interlace 6 fields [16:51] <kierank> av500: interlace the luma but not the chroma :) [16:51] <kierank> michaelni: how? [16:51] <av500> kierank: maybe for a video wall :) [16:52] <kierank> iphone on its side [16:52] <kierank> as a line [16:52] <michaelni> kierank, i dont know but i think ive a sample IIRC so there must have been a way [17:16] <j-b> michaelni: btw, what source did you take for your downmixing matrices? [17:18] <michaelni> j-b, thats a good question, i think google and some audio decoder specs maybe (e)ac3 or aac i dont remember exactly [17:19] <michaelni> note, i assume you talk about swr [17:19] <michaelni> i also tested it [17:19] <michaelni> j-b, why do you ask ? [17:20] <j-b> I do talk about swr [17:25] <durandal_1707> my ffplay broke [17:31] Action: Snaggle is trying to write a patch for the opencl detection I mentioned yesterday. But I'm having problems with getting a HAVE_CL_CL_H token declared. I changed configure to use check_lib to find the right heard/lib combo because the current require2 didn't work when checking different combinations. [17:33] <j-b> michaelni: ITU-R BS.775-2 [17:36] <michaelni> durandal_1707, which commit broke what ? [17:37] <durandal_1707> ffplay crashes when displaying any video [17:37] <durandal_1707> audio seems to work here [17:43] <michaelni> durandal_1707, no crash here [17:43] <michaelni> j-b, is this a feature request ? [17:44] <j-b> michaelni: no. [17:44] <j-b> michaelni: just wondering about doc [17:49] <michaelni> Snaggle, HAVE_CL_CL_H would be defined if theres a CL/cl.h only id guess [17:49] <Snaggle> michaelni: same thought here, but if I flip the logic around (look for OpenCL/cl.h), HAVE_OPENCL_CL_H is not being defined either [17:50] <Snaggle> I'm following the example of the libopenjpeg check since that seems closest [17:53] <michaelni> Snaggle, you did look in config.log ? [18:00] <Snaggle> ahhh.. I probably have to add the token to the HAVE_LIST variable in configure [18:05] <Snaggle> oh good [18:15] <j-b> michaelni: I was looking for 7.1 to 5.1 matrices, to be exact [18:18] <kurosu> michaelni: re apply noise not applying correctly, strange, as I have just pull-rebased against ffmpeg's master without issue [18:20] <kurosu> I'll regenerate them anyway [18:21] <michaelni> kurosu, thx, i think some define NREGS was lost [18:21] <kurosu> right, probably when rebasing from libav to ffmpeg [18:21] <kurosu> and indeed [18:23] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:b59aa007077f: doc/libavutil: remove include of opencl.texi [18:23] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:949c8627b639: doc: merge syntax.texi, eval.texi and opencl.texi into utils.texi [18:23] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:9ae9ae10f9fe: doc/utils: apply misc fixes to OpenCL docs [18:23] <kurosu> I have probably ran the fate-aac test twice in the same build dir... [18:23] <saste> burek, this ffspeak thing could be everything [18:24] <saste> also i suppose this is what ffserver is supposed to be [18:24] <saste> but as you know we lack a maintainer for it [18:28] <kurosu> oh, %elifdef works [18:38] <durandal_1707> i will do convolution filter next [18:43] <saste> durandal_1707, good [18:43] <saste> we have already relevant code in edgedetect and transform [18:44] <durandal_1707> saste: transform? [18:45] <saste> durandal_1707, libavfilter/transform.h [18:45] <saste> used by deshake [18:46] <ubitux> durandal_1707: did you look at the ffplay -f lavfi testsrc problem? [18:47] <durandal_1707> ubitux: what problem? crash for lavfi have been fixed, but ffplay now crashes when displaying any video (expcept internal spectrum) [18:47] <durandal_1707> michaelni could not reproduce it [18:48] <durandal_1707> actually its av_assert crash [18:48] <durandal_1707> saste: its 3x3 only matrix [18:48] <kurosu> michaelni: about the ac3 FP implementation by the mips guy, I feel bad reviewing it while it's not up to me to decide what to do of it [18:49] <durandal_1707> thought that one header looks to not be insttalled, but have av* prefix ,. ..... [18:49] <ubitux> durandal_1707: i have a problem playing videos with ffplay, only with -f lavfi testsrc triggering a lavu/imgutils assert [18:50] <ubitux> life is broken [18:50] <ubitux> well something in lavfi looks broken [18:51] <saste> ubitux, same here [18:51] <durandal_1707> Assertion abs(dst_linesize) >= bytewidth failed at libavutil/imgutils.c:249 [18:51] <ubitux> yes, that one [18:51] <ubitux> with anything involving lavfi [18:51] <saste> something related to the recent refactoring wave [18:51] <durandal_1707> but i have that always .... [18:52] <ubitux> durandal_1707: most likely a filter is inserted all the time [18:52] <durandal_1707> i think its probably because filters are differently created/allocated [18:53] <michaelni> kurosu, if you want to make the final decission on ac3 FP, i dont mind [18:53] <michaelni> mips ac3 FP [18:53] <michaelni> just note that the mips code itself is maintained by the mips guys so [18:53] <MrNice06> Hi, I am a newbie don't know exactly how to start my request regarding a bug. Can I now and here? [18:53] <michaelni> that part is their decission [18:53] <kurosu> I'm a bit uncomfortable about it, I'm not experienced, and was just looking how they did it [18:53] <kurosu> yeah of course [18:55] <MrNice06> OK. I opened a ticket in the track-bug #2367. Can you read it, mainly the second part? [18:55] <durandal_1707> saste: that transform thing have nothing of interest for me [18:56] <saste> durandal_1707, how do you plan to provide the convolution matrix? [18:57] <durandal_1707> saste: binary and/or string [18:57] <saste> durandal_1707, ok [18:57] <MrNice06> It seems that this bug appears with other users [18:57] <saste> i wanted to add a matrix parser, but never found the time [18:57] <durandal_1707> now is there some code for matrix multiplication that optimized? [18:57] Action: saste trying to fix the crop bug [18:58] Action: saste found the bug [18:58] <ubitux> that was fast [18:59] <michaelni> MrNice06, for ffmpeg/v4l2 bugs it might make sense to mail Giorgio Vazzana unless he already said something in the bug report (didnt read the ticket as ive no time ATM so dunno) [19:01] <MrNice06> I have no answer in the ticket so far. Where can I find Giorgio mail? [19:02] <durandal_1707> lol, this shows how auto insterted filters are evil [19:03] <michaelni> MrNice06, git log for sure possibly elsewhere too [19:03] <durandal_1707> but why is crop always inserted in my case, that is just retarded [19:04] <ubitux> a366beaf6a7cbaa43fef09cabc2248225290208f ? [19:05] <durandal_1707> my resolutions of files are not odd [19:05] <kurosu> michaelni, anyway, they want to create a fixedpoint dsp context, but using a predetermined FP format, which may not suit FP implementation for other codecs [19:05] <ubitux> michaelni: can you reproduce the ffplay -f lavfi testsrc bug? (and almost anything involving filters?) [19:06] <kurosu> I don't think that's good but I have no idea what to make of this [19:07] <michaelni> kurosu, i share your feeling on this, it doesnt feel "right" [19:08] <michaelni> ubitux, "Assertion abs(dst_linesize) >= bytewidth failed at libavutil/imgutils.c:249" [19:08] <michaelni> but ive to go afk now for 30-60min [19:08] <kurosu> michaelni, anyway I'd prefer if someone would comment too, because I don't feel like making decisions [19:09] <michaelni> kurosu, ok ill comment too when iam back, need to leave now [19:09] <kurosu> no problem [19:20] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:1d86fe6970fa: lavfi/crop: restore pos constant, and fix "t" variable misplacement in variable array [19:20] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:9fa3b5b8a899: lavfi/crop: log pos in debug message [19:20] <saste> ubitux, what's the status with regards to filtering mergemess? [19:21] <saste> what's still missing? [19:23] <MrNice06> michaelni: Thanks, I found the mail and will send him my questions. [19:23] <ubitux> michaelni: regression since 0acf7e268b2f873379cd854b4d5aaba6f9c1f0b5 [19:23] <ubitux> saste: mostly done afaik [19:23] <ubitux> ...documentation needs some review [19:50] <ubitux> nice the crop bug :) [19:51] <durandal_1707> lol, what was this?: http://ffmpeg.org/pipermail/ffmpeg-devel/2009-October/078962.html [19:54] <saste> durandal_1707, slicify [19:54] <saste> it was removed less than one year ago [19:58] <saste> uhm hue regression? [19:58] <saste> related to copy permission, meat for ubitux [20:01] <ubitux> isn't hue tested with perms? [20:03] <saste> ubitux: do we have a hue test? [20:03] <saste> we should [20:03] <ubitux> tests/lavfi-regression.sh:do_lavfi "hue" "perms=random,hue=s=sin(2*PI*t)+1" [20:03] <ubitux> we should add the propers deps to the lavfi filters btw [20:04] <ubitux> filters tests* [20:06] <saste> ubitux, ok still need to check the report [20:09] <durandal_1707> it assumes linesize[1] == linesize[2] [20:10] <ubitux> indeed [20:10] <durandal_1707> but always wonder why would that not be same [20:11] <cone-607> ffmpeg.git 03Clément BSsch 07master:72ad0d56a135: lavfi/hue: make do while (0) form meaningful. [20:17] <durandal_1707> hah now it starts adding other filters... [20:21] <ubitux> something fishy in the second bug he's reporting [20:21] <durandal_1707> alpha is left uninitialized, how added alpha format? [20:21] <ubitux> sounds like another auto-inserted filter hell [20:25] <durandal_1707> i think i fixed one bug [20:30] <cehoyos> Is anybody able to compile vlc? https://trac.videolan.org/vlc/ticket/8320 [20:34] <cone-607> ffmpeg.git 03Paul B Mahol 07master:047a027d0b55: lavfi/hue: copy alpha plane too [20:35] <durandal_1707> cehoyos: is that really ffmpeg bug? (iirc vlc use l) [20:35] <cehoyos> durandal_1707: Can you compile vlc to find out? [20:36] <durandal_1707> i dont have vlc here [20:36] <cehoyos> Did your last commit fix ticket 2465? [20:37] <durandal_1707> only hue part [20:37] <cehoyos> The other part is not relevant for ticket afaict [20:37] <durandal_1707> i did not download his samples, (it may be yet another bug) [20:37] <cehoyos> I will test later. [20:41] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:e87c1cdbb548: ffplay: replace opaque/params for buffersinks [20:41] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:9b595e86e342: avdevice/lavfi: add error checking for av_opt_set_int_list() [20:48] <durandal_1707> ubitux: fixing bug(s) or? [21:13] <ubitux> durandal_1707: not really [21:20] <durandal_1707> looks like if filters are combined, worst pix fmt is picked [21:29] <durandal_1707> funny if i remove scale filter, colors are preserved [21:31] <langoliers> somebody has a link to the serpent standard? i can not locate it ;/ [21:46] <cone-607> ffmpeg.git 03Paul B Mahol 07master:a47288f6dc4b: lavfi/hue: add support for more alpha formats [22:27] <cone-607> ffmpeg.git 03Diego Biurrun 07master:a89c05500f68: x86: h264qpel: int --> ptrdiff_t for some line_size parameters [22:27] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:52bda1d90395: Merge commit 'a89c05500f68d94a0269e68bc522abfd420c5497' [22:36] <cone-607> ffmpeg.git 03Diego Biurrun 07master:e027032fc6a4: x86: dsputil: ff_h263_*_loop_filter declarations to a more suitable place [22:36] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:43bf4ee9a974: Merge commit 'e027032fc6a49db5a4ce12fc3e09ffb86ff20522' [22:41] <ubitux> #include "libavfilter/libmpcodecs/img_format.h" [22:41] <ubitux> wat [22:42] <durandal11707> where? [22:42] <ubitux> x11 patch [22:42] <durandal11707> yust ignore it [22:42] <durandal11707> i will just take patch and makt it usefull [23:11] <cone-607> ffmpeg.git 03Clément BSsch 07master:0bc34ef02e07: doc/utils: use @code markup for opencl function names. [23:18] <ubitux> btw, what's the reason to prefer a configurable buffersink over inserting a format constraint filter like l. does? [23:22] <saste> (The effect that the bug produced may be worth a separate filter imo.) [23:22] <saste> i agree :) [23:23] <durandal11707> reallz_ [23:28] <michaelni> ubitux, iam in favor or loosing the sink config stuff if its not neccesary and simpler without [23:31] <ubitux> i'm not comfortable with that stuff, so i let the maintainers (saste & nicolas?) deal with it :p [23:31] <ubitux> it was just curiosity [23:31] <saste> michaelni, I agree [23:31] <ubitux> it seems we have to deal with various related troubles everytime we have a lavfi merge [23:32] <saste> ubitux, why, don't you like fixing bugs? [23:33] <cehoyos> Is it known why fate fails some tests with msvc? [23:34] <ubitux> cehoyos: the ebur128 is a msvc bug; i think nevcairiel reported it [23:34] <ubitux> for the rest, no idea [23:34] <ubitux> saste :) [23:35] <nevcairiel> in mxf the code must be faulty, it casues a crash [23:35] <nevcairiel> i didnt bother to look into it yet [23:35] <cehoyos> "msvc bug" as in: A bug in the Microsoft compiler? If yes, is no work-around possible? [23:36] <cehoyos> Is the change known that causes the mxf crash? [23:37] <cone-607> ffmpeg.git 03Diego Biurrun 07master:2004c7c8f763: x86: dsputil: cosmetics: Remove two pointless variable indirections [23:37] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:ed3680bc9b8f: Merge commit '2004c7c8f763280ff3ba675ea21cf25396528fd3' [23:39] <cehoyos> Assuming it is 25b143a - in which function (line) does it crash? [23:41] <nevcairiel> if i knew, i would fix it :P [23:41] <cehoyos> If you knew if the ebur regression is a Microsoft bug, or if you knew in which line mxf crashes? [23:42] <cehoyos> (Wasn't the reason for msvc that debugging is so much easier now?) [23:42] <nevcairiel> it is, still means someone has to do it [23:44] <nevcairiel> i know that its a msvc bug because i debugged it, and then created a small test case and reported the issue to MS, but i didnt have motivation to look into mxf [23:45] <cehoyos> Iirc, the usual response time from MS is >months, so perhaps a work-around should be committed? [23:48] <cone-607> ffmpeg.git 03Diego Biurrun 07master:a3cb865310cc: x86: dsputil: Move some ifdefs to avoid unused variable warnings [23:48] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:34b78ad04fe3: Merge remote-tracking branch 'qatar/master' [23:53] <saste> michaelni, there is a patch from highgod that you can apply (if you are fine with it) [00:00] --- Sat Apr 13 2013

1 0

ffmpeg.log.20130412
by burek 13 Apr '13

13 Apr '13

[00:12] <elkng> what is wrong with mkv container ? I have pentium 500 and it can run video 700xXXX without problem but if its video in mkv forman even if its less than 700xXXX something like 640x480 and it run slowly about 10 fps [00:13] <elkng> s/forman/format [00:13] <klaxa> are you using the same codecs? [00:14] <klaxa> also what player are you using? [00:14] <elkng> so I need to convert it into avi, and now same 640x480 run with 60-70% of CPU, wha is with mkv container that makes it load CPU ? [00:14] <klaxa> <klaxa> are you using the same codecs? [00:14] <elkng> not the same [00:15] <klaxa> there's your problem [00:15] <elkng> but I noticed it many times, video in mkv run slowly [00:15] <klaxa> that's not the fault of mkv though [00:16] <klaxa> i'm going to make a wild guess: the files in the .avi container will be xvid or mpeg4 part 2 with mp3 and the files in the mkv will be h264, probably main or high profile with probably aac as audio track, maybe even a subtitle track [00:16] <elkng> I don't say fault, but is some overheads happens when playing video inside mkv container ? [00:16] <klaxa> there shouldn't be much of a difference [00:16] <klaxa> try putting the avi video into an mkv container and try again [00:17] <elkng> can it be done using ffmpeg ? [00:17] <klaxa> yes [00:17] <klaxa> ffmpeg -i somefile.avi -c copy somefile.mkv [00:18] <elkng> codec is h264 with aac audio and subtitle track [00:18] <klaxa> i doubt that the avi will have those codecs [00:18] <elkng> those are inside mkv video [00:18] <klaxa> yes [04:02] <Mista_D> can progressive scan video in x264 show "intelaced" in the header (mediainfo). Tried -x264opts fake-interlaced" [04:05] <Mista_D> and tried "-x264opts fake-interlaced=1"... still shows as "progressive"/ [04:05] <Mista_D> MBAFF wont do it, just PAFF interlace, dealing with older STB here. [09:53] <bogdanp> I'm streaming mov files to rtmp and every now and then I get audio packets that have negative duration (which fail to send b/c dts doesn't increase monotonically in that case) [09:54] <bogdanp> this causes delay between the audio and video streams, simply dropping those packets doesn't help. Is there anything else I could do to fix it? I'm using libav 0.9. [09:55] <bogdanp> the mov files are h264 with AAC, generated by the native iOS encoder [09:55] <bogdanp> I'm guessing this is due to "multiple edit list entries, a/v desync might occur" but I haven't been able to find any solutions by googling [09:56] <relaxed> bogdanp: try #libav [09:57] <bogdanp> relaxed, I will. I meant ffmpeg 0.9 above though :D. [09:59] <relaxed> oh, then this is the right place [10:02] <Mista_D> if I encode in FFmepg I get "variable" fps, if use x264 with "--fps 24" I get constant framrate of 24 (reported by MediaInfo). Anyway to get constant fps in FFmpeg? [11:09] <arpu> hello @all [11:09] <arpu> is it possible to stream all files in an directory without restarting ffmpeg ? [11:10] <arpu> the output is rtmp [11:21] <BtbN> Anyone else having problems with vaapi in recent ffmpeg versions? For me all applications using latest ffmpeg(1.2 + one fix i found and latest git) just crash [11:21] <BtbN> I can't realy see what's causing this [11:25] <diroots> hi all, anyone knows what means "avcodec encoder error: filling error on fillup" I think this is related to a problem giving me a black screen instead of the video [11:26] <diroots> I don't find any reference in search engine, on this error [11:33] <diroots> Hi fflogger this is an avcodec error when calling ffmpeg within vlc to transcode to either x264 or flv. I prepare the paste [11:38] <diroots> fflogger: http://pastebin.com/PBch1dMg line 298 [11:39] <ubitux> BtbN: https://www.ffmpeg.org/bugreports.html [11:39] <BtbN> Well, i can't pin down what exactly the problem is. Just reporting "vaapi does not work" won't help much. [11:40] <ubitux> if you are able to reproduce with latest git ffmpeg, that's enough [11:40] <ubitux> (ffmpeg, ffplay or any other tool) [11:41] <BtbN> all ff cli tools are not affected [11:41] <BtbN> as they don't use any hw acceleration [11:41] <BtbN> only hits libavcodec [11:41] <BtbN> and apps that use it, like vlc or xbmc [11:42] <diroots> durandal_1707: http://pastebin.com/PBch1dMg line 298 [11:43] <ubitux> BtbN: we need a way to reproduce; maybe just open a ticket without much info, it's better than no ticket i guess [11:44] <ubitux> ideally a test code should be provided, but if you're not one of the developer of those app, that might not be simple [11:44] <durandal_1707> diroots: why you are reporting vlc issue on ffmpeg channel? [11:44] <BtbN> I am, but it's still not simple as even a simple test-case would be multiple thousand lines [11:45] <diroots> because my problem is an avcodec error [11:45] <diroots> and avcodec is part of ffmpeg, no? [11:45] <BtbN> It's not easy to reduce a full HW decoding and rendering pipeline to a simple testcase [11:45] <BtbN> Best Test-Case ir propably vlc git master [11:45] <BtbN> *is [11:46] <durandal_1707> diroots: output is from vlc .... [11:47] <diroots> durandal_1707: yes, vlc using ffmpeg / avcodec to transcode stream. If I ask to vlc, they'll tell me that's an avcodec problem,... [11:48] <BtbN> "avcodec encoder error: filling error on fillup" it totaly a VLC error message. It's coming from the avcodec vlc plugin. [11:49] <durandal_1707> diroots: your vlc really use ffmpeg? what version of libavcodec lib? [11:50] <diroots> durandal_1707: as you can see in the paste, my vlc is compiled with '--enable-ffmpeg' option [11:51] <durandal_1707> diroots: you are not helping yourself [11:51] <durandal_1707> that switch means nothing to me [11:51] <durandal_1707> i need exact lib version that is used [11:52] <diroots> durandal_1707: as I understood, libavcodec is shipped with ffmpeg when you get ffmpeg from git, right? [11:53] <BtbN> that message does NOT come from libavcoded [11:53] <BtbN> it comes from the vlc plugin that handles libavcodec [11:54] <BtbN> So it's clearly a vlc issue, and you're much better off in the vlc channel [11:54] <durandal_1707> diroots: do not repeat yourself you said that several times, and i did not said its incorrect [11:57] <diroots> I did a git clone for ffmpeg this morning. [11:59] <durandal_1707> that is much helpfull if vlc does not use that ffmpeg [11:59] <durandal_1707> *not much helpfull.... [12:02] <ubitux> i remember it's quite a pain to make vlc link against the appropriate lib [12:13] <xlinkz0> how does ffmpeg -i gets its stream info? [13:04] <zap0> from the file header [13:40] <bogdanp> is it normal when doing av_read_frame(oc, &packet) for the packet's duration to vary? (I'm seeing values like 85, 86 which I'm guessing is normal but I'm also seeing 152) [13:41] <brontosaurusrex> what would be the lasest fashion quality based command for fdk_aac ? [13:44] <lentferj> ubitux: I see you fixed the misleading syntax warnings.. thanks [14:05] <thebuggz> Hi guys [14:05] <thebuggz> i am looking for support in configuring ffmpeg for https.. any pointers ? [14:11] <thebuggz> compiled ffmped using http://ffmpeg.org/trac/ffmpeg/wiki/CentosCompilationGuide but i can't run ffmpeg with https ulrs [16:33] <Tabs> Hello [16:34] <Tabs> I am struggling with an issue using ffmpeg timestamps, especially -vsynch and -copyts anyone could help ? [16:35] <Tracerneo> Hello. [16:35] <Tabs> The goal is to cut a video according to a give time and number of frames (i.e. from time t up to 19 frames after) but keeping the same times stamps (or at least the same non constant frame rate) [16:36] <Tabs> here is the command I am currently using: [16:36] <Tracerneo> How can i set audio stream language? There are options like -vlang (for video) and -slang (for subtitles), but I can't find any option for audio. [16:36] <Tabs> ffmpeg -y -i inputVideo.asf -ss 353.035 -vcodec mpeg4 -vframes 100 -vsync 0 -copyts -vf showinfo outputVideob.avi 2> videoinfo.txt [16:37] <Tabs> but still when I look at the PTS of the output using ffprobe I do not get the same timestamps (starting from 0 + the rate changes) [16:38] <Tabs> For instance when I look at the first lines of the ffmpeg info filter output I get: [16:39] <Tabs> n:0 pts:353035000 pts_time:353.035000 pos:156948836 fmt:yuv420p sar:0/1 s:1280x1024 i:P iskey:1 type:? checksum:13764275 plane_checksum:[F47E3FFC 6811D07D 1EC431ED 00000000] [16:39] <Tabs> [showinfo @ 0xc9f3c0] n:1 pts:353161000 pts_time:353.161000 pos:156948836 fmt:yuv420p sar:0/1 s:1280x1024 i:P iskey:1 type:? checksum:0F8DFF59 plane_checksum:[484FE07F E077B316 DA056BB5 00000000] [16:39] <Tabs> [showinfo @ 0xc9f3c0] n:2 pts:353236000 pts_time:353.236000 pos:157052836 fmt:yuv420p sar:0/1 s:1280x1024 i:P iskey:1 type:? checksum:E83B38CB plane_checksum:[23802B1D AAA6C229 F0374B76 00000000] [16:39] <Tabs> [showinfo @ 0xc9f3c0] n:3 pts:353317000 pts_time:353.317000 pos:157068836 fmt:yuv420p sar:0/1 s:1280x1024 i:P iskey:1 type:? checksum:2C92157F plane_checksum:[5D2C0CB0 5CB2C00B BC8C48B5 00000000] [16:39] <Tabs> But when I look at the ffprobe of the output I get as pts (no copying the whole thing): [16:40] <Tabs> 0.000000 ; 0.134078 ; 0.201117 [16:40] <Tracerneo> Ok, I have found -metadata option. :) [16:49] <Tabs> ok, I will post that on the mailing list I think [16:49] <Tabs> cheers ! [17:03] <tmatth> if i have .avi files with arbitrary amounts of audio streams, what's the simplest way of extracting them all into separate files? [17:04] <tmatth> i've used the -map option, but only by explicitly giving stream mappings. [17:19] <diroots> __with_delay__ BtbN: ok, I understand, and right, this problem happened after I recompiled vlc. ----->[vlc] . durandal_1707 Thx. [19:12] <langoliers> microsoft windows 0.1 in C ! http://pastebin.com/BsXquP4S [19:13] <LithosLaptop> :) [19:17] <klaxa> they couldn't even indent correctly [19:22] <Magicking> LithosLaptop: not possible § [19:31] <megaTherion> langoliers: Im sure they used while(1) { fork(); } too :D [19:34] <klaxa> while(fork()); [19:34] <klaxa> not as fast, but shorter [19:41] <langoliers> microsoft windows 0.2 in C ! http://pastebin.com/kdb5qJty [19:41] <langoliers> better? [20:10] <elkng> langoliers: https://www.kernel.org/pub/linux/kernel/Historic/linux-0.01.tar.gz [20:11] <elkng> use it with wisdome [20:43] <langoliers> :) [20:44] <langoliers> elkng<= hehe, never even thinked about reading that [20:54] <langoliers> somebody has a link to the serpent standard? [20:54] <langoliers> i found some nice reading http://csrc.nist.gov/groups/ST/toolkit/block_ciphers.html [22:08] <Iv0ry> Hi guys i streming musicvideos for a party from an usb hdd to an streming server with 720p works great but 1080p will stuck and laaaagggging any ideas? Thats my code: ffmpeg -re -i music.mp4 -vcodec copy -acodec aac -ab 256k -strict -2 -cutoff 15000 -f flv rtmp://192.168.1.10 [22:18] <elkng> Iv0ry: not enough bandwidth ? [22:19] <elkng> or CPU power [22:20] <Iv0ry> http://www.speedtest.net/result/2641715939.png [22:21] <Iv0ry> And the CPU is an Intel Atom Processor 1.6 GHz [22:25] <Iv0ry> bandwidth should be ok the cpu power i hope that is there a way to buffer it or something in ffmpeg? [22:47] <_aeris_> hello everybody ! [22:48] <_aeris_> i need some quick help :D [22:48] <_aeris_> i use ffmpeg on a very old project, for convert from jpeg2000 video to mpg video [22:48] <_aeris_> recently, i try it on Windows 7, it crash& [22:49] <_aeris_> what the first ffmpeg version compatible with 7 ? [22:52] <klaxa> why can't you use a more recent version? [22:52] <_aeris_> #define LIBAVCODEC_VERSION_MAJOR 52 [22:52] <_aeris_> #define LIBAVCODEC_VERSION_MINOR 36 [22:52] <_aeris_> #define LIBAVCODEC_VERSION_MICRO 0 [22:52] <_aeris_> & [22:52] <_aeris_> VERY old ffmpeg [22:52] <_aeris_> with JNI >< [22:52] <_aeris_> a lot of work to port to the current API :( [22:54] <_aeris_> a lot of deprecated, rename, modification& [22:54] <_aeris_> (and not public API method usage >>>><<<<) [23:06] <elkng> why developers make ffmpeg run on windows ? why not to make it unix-only ? [23:06] <_aeris_> elkng > i let you say this to my customers :D [23:07] <elkng> Iv0ry: are you from deutschland ? [23:07] <Iv0ry> yes [23:08] <elkng> Iv0ry: what does "na" means in german ? [23:08] <Iv0ry> like "hi" [23:09] <elkng> some kind of filler, without any translation ? [23:09] <Iv0ry> no its a bit complicated "na und" = i dont care "na wie gehts" hi how are you ... [23:10] <Iv0ry> ahh and its no word its more slang ^^ [00:00] --- Sat Apr 13 2013

1 0

ffmpeg-devel.log.20130411
by burek 12 Apr '13

12 Apr '13

[00:06] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:2ffee3ec14e7: lavfi/smartblur: switch to AVOption-based system [00:06] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:8695d9de4704: lavfi/smarblur: add missing newline at end of verbose message [00:06] <saste> oh typo [00:07] <saste> what if a filter has no options? [00:08] <saste> ubitux: decimate, or do you want to rename it before? [00:09] <ubitux> no option, no init? [00:09] <ubitux> do you have a filter in mind? [00:09] <ubitux> saste: go ahead, i'll push my ivtc stuff later [00:09] <saste> many [00:09] <saste> super2xsai for example [00:12] <saste> field, decimate, and I'll have a look at hue [00:15] <cone-987> ffmpeg.git 03Anton Khirnov 07master:ffea3b00c39c: vf_settb: switch to an AVOptions-based system. [00:15] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:6a2fb5b90cb7: Merge commit 'ffea3b00c39caa8ad78456ae08c8353929974dfd' [00:17] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:2040b428b45f: lavfi/field: switch to AVOption-based system [00:19] Action: mateo` working on vf_il [00:21] <saste> mixed syntax seems to work, or am I dreaming? [00:22] <ubitux> you're dreaming [00:23] <saste> missing compilation... [00:23] <ubitux> compilation is overrated [00:24] <ubitux> saste: you want to work on the mixed syntax right? [00:24] <saste> ubitux, if I want it? no [00:24] <saste> but i'll do it if no-one does, but not tomorrow [00:24] <ubitux> i'm sure you want [00:25] <ubitux> please stop resisting [00:25] <saste> i want to sleep [00:25] <ubitux> :( [00:25] Action: ubitux on histeq [00:26] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:cff8f91ddfbd: lavfi/decimate: switch to an AVOption-based system [00:26] <michaelni> saste, please commit to your own repository [00:26] <saste> michaelni, ok [00:27] <michaelni> thanks [00:27] <michaelni> rebasing & restesting merges is annoying [00:27] <saste> yes sorry for bothering [00:28] <ubitux> saste: please change the option introducing sentence while at it [00:28] <ubitux> i'm using the "The filter accepts the following options:" template [00:28] <ubitux> (for !source) [00:29] <ubitux> (btw, i have various sources done locally) [00:30] <saste> hue is messy [00:30] <ubitux> saste: yes, that's why i avoid it every time [00:38] <cone-987> ffmpeg.git 03Anton Khirnov 07master:0c2466dec719: vf_transpose: switch to an AVOptions-based system. [00:38] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:a4e0defa75b6: Merge commit '0c2466dec719b933d161f5d680a57fde38aa5daa' [00:38] <ubitux> saste: btw... [00:38] <ubitux> histeq looks untested [00:38] <saste> ubitux, yes [00:38] <ubitux> even though an entry exist in lavfi-regression.sh [00:38] <ubitux> (wtf?) [00:39] <saste> floating point issue, it was breaking regression on some platforms [00:39] <ubitux> mmh maybe it's missing in avfilter.mak [00:39] <ubitux> oh [00:39] <ubitux> ok [00:39] <saste> we disabled it just before the release [00:39] <ubitux> ok [00:42] <saste> about hue the only sane choice is to break syntax [00:42] <saste> even if most users won't even realize that [00:42] <cone-987> ffmpeg.git 03Clément BSsch 07master:169880e31d62: lavfi/cellauto: switch to an AVOptions-based system. [00:42] Action: mateo` working on kerndeint [00:42] <cone-987> ffmpeg.git 03Clément BSsch 07master:bd85fee91481: lavfi/life: switch to an AVOptions-based system. [00:42] <cone-987> ffmpeg.git 03Clément BSsch 07master:791da4706cc3: lavfi/mandelbrot: switch to an AVOptions-based system. [00:42] <cone-987> ffmpeg.git 03Clément BSsch 07master:4ecee69b20a7: lavfi/mptestsrc: switch to an AVOptions-based system. [00:42] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:b4ffc35e76fc: Merge remote-tracking branch 'ubitux/shorthand-cleanup' [00:43] <ubitux> histeq looks horribly broken :D [00:43] <ubitux> maybe that's normal but well.. [00:44] <durandal_1707> what's broken, it's regression? [00:44] <ubitux> not sure [00:44] <ubitux> but... [00:44] <ubitux> http://ubitux.fr/pub/pics/_gn-histeq.png [00:44] <ubitux> even if the filter is not good, this doesn't look very normal [00:46] <durandal_1707> hmm it seems to have overflows [00:47] <durandal_1707> or just uninitalized data? [00:48] <ubitux> no idea, i'm moving to the next filter [00:48] <ubitux> i've fixed what i was working on, ENOCARE for the rest [00:48] Action: ubitux on vf noise [00:53] Action: ubitux on stereo3d [00:54] Action: ubitux on histogram [00:57] <mateo`> michaelni: you can merge https://github.com/mbouron/FFmpeg/tree/shorthand-cleanup (il & kerndeint) [00:58] Action: ubitux on idet [00:58] <mateo`> anyone on vf_field ? [00:59] <ubitux> 00:12:04 <@saste> field, decimate, and I'll have a look at hue [00:59] <ubitux> mateo`: go for blend eventually [00:59] <mateo`> ok :) [00:59] <saste> mateo`, field already pushed [01:00] Action: mateo` working on blend [01:00] Action: michaelni integrating all the unrelated changes from libav [01:03] <cone-987> ffmpeg.git 03Anton Khirnov 07master:b83e9efc53e5: vf_unsharp: switch to an AVOptions-based system. [01:03] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:c2696dabbbc6: Merge commit 'b83e9efc53e5491716625aa31c69006b1119b280' [01:05] <ubitux> michaelni: 5 additional filters in my branch [01:08] Action: ubitux on concat [01:10] <cone-987> ffmpeg.git 03Matthieu Bouron 07master:6f46403cb056: lavfi/il: switch to AVOption-based system [01:10] <cone-987> ffmpeg.git 03Matthieu Bouron 07master:531158e6becb: lavfi/kerndeint: switch to AVOption-based system [01:10] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:72ce979f109f: Merge remote-tracking branch 'mbouron/shorthand-cleanup' [01:10] <cone-987> ffmpeg.git 03Clément BSsch 07master:e4bd1db88ec0: lavfi/histeq: switch to an AVOptions-based system. [01:10] <cone-987> ffmpeg.git 03Clément BSsch 07master:a689a6b698b9: lavfi/noise: switch to an AVOptions-based system. [01:10] <cone-987> ffmpeg.git 03Clément BSsch 07master:d5226fc57578: lavfi/stereo3d: switch to an AVOptions-based system. [01:10] <cone-987> ffmpeg.git 03Clément BSsch 07master:b2d589188f36: lavfi/histogram: switch to an AVOptions-based system. [01:10] <cone-987> ffmpeg.git 03Clément BSsch 07master:7668b6832d3c: lavfi/idet: switch to an AVOptions-based system. [01:10] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:8ac7090c233c: Merge remote-tracking branch 'ubitux/shorthand-cleanup' [01:15] <durandal_1707> about histeq; that formula that calculates luma is funny [01:18] Action: mateo` working on afade [01:19] <saste> x11 output... interesting [01:20] <ubitux> re-integrating the mixed syntax is going to require a brain [01:21] <durandal_1707> you mean too add all this shorhands you just removed? [01:22] <durandal_1707> saste: wonderful i wanted to do it, but never found time [01:23] <durandal_1707> but why people cant send patches? [01:23] <cone-987> ffmpeg.git 03Anton Khirnov 07master:7536c671040f: vf_yadif: switch to an AVOptions-based system. [01:23] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:ab6091f8e21d: Merge commit '7536c671040f1f3ebc9f0d3b7831dac71436b775' [01:26] Action: mateo` working on apad [01:29] <ubitux> i'm dead, good night guys [01:31] <durandal_1707> saste: why drop reinit in hue? [01:32] <saste> because setting all the options at the same time, is not very robust [01:32] <saste> for example what if the user specify h and H [01:33] <durandal_1707> makes sense [01:33] <saste> setting each separate option seems more robust [01:33] <saste> the only drawback is more overhead, but i think it's safe to assume that it will be unnoticeable [01:34] <saste> btw we should drop reinint in drawtext as well, which sometimes crash badly [01:34] <saste> (for example if the fontfile is not valid) [01:34] <saste> hue is not tested very well, but i need to sleep now so feel free to apply if you have hurry to swith to anton-based options for whatever reason [01:35] Action: mateo` working on aphaser [01:35] <michaelni> saste, anything i should merge from you ? [01:35] <saste> michaelni, not really [01:36] <saste> read my message above [01:36] <michaelni> ok [01:36] <michaelni> saste, good night! [01:37] <saste> bye :) [01:37] <cone-987> ffmpeg.git 03Anton Khirnov 07master:ac20e3ab8ed4: af_amix: switch to an AVOptions-based system. [01:37] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:314be1933e3a: Merge commit 'ac20e3ab8ed497549ee2a62754f01eee5e7a5066' [01:42] Action: mateo` working on avolume [01:44] <durandal_1707> ught vf_framestep is not using opt at all [01:47] <cone-987> ffmpeg.git 03Anton Khirnov 07master:b2b25b0659fa: af_asyncts: switch to an AVOptions-based system. [01:47] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:ac217bda30a7: Merge commit 'b2b25b0659fa047da6266d2ce165d43011136b30' [01:48] <mateo`> michaelni: additional filters on my branch [01:49] <durandal_1707> if nobody picked vf_framestep, i can [01:56] <mateo`> i'm tired, good night guys. [01:58] <cone-987> ffmpeg.git 03Matthieu Bouron 07master:6afed2aec129: lavfi/blend: switch to AVOption-based system [01:58] <cone-987> ffmpeg.git 03Matthieu Bouron 07master:2b17e5816327: lavfi/afade: switch to AVOption-base system [01:58] <cone-987> ffmpeg.git 03Matthieu Bouron 07master:a5ce5ac9fb49: lavfi/apad: switch to AVOption-based system [01:58] <cone-987> ffmpeg.git 03Matthieu Bouron 07master:f43da3def764: lavfi/aphaser: switch to AVOption-based system [01:58] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:57148122375a: Merge remote-tracking branch 'mbouron/shorthand-cleanup' [01:58] <michaelni> mateo`, good night [02:14] <durandal_1707> michaelni: i pushed framestep patches to repo [02:14] <cone-987> ffmpeg.git 03Anton Khirnov 07master:ba8efac977f4: af_channelmap: switch to an AVOptions-based system. [02:14] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:d9d7c5496080: Merge commit 'ba8efac977f4276f05274947b2b67d144cbc965a' [02:14] <durandal_1707> ugh cancel it for now [02:14] <durandal_1707> i rebase wrong [02:16] <durandal_1707> michaelni: now it should be ok [02:17] <durandal_1707> ught typo there, i hope you can fix it: s/that/than [02:19] <durandal_1707> fixed it (if you still did not fetch it) [02:23] <cone-987> ffmpeg.git 03Anton Khirnov 07master:3f14febbdf7c: af_channelsplit: switch to an AVOptions-based system. [02:23] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:f17651bd160d: Merge commit '3f14febbdf7c93bbd186399da27991180e5916b6' [02:32] <cone-987> ffmpeg.git 03Paul B Mahol 07master:f77db72965f8: lavfi/framestep: switch to an AVOptions-based system [02:32] <cone-987> ffmpeg.git 03Paul B Mahol 07master:7c79ec66b6cc: lavfi/framestep: remove request_frame hack [02:32] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:5e5ef6e8aecd: Merge remote-tracking branch 'richardpl/framestep' [02:37] <durandal_1707> michaelni: same with biquads branch [02:39] <durandal_1707> but its without doc changes.... [02:39] <durandal_1707> so no need to rush so faast [02:42] <cone-987> ffmpeg.git 03Anton Khirnov 07master:dd7fc37c7195: af_join: switch to an AVOptions-based system. [02:42] <cone-987> ffmpeg.git 03Paul B Mahol 07master:e6067acc8da4: lavfi/biquads: switch to an AVOptions-based system [02:42] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:42c70b358418: Merge commit 'dd7fc37c71955b78a2687f29e871f714d18de386' [02:42] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:8905a811b79c: Merge remote-tracking branch 'richardpl/biquads' [02:48] <cone-987> ffmpeg.git 03Anton Khirnov 07master:b13623e18475: af_volume: switch to an AVOptions-based system. [02:48] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:ae6634da8b81: Merge commit 'b13623e184759f37348b8fdb1276b1bb408f3e59' [02:55] <cone-987> ffmpeg.git 03Anton Khirnov 07master:73d5d405d424: split: switch to an AVOptions-based system. [02:55] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:7f2198a2f1cf: Merge commit '73d5d405d424c06f3f354337cfdb24794932094d' [03:16] <cone-987> ffmpeg.git 03Anton Khirnov 07master:7bc1a883c915: vsrc_color: switch to an AVOptions-based system. [03:16] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:8bafd83aaafa: Merge commit '7bc1a883c9158bb6e383d86fc76bdf2541e9a9fe' [03:28] <cone-987> ffmpeg.git 03Anton Khirnov 07master:b7b3302f8f2c: vsrc_nullsrc: switch to an AVOptions-based system. [03:28] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:670e0f7f6620: Merge commit 'b7b3302f8f2c7d9f6e08e0a976890eefe3edcc9e' [03:48] <cone-987> ffmpeg.git 03Anton Khirnov 07master:a42d6e6c4c19: vsrc_movie: switch to an AVOptions-based system. [03:48] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:3a3d984445f0: Merge commit 'a42d6e6c4c19912b73cae8ca9133b4202667c303' [03:56] <cone-987> ffmpeg.git 03highgod0401 07master:db2de94e95c2: avutil/opencl: add opencl device list APIs 20130411 [09:06] <tgreenwood> I was discussing h264 MAX_SLICES on ffmpeg and got the suggestion I continue here. [09:07] <tgreenwood> There are 3 bugs that relate to MAX_SLICES being too low and wanted to know if it would be worth posting a patch increasing the value to 32 or 64. [10:19] <cone-607> ffmpeg.git 03Anton Khirnov 07master:7b3eb745b98b: vsrc_testsrc: switch to an AVOptions-based system. [10:19] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:ef29a95a4d60: Merge commit '7b3eb745b98b04dd8a4970b9fd6c98998e858fc1' [10:25] <ubitux> hello [10:25] <ubitux> i should be done with the shorthand mixed short/long soon [10:25] <ubitux> hopefully [10:32] <cone-607> ffmpeg.git 03Anton Khirnov 07master:c43a7ecad997: lavfi: remove now unused args parameter from AVFilter.init [10:32] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:31428240ec1e: Merge commit 'c43a7ecad997fc527af34b952334f3d030709a1b' [10:36] <ubitux> yepee it works. [10:36] <ubitux> or at least it seems [10:38] <cone-607> ffmpeg.git 03Anton Khirnov 07master:62549f9655c4: lavfi: error out when options are provided to a filter that does not take any [10:38] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:c7b0a84e1fe1: Merge commit '62549f9655c48f0ec061087fa33a96040ce01145' [10:41] <ubitux> michaelni: can you have a look to my github/lavfi-mixed-short-long? [10:41] <ubitux> (possibly not yet ready for merge, need some little more testing) [10:48] <cone-607> ffmpeg.git 03Anton Khirnov 07master:4d1f31ea44f8: lavfi: make AVFilterContext export filter options. [10:48] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:0965be00fe20: Merge commit '4d1f31ea44f82adfb2b712534d71e27233a4f785' [10:49] <ubitux> michaelni: should be ready for merge [10:51] <durandal_1707> tell me when is fine to push new things [11:00] <durandal_1707> the process_command callback should be simplified using AVOption [11:01] <ubitux> do you think every command can be translated to an option? [11:02] <durandal_1707> not, but those that do could take less lines.... [11:03] <michaelni> durandal_1707, i can merge things any time if you have commits [11:03] <ubitux> michaelni: lavfi-mixed-short-long so i can move on! :) [11:04] <michaelni> ubitux, you can move on already [11:04] <michaelni> no need to wait [11:04] <ubitux> ok :) [11:09] <cone-607> ffmpeg.git 03Anton Khirnov 07master:8114c1016078: lavfi: add avfilter_get_class(). [11:09] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:44d4488301c6: Merge commit '8114c101607843a86960bd3e0e9c40487dc2a961' [11:17] <durandal_1707> michaelni: those merges from l become pointless [11:17] <cone-607> ffmpeg.git 03Clément BSsch 07master:b89ce54e7405: lavfi: restore mixed short/long option support with the new system. [11:17] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:d7e8b3d83179: Merge remote-tracking branch 'ubitux/lavfi-mixed-short-long' [11:17] <durandal_1707> it would be far more useful it you commit it as cherry-picks, where one could see real changes in git log [11:19] <michaelni> durandal_1707, the mailing list log is created by a script, it could be changed to show diffs for merges [11:25] <michaelni> if we want such change then someone should talk with thresh probably ... [11:26] <cone-607> ffmpeg.git 03Anton Khirnov 07master:cdac3acb1135: lavfi: add a bump and docs entries for the AVOptions switch [11:26] <cone-607> ffmpeg.git 03Anton Khirnov 07master:837112c0c84d: af_channelmap: fix uninitialized variable use introduced in ba8efac977f4276f05274947b2b67d144cbc965a [11:26] <cone-607> ffmpeg.git 03Christophe Gisquet 07master:2e81acc687e6: x86inc: Fix number of operands for cmp* instructions [11:26] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:cb20ca14a330: Merge commit '2e81acc687e64d15dd93c74793060bb5a233f44d' [11:31] <durandal_1707> michaelni: thing is i want to commit myself and hate merges in git - they are useful only if merger did not changed anything [11:31] <cone-607> ffmpeg.git 03Ronald S. Bultje 07master:6d25c9db11e8: dsputil: Make square put/avg_pixels functions local to h264qpel [11:31] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:3bc3c78a59d3: Merge commit '6d25c9db11e87ec58652de6588448225137b0c57' [11:32] <durandal_1707> and i not going to type git log -p -m for each merge hash [11:33] <michaelni> durandal_1707, like i said the ML log script can be fixed but im not root on videolan so i cannot do it [11:33] <durandal_1707> michaelni: i dont want to waste bandwitch to see what merge does [11:33] <michaelni> and you can commit directly noone stops you from it [11:34] <durandal_1707> but i dont want any merges [11:36] <ubitux> michaelni: avoption merge is done, we can push again? [11:36] <michaelni> i still have 4 commits from yesterday to merge ... [11:36] <ubitux> ok [11:37] <ubitux> concat, tile, tinterlace done here, 'working on deshake [11:42] <ubitux> deshake done, sine done [11:43] <michaelni> durandal_1707, i dont like the merges from libav either [11:44] <cone-607> ffmpeg.git 03Martin Storsjö 07master:e8cafd2773bc: h264: Clear the mb members via memset instead of using dsputil [11:44] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:944ad4618248: Merge commit 'e8cafd2773bc56455c8816593cbd9368f2d69a80' [11:47] <durandal_1707> michaelni: same issue with aformat as yesterday with overlay [11:52] <ubitux> michaelni: you can merge shorthand-cleanup [11:53] <ubitux> oh wait just a sec [11:54] <ubitux> should be ok now [11:58] <cone-607> ffmpeg.git 03Ronald S. Bultje 07master:62844c3fd669: h264: Integrate clear_blocks calls with IDCT [11:58] <cone-607> ffmpeg.git 03Clément BSsch 07master:35a27402ade9: lavfi/concat: switch to an AVOptions-based system. [11:58] <cone-607> ffmpeg.git 03Clément BSsch 07master:64a103138c04: lavfi/tile: switch to an AVOptions-based system. [11:59] <cone-607> ffmpeg.git 03Clément BSsch 07master:c85f56bb4fd8: lavfi/tinterlace: switch to an AVOptions-based system. [11:59] <cone-607> ffmpeg.git 03Clément BSsch 07master:9ceb7bdf6f88: lavfi/deshake: switch to an AVOptions-based system. [11:59] <cone-607> ffmpeg.git 03Clément BSsch 07master:bca4cafaf2d4: lavfi/sine: switch to an AVOptions-based system. [11:59] <cone-607> ffmpeg.git 03Clément BSsch 07master:cb2327e88d66: lavfi/aselect: switch to an AVOptions-based system. [11:59] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:0724b4a16d6b: Merge commit '62844c3fd66940c7747e9b2bb7804e265319f43f' [11:59] <cone-607> ffmpeg.git 03Clément BSsch 07master:73c1a00e73fb: lavfi/asettb: switch to an AVOptions-based system. [11:59] <cone-607> ffmpeg.git 03Clément BSsch 07master:77add967a5e5: lavfi/setfield: switch to an AVOptions-based system. [11:59] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:a87e85fa722e: Merge remote-tracking branch 'ubitux/shorthand-cleanup' [12:01] <ubitux> ok, 3 filters left [12:02] <ubitux> aspect, hue, scale [12:04] <cone-607> ffmpeg.git 03Ronald S. Bultje 07master:85deb51a01f1: h264: Only initialize dsputil if error resilience is enabled [12:04] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:65120f628ca5: Merge commit '85deb51a01f1ecc5ac5faa52ad8ea141c384e23a' [12:16] <ubitux> michaelni: you can merge; i'm not working on the 3 remaining filters [12:17] <ubitux> (hue is done by saste, and aspect and scale well...) [12:17] <durandal_1707> aspect is easy, isn't it? [12:18] <ubitux> might not be that trivial [12:18] <ubitux> i need to look at it properlu [12:18] <ubitux> -u+y [12:19] <durandal_1707> ubitux: old syntax is already depcrecated and should be removed [12:19] <durandal_1707> no point in keeping it [12:20] <ubitux> i'm going to look at it [12:20] <ubitux> unless you want to do it [12:20] <durandal_1707> no i fixing mss4 [12:28] <durandal_1707> michaelni: where is place where frame is allocated in reget_buffer? [12:30] <durandal_1707> the frame format should be set to -1 (AV_PIX_FMT_NONE or AV_SAMPLE_FMT_NONE) [12:30] <durandal_1707> hmm i think now how to fix it properly [12:37] <cone-607> ffmpeg.git 03Ronald S. Bultje 07master:b93b27edb045: dsputil: Make dsputil selectable [12:37] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:742c392885b9: Merge remote-tracking branch 'qatar/master' [12:37] <michaelni> durandal_1707, ubitux, merges done, you can commit directly again [12:38] <ubitux> cool, thx [12:40] <durandal_1707> there should be one paid dev(because it is not fun work) improving fate coverages all the time [12:41] <cone-607> ffmpeg.git 03Paul B Mahol 07master:b6d5a948f916: mss4: use init_get_bits8() [12:41] <cone-607> ffmpeg.git 03Paul B Mahol 07master:310bf283542f: mss4: make code independent of sizeof(AVFrame) [12:41] <cone-607> ffmpeg.git 03Clément BSsch 07master:e6b1d904bb03: lavfi/field: cleanup unused shorthand and doc. [12:41] <cone-607> ffmpeg.git 03Clément BSsch 07master:e3bbfaa5eb34: lavfi/overlay: cleanup unused shorthand. [12:41] <cone-607> ffmpeg.git 03Clément BSsch 07master:52848c3a3cd7: lavfi/smartblur: cleanup unused shorthand and doc. [12:41] <cone-607> ffmpeg.git 03Clément BSsch 07master:3417280b579e: lavfi/field: cleanup unused shorthand and doc. [12:41] <cone-607> ffmpeg.git 03Clément BSsch 07master:fcaea2170f07: lavfi/overlay: cleanup unused shorthand. [12:41] <cone-607> ffmpeg.git 03Clément BSsch 07master:4c9b031559c6: lavfi/smartblur: cleanup unused shorthand and doc. [12:42] <ubitux> wat [12:42] <ubitux> ah haha get it [12:43] <ubitux> push failed but the bot print it anyway [12:44] <durandal_1707> it did not failed for me [12:44] <durandal_1707> also i got your commits [12:44] <durandal_1707> interesting bug btw [12:50] <ubitux> durandal_1707: the first time my push failed because of fast forward (because you just pushed the mss4) [12:50] <ubitux> so i rebased and pushed again [12:51] <durandal_1707> but when it happened for me to push after michaelni it would auto-resolv itself [12:52] <cone-607> ffmpeg.git 03highgod0401 07master:9d442b9cc034: opencl: add spec opencl device APIs 20130411 [12:54] <durandal_1707> getted, .... if Derek see this .. [13:00] <ubitux> haha [13:02] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:c66c92add6fa: doc/opencl: speling and gramer fixes [13:02] <ubitux> haha [13:03] <durandal_1707> ???? [13:03] <durandal_1707> michaelni: leave it to natives [13:04] <michaelni> but me spell good too englis [13:04] <ubitux> :D [13:04] <cptspiff> natives are usually the worst ;) [13:04] <ubitux> native chineses? [13:05] <cptspiff> i assume were are on about native english speak/writers. [13:05] <durandal_1707> michaelni: use simple english from your native tongue, there is google translator [13:06] <durandal_1707> so even chinese can uderstand simple english [13:07] <cone-607> ffmpeg.git 03Clément BSsch 07master:a1e798ef0a5f: lavfi/curves: fix a spacing weirdness. [13:07] <cone-607> ffmpeg.git 03Clément BSsch 07master:7e7dc1489f5e: lavfi/curves: make possible to customize presets. [13:07] <cone-607> ffmpeg.git 03Clément BSsch 07master:2844ea86c8f7: lavfi/curves: introduce "all" field in presets and use it. [13:07] <durandal_1707> or obfuscate code and remove commets/doc so code is universal [13:07] <durandal_1707> *comments [13:10] <ubitux> erm we need to deal with the compat break in aevalsrc [13:11] <ubitux> and some other filters will need some lifting [13:12] <durandal_1707> like? [13:13] <ubitux> durandal_1707: any filter not using a priv_class [13:14] <durandal_1707> like i did with framestep, switched to avopt [13:14] <ubitux> yes, likely [13:14] <ubitux> is aconvert still necessary? [13:15] <durandal_1707> there is better one? [13:15] <ubitux> isn't aformat enough? [13:16] <durandal_1707> yes, why it was added at first place? [13:16] <ubitux> no idea. [13:17] <ubitux> anyway, 'working on evalsrc [13:17] <durandal_1707> mark it for deprecation with warning and remove it on next major bump [13:17] <ubitux> yeah [13:22] <ubitux> ocv is weird [13:22] <ubitux> oh. my bad [13:27] <cone-607> ffmpeg.git 03Paul B Mahol 07master:661e284b4a69: telecine filter [13:27] <cone-607> ffmpeg.git 03Paul B Mahol 07master:10b1cc63c2cc: lavfi: remove double .priv_class initializers [13:27] <cone-607> ffmpeg.git 03Paul B Mahol 07master:32a7b858c917: lavfi: remove telecine libmpcodecs wrapper [13:28] <ubitux> i'm going to inverse the filters list in avfilter [13:29] <durandal_1707> why? [13:29] <ubitux> because any new filter needs to add an entry herer [13:29] <ubitux> i'd better make it a list of filters remaining to switch [13:35] <cone-607> ffmpeg.git 03Nicolas George 07master:0d58bbb28d90: lavfi/buffersrc: update for compatibility with the fork. [13:35] <cone-607> ffmpeg.git 03Nicolas George 07master:08dd2c9df1c1: fate: add tests for unknown channel layouts. [13:35] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:eeb6efcd58bc: Merge remote-tracking branch 'cigaes/master' [13:41] <ubitux> arg nicolas is adding some shorthand [13:42] <cone-607> ffmpeg.git 03Marton Balint 07master:f1e606c762a3: mxfdec: update current edit unit when seeking in mxf files without index tables [13:56] <cone-607> ffmpeg.git 03highgod0401 07master:d83c481dc04c: MAINTAINERS: add opencl maintainer [14:10] <cone-607> ffmpeg.git 03Clément BSsch 07master:38788f2de5d6: lavfi: replace new filter list with filters left to update. [14:12] <durandal_1707> ubitux: why smptebars is listed without testsrc? [14:14] <ubitux> smptebars is the only one having some local priv_class set in the init [14:14] <ubitux> and some weird option thing [14:19] <cone-607> ffmpeg.git 03Clément BSsch 07master:3c2e4c2a9bac: lavfi/showwaves: switch to an AVOptions-based system. [14:20] <durandal_1707> ubitux: tell what you do [14:24] <cone-607> ffmpeg.git 03Paul B Mahol 07master:38d1a5a27024: lavfi/smptebars: switch to AVOptions-based system [14:25] <voidah> hi guys [14:25] <durandal_1707> hi [14:25] <voidah> I'm trying to hack something in ffmpeg for testing purpose [14:26] <voidah> Do you know if there is an easy way of compiling ffmpeg and test it without installing it system wide? [14:26] <av500> sure [14:26] <av500> take make [14:26] <av500> type make [14:26] <av500> then ./ffmpeg [14:27] <av500> build it static [14:27] <voidah> it asks for some missing libs [14:27] <av500> --enable-static --disable-shared [14:27] <ubitux> that should be the default. [14:27] <av500> or set LD_LIBRARY_PATH [14:27] <ubitux> just don't --enable-shared [14:27] <cone-607> ffmpeg.git 03Diego Biurrun 07master:8a776ad90e00: h261: Move shared data tables from a header to a proper C file [14:27] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:567616c1b34a: Merge commit '8a776ad90e00ab2b98e8683ac6182d641a383c3a' [14:27] <voidah> when I build static, configure seems to look for some libs in my system that are'nt available as static lib [14:28] <voidah> I compile ffmpeg with the exact same options as the one installed by arch [14:28] <voidah> av500: your LD_LIBRARY_PATH looks interesting, I test this [14:28] <ubitux> voidah: that's not normal [14:29] <ubitux> the static build isn't related to the link with system libraries [14:29] <voidah> hum ok [14:29] <ubitux> it is related to the libav* only [14:29] <ubitux> just don't --enable-shared [14:29] <ubitux> if you have a build failure, what is it exactly? [14:29] <voidah> so I don't understand why I have errors about missing libs only when static building [14:29] <ubitux> what errors? [14:29] <ubitux> with what configure flags? [14:30] <voidah> https://gist.github.com/voidah/566f8fc8446dd2ae0755 [14:31] <ubitux> --disable-static [14:31] <ubitux> remove this [14:31] <ubitux> install fontconfig [14:31] <voidah> crpa, missed that one [14:31] <ubitux> remove this: [14:31] <voidah> fontconfig is already installed [14:32] <ubitux> --disable-shared --enable-static --extra-libs=-static --extra-cflags=--static [14:32] <ubitux> remove this. [14:32] <av500> -ETOOMUCHSTATIC [14:32] <voidah> ok [14:32] <ubitux> voidah: seriously, just let the default unless you know what you are doing [14:32] <voidah> recompiling... [14:33] <ubitux> as long as your configure line greps something like "static" or "shared", trim it [14:33] <voidah> ok [14:33] <voidah> and this way you're telling me that it should run without installing it system wide? [14:34] <av500> yes [14:34] <av500> try it [14:34] <voidah> ok good, still compiling :) [14:35] <ubitux> voidah: you don't even need --prefix=installhere [14:35] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:dd97736d4f40: avfilter: add missing , [14:35] <voidah> ah? [14:36] <ubitux> voidah: you don't need to make install; when make is done, ./ffmpeg ... [14:36] <voidah> oh ok.. [14:36] <ubitux> michaelni: oups... sorry [14:40] <voidah> oops [14:40] <voidah> got this: [14:40] <voidah> ./ffmpeg: error while loading shared libraries: libavdevice.so.55: cannot open shared object file: No such file or directory [14:40] <durandal_1707> what we will do with aliases? [14:41] <ubitux> voidah: you still have some shared somewhere in your configure line [14:41] <ubitux> voidah: don't forget to make distclean; also, ccache is your friend [14:41] <cone-607> ffmpeg.git 03Clément BSsch 07master:389eb0a91995: lavfi/testsrc: remove useless args/NULL passing. [14:41] <cone-607> ffmpeg.git 03Clément BSsch 07master:7eb1ea2ae672: lavfi/sendcmd: switch to an AVOptions-based system. [14:41] <voidah> oh, you're right [14:41] <voidah> sorry [14:48] <cone-607> ffmpeg.git 03Clément BSsch 07master:ab4afcb85b3a: lavfi/blackdetect: switch to an AVOptions-based system. [14:50] <voidah> av500: ubitux: thank you [14:50] <voidah> it works [14:53] <cone-607> ffmpeg.git 03Diego Biurrun 07master:20429ba96e55: h261: Move encoder/decoder shared table init to common code [14:53] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:5ec016ad0b29: Merge commit '20429ba96e55db17ac57f6c0e330914287f1f614' [15:07] <ubitux> erm astreamsync is broken [15:08] <ubitux> 'need to mail Nicolas. [15:09] <cone-607> ffmpeg.git 03Clément BSsch 07master:2fb8ca7d2c27: lavfi/astreamsync: fix eval memleak. [15:09] <cone-607> ffmpeg.git 03Clément BSsch 07master:7fdebb116c98: lavfi/astreamsync: switch to an AVOptions-based system. [15:10] <cone-607> ffmpeg.git 03Paul B Mahol 07master:11f6174a9104: lavfi/blend: support more yuv pixel formats [15:14] <cone-607> ffmpeg.git 03Clément BSsch 07master:7464b9e31904: lavfi: remove bbox from the filters left to update. [15:19] <ubitux> what is the point of anullsrc? [15:19] <ubitux> generating silence? [15:21] <ubitux> well, whatever. [15:22] <cone-607> ffmpeg.git 03Clément BSsch 07master:1c0feee05930: lavfi/anullsrc: switch to an AVOptions-based system. [15:22] <durandal_1707> ubitux: same as nullsrc [15:24] <durandal_1707> also its init function is useless [15:26] <ubitux> btw, i think we need to add a notice in the filters documentation, saying that shorthands are not reliable [15:26] <ubitux> and the key=value form should be prefered for scripting [15:26] <ubitux> because it's likely to get broken as soon as libav decide to add an option [15:29] <durandal_1707> bbox have hardcoded min value, i thing this should be configurable [15:30] <ubitux> if you want to work on bbox, you should make it inject metadata [15:30] <durandal_1707> to what? [15:30] <ubitux> and eventually make the appropriate following filters honor them [15:30] <ubitux> durandal_1707: "lavfi.crop" "lavfi.drawtext" [15:30] <durandal_1707> ah adding metadata for stuff it printfs out [15:31] <durandal_1707> log should be removed, or? [15:31] <ubitux> drawbox* [15:31] <ubitux> verbose/debug [15:31] <ubitux> -vf bbox,crop=metadata,drawtext=metadata [15:31] <ubitux> same for cropdetect [15:31] <ubitux> blackdetect eventually, etc [15:31] <ubitux> so the filter can be usefuls.. [15:40] <cone-607> ffmpeg.git 03Clément BSsch 07master:e2e992c07761: lavfi/asetnsamples: switch to an AVOptions-based system. [15:42] <michaelni> work today 2 hours rebasing merges [15:43] <ubitux> michaelni: what are you going to do with the resample/aresample thing? [15:43] <ubitux> something similar to 4fa1f52e should be done for aresample [15:44] <cone-607> ffmpeg.git 03Anton Khirnov 07master:2d6edb2b7ebc: lavc: extend / update the [coded_]{width,height} doxy [15:44] <cone-607> ffmpeg.git 03Anton Khirnov 07master:9de9b828ef00: lavc: don't overwrite display dimensions with coded dimensions. [15:44] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:cf7d3846fc86: Merge commit '9de9b828ef005dec37052548c195a6b4f18fc701' [15:47] <michaelni> ubitux, correct design is to set the options into the swr context directly, but the "new" design cant handle that [15:48] <michaelni> so yeah passing unchecked dictionaries thorugh that init_dict hack it will be probably [15:49] Action: michaelni goes back in trying to finish todays merge [15:50] <Snaggle> Should/would the recent OpenCL code work/be useful on ffmpeg on OS X? [15:53] <ubitux> remote: (commit.author, commit.logmsg) = metainfo.split("|") [15:53] <ubitux> remote: ValueError: too many values to unpack [15:53] <ubitux> erm [15:55] <durandal_1707> where you get that? [15:56] <ubitux> pushing my latest commit [15:56] <ubitux> with a '|' in the commit message [15:58] <durandal_1707> perhaps some broken script that fed logger [15:58] <michaelni> ubitux, its the script being angry at you for breaking my merge again :) [15:59] <ubitux> michaelni: ok i stop committing for a while [15:59] <ubitux> :p [15:59] <michaelni> just 5min will be enough :) [15:59] <ubitux> take your time, i'm on a longer work anyway [16:00] <michaelni> if i take my time someone else will break my merge ... [16:01] <cone-607> ffmpeg.git 03Diego Biurrun 07master:ac9362c5d985: Move misplaced file author information where it belongs [16:01] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:580a0600efbb: Merge remote-tracking branch 'qatar/master' [16:01] Action: michaelni is done merging for today [16:03] <cone-607> ffmpeg.git 03Paul B Mahol 07master:567feaafa807: lavfi/bbox: remove unused variables and functions [16:07] <cone-607> ffmpeg.git 03highgod0401 07master:d757f7510bc2: cmdutils: add opencl command line options to ffmpeg [16:07] Action: Snaggle was trying to patch the opencl detection to work on OS X but ran into some issues, so might as well first determine if it'll actually do anything [16:07] <durandal_1707> Snaggle: what opencl version? [16:08] <Snaggle> durandal_1707: 10.7/Lion comes with 1.1 which is part of the problem I think since the ffmpeg code is looking for 1.2 features near as I can tell. [16:09] <durandal_1707> yes [16:09] Action: michaelni knows that 1.0 definitly is not enough [16:10] Action: Snaggle was going to file a trac entry for the overall OS X detection as well as the version check, but wanted to make sure it's not just something due to OS X OpenCL [16:12] <durandal_1707> Snaggle: you could ask highgod when it was still here, left 7 min ago [16:12] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:0a95449d2fe0: cmdutils: remove unneeded #if [16:12] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:fb78b94c3a00: cmdutils_common_opts: fix indention of #if [16:15] <Snaggle> durandal_1707: ok. will stick around in channel for highgod to return [16:31] <cone-607> ffmpeg.git 03Paul B Mahol 07master:d1853436876d: libshine fixed point MP3 encoding support [16:33] <cone-607> ffmpeg.git 03Stephen Hutchinson 07master:ff99b42e045a: avformat/avisynth: Don't insert null frames or samples at the start of a stream. [16:42] <cone-607> ffmpeg.git 03James Almer 07master:94d13df34c2d: Fix compilation with --disable-everything --enable-filter=deshake [16:48] <ubitux> yay almost done with aevalsrc [16:49] <ubitux> the remaining filters are going to be a pain in the ass [16:57] <durandal_1707> https://github.com/FFmpeg/FFmpeg/graphs/code-frequency [16:58] <ubitux> fun [16:58] <ubitux> i wonder how it looks like ignoring the moves [16:59] <Compnn> 900k lines :D [17:00] <durandal_1707> this is unfair: https://github.com/FFmpeg/FFmpeg/contributors [17:05] <Compnn> yeah its unfair, i'm not listed! :P [17:05] <av500> me neither! [17:05] <av500> its a fake [17:06] <nevcairiel> you need to write code to get on there :P [17:06] <Compnn> http://www.ohloh.net/p/ffmpeg/contributors/summary [17:06] <Compnn> pretty sure i'm in the ohloh list [17:10] <voidah> quick question: there is some functionnality already implemented in code that allows to change ffmpeg settings while it runs? [17:10] <ubitux> what kind of settings? [17:10] <av500> output filename [17:10] <voidah> I would like to implement something that would allow me to change draw_mouse in x11grab [17:11] <voidah> to toggle on/off mouse cursor drawing [17:11] <ubitux> not that i know of [17:11] <voidah> x11grab.c:556 [17:11] <voidah> ok [17:12] <voidah> I can hack something quick and dirty, but I was considering making something cleaner [17:12] <ubitux> quick & dirty = signal() handler? [17:12] <voidah> check for presence of a file in /tmp [17:12] <voidah> heh [17:12] <ubitux> haha [17:13] <voidah> my initial problem is that I sceencast to make "tutorials" for my students [17:13] <voidah> I like to be able to zoom in easily with compix [17:13] <voidah> compiz [17:13] <voidah> so I don't have to postprocess [17:14] <voidah> _but_ even if compiz hide de cursor when zoomed (and draw it's own scaled cursor), ffmpeg still render the hidden cursor [17:14] <voidah> the result is that in my video I have 2 cursors [17:14] <ubitux> make them fight [17:14] <voidah> haha [17:15] <voidah> a better fix could be to detect if the X cursor is hidden [17:15] <voidah> in ffmpeg [17:15] <voidah> and then do not call paint_mouse_pointer [17:16] <voidah> but from what I've read on XFIXES extension, someone can hide of show the X cursor, but not detect if it's hidden [17:16] <Compnn> patches are welcome :) [17:16] <Compnn> ah [17:17] <voidah> i'm not an expert at this [17:17] <voidah> I'll continue my reading [17:17] <ubitux> Compnn: what about you sending patches btw? ;) [17:19] <Compnn> you know i like to commit directly :P [17:28] <ubitux> ok, got aevalsrc and its compat layer working [17:33] <cone-607> ffmpeg.git 03Clément BSsch 07master:64ce15b9f433: lavfi/aevalsrc: switch to an AVOptions-based system. [17:40] <ubitux> durandal_1707: where is that format auto inserted? [17:41] <durandal_1707> ubitux: by ffmpeg [17:41] <ubitux> oh ffmpeg_filter.c. [17:41] <durandal_1707> look auto inserted filters and their syntax [17:41] <durandal_1707> just update it to stop confusing people [17:42] <ubitux> easy to fix [17:43] <ubitux> anton didn't do it either [17:43] Action: ubitux wonders why [17:43] <durandal_1707> evil [17:43] <av500> anton cannot do all the evil things in the world [17:45] <ubitux> why that syntax thing isn't even with the others [17:45] <ubitux> ffs [17:45] <ubitux> oh, got it. [17:51] <cone-607> ffmpeg.git 03Clément BSsch 07master:937325f3db31: ffmpeg: insert aformat filter with the appropriate separator. [18:03] <cone-607> ffmpeg.git 03Clément BSsch 07master:23a750c9940e: lavfi: do not warn inappropriately with aevalsrc. [18:03] <ubitux> durandal_1707: any other warnings? [18:04] <ubitux> mmh, format as well. [18:07] <durandal_1707> nobody porting eedi3? [18:09] <cone-607> ffmpeg.git 03Clément BSsch 07master:eabc0c733dee: ffmpeg: insert format filter with the appropriate separator. [18:30] <ubitux> durandal_1707: not me, but i would love to see it [18:33] <cone-607> ffmpeg.git 03Clément BSsch 07master:c74b0eda34f1: lavfi/amerge: switch to an AVOptions-based system. [18:39] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:865e20daf3c9: avfilter/f_select: use AVFrame metadata accessor function [18:39] <cone-607> ffmpeg.git 03Tobias Rapp 07master:7f09b888e88d: avienc: fix overflow of audio sample count [18:51] <cone-607> ffmpeg.git 03Clément BSsch 07master:c1907bd732c0: lavfi/removelogo: switch to an AVOptions-based system. [18:53] <ubitux> michaelni: Anton is going to push some more dict thing for avfilter [18:53] <ubitux> it's certainly relevant to fix aresample asap [18:54] <durandal_1707> aw, its broken? [18:55] <ubitux> not really [18:55] <ubitux> but resample filter switched to the dict init form [18:56] <ubitux> and that might have some consequences over auto-inserted params and stuff like that [18:56] <ubitux> (i didn't look closely) [18:56] <ubitux> anyway, resample is now initialized with a dict, so aresample should do something similar [18:57] <ubitux> i'll send a mail to ffmpeg-devel in a moment about the remaining filters [18:57] <ubitux> because i'm not going to do all of them [19:21] <michaelni> ubitux, if you think it matters i can look into changing aresample [19:21] <ubitux> please do [19:21] <ubitux> i'm sending a summary mail now [19:21] <michaelni> ok, will later (today probably) [19:21] <ubitux> saying what's left, and that i won't work on the remaining ones [19:30] <ubitux> mail sent, hf [19:48] <ubitux> hey saste :) [20:00] <saste> hi [20:00] <saste> michaelni, you ok with eval+syntax merge? [20:08] <michaelni> saste, ive really no oppinon on that, just that if it causes conflicts on merges then someone will need to help integrating improvments [20:09] <saste> michaelni, i'll help as far as i can [20:09] <michaelni> also the docs are already quite different from libav in some areas ... [20:13] <durandal_1707> appy erratic camera efect depending on timestamp example does not work here [20:24] <durandal_1707> quick emboss: format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'" [20:34] <saste> durandal_1707, nice [20:35] <saste> we should have a gallery of such effect combinations [20:35] <durandal_1707> i need contrast filter [20:37] <durandal_1707> i thinking about adding some morphology filters, like dilation and erosion [20:37] <saste> durandal_1707, ocv=erode/dilate [20:38] <saste> about contrast, maybe eq2? [20:38] <saste> eq/eq2 should be ported [20:38] <saste> if michaelni or someone else helps with ASM i'll do it [20:38] <saste> (but someone may be faster than me) [20:41] <durandal_1707> saste: that filter is not listed in that list that need work.... [20:41] <saste> durandal_1707, uh? [20:41] <saste> what filter and what list? [20:42] <durandal_1707> ocv, see filters_left_to_update in avfilter.c [20:43] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:f6bfeef7faf9: avfilter: fix handing of child context options [20:43] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:4f43e9b4af50: avfilter/aresample: use init_dict system and set child AVOption context correctly [20:44] <ubitux> thx michaelni :) [20:45] <michaelni> ffplay -f lavfi -i nullsrc,noise=9:9:t [20:46] <michaelni> ^ shows one flaw in the noise filter [20:47] <michaelni> left side becomes cyan after a while and right green [20:49] <durandal_1707> michaelni: i have again that 00000000000 spam with that command, same that happens with sox muxer [20:49] <durandal_1707> and ffplay is stuck in something when i start it [20:51] <durandal_1707> michaelni: after how many seconds? [20:54] <michaelni> 10-20 sec should be enough to see the issue [20:54] <ubitux> here you go [20:54] <ubitux> second lavfi wave [20:54] <durandal_1707> michaelni: 26 here and everything is still green [20:55] <durandal_1707> michaelni: if you could upload image or something [20:55] <durandal_1707> ubitux: what wave? [20:55] <ubitux> durandal_1707: new merge party [20:55] <ubitux> check libav-commits [20:56] <cone-607> ffmpeg.git 03Clément BSsch 07master:c36ab01cd2ec: lavfi/ebur128: remove now unused av_opt_free call. [20:58] <durandal_1707> saste: ocv filter uses only say 0.01% of libopencv funcionality? [20:58] <saste> durandal_1707, yes [20:59] <saste> durandal_1707, most libopencv works with float images, we don't have them [21:00] Action: durandal_1707 sad [21:01] <ubitux> saste: break flat syntax in hue? [21:01] <ubitux> how so? [21:02] <saste> ubitux, flat syntax = h:s with h and s expressions *constants* [21:02] <saste> so they were not evaluated per-frame [21:03] <saste> a saner solution would be to support an eval mode, like in overlay [21:03] <ubitux> is it a problem to eval at each frame? [21:03] <saste> ubitux, I don't think so, but who can tell [21:03] <saste> anyway as I said we can restore that feature, in a much saner way [21:04] <ubitux> ok [21:05] <saste> ubitux, what's the problem with blackframe? [21:05] <ubitux> did i say something about blackframe? [21:05] <durandal_1707> wtf: http://lists.libav.org/pipermail/libav-devel/2013-April/045691.html [21:06] <saste> ubitux, i mean blackdetect [21:06] <ubitux> did i say something about blackdetect? [21:07] <saste> why was not updated (and is not in the list of filters to update)? [21:08] <ubitux> it has no init [21:08] <ubitux> so no option, so no problem [21:09] Action: durandal_1707 picking ocv [21:09] <ubitux> what's wrong with ocv? [21:09] <durandal_1707> still use args [21:09] <ubitux> durandal_1707: it has a compat layer [21:09] <ubitux> see avfilter.c:avfilter_init_filter() [21:10] <saste> ubitux, are we looking at the same thing? [21:10] <durandal_1707> ubitux: yes, my fail [21:10] <saste> vf_blackdetect.c:85 [21:11] <ubitux> saste: config_input is not init callback [21:11] <ubitux> there is no args parsing whatsoever [21:12] <saste> ubitux, my bad, I was looking at an old file buffer [21:12] <ubitux> oh god, another wave [21:12] <ubitux> seriously. [21:13] <ubitux> [PATCH 07/11] lavfi: add an asetpts filter [21:14] <durandal_1707> trim filters [21:14] <ubitux> yeah just saw that [21:15] Action: ubitux wonder what's the diff with select [21:15] <saste> ubitux, faster? [21:15] <saste> does it matter? [21:15] <ubitux> no idea [21:16] Action: ubitux wonders when anton will stop [21:18] <durandal_1707> this is war [21:18] <ubitux> anyway [21:18] <ubitux> who is working on purging filters_left_to_update? [21:19] <durandal_1707> enoon [21:20] <ubitux> i guess i'll do the aspect filters tonight [21:24] <saste> ubitux, I'll commit flite soon [21:25] <ubitux> cool thx [21:37] <ubitux> aspect is really going to be a pain [21:38] <ubitux> i'd better do something else [21:39] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:e4fd58f47252: lavfi/hue: apply major simplifications, and switch to AVOption-based system [21:39] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:3411aba82c83: lavfi/hue: reindent after last commit [21:39] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:160ed967f459: lavfi/hue: extend debugging message [21:39] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:976e3439ed99: lavfi/flite: switch to AVOption-based system [21:42] <ubitux> cool :) [21:42] <ubitux> saste: tinterlace just got commited [21:42] <ubitux> opinions? :) [21:43] <saste> ubitux, what tinterlace? [21:43] <saste> wasn't it named interlace? [21:43] <ubitux> yeah right, interlace [21:43] <saste> we keep it, as far as I know [21:44] <ubitux> so we have 2x tinterlace [21:44] <ubitux> and interlace? [21:44] <ubitux> :D [21:44] <durandal_1707> tinterlace is GPL afaik [21:44] <saste> yes life is dumb sometimes [21:44] <ubitux> durandal_1707: interlace as well [21:44] <saste> that also [21:44] <durandal_1707> dumb even more [21:44] <ubitux> it's based on tinterlace, but not enough to keep the copyright [21:44] <ubitux> thought, the gpl is kept [21:45] <ubitux> from what i can tell :) [21:45] <saste> ubitux, I asked the copyright holders when I ported tinterlace, but noone replied [21:45] <durandal_1707> if you NIH something do it at least to: make it faster and/or better license and/or more features and/more less bugs [21:45] <saste> durandal_1707, it has less bugs [21:45] <ubitux> they likely dropped some features [21:45] <saste> less features -> less bugs [21:46] Action: saste stops to troll and get back at work [21:46] <ubitux> and btw [21:46] <durandal_1707> what bugs tinterlace have? [21:46] <ubitux> lol @comments in the interlacing function [21:46] <ubitux> exactly the same code as tinterlace [21:46] <ubitux> nice license violation [21:46] <ubitux> or well, copyright violation [21:46] <durandal_1707> sue asap [21:47] <ubitux> well, we'll just copy/paste the copyrights on top of the file [21:47] <saste> do I give a **i* about it? [21:47] Action: durandal_1707 no merge - no care [21:48] <ubitux> can AVRATIONAL take ':' sep? [21:50] <durandal_1707> see lavu/opt.c [21:50] <durandal_1707> or parseutils... [21:57] <ubitux> i might be able to fix aspect filters [22:04] <cone-607> ffmpeg.git 03Anton Khirnov 07master:9e6b5e61a0c9: img2enc: add an option for overwriting one file with subsequent images [22:04] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:647fe2e77733: Merge commit '9e6b5e61a0c91e25e298d0a21b189c4c95443795' [22:05] <ubitux> ok got it working [22:05] <ubitux> michaelni: can i push? [22:05] <ubitux> or you start your merging work? [22:06] <ubitux> 7 filters left, yay. [22:06] <ubitux> actually, 6 [22:07] <ubitux> well, i'm pushing, sorry michaelni :) [22:07] <cone-607> ffmpeg.git 03Clément BSsch 07master:71ef1ec7b482: lavfi/aspect: switch to an AVOptions-based system. [22:08] Action: durandal11707 writes another filter, so future merges will be even more pita [22:09] <ubitux> saste: http://lists.libav.org/pipermail/libav-devel/2013-April/045706.html [22:09] <ubitux> you and your hacks! [22:10] <ubitux> ok i'm almost motivated to do scale now [22:11] Action: saste feels dirty [22:14] <cone-607> ffmpeg.git 03Clément BSsch 07master:9ab221f8d82d: lavfi: deprecate aconvert. [22:17] Action: ubitux going to do vf mp [22:30] <cone-607> ffmpeg.git 03Anton Khirnov 07master:93e65e9a6c08: Cosmetics, restore alphabetic order in atomics lists. [22:30] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:b2175b0f2a7c: Merge commit '93e65e9a6c0895f869752a66731d9b2ec3916467' [22:33] <michaelni> ubitux, you broke -vf setdar=16:9 [22:34] <ubitux> michaelni: that one should work [22:34] <ubitux> setdar=16/9 won't though [22:35] <ubitux> or i am missing something? [22:35] <michaelni> try ffmpeg matrixbench_mpeg2.mpg -vf setdar=16:9 test.avi [22:35] <michaelni> ffmpeg -i ... [22:36] <michaelni> i dunno why its not working, i justtested and found it not workimg [22:36] <ubitux> ok, will look in a minute [22:44] <ubitux> found the problem [22:45] <cone-607> ffmpeg.git 03Anton Khirnov 07master:f223ad1e000d: doc: document the prores encoder options [22:45] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:97104aa011da: Merge commit 'f223ad1e000d56ef5231a3b1fc00495b538a9ed6' [22:48] <cone-607> ffmpeg.git 03Clément BSsch 07master:7bd014eaecef: lavfi/aspect: fix d2q convert in init. [22:48] <ubitux> michaelni: should be fixed [23:05] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:9ad27cd44758: doc/filters: update hue command syntax after recent update [23:05] <cone-607> ffmpeg.git 03Stefano Sabatini 07master:d980641ec4a9: doc/filters: fix fade-out hue saturation expression in sendcmd example [23:07] <ubitux> yepee mp working :) [23:09] <durandal11707> i have found funny bug in drawutils (converting rgb color to yuv) [23:11] <ubitux> michaelni: do you have a minute to have a look to http://b.pkh.me/0001-lavfi-mp-switch-to-an-AVOptions-based-system.patch ? [23:13] <saste> I need some way to send commands to a filter in an interactive way [23:13] <saste> controlled by an external process [23:14] <saste> one way could be to make use of the ffmpeg undocumented feature of reading commands from stdin [23:14] <ubitux> ømq ? [23:14] <saste> an alternative could be to extend sendcmd, and make it re-read the file for sendcmd [23:14] <saste> but in this case i would need to extend the syntax, to support "now" commands [23:15] <saste> the clean solution of course would be scripting, but since we lack that what could be a sane temporary solution? [23:15] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:b3c399621267: avcodec: rename prores encoders [23:15] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:70b72ca6f3fa: doc/encoders: Clarify that there are 2 prores encoders [23:16] <saste> and another solution could be a variant of sendcmd (which reads a sequence of commands from a file, and executes them immediately) [23:18] Action: ubitux doesn't get it [23:19] <saste> ubitux, suppose you have an external process which says that overlay should disable overlaying at some point [23:20] <saste> how can you control that? [23:20] <ubitux> at runtime? in ffmpeg? in ffplay? [23:20] <saste> yes [23:20] <ubitux> 23:14:11 <@ubitux> ømq ? [23:20] <ubitux> :p [23:21] <saste> ideally with scripting you can implement the required custom logic (read a file, read commands from a socket, poll a web page, process incoming twitters etc.) [23:21] <saste> failing that we can still read from a file [23:21] <michaelni> ubitux, the mp patch seems to work [23:21] <ubitux> michaelni: ok, will push [23:21] <saste> the process updates the file, ffmpeg realizes that, reads the file and sends the command [23:22] <ubitux> ./ffmpeg -open0mqsock ... [23:22] <ubitux> and you have a 0mq interface [23:22] <ubitux> did i misunderstand? [23:23] <saste> links to this 0mq thing? [23:23] <ubitux> http://www.zeromq.org/ [23:23] <saste> i believed that was a typo, or some weird geek lingo [23:24] <ubitux> it's a messaging lib [23:24] <saste> nice, I want that design for the ffmpeg web page [23:25] <ubitux> go ahead [23:26] <ubitux> with BUY and PAID in first [23:26] <saste> and we need some weirdo tale about how ffmpeg got superpowers [23:26] <ubitux> let's get the superpowers first [23:28] <ubitux> anyway, that's a cool library, afaik [23:29] <durandal11707> ubitux: push ivtc so can remove several useless filters [23:29] <cone-607> ffmpeg.git 03Clément BSsch 07master:838d803155dd: lavfi/mp: switch to an AVOptions-based system. [23:29] <ubitux> durandal11707: i'll do it when we're done with the filters left [23:29] <saste> ok we'll be soon able to control ffmpeg via twitter [23:30] <ubitux> durandal11707: give me a hand and it will be pushed faster [23:30] <ubitux> atempo should be done by its maintainer, abuffer and buffer i hope nicolas will do them, i can do pan eventually, and there is also scale to merge properly (tricky) [23:32] <ubitux> and we might do the switch for aconvert even if deprecated too [23:33] <durandal11707> ubitux: i writing new source filter [23:33] <ubitux> source of what? [23:35] <durandal11707> smptehdbars [23:35] <cone-607> ffmpeg.git 03Anton Khirnov 07master:91d2efa7d6b4: lavfi: add const to AVFilterContext.filter. [23:35] <cone-607> ffmpeg.git 03Anton Khirnov 07master:4a37d4b3f813: lavfi: add const to the pads parameter of avfilter_pad_get_name/type [23:36] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:9f3ae7aafd5f: Merge commit '4a37d4b3f8137a4c2bbbca043de076af966b9446' [23:36] <ubitux> have fun [23:47] <cone-607> ffmpeg.git 03Clément BSsch 07master:9110f95985a0: doc/filters: use -af and -filter_complex instead of amovie for two examples. [23:55] <cone-607> ffmpeg.git 03Anton Khirnov 07master:38f0c0781a6e: lavfi: merge avfiltergraph.h into avfilter.h [23:55] <cone-607> ffmpeg.git 03Michael Niedermayer 07master:231fd4411ff3: Merge commit '38f0c0781a6e099f11c0acec07f9b8be742190c4' [00:00] --- Fri Apr 12 2013

1 0

ffmpeg.log.20130411
by burek 12 Apr '13

12 Apr '13

[00:00] <shambala> yep [00:00] <saste> shambala, the first or the second? [00:00] <shambala> family history [00:00] <saste> ah ok [00:01] <shambala> so, the format needs to be changed, huh? [00:01] <llogan> what are you tryng to do? why do you need the transparency preserved? [00:02] <shambala> this was a sample movie.. there'll be other movies that will have all kinds of zoom effects and black fill looks ugly [00:02] <shambala> doesn't look natural [00:02] <llogan> how about a white fill? [00:03] <shambala> hmm& can try that, do you know the option, by any chance? [00:04] <llogan> see the "color" filter source and "overlay" for one method [00:04] <llogan> although the seems more like work for a video editor or compositor [00:09] <llogan> shambala: as in "ffmpeg -i input -filter_complex color=c=white:s=640x480,overlay output" [01:10] <shambala> llogan: thanks for the help, it didn't work though [01:10] <shambala> had to get into an impromptu meeting, will keep trying [01:14] <llogan> shambala: "it doesn't work" is a common error message. [01:15] <shambala> sorry, I meant it behaves exactly same as before (i.e. fill is still black) [01:21] <llogan> works for me [01:42] <shambala> hmm& will try again tonight, thanks a bunch [04:10] <schtinky> burek, I saw your bug post here: https://ffmpeg.org/trac/ffmpeg/ticket/614?cversion=0&cnum_hist=4 [04:11] <schtinky> I'm experiencing something similar, intermittent failure of the hardware to properly lock and spit out data to ffmpeg [04:11] <schtinky> was your device analog or digital? [07:30] <kristopolous> HI! [07:31] <kristopolous> ffmpeg -i out.mkv -an -vcodec libx264 -vpre slow_firstpass -y -pass 1 -f h264 -y /dev/null && ffmpeg -i out.mkv -vcodec libx264 -vpre slow -pass 2 -f h264 video.mp4 I'm doing that [07:32] <kristopolous> and it doesn't work [07:32] <kristopolous> it's about as simple and supported as it should be [07:32] <kristopolous> but no chance@! [07:34] <drv> you probably don't want to say -f h264, the format will be autodetected from the .mp4 filename [07:35] <drv> (h264 is the raw h264 format, not mp4) [07:35] <kristopolous> ok I will drop that and try again [07:36] <drv> you probably still need something like it in the first pass command since it's outputting to /dev/null, but not in the 2nd pass [07:36] <kristopolous> surely. [07:39] <kristopolous> nope [07:39] <kristopolous> certainly broken [07:39] <kristopolous> "constant rate-factor is incompatible with 2pass." [07:39] <kristopolous> woopdee [07:40] <kristopolous> yeah, this stuff never works for me [07:41] <kristopolous> even when I copy and paste from the manpage [07:44] <drv> i think you need to specify bitrate explicitly, e.g. -b:v 400k [07:44] <drv> the preset probably has crf by default [07:45] <kristopolous> what's with the new colon syntax [07:45] <drv> http://ffmpeg.org/trac/ffmpeg/wiki/x264EncodingGuide#twopass [07:45] <kristopolous> how long has that been around [07:46] <drv> dunno, switched some time ago, more consistent between audio and video now [07:48] <kristopolous> yeah I'm just going to try 1 pass [07:49] <kristopolous> as it turns out, compressing complex math graphs yields a bunch of compression blocks with most ways I approch [07:54] <relaxed> kristopolous: try, ffmpeg -i input -c:v libx264 -crf 18 -preset veryslow -tune grain output [07:56] <kristopolous> ok sure [07:56] <kristopolous> I just can't be passing around a 14GB video [07:58] <relaxed> -preset veryslow will yeild the best compression. [08:00] <relaxed> if it's still too large increase the -crf value until you're happy with the size/quality. [08:01] <kristopolous> oh wow, it's at 7 seconds in [08:01] <kristopolous> this will take a while [08:01] <kristopolous> I'm at -thread 0 [08:02] <kristopolous> is there an mpi or gpu version of ffmpeg? [08:04] <kristopolous> oh, and the bitrate is over 900000kbit/s [08:04] <kristopolous> this won't work [08:10] <relaxed> add -t $seconds to encode some samples and figure out what -crf to use [08:16] <kristopolous> oh good call [08:16] <lentferj> kristopolous: I ususually use -crf in the range of 20 - 23 [08:17] <kristopolous> im trying with 22 right now [08:17] <lentferj> I encountered artifacts when using preset lower than medium on some FullHD content [08:18] <lentferj> so usually I try slow first, if that makes problems revert to medium [08:18] <kristopolous> fullhd? you mean 1920x1080? [08:18] <lentferj> yes [08:18] <kristopolous> oh ok [08:19] <lentferj> relaxed: why "-tune grain"? I usually use film. What's grain doing? [08:20] <kristopolous> these are abstract mathematical graphs ... I think grain is right [08:21] <lentferj> ah... ok [08:21] <lentferj> never came to my mind you could use it for something else than.. well. movies ;) [08:23] <kristopolous> most of my ffmpeg use is generating frames through some C or R program and then stitching them together [08:23] <kristopolous> or doing screencasting [08:24] <kristopolous> I've *heard* of people watching movies on computers ... but I'm far too nerdy for that I think [08:26] <lentferj> kristopolous: haha :) [08:59] <kristopolous> yay, down to 140MB and still watchable [08:59] <kristopolous> good [09:42] <bogdanp> can anyone think of a reason why doing http://pastebin.com/raw.php?i=UMvFQPiW would result in http://i.imgur.com/2EmLbPj.png ? I'm trying to convert RGB32 frames received from a camera to h264 video but this has me stumped. [09:45] <viric> I've the impression that I can't beat the quality/bitrate of youtube webm videos [09:45] <viric> does anybody know their coding settings? [09:45] <viric> I even think that I'm *far* to achieve their quality/bitrate [15:15] <bonana|dlr> Heho! In my application I try to capture an RTSP-stream (audio + video). Following this example http://stackoverflow.com/a/13457773/1228324 I can capture the video. Where do I find information how to write the audio simultaneously into the same file? [15:16] <bonana|dlr> is the muxing-example, coming with the dev-package, the right way to start or is it easier to deal with? [17:21] <lentferj> with this command line [17:21] <lentferj> ffmpeg -probesize 50M -analyzeduration 50M -i 11303_20130409181200.mpg -map_metadata 0 -map 0:v -map 0:a -c:a copy -b:a 384k -c:v libx264 -crf 23 -preset slow -threads 10 -tune film -stats .ffmpeg_tmp_19173/11303_20130409181200.mpg [17:21] <lentferj> I get "This syntax is deprecated. Use '|' to separate the list items." [17:21] <lentferj> on latest ffmpeg from git [17:22] <lentferj> I understood that ',' should not be used anymore [17:22] <lentferj> but I don't have any in this command [17:22] <lentferj> where else should I use '|' ? [17:23] <durandal_1707> lentferj: that warning comes from filter that is auto insterted [17:23] <durandal_1707> there is nothing wrong on your side [17:23] <durandal_1707> this is known problem and will be fixed eventually [17:23] <lentferj> ok, thought something like that :) [17:24] <lentferj> one more regarding this, atm I have something like [17:24] <lentferj> -vf "crop=1276:716:2:2",yadif [17:24] <lentferj> in my scripts [17:25] <lentferj> how should that be written in new syntax? [17:25] <lentferj> -vf "crop=1276:716:2:2"\|yadif [17:25] <lentferj> gives me error [17:25] <durandal_1707> the : is "changed" by fork [17:26] <durandal_1707> so instead of : using to separate options in filter like crop, you use | [17:26] <lentferj> oh, for the values too [17:26] <lentferj> I thought just for lists of options/filters, instead of , use | [17:27] <durandal_1707> nope [17:28] <durandal_1707> but nothing yet changed [17:28] <durandal_1707> because unlike fork we care for backward compatibility [17:29] <lentferj> ;) [17:29] <lentferj> yeah, I see that the old syntax is still working [17:29] <lentferj> but while I upgraded ffmpeg I thought I could also look in to this so it doesnt bite me later :) [17:30] <lentferj> so, ok with -vf "crop=1276\|716\|2\|2" I get "[Parsed_crop_0 @ 0x2ab7400] [Eval @ 0x7ffffa5c3cd0] Invalid chars '|716|2|2' at the end of expression '1276|716|2|2'" [17:31] <durandal_1707> well you could switch to named ones, those will always work, they are still separated by : [17:31] <durandal_1707> even to me its not clear what that warning about | actually means .... [17:31] <ubitux> huh? [17:32] <ubitux> why are you putting '|'? [17:32] <ubitux> no no no there is no reason to put '|' [17:32] <lentferj> ubitux: dunno... trying to "learn" the new syntax [17:32] <ubitux> '|' are used only for list of items [17:32] <durandal_1707> ubitux: you did not notice warning by auto insterted filters that talk about | thin for list of things of same type? [17:33] <lentferj> ubitux: so, what I thought [17:33] <ubitux> yes, for lists, with a few little filters [17:33] <ubitux> durandal_1707: any problem with it? [17:33] <lentferj> ubitux: no problem, just a little confusion [17:33] <durandal_1707> it just spill warning and confusing users, who than gona change all scripts ... [17:34] <ubitux> durandal_1707: with what filtergraph is it warning? [17:34] <lentferj> ubitux: so, when I use e.g. "-vf "crop=1276:716:2:2"\|yadif" I get "Invalid chars '|yadif' at the end of expression '2|yadif'" [17:34] <ubitux> don't use '|' here [17:34] <ubitux> it makes no sense [17:34] <durandal_1707> ubitux: random conversions, aformat [17:34] <ubitux> use a comma, just like before [17:34] <lentferj> oh, ok.. [17:35] <lentferj> then it seems I didn't understand what the warning is about [17:35] <ubitux> lentferj: with what command did you see a warning? [17:35] <lentferj> ubitux: do you have backlog? [17:35] <lentferj> 14 mins ago [17:36] <ubitux> i don't see any filtering info in your cmd line [17:37] <durandal_1707> [audio format for output stream 0:0 @ 0x2984f600] This syntax is deprecated, use '|' to separate sample rate. [17:37] <durandal_1707> when doing simple mp3 encoding [17:37] <ubitux> do we auto insert aformat somewhere? [17:37] <durandal_1707> ffmpeg does [17:38] <lentferj> http://pastie.org/7454670 [17:39] <ubitux> mmh strange [17:39] <ubitux> lentferj: note that we are currently merging such stuff, so it's a bit unstable [17:39] <lentferj> well, it's latest git version .. I can live with that :) [17:40] <lentferj> I just I didn't understand the new synatx right (from the little I could draw from the commit logs) [17:40] <lentferj> thought* [17:40] <ubitux> lentferj: please consider waiting a day or two when the chaos calm down [17:40] <ubitux> it's still a little bit fresh [17:41] <lentferj> ja, I saw like 2 dozen commits of "new synatx".. I pulled last 2 days ago I think :) [17:41] Action: lentferj is a early adopter [17:42] <lentferj> ok, gotta run to pick my kids. If I can test something or the like just ping me. [17:42] <bonana|dlr> Capturing an RTSP-stream how can I write video in audio simultaneously into one file? [17:43] <bonana|dlr> in my custom application. [17:44] <ne555> ¿what should I put in `input' in order to capture the audio output? [17:45] <ne555> (linux) [17:51] <dv-> i'm trying to record my desktop with ffmpeg (libx264 & crf 0) but the result is kind of meh... the colors look noticably washed out and it's not as sharp. is there anything I can do to improve it? [17:52] <ubitux> lentferj: you can try again eventually [17:59] <dv-> this describes the problem pretty well: http://ubuntuforums.org/showthread.php?t=2124591 [17:59] <LithosLaptop> maybe try 4:4:4 chroma format? [18:01] <LithosLaptop> I don'y know the switch [18:01] <LithosLaptop> *dob't [18:02] <LithosLaptop> don't [18:04] <dv-> LithosLaptop: ok, using yuv444p improved it a lot. [18:05] <dv-> anything else I can do? [18:06] <JEEBsv> Just capture in RGB (4:4:4 with RGB coefficients) instead of YCbCr? But to be honest your video will end up being 4:2:0 in the end if you want it to be widely decode'able [18:06] <JEEBsv> 4:4:4 profile is just not supported in most things [18:06] <JEEBsv> (except for libavcodec, naturally) [18:06] <dv-> does youtube support it? [18:07] <JEEBsv> youtube will re-encode your content to 4:2:0, and whether or not it can decode 4:4:4 depends on how old of a libavcodec they're using [18:08] <JEEBsv> If you want to be able to view the thing in flash, you'd have to either host it yourself or use a video hoster that doesn't re-encode, as well as use 4:2:2 YCbCr at most [18:08] <JEEBsv> Flash uses MainConcept's decoder and that surprisingly has support for 4:2:2 as well [18:09] <JEEBsv> most end-user decoders only support 4:2:0 [18:09] <dv-> hm [18:09] <ubitux> lentferj: every unwanted warning should be fixed; if you still have some, please report them [18:10] <dv-> so would it be better in the end to record to 4:2:0 or 4:4:4 and let youtube convert it? [18:12] <dv-> if they do at all. i've had some problems with this in the past [18:12] <LithosLaptop> I think it is best to keep it at 4:2:0 for youtube [18:12] <JEEBsv> Personally I always record in RGB as-is with some fast encoder, and then convert to 4:2:0 with whatever I happen to prefer at that point and upload that in either lossless (youtube etc. that always re-encode), or lossy (non-re-encoding services) [18:12] <LithosLaptop> My video colors are messed up in my computer's media player, or YouTube is only showing gray video. Now what? [18:12] <LithosLaptop> [18:12] <LithosLaptop> Add -pix_fmt yuv420p as an output option. YouTube and most (or perhaps all) non-FFmpeg based players do not support proper decoding of YUV 4:2:2 or YUV 4:4:4. [18:13] <JEEBsv> LithosLaptop: youtube does use libavcodec for decoding, it just means that 'tube uses an old version of it :P [18:13] <JEEBsv> they update very rarely [18:13] <LithosLaptop> oh ok [18:13] <JEEBsv> but yes, in general very, very few decoders support something else but 4:2:0p [18:14] <dv-> that's too bad [18:14] <JEEBsv> MainConcept (Flash) happens to support 4:2:2 as well, but that is closer to being tan exception than the rule [18:14] <dv-> now what about aliasing? is there anything I can tell ffmpeg over -q 0 so that it will keep more detail? [18:14] <JEEBsv> if you are setting a qp of zero for x264, you should already be getting a lossless picture [18:15] <JEEBsv> so if something is happening, it's happening before the picture comes to x264, or after it has been decoded [18:15] <JEEBsv> if you are talking about aliasing between two very different colors [18:16] <JEEBsv> then you can't really do anything 'bout it in general [18:16] <LithosLaptop> what do the html5 browsers vp8 decoders support? [18:16] <JEEBsv> I think VP8 only has a 4:2:0 mode to begin with [18:16] <LithosLaptop> ah ok [18:16] <JEEBsv> since 4:2:0 has full luma ("grayscale"), and a single value for a 2x2 area of chroma (color) [18:17] <JEEBsv> you are bound to get some derpiness that can be more or less apparent depending on the content [18:18] <dv-> even with 4:4:4 it doesn't look as sharp as what I see on screen [18:18] <JEEBsv> first of all try with the RGB encoding mode [18:19] <JEEBsv> I think ffmpeg made that into a separate video codec [18:19] <JEEBsv> although it uses the same x264 in the background [18:20] <JEEBsv> libx264rgb it seems [18:20] <JEEBsv> that should keep the picture in RGB if the input is RGB as well [18:21] <dv-> i assume x11grab is [19:35] <ljudevit> hello, need help with subtitle encoding [19:35] <ubitux> explain your problem [19:36] <ljudevit> works on xubuntu, does not work on red hat. Unable to open iconv context with input character encoding "windows-1250" [19:36] <ljudevit> interesting is, that my own little program that uses icon with 1250 works ok.. [19:36] <ljudevit> ok, will upload [19:37] <ljudevit> command im using: http://pastebin.com/j1ESN9dU [19:39] <ubitux> "and the COMPLETE console output" [19:39] <ljudevit> jep, its here http://pastebin.com/jaznkaRz [19:41] <ubitux> can you try "cp1250" instead? [19:43] <ljudevit> yes, I already did. The same error [19:43] <ubitux> ljudevit: can you pastebin a iconv -l? [19:45] <ljudevit> i can, but it lists all the ever-known-to-man encodings, including the windows-1250 :) [19:45] <ubitux> mmh [19:45] <ubitux> that's strange then [19:45] <ubitux> ljudevit: oh i think i know [19:46] <ubitux> ljudevit: this static build was probably configured with a iconv where these are in a different location [19:47] <ljudevit> that makes sense.. can I somehow help myself without recompiling it on my own? [19:48] <ubitux> burek is building those on ubuntu, so the path where those encodings are is probably different [19:48] <ubitux> i don't know if you can change that at runtime with some magic iconv env variables [19:48] <ubitux> if not, you probably need to build ffmpeg yourself [19:49] <ljudevit> well, that was helpful. thank you very much [19:52] <LithosLaptop> http://www.motobit.com/util/charset-codepage-conversion.asp [19:53] <LithosLaptop> could maybe convert it to utf8? [19:54] <ljudevit> i could, but this is executed as a job and im trying to do everytinh in one process [19:54] <ljudevit> within ffmpeg [19:56] <ubitux> just build it yourself :p [19:57] <ljudevit> will do, using burek's build was a quick shortcut :p [20:27] <Sc0rc3d> Hi folks i got a hopefully smal question for you. :) i wannt to use ffmpeg to stream an local mp4 file to my Red5 Vidoe Server. But i got an error the stream starts 1sec and stops. this my code: ffmpeg -re -i xx.mp4 -vcodec copy -acodec copy -f mp4 rtmp://111.222.333.444:1935/live/ [20:28] <klaxa> hmm... that's a tough one, no idea where it could go wrong [20:28] <klaxa> are you sure red5 is set up correctly? [20:28] <Sc0rc3d> yes i can stream over open broadcast software [20:28] <Sc0rc3d> that works pretty well [20:28] <klaxa> then paste complete console output [20:28] <klaxa> on pastebin [20:28] <klaxa> or a similar site [20:28] <Sc0rc3d> on my way [20:29] <Sc0rc3d> http://pastebin.com/E4w286Cf [20:30] <LithosLaptop> ac3 audio might be a problem? [20:30] <Sc0rc3d> yes i think so too so i tryed with : [20:31] <Sc0rc3d> damn putty closed ^^ [20:31] <LithosLaptop> haha [20:32] <Sc0rc3d> so: ffmpeg -re -i xx.mp4 -vcodec copy -ab 192k -ac 2 -ar 44100 rtmp://xxx.x:1935/live [20:33] <klaxa> you would want to specify an audio codec [20:33] <Sc0rc3d> and then i get an Unable to find a suitable output format for 'rtmp://81.169.184.174:1935/live/test' [20:33] <klaxa> you need -f mp4 or something [20:33] <klaxa> to force an output format [20:33] <Sc0rc3d> yeah tryed that too [20:33] <Sc0rc3d> gave me the same "incorect codec params" error [20:34] <klaxa> well try to specify like... -acodec libmp3lame or something [20:34] <klaxa> you haven't set an audio codec i don't know what the default codec is, you could obtain that info from the log [20:34] <Sc0rc3d> no lame [20:34] <Sc0rc3d> i think i can check by ffmpeg -acodec ? [20:35] <klaxa> would vorbis be okay? [20:35] <Sc0rc3d> quality > * [20:35] <klaxa> ffmpeg -codecs [20:35] <LithosLaptop> -c:a aac -b:a 256k -strict -2 -cutoff 15000 [20:36] <LithosLaptop> or [20:36] <klaxa> mhh >inb4 -c:a is too new of a syntax for his ffmpeg-build [20:36] <LithosLaptop> -acodec aac -ab 256k -strict -2 -cutoff 15000 [20:37] <LithosLaptop> oh ok [20:37] <LithosLaptop> is the experimental aac encoder included in his build? [20:37] <Sc0rc3d> like this? : ffmpeg -re -i xx.mp4 -vcodec copy -acodec aac -ab 256k -strict -2 -cutoff 15000 -f mp4 rtmp://111.222.333.444:1935/live/ [20:37] <klaxa> looking good i guess [20:37] <LithosLaptop> yeah [20:38] <LithosLaptop> oh wait [20:38] <LithosLaptop> how many channels is the ac3 [20:38] <klaxa> 2 [20:39] <klaxa> >Stream #0.1(ger): Audio: ac3, 48000 Hz, 2 channels, s16, 192 kb/s [20:39] <LithosLaptop> ah ok [20:39] <LithosLaptop> thats ok then [20:39] <Sc0rc3d> hmm same error incorect codec params [20:39] <LithosLaptop> paste quickly in pastebin [20:39] <LithosLaptop> does it sy ehich params [20:39] <LithosLaptop> oops [20:39] <LithosLaptop> my wireless keyboard skips letters [20:40] <klaxa> (from the first paste: >Could not write header for output file #0 (incorrect codec parameters ?) ) [20:40] <Sc0rc3d> http://pastebin.com/vrcysnRe [20:42] <klaxa> according to wikipedia mp4 should support ac3 even [20:42] <klaxa> can you as a test write it to a temporary .mkv file? [20:43] <Sc0rc3d> like this: ffmpeg -re -i xx.mp4 -vcodec copy -acodec aac -ab 256k -strict -2 -cutoff 15000 -f mkv test.mkv [20:43] <LithosLaptop> yeah [20:43] <klaxa> not really [20:43] <klaxa> either use -f matroska, or omit -f completely [20:43] <LithosLaptop> wthout -f mkv [20:43] <klaxa> -f mkv will not be recognized [20:44] <Sc0rc3d> he does it [20:44] <Sc0rc3d> with -f matroska [20:44] <Sc0rc3d> writing... [20:44] <LithosLaptop> oo ok [20:45] <Sc0rc3d> i can quit it right? [20:45] <klaxa> yeah [20:45] <Sc0rc3d> last line: frame= 985 fps= 24 q=-1.0 Lsize= 29417kB time=00:00:41.00 bitrate=5877.6kbits ^^ [20:45] <LithosLaptop> and if you try -f mp4 test.mp4 [20:45] <Sc0rc3d> he wirtes too [20:45] <Sc0rc3d> *writes [20:45] <Sc0rc3d> frame= 345 fps= 24 q=-1.0 Lsize= 14423kB time=00:00:14.27 bitrate=8278.8kbits [20:45] <LithosLaptop> hmmm [20:46] <klaxa> hmm.. i can write h264 video with ac3 audio [20:47] <Sc0rc3d> maybee red5 just says that thes support mp4 in some forum threads i just find use flv but i try to convert mkv -> flv it was just a pixelblock at all -.- [20:47] <klaxa> you um... [20:47] <klaxa> you're not supposed to re-encode, also flv only supports certain sample rates [20:47] <klaxa> you'd probably have to resample audio [20:47] <klaxa> but you can try to simply specify -f flv [20:47] <Sc0rc3d> yes i does it with :v copy [20:47] <LithosLaptop> ffmpeg -re -i xx.mp4 -vcodec copy -acodec aac -ab 256k -strict -2 -cutoff 15000 -f flv rtmp://111.222.333.444:1935/live/ [20:48] <Sc0rc3d> omg he does it [20:48] <Sc0rc3d> but it lags ^^ [20:48] <LithosLaptop> haha [20:48] <LithosLaptop> ahh [20:48] <Sc0rc3d> is there a buffer param? [20:49] <Sc0rc3d> awsome :> [20:49] <LithosLaptop> I am not sure, are you streaming over the internet? [20:49] <Sc0rc3d> yep from my qnap nas -> root server [20:49] <Sc0rc3d> t hink not it runs without lag [20:50] <Sc0rc3d> i got enough speed i think http://www.speedtest.net/result/2561684204.png :> [20:50] <LithosLaptop> wow [20:50] <LithosLaptop> want to see mine? [20:51] <LithosLaptop> :) [20:51] <Sc0rc3d> sure :> [20:52] <Sc0rc3d> so there is a -bufsize param expamle "-bufsize 1835k" [20:52] <Sc0rc3d> but cant find on the ffmpeg wiki ^^ [20:52] <LithosLaptop> http://www.speedtest.net/result/2639235177.png [20:53] <Sc0rc3d> uff thats pretty slow bro :) [20:53] <LithosLaptop> hehehe [20:53] <LithosLaptop> mabe I can stream a 160x120, 12fps video :) [20:54] <klaxa> you could stream 1280x720 30 fps video if the bitrate is low enough [20:54] <klaxa> it's not a quesiton of the resolution and framerate :P [20:55] <klaxa> http://x264dev.multimedia.cx/archives/98 [20:55] <uffo> does anyone knows what this means http://pastebin.com/rYM14sRS [20:56] <klaxa> it means "This syntax is deprecated, use '|' to separate sample rate." [20:56] <klaxa> however, if we don't know what you entered to trigger that we can't really help [20:56] <klaxa> (or at least i can't) [20:57] <uffo> ffmpeg.exe -i i.mts -map 0 -map_metadata 0:s:0 -codec:a libfdk_aac -profile:a aac_low -vbr 2 -afterburner 1 -flags +qscale -vn OUTPUT_AAC.m4a [20:57] <uffo> what i should replace here [20:57] <uffo> to lose message [20:59] <klaxa> hmm good question, no idea though... [21:00] <LithosLaptop> yeah, it looks ok [21:00] <klaxa> put the complete console log on pastebin [21:00] <klaxa> that might give more hints [21:01] <durandal_1707> nothing, message is from inserted filter [21:01] <durandal_1707> updating to latest master should silence that nonsense spam [21:01] <uffo> http://pastebin.com/965aGrG0 [21:02] <uffo> even with other codecs it apears [21:02] <durandal_1707> as i already said, nothing is wrong at your side [21:03] <uffo> thanks, then i just ignore it [21:05] <Sc0rc3d> oh i forgot thanks a lot to klaxa and LithosLaptop [21:05] <Sc0rc3d> :) [21:09] <LithosLaptop> no prob :) [21:24] <uffo> looks like my last problem is connected with this [21:24] <uffo> http://permalink.gmane.org/gmane.comp.video.libav.devel/44209 [21:24] <uffo> what this means [21:28] <durandal_1707> uffo: uffo what problem with ffmpeg do you have? [21:29] <uffo> i tried to engode, encoding fine just warning messages http://pastebin.com/965aGrG0 [21:29] <uffo> encode* [21:30] <durandal_1707> uffo: when you compiled ffmpeg from source? [21:31] <uffo> today [21:31] <durandal_1707> why? [21:31] <durandal_1707> anyway you could fetch and recompile again and that warning should dissapear [21:32] <uffo> http://permalink.gmane.org/gmane.comp.video.libav.devel/44209 how this is connected because this has something to do with this i presume [21:32] <uffo> This syntax is deprecated", use '|' to separate channel layout [21:33] <uffo> what layout? i dont see anything similar [21:33] <durandal_1707> uffo: because converions is done in background [21:33] <durandal_1707> see -v debug [21:34] <uffo> so i just should ignore this, that this do not cause anything [21:35] <durandal_1707> is runs this filter for example aformat=sample_fmts=u8,s16p [21:35] <durandal_1707> but now it it is: aformat=sample_fmts=u8|s16p [21:38] <uffo> weird that it displays to me if i do not use this, is this just a todays build bug? [21:40] <durandal_1707> uffo: it is already fixed, (i'm repeating myself several times) [21:40] <uffo> i understood [21:40] <uffo> thanks [21:46] <brontosaurusrex> whats the procedure to get certain file to work in ffmpeg? is that a bug, request or what exactly? mailing list or something else? [22:30] <stefan_> Did anybody see Mans conversation with Mr. Russell King over at the linux mailing lists? [22:30] <ubitux> i did :) [22:31] <stefan_> I felt so remided of the old days on the ffmpeg-devel mailing list. [22:34] <stefan_> http://thread.gmane.org/gmane.linux.ports.arm.kernel/230304/focus%3D230523 for the one interested in doing some reading up [00:00] --- Fri Apr 12 2013

1 0

ffmpeg-devel.log.20130410
by burek 11 Apr '13

11 Apr '13

[00:00] <durandal_1707> that will make shorthand gone? :( [00:00] <ubitux> shorthand code yes, not the feature [00:01] <durandal_1707> then we will get bunch of reports that random thing breaks [00:01] <ubitux> but we need to do something about aliases [00:21] <megaTherion> Hello, Im ffmpeg API newbie [00:21] <megaTherion> what can I do if avformat_open_input is not able to determine the codec used? [00:49] <elkng> is opencv support currently broken as it was in 0.11.1 ? [00:49] <durandal_1707> what is broken? [00:50] <elkng> durandal_1707: libopencv [00:51] <durandal_1707> i know that, you told it already [00:51] <durandal_1707> but i need more info [00:51] <elkng> is it safe to use --enable-libopencv ? [00:51] <durandal_1707> safe in what sense? [00:52] <durandal_1707> michaelni: i dont see curves patch anywhere [00:52] <durandal_1707> actually it is another thread - i really dislike that .... [00:52] <durandal_1707> because i make pointless replies [00:53] <elkng> durandal_1707: http://code.opencv.org/issues/1925 [01:09] <durandal_1707> ubitux: there is smartdecimate [01:18] <saste> uhm the select patch is broken [01:18] <saste> in a subtle way [01:19] <ubitux> saste: i think working on a generic option for filters would make more sense [01:19] <ubitux> or something in that spirit [01:20] <llogan> fun. i get a segfault using libx264 [01:20] <saste> ubitux, yes, but splice is still somehow useful *for other purposes* [01:20] <ubitux> saste: what are you thinking about? [01:20] <saste> e.g. for inserting subliminal frames [01:21] <ubitux> :D [01:21] <ubitux> ok :) [01:26] <durandal_1707> i gonna push separatefields and telecine if nobody object [01:31] <ubitux> durandal_1707: so you were able to get something working with the pts? [01:31] <saste> durandal_1707, give me a day for review, or push it [01:31] <ubitux> durandal_1707: did you check if it works fine with mpeg and its huge tb? [01:31] <durandal_1707> you talking about telecine? [01:32] <ubitux> no, separatefields [01:32] <durandal_1707> ubitux: i replaced *2 with ++ [01:33] <durandal_1707> and doubled framerate nominator and timebase denominator, [01:33] <ubitux> yes right, now you have a frame counter in the frame pts [01:33] <ubitux> but is it correct? [01:33] <durandal_1707> it works with avi/mkv/lavfi [01:34] <ubitux> ok.. [01:34] <ubitux> i'm not a pts expert, so i let that for others :p [01:35] <ubitux> durandal_1707: are you able to separate and reconstruct frames without pts messing? [01:35] <durandal_1707> what you mean? [01:35] <llogan> eh, the seg fault was a pebkac i guess [01:36] <ubitux> durandal_1707: the idea is to have -vf "separatefields, randomvideoprocessing, joinfields", and no pts messing in output [01:38] <durandal_1707> how could i do that? [01:38] <durandal_1707> leave pts as is? [01:39] <ubitux> generating pts according to the fps and tb [01:39] <ubitux> just like you did in telecine [01:39] <durandal_1707> that did not work [01:39] <ubitux> ah? [01:40] <durandal_1707> bunch of frame droping in ffplay [01:40] <ubitux> what caused this? [01:40] <durandal_1707> perhaps all pts become same [01:43] <ubitux> so fps = (AVRationnal){inlink->frame_rate.num * 2, inlink->frame_rate.den} and multiplying the frame count for each frame by av_q2d(av_inv_q(av_mul_q(fps, outlink->time_base))) didn't work? [01:46] <cone-724> ffmpeg.git 03Michael Niedermayer 07master:eedcac68f31c: ffmpeg: dont exit 0 if fewer than 1/3 of the input could be decoded. [01:47] <ubitux> maybe something is wrong in the calculation, but i remember checking it [01:49] <durandal_1707> i get bunch of fd increasing with ffplay and -f lavfi -i testsrc [01:50] <durandal_1707> but that looks to be yet another stupid ffplay bug [01:50] <ubitux> durandal_1707: try to revert de81d8da [01:50] <ubitux> if that's still possible [01:51] <durandal_1707> i test with ffmpeg now [01:51] <ubitux> and feel free to +1 ticket #2446 if it solves your problem [01:52] <ubitux> saste: mans repo is forked from yuvi so the patch doesn't look like a good idea [01:52] <ubitux> instead we should make sure yuvi repo pulls from it [01:53] <Yuvi> hm? [01:53] <llogan> is there a downside to using -filter_complex for a simple filtergraph? [01:53] <ubitux> Yuvi: http://ffmpeg.org/pipermail/ffmpeg-devel/2013-April/142075.html [01:53] <ubitux> Yuvi: are you github/yuvi? [01:54] <Yuvi> yeah [01:54] <ubitux> then it seems something in mansr repo fixed something you didn't merge [01:54] <durandal_1707> llogan: i dont think so [01:54] <ubitux> at least according to this mail [01:55] <saste> llogan, "simple filtergraph" -> define it [01:55] <cone-724> ffmpeg.git 03Nicolas George 07master:de38e7c8c3df: ffmpeg: use a rational for -aspect option. [01:55] <cone-724> ffmpeg.git 03Michael Niedermayer 07master:6a78ec180eb9: ffmpeg: print frame decoding error statistic [01:55] <cone-724> ffmpeg.git 03Michael Niedermayer 07master:f1c0c6b7de6f: Merge remote-tracking branch 'cigaes/master' [01:56] <llogan> saste: one input and one output as defined in http://ffmpeg.org/ffmpeg.html#Simple-filtergraphs [01:57] <saste> llogan, not that i know of [01:57] <durandal_1707> saste: that example with pr0n is .... [01:57] <llogan> saste: i see users swapping -vf and -filter_complex often [01:58] <saste> durandal_1707, evil? [01:58] <Yuvi> bah, no pull request [01:59] <llogan> durandal_1707: no tentacles? [02:00] <saste> durandal_1707, it's a hidden reference to a popular movie [02:00] Action: saste sleeps [02:00] <Yuvi> there [02:01] <ubitux> Yuvi: cool; care to send a mail on the ml to say so to the guy? [02:04] <Yuvi> once I get home [02:14] <cone-724> ffmpeg.git 03Paul B Mahol 07master:d0073c7a0bc5: separatefields filter [02:20] <durandal_1707> on one of fourm: "What is the syntax to get ffmpeg to encode with vapoursynth?" [02:27] <cone-724> ffmpeg.git 03d s 07master:ca3cef719e38: avformat/avisynth: Fix off-by-one error in avisynth demuxer. [02:27] <cone-724> ffmpeg.git 03d s 07master:685617ac29d7: avformat/avisynth: Cosmetics [02:29] <ubitux> durandal_1707: indeed, looking at showinfo the pts are the same after the op [02:31] <ubitux> i wonder if doubling the tb wouldn't help [02:32] <ubitux> durandal_1707: you working on weave/joinfilters btw? [02:32] <durandal_1707> ubitux: i reading that flame again [02:33] <durandal_1707> ubitux: doubling tb and ussing ++ caused unusual frame dropping for mpeg sample that we have in fate [02:35] <ubitux> even using the outlink tb for the ts unit computation? [02:35] <ubitux> i got it fixed locally here [02:36] <ubitux> durandal_1707: http://pastie.org/7397150 [02:36] <durandal_1707> that good (i suck at it) [02:36] <ubitux> can you try this? [02:36] <durandal_1707> ubitux: for what? it it fixes problem push it (i had pts warning with -f sdl but did not cared ....) [02:37] <durandal_1707> *if it fixes .... [02:39] <cone-724> ffmpeg.git 03Clément BSsch 07master:d2e051e30c0f: lavfi/separatefields: adjust tb to avoid rounding errors. [02:39] <ubitux> can you retry with the mpeg sample, -f sdl, etc? [02:41] <durandal_1707> ubitux: seems ok, sdl warning is gone [02:42] <durandal_1707> guess telecine needs updating too? [02:42] <ubitux> did you push it already? [02:42] <durandal_1707> noo [02:43] <ubitux> telecine is increasing frame rate, so yeah you'll need to adjust the tb [02:43] <ubitux> and use it properly [02:43] <cone-724> ffmpeg.git 03Christophe Gisquet 07master:2383068cbfac: x86: sbrdsp: implement SSE2 qmf_pre_shuffle [02:48] <llogan> separatefields: s/use/uses. yeah, i'm too late and it's just a nit. [02:50] Action: ubitux would like to propose ohgodno.wmv instead of pr0n.mkv [04:44] <cone-724> ffmpeg.git 03Christophe Gisquet 07fatal: ambiguous argument 'refs/tags/n0.10.7': unknown revision or path not in the working tree. [04:44] <cone-724> Use '--' to separate paths from revisions [04:44] <cone-724> refs/tags/n0.10.7:HEAD: x86: sbrdsp: implement SSE2 qmf_pre_shuffle [08:29] <MrNaz> i have 3 h.264 files in .mp4 containers which I need to join without changing their codec or container format. the method in the docs http://libav.org/faq.html#How-can-I-join-video-files_003f is about as ugly as things get. surely there is a way to achieve video joins without transcoding twice ? [08:30] <MrNaz> oops wrong chan... i'm actually using libav [08:30] <MrNaz> sorry [09:23] <ubitux> MrNaz: libav doesn't support it, but we do; https://ffmpeg.org/faq.html#How-can-I-join-video-files_003f [09:23] <ubitux> also, this is a user question, so it belongs to #ffmpeg [09:24] <ubitux> (see the concat demuxer) [09:32] <Tjoppen> here's a thought: why aren't more people using melt? [10:02] <cone-987> ffmpeg.git 03James Almer 07master:89388a953a51: Replace all occurrences of PRI in sscanf() calls with SCN [12:12] <cone-987> ffmpeg.git 03Clément BSsch 07master:2a7f885fe1cf: lavf: add libquvi demuxer. [12:17] <ubitux> now we need some overlay api. [12:17] <ubitux> oh well, or maybe i'll work on injecting avsubtitles in lavfi [12:35] <saste> ubitux, yes [12:36] <saste> a good step in the right direction would be to move the overlay code from vf_overlay.c to a separate file [12:36] <saste> and maybe write a public API for that [12:36] <saste> or we could consider a libswsblend lib [12:36] <saste> (or move it to libavutil) [12:38] <ubitux> the other way is to make sure every blending code can be done with overlay filter [12:38] <ubitux> but that means injecting subtitles into lavfi [12:38] <ubitux> but anyway, we already have some blending code somehow [12:38] <ubitux> such as drawutils [12:38] <ubitux> but private :( [12:39] <saste> ubitux: are you aware of any drawing library with scripting support? [12:39] <saste> which allows to load a script to write custom graphics? [12:43] <ubitux> nope [12:44] <saste> ubitux: what do you think of splice? [12:44] <ubitux> didn't have a look yet [12:44] <saste> it has many limitations, but can still be useful [12:44] <saste> i tried to explain limitations in the docs, so the users should be aware of them [12:45] <saste> it is an advanced filter anyway, so I don't expect it will be widely deployed [12:45] <ubitux> i'll have a look this afternoon or tonight [12:47] <ubitux> so who's gonna write libswblend? [12:55] <cone-987> ffmpeg.git 03Kostya Shishkov 07master:56c1b9257698: dfa: implement missing TDLT coding method [12:55] <cone-987> ffmpeg.git 03Anton Khirnov 07master:b439c992c23f: lavfi: switch to an AVOptions-based system. [12:55] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:0594ef0deab7: Merge commit 'b439c992c23f3e0f3832fffd2a34a664b236c525' [13:09] <michaelni> ubitux, saste if you have bugfixes for antons options, or suggestions / comments on merging, dont hesitate to commit/say [13:20] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:8afcaaeb75bc: lavu/eval: rename "new_eval_expr()" to "make_eval_expr()" [13:21] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:cc3edd99a678: lavfi/overlay: add dynamic expression evaluation support [13:21] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:006e20bd3655: lavfi/overlay: add enable expression [13:21] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:d2752ef0615d: lavfi/overlay: add process_command callback [13:37] <cone-987> ffmpeg.git 03Anton Khirnov 07master:4fa1f52e33b7: af_resample: switch to an AVOptions-based system. [13:37] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:828044aca289: Merge commit '4fa1f52e33b70029e2b621852f3af7c1ef9aecff' [13:47] <langoliers> 8 keys are enough for everything, you just need to press the right combination, or use 10 and one of your fingers can be clock, and 512 keys with one clock. nah, cool? [13:49] <cone-987> ffmpeg.git 03Anton Khirnov 07master:d28cb849899a: buffersrc: switch to an AVOptions-based system. [13:49] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:451315830a3b: Merge commit 'd28cb849899abd98bf2dd72f5a3dd56d441aeb27' [14:06] <cone-987> ffmpeg.git 03Anton Khirnov 07master:0af7fe1f95a9: af_aformat: switch to an AVOptions-based system. [14:06] <cone-987> ffmpeg.git 03Anton Khirnov 07master:e67a87eac814: vf_(no)format: switch to an AVOptions-based system. [14:06] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:2a2a643c90ff: Merge commit 'e67a87eac814c7805d18c983c43033a8a1bd62af' [14:08] <michaelni> and again the commits from libav are full of unrelated changes "a hack for compatibility with the old syntax replace colons with |s" [14:20] <saste> michaelni, what if we merge doc/eval.texi and syntax.texi -> utils.texi? [14:28] <michaelni> saste, dunno, no oppinion from me, iam not maintaining the docs [14:28] <saste> michaelni, main problem is merge issues, that's why I'm asking [14:29] <saste> i want to avoid to have too many files, for example we may need to add an opencl.texi [14:31] <michaelni> saste, if we run in a problem with docs merges, then you would have to help [14:32] <michaelni> also about avfilter if someone wants to help merging that options stuff, commiting cherry picks / rebased commits is welcome! [14:33] <michaelni> or maybe asking me to merge some branch with rebased / cherry picked commits [14:39] <saste> michaelni, I can try to allocate a few hours, you could create a merge branch and we work on it, like we did for TEM [14:40] <elkng> since when it became possible to extract subtitles from video ? since 1.2 version ? [14:41] <michaelni> saste, no need for a merge branch, you can just branch git master and work on that [14:41] <michaelni> individual commits should not break anything this time [14:42] <saste> michaelni, ah yes [14:42] <michaelni> so individual commits can simply be merged into master [14:43] <michaelni> now i need to fix alternative option names with the new system [14:44] <michaelni> well "new" is maybe not the right term, "different" is maybe a better term [15:42] <cone-987> ffmpeg.git 03Anton Khirnov 07master:2831b307e65f: vf_aspect: switch to an AVOptions-based system. [15:42] <cone-987> ffmpeg.git 03Anton Khirnov 07master:62dcdb028cc8: vf_blackframe: switch to an AVOptions-based system. [15:42] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:43bac121d3de: Merge commit '62dcdb028cc84845fd263bb09304c4c6500bda7a' [15:53] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:90efdf98b1ce: avfilter: support alternative keys in the alternative shorthand system [15:53] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:a3cb303a4c22: avfilter/blackframe: add "thres" back as alternative [16:05] <saste> ubitux: how to use quvi? [16:05] <saste> an example command would be nice [16:05] <saste> also libquvi is listed before image2 [16:39] <cone-987> ffmpeg.git 03Anton Khirnov 07master:51def31dbe5b: vf_boxblur: switch to an AVOptions-based system. [16:39] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:eb6d58d23c55: Merge commit '51def31dbe5b6e857536de8fa428f263d64f3ae5' [17:02] <cone-987> ffmpeg.git 03Anton Khirnov 07master:fba0156af77b: vf_crop: switch to an AVOptions-based system. [17:02] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:200e04c70f9a: Merge commit 'fba0156af77b11ec99edf4ee8f511b7aaa6b1891' [17:17] <durandal_1707> libswblend - are you serious? [17:27] <saste> durandal_1707, libswsdraw? [17:27] <durandal_1707> that is pointless [17:27] <durandal_1707> where would you use libswwhatever? in lavfi only? [17:28] <saste> durandal_1707, no [17:28] <saste> do you know a C library for low level drawing utilities? [17:28] <saste> but i'm not saying we should implement it for real, just the idea doesn't look so crazy [17:29] <durandal_1707> just do in in lavu, we do not need yet another lib [17:30] <durandal_1707> i think that would also make mng decoder possible [17:30] <durandal_1707> (with less work....) [17:31] <durandal_1707> maybe separate lib would make sense for vector(instead of pixels) based stuff [17:33] <saste> durandal_1707, NIH [17:33] <durandal_1707> users enjoy broken scripts [17:33] <durandal_1707> why lavfi major is not bumped? [17:34] <durandal_1707> saste: NIH NIH what? [17:34] <saste> durandal_1707, because l likes big bunches of patches which broke everything, and bumps major at the end of it [17:35] <saste> NIH some vector graphics library [17:35] <saste> not sure ffmpeg would be the place for that [17:39] <durandal_1707> so i still keep shorthand or? [17:44] <cone-987> ffmpeg.git 03Anton Khirnov 07master:460e7b4f6d47: vf_cropdetect: switch to an AVOptions-based system. [17:44] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:1c3cdf533ba0: Merge commit '460e7b4f6d473d9f03ed45501221f9cb209b28fd' [17:59] <cone-987> ffmpeg.git 03Anton Khirnov 07master:63e58c55c17d: vf_delogo: switch to an AVOptions-based system. [17:59] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:eebe0b02245c: Merge commit '63e58c55c17d7f8b5eec9c082fe0f8edc305a24e' [18:01] <durandal_1707> how many more to go? [18:03] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:270217908b8c: ffmpeg: update error message, and make use of av_err2str() to simplify [18:03] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:e62fca11a84d: lavfi/amerge: drop duplicated "Buffer queue overflow" message [18:03] <cone-987> ffmpeg.git 03Stefano Sabatini 07master:215ca864759a: doc/filters: review introductory example and explanation [18:07] <michaelni> durandal_1707, dunno, i didnt count but there are plenty left [18:16] <cone-987> ffmpeg.git 03Anton Khirnov 07master:335c31293bae: vf_drawbox: switch to an AVOptions-based system. [18:16] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:0726d516dcd1: Merge commit '335c31293baec6e6cf5907bd29840af3de8ff735' [18:20] <durandal_1707> this page is not maintained: http://wiki.multimedia.cx/index.php?title=Libavfilter [18:20] <saste> durandal_1707, destroy it [18:20] <saste> better nothing than outdated docs [18:21] <saste> also the multimedia wiki is not the place for ffmpeg library docs [18:36] <cone-987> ffmpeg.git 03Anton Khirnov 07master:8ec464c61c46: vf_drawtext: switch to an AVOptions-based system. [18:36] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:d70b49442238: Merge commit '8ec464c61c46a5d088b43d24e384d6c37b7ef14a' [18:41] <saste> f**k why is anton removing the option help messages?? [18:42] <durandal_1707> because he is evil [18:44] <iive> probably because no program needs built-in help, when external one is available and even better.~~ [18:46] <saste> iive: what external help? [18:49] <iive> good question. ;) aren't all options covered in man ffmpeg/avconv ? . (btw ~~ is sign for sarcasm). [18:53] <cone-987> ffmpeg.git 03Anton Khirnov 07master:b9dfee9fa259: vf_fade: switch to an AVOptions-based system. [18:53] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:0c3d706bdb22: Merge commit 'b9dfee9fa259dfc885508179a359dccc9e7840bd' [18:59] <ubitux> saste: ffplay 'http://youtube.com/...' [19:00] <ubitux> michaelni: i'll have a look to the option thing, thx [19:00] <saste> ubitux, I got: [libquvi @ 0x7fb5340008c0] ý [19:00] <ubitux> huh? [19:01] <ubitux> what version? [19:01] <saste> but that's maybe fault of the possibly cr@@y packaging provided here [19:01] <ubitux> saste: what help messages are you talking about? [19:02] <ubitux> they are consistent over the filters iirc now [19:02] <ubitux> so they can be removed [19:03] <durandal_1707> i get 0000..... with soxenc and i still dont know why..... [19:03] <saste> ubitux, libquvi7.0.0 [19:03] <ubitux> dafuck [19:03] <ubitux> latest is 0.4.1 [19:04] <ubitux> .14 sorry [19:04] <saste> 0.4.0-1 [19:04] <saste> debian funny numbering, to confuse and dominate over the masses [19:05] <ubitux> do you have warnings at build time? [19:05] <saste> anyway, also you're returning the quvi error messages, while i would expect an AVERROR_EXTERNAL [19:05] <saste> is that intended? [19:05] <saste> let me check [19:06] <ubitux> maybe AVERROR_EXTERNAL is more relevant indeed [19:06] <ubitux> or AVERROR(ret) [19:07] <ubitux> saste: btw, we need to review the commits from anton [19:07] <ubitux> typically the drawtext one [19:07] <ubitux> where he removed for no reason the (admitely incomplete) option description [19:08] <ubitux> i'll try to have a look to every filter tonight [19:09] <cone-987> ffmpeg.git 03Anton Khirnov 07master:a39c154049a2: vf_fieldorder: switch to an AVOptions-based system. [19:09] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:b8a1f8b4e28b: Merge commit 'a39c154049a2d0c4fb02a5c74f58d6986ec21cec' [19:13] <saste> ubitux, most option help messages are set [19:13] <ubitux> yes seems michael merged them properly [19:13] <ubitux> but that was just an example [19:14] <saste> also I don't know why l insists with using grammatically inconsistent forms for the help field [19:15] <ubitux> let's not talk about that so loudly [19:15] <ubitux> when they realize they will use a different consistency than ffmpeg [19:15] <ubitux> and it will be a pain to merge again [19:15] <durandal_1707> when auto-merging will stop? [19:16] <saste> ubitux, anyway something is fishy with libquvi here [19:16] <ubitux> durandal_1707: the introduced feature is nice [19:16] <ubitux> saste: it's possibly only supporting 0.4.1 [19:16] <saste> ubitux, how's that possible? [19:16] <saste> shouldn't it fail to compile then? [19:17] <saste> or did they break ABI? [19:17] <ubitux> i can only test 0.4.1 here [19:17] <saste> uhm.. i'll try later with an updated version [19:20] <ubitux> michaelni: can you push the curves patch btw? [19:23] <ubitux> huh, '|' is the option separator now? oO [19:24] <ubitux> saste: i'll fix libquvi in a day or two, i'd like to fix lavfi asap firs [19:24] <ubitux> first [19:25] <saste> ubitux, sure [19:27] <cone-987> ffmpeg.git 03Anton Khirnov 07master:f13ab2992588: vf_fps: switch to an AVOptions-based system. [19:27] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:81c1ed748be5: Merge commit 'f13ab29925883b4245da4129694af3af378d67be' [19:33] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:9280fc7da1fc: vf_curves: add option to set all curves at once [19:56] <cone-987> ffmpeg.git 03Anton Khirnov 07master:5aa1a668cfae: vf_frei0r: switch to an AVOptions-based system. [19:56] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:da3f89988f03: Merge commit '5aa1a668cfae7f617e1a06efad20f87283badd8a' [20:26] <burek> hm, this is strange [20:27] <burek> i got a situation where ffmpeg compiled fine with libx264 [20:27] <burek> but when i run it using ./ffmpeg [20:27] <burek> (didnt do make install yet) [20:27] <burek> it says ./ffmpeg: error while loading shared libraries: libx264.so.130: cannot open shared object file: No such file or directory [20:27] <burek> how did it find it during make and now it cant find it? [20:28] <burek> when i type: LD_LIBRARY_PATH=/usr/local/lib ./ffmpeg [20:28] <burek> it works fine, of course [20:28] <nevcairiel> depending how the configure check is setup, it may be enough to have the headers during build [20:30] <burek> build goes fine, without problems [20:30] <burek> compiled libx264 and make install, then just configured ffmpeg and make [20:32] <burek> oh... im so stupid.. [20:32] <burek> it was /etc/ld.so.conf [20:32] <burek> missing /usr/local/lib entry.. sorry guys :S [20:33] <JEEB> yes, shared libs are fun :) [20:33] <JEEB> if you are only going to use that library for ffmpeg, I recommend just using static libraries [20:41] <burek> yeah, i just wanted to test a 16 core machine :D [20:41] <burek> ^^ [20:41] <burek> it compiled ffmpeg in 1:30 min [20:41] <burek> :))) [20:41] <burek> with libx264 and libmp3lame :) [20:42] <nevcairiel> i was surprised how small libx264 really is, i always figured it would be bigger, but i build it last week for the first time... :p [20:42] <JEEB> yeah, ~1MB or so [20:43] <kierank> nevcairiel: you can write mixed 8/10-bit support perhaps :) [20:43] <kierank> if you want it bigger [20:43] <nevcairiel> nah thanks [20:43] <nevcairiel> i dont usually encode stuff [20:43] <cone-987> ffmpeg.git 03Anton Khirnov 07master:c334c113d4d9: vf_scale: switch to an AVOptions-based system. [20:43] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:85f115b5d8d7: Merge commit 'c334c113d4d9e9a41bc38a3e4458d7ab21010401' [21:03] <durandal_1707> ubitux: fieldorder query_formats looks wrong [21:14] <cone-987> ffmpeg.git 03Anton Khirnov 07master:7ed833d78ea6: vf_gradfun: switch to an AVOptions-based system. [21:14] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:1ee9eacaa492: Merge commit '7ed833d78ea661d619124fd898547a900f6480bc' [21:24] <cone-987> ffmpeg.git 03Anton Khirnov 07master:8c747d46f721: vf_hqdn3d: switch to an AVOptions-based system. [21:24] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:f780a90355aa: Merge commit '8c747d46f721cffa8ea51990805ad1d3a3a4fd0a' [21:30] <cone-987> ffmpeg.git 03Anton Khirnov 07master:ee0e8d4b15a8: vf_libopencv: switch to an AVOptions-based system. [21:30] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:94ec709e3bdd: Merge commit 'ee0e8d4b15a87932ab6066cd6eae3cab08726319' [21:43] <ubitux> durandal_1707: what's wrong? [21:44] <durandal_1707> subsampling check [21:44] <ubitux> sorry i don't get it [21:45] <durandal_1707> also it calls ff_formats_ref twich which is not needed [21:46] <ubitux> i don't see any ff_formats_ref call [21:46] <durandal_1707> libavfilter/vf_fieldorder.c [21:47] <ubitux> ohh sorry [21:48] <ubitux> oh well, i'm not responsible from that filter ;) [21:56] <durandal_1707> also it still have .needs_writable for some reason [21:56] <durandal_1707> ignore that.... [21:56] <durandal_1707> it calls ff_formats_unref, i dont think many filters call this [22:14] <cone-987> ffmpeg.git 03Anton Khirnov 07master:20b46f8f4fff: vf_lut: switch to an AVOptions-based system. [22:14] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:8bdb119197ad: Merge commit '20b46f8f4fff6aeeab9ea418dc359eda8887ced6' [22:32] <cone-987> ffmpeg.git 03Anton Khirnov 07master:9087eaf193b8: vf_overlay: switch to an AVOptions-based system. [22:32] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:e434ddd4c699: Merge commit '9087eaf193b8ce99c41352064a81916fa66adb49' [22:43] <durandal_1707> michaelni: your compiler does not issue warning for multiplier initializers? [22:45] <cone-987> ffmpeg.git 03Paul B Mahol 07master:16f1e832badd: lavfi/overlay: remove dupe initializer [22:52] <michaelni> durandal_1707, either that or more likely i missed it [22:57] <durandal_1707> huh what is this -filter_script thing, i missed this feature [22:57] <cone-987> ffmpeg.git 03Anton Khirnov 07master:40c885c58980: vf_pad: switch to an AVOptions-based system. [22:57] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:7e99ccf5d8c2: Merge commit '40c885c589808455a1c4b473509f1e6cd4908f55' [22:57] <nevcairiel> you can put complex filters into a file [22:58] <nevcairiel> dont need to pass them on the commandline [22:59] <durandal_1707> but its worded wrong, its far from script .... [23:00] Action: durandal_1707 wish to make michaelni faster [23:04] <durandal_1707> what to do with rgb2xyz filters? [23:29] <cone-987> ffmpeg.git 03Anton Khirnov 07master:95f1f56a21bc: vf_select: switch to an AVOptions-based system. [23:29] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:b3fb2d8cec91: Merge commit '95f1f56a21bc2f824af6fb97ca7ab35cdd0c401e' [23:33] <ubitux> hey saste [23:33] <saste> yo [23:33] <ubitux> saste: i'm switching a bunch of our filters to the new system [23:33] <ubitux> care to give me a hand? [23:34] <saste> ubitux, what's needed? [23:34] <ubitux> i've done those: https://github.com/ubitux/FFmpeg/compare/shorthand-cleanup [23:34] <cone-987> ffmpeg.git 03Clément BSsch 07master:7c11c216690c: lavfi/thumbnail: switch to an AVOptions-based system. [23:34] <cone-987> ffmpeg.git 03Clément BSsch 07master:1b1688bfc640: lavfi/silencedetect: switch to an AVOptions-based system. [23:34] <cone-987> ffmpeg.git 03Clément BSsch 07master:4964ec2cd482: lavfi/showspectrum: switch to an AVOptions-based system. [23:34] <cone-987> ffmpeg.git 03Clément BSsch 07master:5afd58fbb457: lavfi/ebur128: switch to an AVOptions-based system. [23:34] <cone-987> ffmpeg.git 03Clément BSsch 07master:8388e1e2b343: lavfi/perms: switch to an AVOptions-based system. [23:34] <cone-987> ffmpeg.git 03Clément BSsch 07master:8da1f1f2081c: lavfi/colormatrix: switch to an AVOptions-based system. [23:34] <cone-987> ffmpeg.git 03Clément BSsch 07master:34610e11e7d5: lavfi/curves: switch to an AVOptions-based system. [23:34] <cone-987> ffmpeg.git 03Clément BSsch 07master:1cdb9f48dbc3: lavfi/edgedetect: switch to an AVOptions-based system. [23:34] <cone-987> ffmpeg.git 03Clément BSsch 07master:f8eabab04d1b: lavfi/geq: switch to an AVOptions-based system. [23:34] <cone-987> ffmpeg.git 03Clément BSsch 07master:8da1fff85a90: lavfi/pp: switch to an AVOptions-based system. [23:34] <cone-987> ffmpeg.git 03Clément BSsch 07master:e2b0edb6ac9c: lavfi/subtitles: switch to an AVOptions-based system. [23:34] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:270d3c96ec49: Merge remote-tracking branch 'ubitux/shorthand-cleanup' [23:35] <ubitux> saste: well, just like that ^ [23:38] <saste> so basically we just need to remove the shorthands, and add the filters to the anton_options list? [23:38] <saste> what's the problem with aliases? [23:39] <durandal_1707> s/anton_options/evil_options/ next time [23:39] <durandal_1707> i said using personal name is unprofessional [23:40] <saste> yes [23:40] <durandal_1707> nit+++++++++++++++++++++: vertical align [23:42] <saste> ubitux: what about aliases? [23:42] <ubitux> saste: michael fixed that [23:42] <ubitux> (aliases need to be following each other, but it works) [23:43] <saste> uhm let me try (smartblur) [23:46] <saste> no more mixed short and long options? [23:46] <ubitux> yeah this needs to be fixed [23:46] <ubitux> if you could do that, that would be a great help... [23:47] <ubitux> i need it for tile notably (or i need to change tests, examples, ..) [23:47] <ubitux> (and it sucks to drop that feature) [23:47] <saste> ubitux: I can't before a few days [23:48] <ubitux> i can give you 15 minutes and 41 seconds [23:48] <saste> today i spent mostly on ffmpeg, but tomorrow i'll have other stuff to do [23:48] Action: llogan is reminded to do taxes [23:49] <ubitux> saste: what about making hue sane then? [23:50] <saste> what's the plan w.r.t. docs? [23:50] <saste> options should be listed in the same order of shorthand? [23:50] <saste> that's not always convenient, but let's see what we can do... [23:51] <ubitux> yes we need to check the orders [23:53] <ubitux> saste: the filters doc will need a complete review after this mess [23:53] <ubitux> but that's imo not the priority :p [23:53] <saste> the keys are assumed to be the option names in the order they are declared [23:53] <ubitux> yes [23:53] <saste> declared in the docs, or in the code? [23:53] <ubitux> we need to make the options list match our shorthand in both code and doc [23:54] <saste> also we need to document how we deal with the aliases [23:54] <ubitux> that's intuitive, so far [23:54] <ubitux> saste: the priority is to get every filter working with dict init [23:54] <cone-987> ffmpeg.git 03Anton Khirnov 07master:33b97faaba27: vf_setpts: switch to an AVOptions-based system. [23:54] <cone-987> ffmpeg.git 03Michael Niedermayer 07master:f42635a50b7a: Merge commit '33b97faaba2744f0a2fd65c0ef9ecc2de3fad7ff' [23:55] <ubitux> saste: fix the short/long limitation in the meantime [23:55] <ubitux> and then deal with the doc [23:55] <saste> ok [23:55] <saste> i prefer to do the doc together with the filters [23:55] <ubitux> yes that's what i do [23:55] <ubitux> i was talking about a second pass [23:55] <ubitux> anyway.. [23:55] Action: ubitux git grep shorthand & go to work [23:56] <saste> and it's not like it's rocket science [00:00] --- Thu Apr 11 2013

1 0

ffmpeg.log.20130410
by burek 11 Apr '13

11 Apr '13

[00:31] <elkng> if I want to extract subtitles from mkv, should I use some option --enable- when compiling ffmpeg to add that ability ? [00:32] <elkng> something like --enable-libass or other ? [00:33] <elkng> will option --enable-gpl restrict of using additional libraries or allow to use them ? [00:34] <klaxa> it will allow more libraries [00:34] <elkng> it will allow to use gpl or not-gpl libraries ? [00:34] <klaxa> it will allow both afaik [00:35] <klaxa> default build is lgpl only so you can use ffmpeg in a commercial product [00:35] <elkng> so its like --disable-gpl-restrictions ? [00:36] <elkng> because --enable-gpl sounds like: "compile code with gpl license enabled so no non-gpl code will be compilled" [00:36] <elkng> s/will be compilled/will not be compilled [00:36] <relaxed> no, it means enable things that are only gpl [00:37] <elkng> its like: "compile, but with gpl license taking into account" [00:37] <elkng> thats why its frustrating [00:37] <elkng> it would sound more clear if it was --enable-gpl-code [00:37] <relaxed> as long as you distribute the binary it makes no difference [00:37] <relaxed> you don't* [00:38] <relaxed> and this in not #explain_the_difference_between_gpl_and_lgpl [00:38] <relaxed> so go read up until you understand [00:41] <elkng> what was that ? [00:43] <elkng> I've heard some squeaky voice [00:45] <llogan> elkng: you'll want --enable-libass if you're converting to ASS or if you want to use the subtitles or ass filters. [00:46] <elkng> is opencv support currently broken ? [00:46] <llogan> ...or at least if you're planning on using the filters [00:46] <elkng> as it was in 0.11.1 [00:47] <llogan> is there an open bug ticket? [03:47] <sinusss> hi does anyone have a Gingerbread device? Need to test a video :P [03:48] <sinusss> *android [03:55] <bourbon> I've got one in front of me [03:55] <bourbon> that said [03:56] <bourbon> do you have a way for me to ignore the input file's SAR/DAR? [04:05] <relaxed> bourbon: why would you ignore it? [04:05] <relaxed> do you want to change the output's dar/sar? [04:06] <elkng> I have video file with subtitles: http://sprunge.us/BCPU and trying to extract subtitles from it I use command: "ffmpeg -i 1.mkv -map 0:4 -scodec copy out.ass" and that is the output: http://sprunge.us/ELia [04:06] <bourbon> nope, I want to pretend that the input's SAR/DAR is valid [04:06] <bourbon> lol... ass [04:08] <elkng> or what is the right way to extract subtitles using ffmpeg ? should they be in ass format inside mkv or it can extract any subtitles ? [04:08] <relaxed> elkng: "Stream #0:4(eng): Subtitle: dvd_subtitle" is not ass subs [04:08] <elkng> in place of "dvd_subtitle" there should be "ass" word ? [04:08] <relaxed> dvd subs are images, where as ass subs is a text file [04:09] <elkng> other subtitles are incompatible with ffmpeg ? [04:09] <relaxed> I believe you can copy dvd subs but not demux them. [04:09] <bourbon> the issue I'm running into is that same one, man [04:09] <rager> damn SAR/DAR are negative on the file [04:10] <rager> it just seems like with options for everything else [04:10] <rager> there should be options for explicitly declaring the DAR and SAR that ffmpeg will use [04:10] <rager> and dealing with the consequences of such with filters [04:10] <elkng> what if there is string "Stream #0:3(eng): Subtitle: subrip (default)" can they be extracted ? [04:12] <relaxed> elkng: use mkvextract [04:12] <relaxed> it can extract anything from the matroska container. [04:14] <elkng> again that is the file: http://sprunge.us/bLaj I use that command: "ffmpeg -i 2.mkv -map 0:3 -scodec copy out.ass" and that is the output: http://sprunge.us/FaWf [04:14] <relaxed> rager: I'm guessing negative sar/dar vaules indicate a broken, or non-standard, video. [04:15] <rager> oh [04:15] <rager> it's totally non-standard [04:15] <rager> but it manages to play in pretty much every player, as you saw a while back [04:15] <rager> just... [04:15] <rager> quicktime flips it upside down (the desired orientation) [04:15] <relaxed> well, you can't expect ffmpeg to handle broken video correctly. [04:15] <rager> and ffplay/vlc/whatever plays it "correct" [04:15] <elkng> relaxed: is mkvextract goes inside mkvtoolnix ? [04:15] <rager> true [04:15] <rager> bleh... was just hoping there was a way around this [04:15] <relaxed> elkng: yes [04:15] <rager> I guess I could just make the iOS guy's life more fun [04:16] <relaxed> pay him to not be retarded [04:17] <rager> I hope he's not being paid hourly [04:17] <rager> for being a tard [04:17] <rager> the other issue I've got is in building ffmpeg for android :\ [04:18] <rager> the end of the output from "make" [04:18] <rager> http://hastebin.com/haranedagu.rb [04:18] <rager> and my options for building: http://hastebin.com/dusosebita.rb [04:18] <rager> and the options for building x264 [04:18] <rager> http://hastebin.com/nubodedula.rb [04:20] Action: rager shrugs [04:20] <rager> I'll deal with it later [04:20] <rager> it's the end of the day [04:20] <rager> I need to go home [04:22] <elkng> that is info from video file: http://sprunge.us/dXac does anyone know what does those strings: "Stream #0:5: Attachment: unknown_codec" means ? is that video,audio,subtitles or is it some useful data at all or not ? [04:24] <elkng> that is a full output of ffprobe for that file: http://sprunge.us/UPbS what are those attachments ? [04:33] <relaxed> elkng: mkvinfo, included with mkvtoolnix should be able to tell you. [05:21] <schtinky> hey folks, got an intermittent bug/issue I'm dealing with. I'm currently doing "gnutv | ffmpeg " to capture video from a digital tuner card on a regular schedule... [05:22] <schtinky> ... Most of the time, the command works just fine. But occasionally, it just hangs. No output is produced. gnutv and ffmpeg appear to be running according to "ps" but nothing is actually getting recorded [05:23] <schtinky> would anyone have any insight into what might be causing this? (I'd say it "fails" 10% of the time on a set schedule.) [05:25] <schtinky> the only thing /var/log/messages says at the time of the last failure is xc5000 (my tuner card): "waiting for firmware to upload" "firmware uploading" and "firmware upload complete" [05:25] <schtinky> I don't have any reason to suspect that behavior is any different than when the "gnutv | ffmpeg" command succeeds as expected [05:26] <relaxed> I doubt this is an ffmpeg problem. [05:26] <schtinky> yeah I was just thinking that [05:26] <schtinky> probably need to ask #v4l2 [05:26] <schtinky> well #v4l anyway [05:27] <relaxed> ffmpeg can caputure from v4ls devices. why not use it? [05:33] <schtinky> I've tried to figure that out with no success, relaxed [05:34] <schtinky> it seems that "ffmpeg -f v4l2 -i /dev/video0" only works on analog devices [05:34] <schtinky> when I try it on /dev/dvb/adapter0/frontend0 or dvr0, I get operation not allowed or something [05:35] <schtinky> I have tried to get an answer to this both on the wider internet and on this channel, but nothing yet [05:35] <relaxed> schtinky: did you look here: https://ffmpeg.org/trac/ffmpeg/wiki/How%20to%20capture%20a%20webcam%20input [05:36] <schtinky> that's analog [05:36] <relaxed> maybe you need to use `v4l2-ctl` to set it up? [09:06] <Vardan> hi all [09:07] <Vardan> people how can I get ffmpeg version from code? [09:33] <Vardan> people I got this error "framerate not set" when trying to open encoder (CODEC_ID_MJPEG) [10:14] <phuzE> Hey guys, does anyone have experience encoding screencapture on windows? I'm currently stuck on how to properly encode an HBITMAP into my video (my output is just a blank gray/white screen right now) [10:26] <bogdanp> can anyone think of a reason why doing http://pastebin.com/raw.php?i=qzH0JaSu would result in http://i.imgur.com/2EmLbPj.png [10:27] <bogdanp> I'm trying to convert RGB32 frames received from a camera to h264 video [10:27] <bogdanp> but this has me stumped [10:47] <lentferj> I would like to 1:1 map the metadata of all audio streams in a DVB video recording when converting with ffmpeg. But I can't seem to get this work (with current ffmpeg from git). [10:48] <lentferj> http://pastie.org/7407658 [10:49] <lentferj> I tried with -map_metadata:s:a 0:s:a but that seems to copy the metadata from the 1st audio stream onto all audio streams [10:50] <lentferj> when I map all streams one by one, it works (2nd part of the pastie), but this is not "generic" (or easy :) ), since I want to use it in a generic conversion script, where the input can have varying numbers of audio streams [10:50] <lentferj> any idea? [11:11] <lentferj> oh boy... that was too easy.... -map_metadata 0 ... [13:11] <Vardan> people how to get video stream duration in seconds using C++ [13:50] <megaTherion> Vardan: did you check docs on avformat_open_input? [13:50] <megaTherion> Vardan: beside's there is an seperate channel for this: #ffmpeg-devel [14:25] <bogdanp> can anyone think of a reason why doing http://pastebin.com/raw.php?i=qzH0JaSu would result in http://i.imgur.com/2EmLbPj.png ? I'm trying to convert RGB32 frames received from a camera to h264 video but this has me stumped. [14:35] <elkng> since when it became possible to extract subtitles from video ? since 1.2 version ? [14:45] <zap0> bogdanp, thats YUV colours [14:45] <bogdanp> zap0, what is? [14:45] <zap0> the picture [14:46] <bogdanp> zap0, yes but shouldn't sws_scale convert the rgb pixel buffer to yuv? making the two frames look the same? [14:48] <zap0> if you input r8g8b8 = 24bit, and a convertion happened, the output would be y8v2u2=12 bits. so did the output bytes be 1/2 the input bytes? [14:48] <zap0> you might just be getting your U and V channels backwards [14:49] <zap0> a YUV signal in yv12 format is sometimes stored as Y,V,U ordering. instead of Y,U,V ordering. [14:49] <zap0> how did you generate that frame? [14:49] <zap0> can you switch U V ? [14:57] <bogdanp> zap0, http://pastebin.com/raw.php?i=UMvFQPiW that's basically what I'm doing. I don't know how to switch U and V, is there some builtin function that can do it for me or do I have to do it manually? (if it's not already obvious, I'm a complete noob to video& stuf) [14:57] <bogdanp> *stuff [14:59] <zap0> have you tried doing basically the same with the command line app? [15:06] <bogdanp> zap0, I haven't but I am going to do it now [15:10] <bogdanp> zap0, it seems that capturing camera input on OS X isn't really possible from the command line [15:10] <bogdanp> zap0, that or my google-fu is failing me hard [15:16] <Snaggle> weee. ffmpeg-next will have the ability to view/edit/transcode/* youtube videos just by passing it the url [15:16] <retard> quvi? [15:16] <Snaggle> yup [15:17] <retard> that is super rad [15:17] <Snaggle> pretty neat. I just tried it, and even funner was passing it through libcaca [21:58] <viric> Can ffmpeg use the vdpau deinterlacers? [23:02] <tgreenwood> I was looking at https://ffmpeg.org/trac/ffmpeg/ticket/568 but checking the latest git it seems MAX_SLICES is still 16. [23:08] <durandal_1707> tgreenwood: what would it need to change? [23:09] <tgreenwood> It's just a 1 liner in h264.h [23:09] <durandal_1707> you still get that bug or? [23:09] <tgreenwood> I get visual artefacts if it's less than 64 [23:09] <durandal_1707> with what version? [23:09] <tgreenwood> I am running 2 RTP endpoints [23:09] <tgreenwood> I've been using gstreamer 1.0 I can get the git version if you like [23:10] <tgreenwood> However I did check latest master for the value. [23:10] <tgreenwood> I also work with some people who have increased the value in other RTP use cases. [23:11] <durandal_1707> its very simple: if latest ffmpeg still have issue, reopen bug [23:11] <tgreenwood> OK - will do thank you! [23:12] <tgreenwood> I can post a patch as well if it would be helpful. [23:12] <durandal_1707> tgreenwood: then send it to mailing list, but make sure it actually helps [23:12] <durandal_1707> changing random line and not testing is ugly and bad [23:13] <tgreenwood> Totally appreciate that I am testing with 2 H.264 RTP endpoints. [23:14] <durandal_1707> if its so relevant than put it into log of patch you sent (eg. with git format-patch......) [23:14] <durandal_1707> tgreenwood: if you want to discuss problem and solution for it you can ask on #ffmpeg-devel i guess [23:17] <tgreenwood> Thanks - I may do that [23:25] <shambala> Hi, I am creating a video from stills.. stills have transparent canvas, but when the video gets made, it has black fill, don't know why.. here is the command I am using [23:25] <shambala> ffmpeg -y -f image2 -r 7 -i /Users/shambala/tabrez/rgplyrx1/tmp/video_creator/zi_1365622727_699172_572887_%04d.png -vf scale=520:600 -crf 0 -preset ultrafast -b:v 25000k -minrate 25000k -maxrate 25000k -bufsize 5000k -tune stillimage -pix_fmt yuv420p -r 30000/1001 /Users/shambala/tabrez/rgplyrx1/tmp/video_creator/manoj_test.mpg [23:26] <shambala> any issue with the options? [23:29] <shambala> durandl_1707, here you go: http://pastebin.com/UxtpK9Pe [23:37] <durandal_1707> shambala: it looks liky you encode this to mpeg1video but use options for libx264 encoder [23:37] <durandal_1707> exactly what is mentioned 3 times in output you pasted [23:40] <shambala> I removed those options, so errors are gone, but the fill issue remains [23:43] <shambala> durandal_1701 I am loading pic and video in google drive, so you can see [23:43] <shambala> sorry, durandal_1707 [23:48] <shambala> https://docs.google.com/file/d/0B6Pi2zJeMhv7UTQza1EyVzlrMjg/edit?usp=sharing [23:48] <shambala> that's the pic [23:49] <shambala> https://docs.google.com/file/d/0B6Pi2zJeMhv7RXY1N2dHR2xFMXM/edit?usp=sharing [23:49] <shambala> that's the video& ffmpeg is adding a black fill [23:49] <shambala> anyway to make it transparent [23:49] <shambala> ? [23:57] <llogan> shambala: you can't. mpeg1video doesn't support an alpha channel [23:58] <llogan> what are you tryng to do? why do you need the transparency preserved? [23:59] <saste> shambala, are you implementing a virtual cemetery? [23:59] <llogan> i'm guessing it's a family history project [00:00] --- Thu Apr 11 2013

1 0