Ffmpeg-devel-irc

Download

ffmpeg-devel-irc@ffmpeg.org

May 2017

1 participants
62 discussions

ffmpeg-devel.log.20170525
by burek 26 May '17

26 May '17

[01:06:44 CEST] <thardin> annnd a minimal configuration that runs twice as fast. going to leave this running overnight [01:28:18 CEST] <thebombzen> is there a reason the trac server is significantly slower than the rest of the website? seems like an odd speed differential [01:29:24 CEST] <kierank> hosted in a different place iirc [01:31:46 CEST] <cone-223> ffmpeg 03James Almer 07master:f8c73e87532a: avformat/movenc: always check for new extradata on a packet [01:31:46 CEST] <cone-223> ffmpeg 03James Almer 07master:7631f14bb35e: avformat/matroskaenc: check packet side data for AAC extradata updates [01:31:46 CEST] <cone-223> ffmpeg 03James Almer 07master:8b3ec51de8a0: avformat/latmenc: check packet side data for AAC extradata updates [01:31:46 CEST] <cone-223> ffmpeg 03James Almer 07master:210388a1979d: avcodec/adtsenc: check packet side data for AAC extradata updates [01:31:46 CEST] <cone-223> ffmpeg 03James Almer 07master:f63c3516577d: avcodec/aac_adtstoasc: propagate new extradata using packet side data [01:31:46 CEST] <cone-223> ffmpeg 03James Almer 07master:437ad467c250: avformat/mux: remove autobsf extradata propagation hack [01:31:47 CEST] <cone-223> ffmpeg 03James Almer 07master:f1cdc01e7208: ffmpeg: remove bsf extradata propagation hack [01:31:48 CEST] <cone-223> ffmpeg 03James Almer 07master:94ec89eb6732: doc/libav-merge: remove line about aac_adtstoasc [01:31:51 CEST] <jamrial> ubitux: ^ another one down [01:36:30 CEST] <thebombzen> another question about the trac: is there any sort of guidelines on what bugs are "important" or "normal" or do you just play it by ear [01:39:51 CEST] <jamrial> critical = data loss (like input file being overwritten and such). important = regressions. normal = not regressions. minor = doc mistakes or stuff like that i guess [01:40:04 CEST] <jamrial> just choose normal and let whoever handles the ticket choose that for you [01:41:20 CEST] <nevcairiel> feature requests = always wish [01:47:50 CEST] <thebombzen> what exactly is a "regression"? is that "used to work and now doesn't" or something? [01:48:19 CEST] <nevcairiel> basically [03:11:06 CEST] <llogan> thebombzen: some info here http://git.videolan.org/?p=ffmpeg.git;a=blob_plain;f=doc/issue_tracker.txt;… [03:11:22 CEST] <thebombzen> thx [04:26:04 CEST] <jamrial> rcombs: ping [04:33:27 CEST] <cone-223> ffmpeg 03James Almer 07master:8ea5ee10a2d6: avcodec/libopenh264dec: fix return error value when h264_mp4toannexb_bsf is not found [04:44:19 CEST] <stevenliu> Hi guys [04:44:27 CEST] <stevenliu> I have a question [04:44:54 CEST] <stevenliu> It's Summer here, can i print ffmpeg picture on T-shirt? [04:47:03 CEST] <stevenliu> Hi guys, I have a question, It's Summer here, can i print ffmpeg picture on T-shirt? [04:47:24 CEST] <stevenliu> for example, the logo of the ffmpeg.org :D [05:38:58 CEST] <rcombs> jamrial: pong [05:40:09 CEST] <jamrial> rcombs: could you test https://pastebin.com/raw/4S4zuTRL? [05:41:04 CEST] <rcombs> hmm, what demuxers AV_PKT_DATA_NEW_EXTRADATA for AAC? [05:41:20 CEST] <rcombs> or would that be on the output of the BSF [05:41:32 CEST] <jamrial> the bsf [05:41:38 CEST] <rcombs> ah, makes sense [05:41:42 CEST] <rcombs> sure, I'll give it a go [05:52:25 CEST] <jamrial> michaelni: "src/libavcodec/jpeg2000dec.c:304:40: warning: too many arguments for format [-Wformat-extra-args]" since 89325417e7 [06:14:34 CEST] <rcombs> jamrial: seems to work as expected [06:14:59 CEST] <jamrial> rcombs: ok, thanks [06:15:25 CEST] <rcombs> (ffplay'd an mpegts file with -acodec aac_at) [06:21:16 CEST] <jamrial> as long as it was adts, that should do it [06:27:29 CEST] <cone-223> ffmpeg 03James Almer 07master:954e2b3d34b7: avcodec/audiotoolboxdec: check packet side data for AAC extradata updates [10:42:13 CEST] <ubitux> how much can i overread in aac ps stereo_interpolate? [10:43:15 CEST] <ubitux> and actually overwrite (the read should be "safe") [10:44:54 CEST] <atomnuker> you should be able to read as much as you need, since this appears to be done on coefficients which are padded and aligned [10:45:22 CEST] <ubitux> it doesn't seem that i can overwrite [10:45:29 CEST] <atomnuker> as for overwriting, none since you'll mess up coefficients from other bands [10:45:38 CEST] <ubitux> that's too bad :( [10:45:48 CEST] <atomnuker> vmaskmov? [10:46:08 CEST] <atomnuker> (I think aarch64 had an alternative to that) [10:46:34 CEST] <ubitux> yeah well, it would have been much better if i could just write 1 more float im/re pair [10:47:23 CEST] <ubitux> i have a working function, but it's just reading 1 pairs of float per channel [10:47:30 CEST] <ubitux> (and writing fwiw) [10:48:00 CEST] <ubitux> if i could read 2 pairs of float per channel instead... [10:48:33 CEST] <ubitux> anyway, i'll keep it simple for now, but it's doing the same as the c basically [10:49:01 CEST] <atomnuker> there is a way but the gains would be less than 2x since the extra instructions needed to handle overwriting [10:49:41 CEST] <ubitux> couldn't we just +1 on the buffer declaration? [10:52:35 CEST] <atomnuker> why do you want to overwrite, isn't len always a multiple of 4? [10:54:04 CEST] <ubitux> ah? is it? didn't actually check [10:54:28 CEST] <ubitux> if it's a multiple of 2 that's good enough for me [10:54:49 CEST] <atomnuker> do check, I really doubt it isn't (since there's x86 asm for that function) [10:55:58 CEST] <ubitux> Assertion !((stop - start) & 1) failed at src/libavcodec/aacps.c:977 [10:56:06 CEST] <ubitux> doesn't seem to be a multiple of 2 [10:57:37 CEST] <ubitux> the x86 asm does only 1 float pair writing per channel AFAICT [11:22:39 CEST] <atomnuker> ubitux: you could still overwrite though [11:22:55 CEST] <atomnuker> check out stereo_processing() in aacps.c [11:23:06 CEST] <atomnuker> INTFLOAT (*l)[32][2], INTFLOAT (*r)[32][2] [11:23:40 CEST] <atomnuker> just increase the second dimension [11:23:56 CEST] <ubitux> yes that's what i was suggesting by increasing the buffer [11:24:23 CEST] <ubitux> but i'll go for a simple version of the asm for now anyway [11:29:40 CEST] <cone-722> ffmpeg 03Michael Niedermayer 07master:5782e0ba8cc3: avcodec/jpeg2000dec: Fix copy and paste error [14:46:05 CEST] <Tehme> Hello. I have a question about avcodec and CUVID. When I do this: https://pastebin.com/iVv1Yc2w I have the following error: https://pastebin.com/dHyquQz2 . Looks like my libnvcuvid.so does not have cuvidGetDecoderCaps() function. Why does avcodec try to load it then? I have a proprietary NVIDIA driver 375.39 and CUDA SDK 8, ffmpeg version is 3.3. [14:47:06 CEST] <nevcairiel> you need a newer driver [14:47:54 CEST] <Tehme> I saw somewhere that 340+ is ok, let me check it. [14:49:56 CEST] <nevcairiel> the video sdk 8.0 has only been supported in drivers 378 [14:50:52 CEST] <Tehme> thank you nevcairiel [15:10:50 CEST] <cone-006> ffmpeg 03Michael Niedermayer 07master:5f91786fc8ad: avcodec/wavpack: Fix: runtime error: signed integer overflow: 3 * -2147483648 cannot be represented in type 'int' [15:10:50 CEST] <cone-006> ffmpeg 03Michael Niedermayer 07master:ea71a48c7e8a: avcodec/wavpack: Fix runtime error: left shift of negative value -14778 [15:10:50 CEST] <cone-006> ffmpeg 03Michael Niedermayer 07master:4dc3714c48e7: avcodec/tscc2: Skip duplicate frames [15:13:25 CEST] <Tehme> So, if the latest driver for linux is 375.66, does it mean I can't use ffmpeg's h264_cuvid decoder? [15:14:28 CEST] <RiCON> it's not the latest for linux though [15:14:34 CEST] <RiCON> maybe it's the latest in your distro [15:14:49 CEST] <Tehme> what is the latest then? [15:29:24 CEST] <RiCON> 381.22 apparently [15:34:21 CEST] <DHE> no nvenc? [15:35:32 CEST] <DHE> oh, decoder, n/m [15:43:48 CEST] <RiCON> Tehme: don't send unrequested private messages [15:43:49 CEST] <RiCON> https://git.archlinux.org/svntogit/packages.git/tree/trunk/PKGBUILD?h=packa… [15:50:56 CEST] <Tehme> RiCON: thank you [16:17:12 CEST] <BBB> does anyone want to review the hevc patch? [16:18:37 CEST] <durandal_170> no [16:22:20 CEST] <jamrial> BBB: there's no maintainer for hevc, so if you know it works and ubitux confirmed it, then just push it [16:23:15 CEST] <BBB> I didnt say I knew hevc [16:23:23 CEST] <BBB> but I feel bad about keeping it broken :-p [16:25:18 CEST] <jamrial> nobody really knows it well i guess. the core devs for it have been kinda awol for a while :p [16:25:57 CEST] <jamrial> maybe jkqxz does, seeing he's writing that neat raw bitstream parsing feature [16:29:42 CEST] <BBB> uhm [16:29:43 CEST] <BBB> @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ [16:29:44 CEST] <BBB> @ WARNING: POSSIBLE DNS SPOOFING DETECTED! @ [16:29:45 CEST] <BBB> @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ [16:29:48 CEST] <BBB> what did you guys do [16:30:23 CEST] <jamrial> got that while trying to push? [16:30:33 CEST] <BBB> yes [16:30:39 CEST] <atomnuker> nope, dry-run works fine for me [16:31:11 CEST] <atomnuker> both ipv4 and ipv6 [16:31:21 CEST] <BBB> my push and fetch url are different, maybe thats why? [16:31:33 CEST] <BBB> one is https, one is ssh [16:31:42 CEST] <atomnuker> that's okay here [16:31:46 CEST] <nevcairiel> when did you push last [16:31:55 CEST] <nevcairiel> the server changed a few weeks back [16:32:02 CEST] <nevcairiel> which would trigger this [16:32:06 CEST] <atomnuker> that was a month ago [16:32:16 CEST] <BBB> Thu Apr 6 13:58:59 2017 -0400 [16:32:23 CEST] <nevcairiel> might be then [16:32:54 CEST] <jamrial> BBB: https://mailman.videolan.org/pipermail/vlc-devel/2017-April/112787.html [16:35:02 CEST] <BBB> ty [16:35:09 CEST] <cone-006> ffmpeg 03Ronald S. Bultje 07master:ca2209d67af0: hevc: fix race condition in max_ra/seq_decode. [16:35:10 CEST] <cone-006> ffmpeg 03Ronald S. Bultje 07master:d98f34d7d440: frame_thread_encoder: extend critical code covered by finished_task_mutex. [16:35:29 CEST] <BBB> ok, so 2 bugs remaining: one for field decoding where owner doesnt match the processing AVCodecContext, and mpegvideo* [16:35:38 CEST] <BBB> whoooooo wants to fix mpegvideo? [16:35:56 CEST] <BBB> I promise free ice cream at the next vdd to whoever fixes mpegvideo [16:36:19 CEST] <BBB> <and there was mere silence :(> [16:37:45 CEST] <jamrial> isn't mpegvideo the one with the nightmarish all-purpose context? [16:38:11 CEST] <BBB> yes [16:38:18 CEST] <BBB> long live mpegvideo and dsputil [16:41:55 CEST] <j-b> 'morning [16:45:42 CEST] <BBB> aha! j-b will fix mpegvideo [16:52:09 CEST] <j-b> haha [16:52:17 CEST] <j-b> mpegvideo? is that still a thing? [16:52:29 CEST] <nevcairiel> of course, for mpeg124 [16:53:23 CEST] <ubitux> so if anyone is ever confused at wtf these freaking aarch64 do based on the shitty doc, this may help: http://sprunge.us/TCdM [16:53:58 CEST] <ubitux> (only the last lines matter) [16:55:56 CEST] <iive> ubitux: NICE :D [17:03:23 CEST] <ubitux> btw, i'm running a fate valgrind instance on a BBB since about 24h [17:03:27 CEST] <ubitux> it's ofc not done yet [17:03:36 CEST] <BBB> lol [17:03:37 CEST] <ubitux> one of my many brillant idea [17:03:53 CEST] <ubitux> heh sorry i meant beaglebone black by "BBB" [17:03:54 CEST] <BBB> and Im trying to encode a 4K video on my android phone [17:04:04 CEST] <BBB> also a brilliant idea [17:04:18 CEST] <ubitux> in software? [17:04:23 CEST] <BBB> its a joke :D [17:04:34 CEST] <ubitux> it's not a joke on my side :( [17:04:45 CEST] <BBB> is there any reason you did that? do you expect that itll find issues that the x86 instance wont find? [17:04:55 CEST] <BBB> or did you encounter bugs that warranted this? [17:04:59 CEST] <BBB> or is this just exploratory? [17:05:06 CEST] <ubitux> i like to do what people never did, especially if it's stupid [17:05:11 CEST] <ubitux> you can call me an artist [17:05:16 CEST] <BBB> uhm... [17:05:24 CEST] <BBB> okaaaaaay... [17:05:25 CEST] <ubitux> anyway, yeah i just wanted to do run valgrind on an arm board [17:05:31 CEST] <BBB> :D [17:05:36 CEST] <ubitux> and that was the only one i had left [17:05:46 CEST] <ubitux> i mean, i also have a rpi1 [17:05:48 CEST] <BBB> so uhm [17:05:51 CEST] <J_Darnley> "Unexpected." "Never seen before." "Unique." "5/7" [17:05:52 CEST] <BBB> while its running [17:05:53 CEST] <ubitux> but i couldn't get fame until next year [17:06:09 CEST] <BBB> does it fry chips and fish? [17:06:20 CEST] <ubitux> while with the BBB i may have hope for next week [17:06:21 CEST] <jamrial> BBB: i mean, x264 has arm optimizations, so maybe some people do encode videos on their phones :p [17:06:24 CEST] <BBB> I mean, if youre an artist, you could just as well be pragmatic about it and use it to prep dinner [17:06:33 CEST] <atomnuker> ubitux: speaking of aarch64, I compiled a kernel for my odroid c2 last night and had to have a water bottle filled with ice above to cool it [17:06:41 CEST] <BBB> anyway [17:06:45 CEST] <BBB> sorry Im teasing too much [17:07:03 CEST] <ubitux> BBB: http://b.pkh.me/homelab.jpg see the hanging one with the blue eth cable? that's the one [17:07:06 CEST] <atomnuker> then it worked fine this morning with the mainline kernel and it doesn't boot now :/ [17:07:09 CEST] <BBB> I hope it finds issues, not that I like issues, but it would basically mean it was worth it [17:07:30 CEST] <BBB> ubitux: you need to clean that up :D [17:07:35 CEST] <ubitux> hehe, yeah [17:08:11 CEST] <ubitux> BBB: i ran this with valgrind stable, so it will likely raise the same issue we hit from last time [17:08:23 CEST] <ubitux> especially since i couldn't build ffmpeg with --disable-optimization, it will raise more issue [17:08:42 CEST] <BBB> I thought that issue was avx related? [17:08:46 CEST] <BBB> so that shouldnt occur on arm [17:09:06 CEST] <ubitux> we have some uninitialized mem issue unrelated to avx [17:09:16 CEST] <ubitux> https://bugs.kde.org/show_bug.cgi?id=378622 [17:09:18 CEST] <ubitux> https://bugs.kde.org/show_bug.cgi?id=378627 [17:09:23 CEST] <ubitux> these two will likely show up [17:09:29 CEST] <ubitux> svn version wouldn't help though indeed [17:10:23 CEST] <ubitux> maybe i should also run a tsan instance on that board& [17:11:09 CEST] <ubitux> atomnuker: due to overheating or boot/kernel bug? [17:11:25 CEST] <atomnuker> no, I think its the image [17:11:41 CEST] <atomnuker> I used a nightly image and I guess it only *occasionally* works [17:12:45 CEST] <ubitux> still 3.14.79 here [17:13:08 CEST] <ubitux> it has "sensitive" stuff running, so i'm not experimenting on it [17:13:21 CEST] <atomnuker> it worked great on mainline until it didn't though [17:14:13 CEST] <atomnuker> personally I would never trust anything without a bios and x86 cpu [17:14:51 CEST] <atomnuker> I bought a rpi v3 in october and guess what? its memory is crap [17:15:38 CEST] <ubitux> yeah, these boards really suck [17:15:39 CEST] <atomnuker> it kept rebooting when I tried to compile ffmpeg and if it didn't the compiler itself printed out some machine bug [17:16:01 CEST] <ubitux> the only thing they don't suck is power, so i have a bunch of them to run all kind of different services instead of a single x86 machine [17:16:17 CEST] <atomnuker> yeah. that's true, they're cheap to run [17:16:54 CEST] <atomnuker> NUCs are pretty good too though, I hear atoms are not power hungry [17:17:06 CEST] <atomnuker> though they're expensive, min 100 euros or so [17:17:36 CEST] <ubitux> yeah i have two of them, but i use them as desktop and mediacenter [17:17:37 CEST] <atomnuker> (about the same as an odroid c2 because of the huge tax bill you get if you order one) [17:17:49 CEST] <ubitux> more appropriate usage than running single services [17:18:23 CEST] <atomnuker> do you have any mips based devices? [17:18:55 CEST] <ubitux> nope [17:19:24 CEST] <ubitux> mmh maybe the edgerouter is running on mips, i should check [17:20:02 CEST] <ubitux> Linux ubnt 3.10.14-UBNT #1 SMP Mon Nov 14 03:56:39 PST 2016 mips GNU/Linux [17:20:06 CEST] <ubitux> yeah that's actually a mips [17:20:16 CEST] <ubitux> but don't expect me to compile and run ffmpeg on my router [17:23:19 CEST] <jamrial> your art has limits then :p [17:23:35 CEST] <kierank> atomnuker: the point is to cross-compile ffmpeg [17:23:37 CEST] <kierank> then run it on rpi [17:23:47 CEST] <kierank> not that i am capable of that [17:25:11 CEST] <atomnuker> kierank: even if it did run it would still cause the kernel to oops because the memory is faulty [17:25:19 CEST] <kierank> that i have never seen [17:25:44 CEST] <kierank> 15:25:46 up 56 days, 19:33, 1 user, load average: 0.00, 0.01, 0.05 [17:25:47 CEST] <kierank> not bad for a rpi2 [17:26:00 CEST] <atomnuker> mine is okay if you don't load it as well [17:26:37 CEST] <atomnuker> the 4 year old rpi I have is still running without issues, but those new models are shit [17:27:03 CEST] <atomnuker> those wifi and bluetooth interface drivers on old stock kernels are horrible and I had to turn them off [17:29:56 CEST] <ubitux> atomnuker: 4yr old rpi? rpi1 ? [17:30:04 CEST] <atomnuker> yep, the first one [17:30:10 CEST] <ubitux> so without any neon [17:30:40 CEST] <ubitux> don't wanna try to add a fate tsan instance on it? :) [17:30:43 CEST] <atomnuker> Features : half thumb fastmult vfp edsp java tls [17:30:54 CEST] <atomnuker> sure, I have trust in it [17:32:33 CEST] <atomnuker> I'll start compiling ffmpeg, hopefully won't take longer than a day [17:32:40 CEST] <jamrial> isn't the first rpi like a pentium 2 performance wise? tsan could take days there... [17:33:00 CEST] <kierank> get atomnuker to buy an rpi3 on prime now [17:33:40 CEST] <atomnuker> I had an old pentium 2 NEC for 10 years but threw it away 4 or so years ago, wish I kept it [18:08:16 CEST] <atomnuker> ubitux: tsan? what would be the point, its only a single cpu? [18:08:55 CEST] <ubitux> i suppose you still have context switches :p [18:09:22 CEST] <ubitux> the point is to test that are atomics in x86 also actually are on arm [18:09:25 CEST] <ubitux> that kind of stuff [18:09:38 CEST] <ubitux> test [instr] that are atomics [18:23:36 CEST] <atomnuker> ubitux: how do I run fate tsan? just compile with --cc="clang -fsanitize=thread"? [18:24:04 CEST] <jamrial> use --toolchain [18:24:29 CEST] <ubitux> --toolchain=clang-tsan or gcc-tsan [18:24:57 CEST] <atomnuker> and after it gets compiled? [18:25:07 CEST] <ubitux> just run fate normalliz [18:25:09 CEST] <ubitux> normally [18:25:29 CEST] <ubitux> i use --extra-cflags=-fsanitize=thread --extra-ldflags=-fsanitize=thread though personally [18:25:34 CEST] <ubitux> (instead of the toolchain) [18:25:42 CEST] <ubitux> because of https://gcc.gnu.org/bugzilla/show_bug.cgi?id=67308 [18:29:59 CEST] <ubitux> atomnuker: oh i forgot, make fate with THREADS [18:30:21 CEST] <ubitux> i currently use THREADS=2 to limit reports [18:31:41 CEST] <ubitux> atomnuker: this is my config: http://b.pkh.me/gcc-tsan.cfg [18:37:33 CEST] <atomnuker> error: unsupported option '-fsanitize=thread' for target 'armv7l-unknown-linux-gnueabihf' [18:37:46 CEST] <ubitux> ah, too bad :) [18:37:51 CEST] <atomnuker> that's clang [18:37:58 CEST] <atomnuker> I think the gcc -fsanitize one works [18:38:05 CEST] <ubitux> ah? cool [18:45:30 CEST] <atomnuker> libtsan0 is missing from raspbian's repo [18:45:52 CEST] <atomnuker> based on debian 8.0 yet still hasn't seen packages newer than 2012 [18:47:11 CEST] <atomnuker> this is another reason why these arm boards suck, your choices are either outdated ubuntu with proprietary crap, outdated debian with some proprietary crap or outdated diy arch images with some proprietary crap [18:47:19 CEST] <ubitux> https://archlinuxarm.org/platforms/armv6/raspberry-pi [18:47:21 CEST] <ubitux> :° [18:48:11 CEST] <atomnuker> 2500 km away from the sd card and the machine to do anything about it :/ [18:49:16 CEST] <ubitux> did you send into space? [18:49:30 CEST] <ubitux> +it [18:50:35 CEST] <ubitux> anyway, i can do it myself, i also have that board :p [19:32:14 CEST] <jamrial> rcombs: i sent a couple more patches for audiotoolbox to the ml [19:34:52 CEST] <cone-006> ffmpeg 03Michael Niedermayer 07master:6d6fc4105b87: avcodec/diracdec: Factor quant matrix reads [19:34:53 CEST] <cone-006> ffmpeg 03Michael Niedermayer 07master:b946bd8ef2c7: avcodec/diracdec: Fix off by 1 error in quant check [21:20:10 CEST] <cbsrobot> durandal_1707: in the vf_upmixer sholdn't there be a delay on the ls and rs (or lb and rb) channels for the 5.1 upmix? [21:20:43 CEST] <durandal_1707> cbsrobot: no [21:21:45 CEST] <durandal_1707> delay is when it comes to your ears [21:23:53 CEST] <cbsrobot> I tested a few upmix examples on http://forum.doom9.org/archive/index.php/t-152034.html [21:24:24 CEST] <cbsrobot> and I find grezen sounds ok most of the times [21:24:35 CEST] <cbsrobot> search fro grezen on that page [21:25:00 CEST] <cbsrobot> in these upmixes the rear channels always get 0.02 seconds delay [21:25:10 CEST] <durandal_1707> that commands are flawed [21:25:17 CEST] <cbsrobot> and also a frequency cut for the rear channels [21:25:19 CEST] <cbsrobot> ok [21:25:31 CEST] <cbsrobot> I'll try your filter when I have time [21:25:36 CEST] <durandal_1707> they do not phase shift rear channels with hilbert [21:28:11 CEST] <durandal_1707> the filter works with single sample i could find [21:28:46 CEST] <durandal_1707> it clearly separates sounds that is originating from rear channels [21:29:20 CEST] <durandal_1707> using it on ordinary stereo files you get mediocre results [21:30:17 CEST] <durandal_1707> if you can find more pro logic enxoded files let me know [22:07:49 CEST] <rcombs> jamrial: oh sweet, I hadn't heard about that new filtering mechanism; looks super nice [22:08:46 CEST] <jamrial> yeah, simplifies decoders like this a lot [22:20:29 CEST] <atomnuker> ubitux: man what the hell, I tried a ton of different images and yet this damned blue light refuses to do its job [22:20:35 CEST] <atomnuker> this thing seems utterly broken [22:21:25 CEST] <atomnuker> it was behind a door, still in its cover and I was ssh'd typing when it just disconnected [22:22:39 CEST] <atomnuker> well, so much for "made in korea", this thing seems straight out of their southern neighbours [22:27:58 CEST] <atomnuker> piece of crap [22:29:40 CEST] <jamrial> the odriod? [22:29:48 CEST] <atomnuker> yeah [22:33:06 CEST] <iive> atomnuker: btw, are you using the original power supply for the rpi3? i've heard that they are quite sensitive about their voltage levels. [22:33:40 CEST] <atomnuker> yep, got a starter pack which had a 5v 2a supply [22:35:01 CEST] <iive> :| [22:36:38 CEST] <BBB> crap, koda is going zscale :( [22:55:51 CEST] <ubitux> atomnuker :( [00:00:00 CEST] --- Fri May 26 2017

1 0

ffmpeg.log.20170525
by burek 26 May '17

26 May '17

[00:17:30 CEST] <Ozgr> BtbN, one way to limit bitrate with x264 is to specify a level. the options are limited though with the max bitrates defined by avc/h264 for each level [00:18:36 CEST] <BtbN> you can just give it a maxrate, that's not a problem at all [00:18:43 CEST] <BtbN> just not a minrate [00:18:54 CEST] <BtbN> if it's just showing a still image, it won't use much/any [00:53:59 CEST] <Gazoo> I'm trying to convert an FFV1 encoded file to x264 (or any other lossy modern format), but the blacks in my original video become this lighter saturated gray :( The other colors seem less affected. I'm wondering if this is something I can work to avoid...? [01:08:47 CEST] <Threads> stop paying the blacks [01:14:45 CEST] <Gazoo> ... O_o [01:16:33 CEST] <c_14> Gazoo: is your ffv1 source rgb? [01:17:40 CEST] <Gazoo> According to VLC, yes. It says Decoded format: 32 bits RGB under codec info [01:17:51 CEST] <Gazoo> Is there another way to check that perhaps? [01:17:56 CEST] <c_14> ffprobe filename [01:24:39 CEST] <Gazoo> will do - just a sec [01:25:17 CEST] <Gazoo> Stream #0:0: Video: ffv1 (FFV1 / 0x31564646), bgr0, 384x272, 4255 kb/s, 50.12 fps, 50.12 tbr, 50.12 tbn, 50.12 tbc [01:25:31 CEST] <Gazoo> So... I suppose 'bgr0' is the color codec? [01:25:38 CEST] <Gazoo> wait - no - color space? [01:26:11 CEST] <c_14> pixel format/colorspace yeah [01:27:06 CEST] <c_14> The rgb -> yuv conversion is probably what's messing with your blacks, I'd recommend trying to use the colorspace filter or the zscale filter to do the conversion instead and see if that's better [01:27:22 CEST] <Gazoo> So when I ffprobe the output it indeed says yuv444p [01:27:51 CEST] <Gazoo> I'll google colorspace filter / zscale filter re. ffmpeg. You don't happen to have a flag you'd recommend of the top of your head? [01:28:14 CEST] <c_14> with zscale it'd be -vf zscale,format=pix_fmt=yuv444p [01:32:04 CEST] <Gazoo> I'll give that a shot [01:32:22 CEST] <Gazoo> Regardless of whether or not it works, I very, very much appreciate you engaging me and helping me with some advice c_14 [01:32:30 CEST] <Gazoo> Thank you - a lot! [01:33:02 CEST] <brian___> hi, i'm calling libavcodec/libavformat from some code i'm writing, and i've noticed that when i call avformat_find_stream_info(container, NULL);, it allocates a 688KiB buffer that isn't cleaned up after calling avformat_close_input. is there possibly another free/close call i'm missing? and more broadly, is there a guide on which free/close to use for any given alloc? [01:36:49 CEST] <brian___> the process of which thing is supposed to own which pointer is hard to follow :( [01:40:37 CEST] <brian___> ah nevermind, seems i just need to do av_free on my aviocontext->buffer [01:40:47 CEST] <Gazoo> c_14 - It seems the pix_fmt option is not enjoyed by the zscale filter [01:41:21 CEST] <c_14> the zscale filter doesn't have a pix_fmt option, the format filter should (though it might be called pix_fmts) [01:42:00 CEST] <Gazoo> ah. Let me give that a try [01:42:31 CEST] <brian___> im somewhat confused why avio->buffer = buf; av_free(avio->buffer); av_free(avio); succeeds but avio->buffer = buf; av_free(avio); av_free(buf); fails [01:42:47 CEST] <brian___> does av_free walk the thing it's freeing, looking for pointers? [01:43:37 CEST] <c_14> brian___: no, av_free is just a wrapper around free() or _aligned_free() [01:44:10 CEST] <brian___> ohhh i bet i know what's going on [01:44:15 CEST] <brian___> i bet avio reallocs its buffer [01:44:20 CEST] <brian___> meaning the pointer im holding is now invalid [01:44:26 CEST] <c_14> possible [01:45:07 CEST] <c_14> Also, I don't think you need anything besides avformat_close_input after avformat_find_stream_info [01:45:08 CEST] <Gazoo> c_14 - It executes, but with quite a few 'code 3074: no path between colorspaces' messages and a 1kb output file. I tried just enforcing the bgr0 colorspace for h264, but that's apparently incompatible with h264 [01:45:10 CEST] <brian___> that's rather strange, i wonder why it accepts a buffer pointer at all [01:45:43 CEST] <brian___> c_14: good to know. thanks! i am seeing now that freeing the avio->buffer takes care of the leak, seems calling avformat_find_stream_info caused the avio->buffer to grow [01:45:56 CEST] <c_14> Gazoo: x264 doesn't support bgr0. If you don't care much about support you can use libx264rgb (as the encoder) [01:46:08 CEST] <c_14> (Support as in decoding support in various players) [01:46:28 CEST] <iive> what is bgr0 ? [01:46:41 CEST] <c_14> pixel format [01:46:54 CEST] <brian___> actually it seems i can just safely give it a NULL ptr and 0 size and it just goes [01:47:05 CEST] <iive> with 0 bits? [01:47:34 CEST] <c_14> bgr0 vs bgra [01:47:38 CEST] <c_14> there's no alpha channel in bgr0 [01:47:43 CEST] <iive> aha [01:48:09 CEST] <iive> bgrx [01:48:46 CEST] <Gazoo> c_14 - It does indeed produce a file VLC can handle and the blacks are just as black as they're supposed to be [01:49:34 CEST] <Gazoo> The end goal is decoding via an application in Windows Media Foundation, so let me see how it likes it :o [01:49:45 CEST] <Gazoo> sorry - using WMF, not in WMF [01:54:36 CEST] <Gazoo> Ok... It likes neither the x264 encoded one, nor the x264rgb encoded one. I wonder if it's because I removed the audiotrack [01:54:45 CEST] <Gazoo> Perhaps WMF dislikes 0 audio [01:56:40 CEST] <c_14> or it just can't handle bgr/yuv444 [01:57:39 CEST] <Gazoo> Or maybe it can't handle x264, but can handle h264 [01:57:45 CEST] <c_14> https://msdn.microsoft.com/en-us/library/windows/desktop/dd797815(v=vs.85).… [01:57:45 CEST] <furq> they're the same thing [01:57:46 CEST] <Gazoo> Let me start by verifying that... [01:57:52 CEST] <c_14> 4:2:0 chroma or monochrome [01:57:57 CEST] <Gazoo> x264 is the same as h264? [01:58:01 CEST] <furq> x264 is an h264 encoder [01:58:19 CEST] <Gazoo> ah [01:58:29 CEST] <furq> and yeah a lot of bad decoders can only handle yuv420p [01:58:43 CEST] <Gazoo> Well it seems to me that the easiest thing here is just to narrow it down for me. I'll take it one step at a time. [01:59:08 CEST] <furq> read what c_14 just said [01:59:14 CEST] <c_14> so you'll have to use yuv420p as an output pixel format for ffmpeg [01:59:19 CEST] <furq> yeah [01:59:34 CEST] <furq> i've actually had the same issue trying to get zscale to go from rgb to yuv [01:59:40 CEST] <furq> i never did figure out why it wouldn't do it [02:03:53 CEST] <Gazoo> You're saying yuv420p is the only thing WMF is likely to chew? [02:04:41 CEST] <c_14> not likely, according to the docs it's the only thing it will take [02:04:48 CEST] <Gazoo> ok [02:09:32 CEST] <Gazoo> Do it seems the only way forward - if possible - is to find some sort of conversion between bgr0 and the color space in yuv420p [02:09:35 CEST] <Gazoo> So even [02:09:54 CEST] <brian___> why is it that avcodec_close inverts avcodec_open2 but doesnt free the context, while avformat_close_input inverts the open and does close the context? [02:11:03 CEST] <c_14> brian___: @note Do not use this function. Use avcodec_free_context() to destroy a codec context (either open or closed). Opening and closing a codec context multiple times is not supported anymore -- use multiple codec contexts instead. [02:11:26 CEST] <c_14> older api where you could open and close a context multiple times [02:11:28 CEST] <c_14> *remnant of a [02:11:46 CEST] <brian___> oh wow, i didnt even notice. theres no deprecation warning [02:11:48 CEST] <brian___> thank you [02:12:21 CEST] <brian___> if libavcodec/format had a guide to how to free things that'd be super handy :D [02:12:38 CEST] <furq> Gazoo: just -pix_fmt yuv420p will do that [02:12:39 CEST] <brian___> im now only leaking 80 bytes of memory per video open/read :D [02:12:44 CEST] <furq> but obviously it's not doing a good job of the conversion [02:13:04 CEST] <c_14> Gazoo: yeah, you could try the colorspace filter with -vf colorspace=iall=bt709:all=bt709:format=yuv420p, but that's making the assumption that your input matches bt709 in transfer, properties and primaries. If it doesn't it might look worse (but there's really no harm in trying) [02:13:22 CEST] <brian___> oh wait, i think avcodec_free_context cleaned up the last 80 bytes [02:13:28 CEST] <brian___> c_14: you are amazing :) [02:13:39 CEST] <Gazoo> Yeah I guess I was implying find a better conversion. But I think the issue is also perhaps the brightness of my main monitor as I don't see the 'issue' as apparently on my alternate screen [02:14:01 CEST] <furq> Gazoo: try -vf scale=in_range=full:out_range=tv:out_color_matrix=bt709,format=yuv420p [02:14:34 CEST] <Gazoo> I will - gotta get some food in me [02:16:36 CEST] <Gazoo> I've run into some issues of getting the output to be in an mp4 container (as opposed to mkv) and I need to fix that in order to get the file loading at all via WMF [02:16:53 CEST] <furq> what issue [02:17:16 CEST] <c_14> probably codec related [02:17:30 CEST] <furq> yeah i'm guessing the audio codec is wrong [02:20:41 CEST] <Gazoo> Yes - most likely. Is there a default audio codec that's preferrable for the mp4 container? [02:21:27 CEST] <furq> aac or mp3 [02:21:42 CEST] <Gazoo> gotcha [02:42:20 CEST] <Threads> Gazoo aac is the best to go with for mp4 [02:42:48 CEST] <Gazoo> Any particular reason? [02:44:20 CEST] <Threads> i like to say mp3 is old as time [02:44:39 CEST] <Threads> best todo tests with mp3 and aac and see what you like [03:09:43 CEST] <JEEB> aac is what most devices are going to support [03:16:43 CEST] <gurki> JEEB: id differ on that one. literally every device supports mp3 [03:16:48 CEST] <gurki> aac ... not so much [03:18:31 CEST] <JEEB> gurki: ...in mp4 [03:19:06 CEST] <JEEB> (there was a discussion about that in the backlog) [03:19:09 CEST] <gurki> if a device supports mp4 it is totally unlikely not to support either aac nor mp3 [03:19:13 CEST] <gurki> i read that. [03:22:24 CEST] <furq> https://msdn.microsoft.com/en-us/library/windows/desktop/dd742784(v=vs.85).… [03:25:54 CEST] <JEEB> gurki: i find things that don't parse mpeg-1 layer x audio from mp4 [03:26:03 CEST] <JEEB> but do parse aac [03:26:35 CEST] <JEEB> that's why I noted that in that specific case aac works better [04:20:37 CEST] <Masshuu_> So I don't know what i'm doing wrong. Basically ffmpeg stops outputing data randomly after ~280 franes [04:21:56 CEST] <Masshuu_> this is a log that I have https://git.io/vHOo7 [04:22:49 CEST] <Masshuu_> right now I use a headless vlc but I was hoping to use ffmpeg [04:32:28 CEST] <rrodriguez> is there a way to use ffprobe to get the video bitrate from an rtmp stream? [04:37:13 CEST] <dystopia_> try add -re to your line Masshuu_ [04:37:53 CEST] <dystopia_> it's encoding faster than realtime from a live source, so maybe it's getting to the end of the input [04:46:21 CEST] <marcurling> how can I tell a program or Desktop shortcut to launch on a certain monitor/display? [04:47:12 CEST] <marcurling> ^wrong channel, sorry [04:52:44 CEST] <Masshuu_> dystopia_:didn't work, still stops [05:02:48 CEST] <Gazoo101> Before I forget, thanks again to c_14, furq, and Threads [05:02:54 CEST] <Gazoo101> You guys have helped me understand a lot! [05:09:51 CEST] <Masshuu_> https://pastebin.com/kzLN0xFf slightly more verbose? [07:03:44 CEST] <thebombzen> LOL [07:03:58 CEST] <thebombzen> so I just ran this year's worth of channel logs through a neural network gabbler [07:04:46 CEST] <thebombzen> it generated a few pretty hilarious sentences [07:04:55 CEST] <thebombzen> xtina Yeah, same settings I can't get ffmpeg working with desynced audio and video! [07:05:58 CEST] <thebombzen> kepstin so output files from a YUV420P image which I want to rescale pts av_gettime_relative to date her furq you could have telecined then bobbed kerio it can potentially use try out that should possibly work? [07:07:02 CEST] <thebombzen> oh sorry i don't really care that much in SDR land at home i am all about quantum mechanics [07:08:00 CEST] <thebombzen> "Just start my own small hook and inject it into parts roughly 15 minutes long why do fast and trivial." [07:08:26 CEST] <thebombzen> "what do you mean overruns?" [07:08:42 CEST] <kepstin> of course I'm mentioned in something that talks about telecine :/ [07:09:22 CEST] <thebombzen> lol, it is based on the 2017 logs [07:09:28 CEST] <thebombzen> of this channel I mean [07:09:29 CEST] <thebombzen> there ok finally found the least problematic way of demuxing packets [07:11:26 CEST] <thebombzen> "just cross compile it on iDevices which definitely shouldn't hit the cpu being too slow" [07:12:55 CEST] <thebombzen> "if you need to enable less cpu intensive stuff on stderr you should look fine" [07:14:17 CEST] <thebombzen> "and the COMPLETE console output" [09:28:45 CEST] <stdint> I have a problem with movenc, I want to record a offset of a data into the udata box, it works [09:29:23 CEST] <stdint> but I use the avio_tell to get the offset, the offset don't match the offset in file result [09:29:29 CEST] <stdint> any idea about that? [09:56:51 CEST] <raboof> I found https://trac.ffmpeg.org/attachment/wiki/RemapFilter - it looks like projection.c only supports 'single fisheye' format, but AFAICS the samsung gear produces 'dual fisheye' output [09:57:19 CEST] <raboof> the fact that that page also contains mappings for samsung gear suggests "projection.c" has been updated to support dual fisheye - where could I find this updated version? [10:08:16 CEST] <thebombzen> raboof: if it's been updated it'll be on git master. if the way to use it isn't documented you'd have to actually open the source file projection.c and look at it [10:08:34 CEST] <thebombzen> and if you figure it out and you are feeling nice submit a patch documenting it [10:09:34 CEST] <raboof> doesn't seem to be in https://git.ffmpeg.org/ffmpeg.git (no projection.c and grepping for a couple of function names didn't turn up anything either) [11:05:06 CEST] <professoroptics> when I cast my screen with compton, the resulting video has intense flickering. how do I get ffmpeg to play nice with compton? [11:11:27 CEST] <styler2go> Hi. I am trying to convert a few files in a folder. I am using the command find and then -exec to use ffmpeg. Now ffmpeg is using the wrong filename and doesn't find the files. any ideas why? [11:11:36 CEST] <styler2go> my command is: find . -iname "Counter*.mp4" -exec ffmpeg -i {} -strict -2 -b:v 10000k {}.smol \; [11:12:03 CEST] <styler2go> it tried to use the .smol file as input.. but i've set it to output? [11:13:30 CEST] <c_14> command output? [11:14:16 CEST] <styler2go> it runs through the folder, calling ffmpeg, but ffmpeg ... oh i found the problem, ffmpeg doesn't know which format .smol is.. so i should add something like .smol.mp4 or so [11:15:10 CEST] <c_14> or use -f mp4 [11:19:03 CEST] <styler2go> thanks, that worked too [11:20:36 CEST] <professoroptics> how do I prevent video flickering in the output video when recording screen while compton is running? [12:12:27 CEST] <Tehme> Hello. I am trying to use avcodec's CUVID decoder. I do the following: [12:12:27 CEST] <Tehme> https://pastebin.com/iVv1Yc2w [12:12:27 CEST] <Tehme> And I get this: [12:12:27 CEST] <Tehme> Loaded lib: libnvcuvid.so.1 [12:12:27 CEST] <Tehme> Cannot load cuvidGetDecoderCaps [12:12:28 CEST] <Tehme> I looked into nvcuvid.h from CUDA SDK 8, and there was no cuvidGetDecoderCaps() function. Why does ffmpeg try to load it then? What am I doing wrong? [13:34:15 CEST] <smooth_penguin> hey, I'm using ffmpeg to concat multiple .ts files, it seems to work out fine, [13:34:29 CEST] <smooth_penguin> except the video seems to be playing much quicker that the audio [13:34:42 CEST] <smooth_penguin> the audio is playing at the normal rate [13:34:53 CEST] <smooth_penguin> any clues on how I can fix this? [13:35:35 CEST] <smooth_penguin> ffmpeg -f concat -i list.ts -c copy workdir/all.ts [13:35:40 CEST] <smooth_penguin> is what I used [14:04:52 CEST] <thebombzen> smooth_penguin: I assume you mean list.txt. and with mpegts you can actually just cat them togther [14:05:13 CEST] <thebombzen> cat first.ts second.ts third.ts fourth.ts >all.ts [14:05:25 CEST] <thebombzen> like that actually works with mpegts [14:25:39 CEST] <marquisor> furq: still struggling with the server setup [14:26:12 CEST] <marquisor> furq: i btw have apache running [14:27:33 CEST] <Tehme> Where should I ask avcodec related questions? Here or ffmpeg-devel? [14:29:08 CEST] <iive> Tehme: if it is about using avcodec, ask here, if it is about fixing/improving avcodec then ask in devel [14:31:11 CEST] <Tehme> Well, I'm not sure if this is using or fixing. I'll try there too. Thank you for your answer. [15:16:03 CEST] <smooth_penguin> thebombzen: yeah, I tried that as well, the video still plays quicker than the audio [16:15:36 CEST] <kepstin> smooth_penguin: is the audio/video in sync if you play each of the segments individually? [16:15:54 CEST] <kepstin> because if that's wrong, concatenating them won't change anything... [16:49:27 CEST] <SolidusAbi> Hi, im trying to generate a smooth streaming. Can someone help checking whether my commands are ok? [16:49:54 CEST] <SolidusAbi> ffmpeg -i ../seg0.mov -s 1920x1080 -f ismv -c:v libx264 -movflags frag_keyframe -pix_fmt yuv420p -b:v 4.2M -map 0:v video.ismv -y [16:50:15 CEST] <SolidusAbi> Im trying to generate a ismv file confied in h264 [16:50:56 CEST] <SolidusAbi> ffmpeg -i seg0.mov -vn -r:a 48000 -acodec mp3 audio.mp3 [16:51:04 CEST] <SolidusAbi> here i want to extract audio [16:54:09 CEST] <SolidusAbi> the next step is generate a video in 720p and bitrate of 2.9M, inserting audio extratecd. [16:55:17 CEST] <SolidusAbi> can someone help me? [18:08:19 CEST] <professoroptics> how do i get ffmpeg to play nice with compton? it causes video flicker when screencasting [18:11:28 CEST] <marquisor> finally got the damn nginx-rtmp-module working [18:11:41 CEST] <marquisor> now i have jerky stream to clients... [18:17:47 CEST] <marquisor> and disconnects [18:17:52 CEST] <marquisor> from input stream [18:47:12 CEST] <rjp421> marquisor, is the problem nginx or ffmpeg [18:48:12 CEST] <marquisor> i really don't know yet [18:48:38 CEST] <marquisor> maybe bandwidth or more a buffer issue also [18:53:17 CEST] <lacrymology> how can I create, say, 10 seconds of white noise or something? [18:55:47 CEST] <DHE> lacrymology: https://ffmpeg.org/ffmpeg-filters.html#aevalsrc there's an example that generates random noise [19:00:18 CEST] <lacrymology> DHE: while I'm looking at it, I found this in stackoverflow, the original was dumping to an mkv but I need an mp4 [19:00:24 CEST] <lacrymology> 13:58 $ ffmpeg -f rawvideo -video_size 1280x720 -pixel_format yuv420p -framerate 25 -i /dev/urandom -ar 48000 -ac 2 -f s16le -i /dev/urandom -codec:a copy -t 5 output.mkv [19:00:49 CEST] <lacrymology> it's failing with 'Could not find tag for codec pcm_s16le in stream #1, codec not currently supported in container' [19:01:42 CEST] <lacrymology> DHE: sorry, I'm trying to generate random video [19:01:51 CEST] <lacrymology> although.. anything goes, really, I just need unique video files [19:01:59 CEST] <furq> !filter noise [19:02:00 CEST] <nfobot> furq: http://ffmpeg.org/ffmpeg-filters.html#noise [19:02:02 CEST] <furq> !filter anoisesrc [19:02:05 CEST] <furq> er [19:02:09 CEST] <furq> !source anoisesrc [19:02:09 CEST] <nfobot> furq: http://ffmpeg.org/ffmpeg-filters.html#anoisesrc [19:02:11 CEST] <furq> use those [19:02:28 CEST] <furq> and encode them to codecs that mp4 supports [19:02:30 CEST] <furq> presumably h264 and aac [19:02:36 CEST] <Mavrik> Hmm, is there any good way of detecting corrupted frames (the ones that reported "ac-tex damaged" etc. when decoding? [19:02:41 CEST] <Mavrik> Via API? [19:06:13 CEST] <rjp421> will videos using anoisesrc and amplitude 0 still all be unique? [19:06:41 CEST] <furq> you can set the seed with noise [19:06:44 CEST] <furq> idk about anoisesrc [19:08:40 CEST] <furq> yeah you can set it with both [19:09:10 CEST] <furq> lacrymology: ffmpeg -f lavfi -i color=black:s=1280x720:d=10,noise=alls=100:allf=t:seed=12345 -f lavfi -i anoisesrc=d=10:s=67890 -c:v libx264 -c:a aac out.mp4 [19:09:43 CEST] <furq> also this is pretty much the absolute worst-case scenario for x264 so it's going to take ages and be enormous [19:11:50 CEST] <lacrymology> furq: where's the length defined? [19:12:03 CEST] <furq> d=10 [19:12:48 CEST] <lacrymology> I thought that was colordepth somehow, thakns [19:14:30 CEST] <lacrymology> 8M for 2 seconds damn [19:14:48 CEST] <lacrymology> at 640x480 [19:16:42 CEST] <furq> if you just need unique files and you don't care about the fidelity of the noise then bump the crf up [19:16:45 CEST] <furq> or set a max bitrate [19:16:50 CEST] <furq> it'll be a horrible blocky mess but it'll still be unique [19:21:51 CEST] <lacrymology> thanks [19:21:57 CEST] <lacrymology> I need to learn the basics [19:22:06 CEST] <lacrymology> I've used ffmpeg for a number of things, but I don't know the syntax [19:25:58 CEST] <james999> lacrymology: would you appreciate a detailed tutorial and walkthrough [19:26:06 CEST] <james999> complete with audio and video? [19:26:29 CEST] <lacrymology> james999: sure would. Something that goes through the syntax, not just formulas, which is what I've used, historically [19:26:52 CEST] <james999> well. i wonder why there isn't a youtube channel like that already [19:26:55 CEST] Action: james999 jots it down for future reference [19:27:27 CEST] <lacrymology> are you being ironic, and is there a very famous one I'm missing? [19:28:02 CEST] <james999> no but you are right i'm being sort of tongue in cheek [19:28:35 CEST] <james999> for example when I type linux ffmpeg tutorial into youtube I get mostly trash. one vid about basics of ubuntu; and this: https://www.youtube.com/watch?v=c9YLTNGVdPU [19:35:04 CEST] <furq> that's me in that video [19:35:07 CEST] <furq> i am "skate.avi" [19:36:45 CEST] <james999> you are the speaker or the video file? :D [19:38:40 CEST] <furq> i'm skate.avi [19:40:20 CEST] <james999> ok [19:42:51 CEST] <pbgswd> is it ok to talk about moviepy here? [22:53:43 CEST] <xhip> I was trying to search a cvlc to ffmpeg convert.. but no luck [22:54:04 CEST] <xhip> someone knows how can I convert this cvlc -vvv v4l2:///dev/video1:chroma=mjpg :input-slave=alsa://hw:2,0 --sout "#transcode{vcodec=mp1v,vb=4096,acodec=mp3,ab=128,scale=2,channels=2}:rtp{dst=127.0.0.1,port=1234,sdp=rtsp://127.0.0.1:1234/stream.sdp} .. to ffmpeg [23:05:50 CEST] <xhip> I'm testing some commands.. ffmpeg -re -i /dev/video1 -f rtsp -muxdelay 0.1 rtsp://127.0.0.1:1234/stream.sdp [23:07:23 CEST] <kepstin> xhip: don't use -re when your input is realtime (e.g. a camera). It's only useful for if your source is e.g. a file or something [23:07:34 CEST] <xhip> oh ok [23:07:46 CEST] <xhip> make sense [23:14:39 CEST] <xhip> I can't watch the stream on vlc :/ [23:15:08 CEST] <xhip> Im also geting this: [23:15:11 CEST] <xhip> [mp3 @ 0x92e9c60] Format detected only with low score of 1, misdetection possible! [23:15:12 CEST] <xhip> [mp3 @ 0x92ebb00] Header missing [00:00:00 CEST] --- Fri May 26 2017

1 0

ffmpeg-devel.log.20170524
by burek 25 May '17

25 May '17

[00:00:46 CEST] <jamrial> am i missing something, or is https://github.com/FFmpeg/FFmpeg/blob/master/libavutil/pixfmt.h#L470 superfluous judging by the enum above? [00:02:20 CEST] <rcombs> jamrial: huh, wacky [00:02:24 CEST] <rcombs> sure looks like it [00:02:37 CEST] <rcombs> maybe somebody #ifdefs it? [00:03:10 CEST] <nevcairiel> might be a merge artifact [00:03:21 CEST] <jamrial> most likely that, yeah [02:51:01 CEST] <J_Darnley> Fuck! I forgot I was supposed to not upgrade Thunderbird! [02:59:50 CEST] <J_Darnley> BBB: Sorry that I sent that mail directly to you. [03:00:09 CEST] Action: J_Darnley goes to bed. [11:36:26 CEST] <wm4> #define ALLOW_INTERLACE [11:36:36 CEST] <wm4> didn't know you could disable all deinterlacing in h264 lavc [11:47:19 CEST] Action: wm4 stares at the use of the "state" variable in h264_parser.c [12:16:01 CEST] <wm4> what's this mode called where h264 codes separate fields as separate slices, why is BBC switching between this and non-field coded slices, how does interlacing without separate fields even work, and why does the h264 parser not merge both field slices into a single packet (making remuxing from .ts generally fail)? [12:16:12 CEST] <wm4> also did I mention yet that I fucking hate interlacing? [12:16:26 CEST] <wm4> kierank: maybe you know something? [12:17:30 CEST] <wm4> my current impulse would be just making h264_parser.c merge those slice NALs into a single packet, but I'm afraid of this dumb state machine [12:18:33 CEST] <wm4> it's basically Open Binary code [12:19:23 CEST] <wm4> state ^= 5; // 2->7, 1->4, 0->5 [12:19:46 CEST] <wm4> no explanations what the states mean [12:19:58 CEST] <wm4> random magic [12:20:30 CEST] <wm4> return i - (state & 5) - 5 * (state > 7); [12:20:32 CEST] <wm4> "fuck you too" [12:25:13 CEST] <J_Darnley> Oh my. Reminds me of the overloaded variables in AVS. Sorry I can't help. [12:25:19 CEST] <jkqxz> The state field is documented: "contains the last few bytes in MSB order". I totally believe that from the code :P [12:26:03 CEST] <wm4> not sure where you see this [12:26:13 CEST] <nevcairiel> thats a different state field [12:26:59 CEST] <jkqxz> It's used to store the state value. [12:27:10 CEST] <wm4> I'm talking about the local variable in h264_find_frame_end [12:27:39 CEST] <wm4> oh. [12:27:57 CEST] <wm4> should have the same semantics [12:28:20 CEST] <nevcairiel> the state jkqxz talks about is typically used to find startcodes [12:28:45 CEST] <nevcairiel> the other statemachine in h264_parser is quite a bit more crazy [12:28:58 CEST] <nevcairiel> anyway merging the fields into one packet is likely going to break all the things [12:29:06 CEST] <wm4> which other things? [12:29:12 CEST] <wm4> well fine, so how do I fix remuxing [12:29:24 CEST] <nevcairiel> tell people to shove off if they report interlaced problems [12:29:25 CEST] <nevcairiel> :) [12:29:28 CEST] <jkqxz> By mixed frames and fields, do you mean like in camp_mot_picaff0_full.26l? [12:30:37 CEST] <wm4> don't know, this is a stream captured by the BBC, and it appears field_pic_flag goes on and off randomly [12:31:43 CEST] <wm4> (probably some sort of encoder decision) [12:34:37 CEST] <wm4> nevcairiel: I'm not sure what's wrong with merging them - shouldn't each packet contain a complete frame? [12:35:02 CEST] <jkqxz> Yeah, it does that. (That stream is the "haha fuck you" conformance test for PicAFF: <http://sprunge.us/AFAj>.) [12:35:47 CEST] <nevcairiel> wm4: a field per packet is fine as well, in ts they can even have distinct timestamps per field [12:36:38 CEST] <nevcairiel> in another note, how do you remux to ts then? :p [12:36:40 CEST] <wm4> in this case they don't (or at least neither the lavf ts demuxer nor the utils.c fuckmagic is setting any) [12:37:08 CEST] <wm4> I think we're trying to remux to mkv [12:49:11 CEST] <ubitux> src/libavutil/arm/intmath.h:77:5: error: impossible constraint in asm [12:49:12 CEST] <ubitux> __asm__ ("usat %0, %2, %1" : "=r"(x) : "r"(a), "i"(p)); [12:49:14 CEST] <ubitux> :( [12:51:39 CEST] <ubitux> maybe that's because of disable-optimization [12:54:22 CEST] <ubitux> no, doesn't seem to make a difference [12:54:44 CEST] <ubitux> i don't understand why it wouldn't be able to solve that given that the argument looks constant [12:54:58 CEST] <ubitux> (it comes from libavfilter/colorspacedsp_template.c) [12:55:01 CEST] <BtbN> Because it's told to not even try? [12:55:25 CEST] <wm4> and the compiler can't generate that code? [12:55:52 CEST] <ubitux> doesn't seem so [12:56:10 CEST] <ubitux> (gcc 6.3.1, armv7l) [13:04:04 CEST] <wm4> I mean, using pure C [13:04:16 CEST] <wm4> ideally a good compiler would use usat instead of generating more stupid code [13:06:30 CEST] <nevcairiel> thats why inline asm is a problem, it puts severe constraints on compiler behavior resulting in such errors [13:07:07 CEST] <ubitux> wm4: the generated code is like 8 instructions here [13:07:38 CEST] <ubitux> maybe it's smaller with a constant though [13:08:15 CEST] <ubitux> yeah 5 instruct with a constant [13:08:23 CEST] <ubitux> still no usat [13:08:30 CEST] <ubitux> anyway [13:08:34 CEST] <ubitux> ffmpeg doesn't compile :p [13:09:27 CEST] <ubitux> i wonder why it doesn't been detected before [13:48:26 CEST] <JEEB> ubitux: does clang do the same? [13:54:16 CEST] <ubitux> JEEB: the number of instructions? pretty much yes [13:54:29 CEST] <ubitux> haven't tried for ffmpeg compilation yet, because it takes& a long time [13:55:16 CEST] <JEEB> k [13:55:22 CEST] <ubitux> i'm retrying compilation with optimizations enabled first [13:55:41 CEST] <ubitux> (i forgot to clean previously so it may have derped) [15:28:43 CEST] <cone-977> ffmpeg 03James Almer 07master:11827ea3f041: fate: add aac-al_sbr_ps_06_ur test [15:39:08 CEST] <ubitux> lol at this comment: //l is s, r is d [15:39:46 CEST] <ubitux> jamrial: nice, i guess i'll be able to test the ipodpdpdpd thing after the simple stereo interpolate :) [15:40:40 CEST] <jamrial> oh wtf, i wrote 06 instead of 04 in the commit message [15:41:52 CEST] <jamrial> ubitux: heh, you reminded me how google would refuse to search ipdopd and change it into ipod for me :p [15:42:34 CEST] <kierank> jamrial: you use the + modifier [15:42:36 CEST] <kierank> and it should work [15:52:03 CEST] <cone-977> ffmpeg 03Michael Niedermayer 07master:4bc3008d0445: avcodec/jpeglsdec: Check get_bits_left() before decoding a picture [15:52:04 CEST] <cone-977> ffmpeg 03Max Justicz 07master:ca616b0f72c6: avcodec/sanm: Fix uninitialized reference frames [16:34:56 CEST] <BBB> ubitux: have you checked tsan reporting lately? has it been clear-ish? [16:36:10 CEST] <ubitux> afaik it's in the same state as when you "left" it [16:36:26 CEST] <ubitux> so hevc/h264 fails sometimes and mpeg* is broken [16:36:50 CEST] <BBB> mpeg* and frame_thread_encoder [16:36:56 CEST] <ubitux> yeah [16:37:02 CEST] <BBB> ok thats not terrible [16:37:13 CEST] <ubitux> + slices [16:37:17 CEST] <BBB> oh right slices [16:37:21 CEST] <ubitux> which are broken in different ways [16:37:43 CEST] <BBB> h264 slices is pretty broken [16:38:02 CEST] <BBB> I should probably look at that [16:38:06 CEST] <BBB> I thought I fixed some of that [16:38:09 CEST] <BBB> maybe just not enough [16:38:19 CEST] <ubitux> i'd rather have the the !slices first, but sure [16:38:30 CEST] <BBB> I agree [16:38:40 CEST] <ubitux> so we can have at least one green thread instance [16:38:42 CEST] <BBB> if I leave it in a 90% state nobody will ever care to get it up to 100% [16:38:47 CEST] <ubitux> which is much easier to monitor [16:38:50 CEST] <BBB> mpeg* is hard though [16:38:58 CEST] <BBB> but I guess someone has to care [16:38:59 CEST] <BBB> hm [16:39:00 CEST] <BBB> ... [16:39:00 CEST] <BBB> ok [16:39:25 CEST] <BBB> still pretty good, only 65 failing tests [16:39:29 CEST] <BBB> it used to be several 100s [16:39:40 CEST] <BBB> not saying its ok to see 65 tests fail [16:39:44 CEST] <BBB> just saying there was some improvement [16:41:04 CEST] <ubitux> yes, it's definitely much better [16:41:33 CEST] <ubitux> and i love the fact that we're pretty close to be able to use it to detect new races :) [16:42:21 CEST] <BBB> oh I remember the hevc one [16:42:28 CEST] <ubitux> jamrial: shouldn't we have the aac test in checkasm? [16:42:32 CEST] <BBB> so, theres this thing called end of sequence slices [16:42:50 CEST] <BBB> which update one variable and otherwise dont do anything [16:42:51 CEST] <ubitux> jamrial: typically testing the expected boundary constraints [16:43:15 CEST] <BBB> but because theyre after a slice, they interfere with frame threading in the sense that they update a variable after finish_setup() was called [16:43:30 CEST] <BBB> so maybe the eob slice should be split and decoded in its own thread even though its not technically a video frame [16:43:31 CEST] <BBB> WDYT? [16:43:45 CEST] <BBB> (by the parser) [16:43:45 CEST] <jamrial> ubitux: i guess, but i'm not going to write it. last time i tried to add float dsp tests it had problems everywhere [16:46:35 CEST] <ubitux> BBB: hevc, h264 or mpeg? [16:46:41 CEST] <BBB> hevc [16:46:42 CEST] <ubitux> jamrial: what kind of problem? [16:46:58 CEST] <ubitux> fuzz-related issues? [16:48:08 CEST] <jamrial> ubitux: no, some tests failing if ran after others, x86_32 failing no matter what [16:48:10 CEST] <jamrial> seemed like something related to emms not being handled right by the generic checkasm code [16:48:35 CEST] <graphitemaster> << -1 * 2 - 1 [16:48:41 CEST] <graphitemaster> er, wrong channel [16:49:18 CEST] <ubitux> jamrial: so we don't have any float tests in checkasm currently? oO [16:49:26 CEST] <jamrial> yes, we do [16:49:33 CEST] <jamrial> synth_float [16:49:51 CEST] <jamrial> what we don't have is tests that return a float instead of void [16:50:55 CEST] <ubitux> BBB: the code setting that var can not be taken out of the slice thread? [16:51:12 CEST] <ubitux> BBB: it's very vague to me what you're trying to accomplish :p [16:51:43 CEST] <ubitux> jamrial: ooh, ok, well the aac dsp doesn't seem to do something else so it should be fine :) [16:52:08 CEST] <ubitux> at least the ps code, i think the sbr code mateo` is working on has some return (but only ints iirc) [16:56:54 CEST] <BBB> ubitux: frame thread, not slice thread [16:57:22 CEST] <BBB> jamrial: someone added emms handling to mmx code at some point [16:57:36 CEST] <BBB> jamrial: but I dont remember how it works grrrrmbl [16:58:11 CEST] <jamrial> yes, but it didn't fix the issues i saw [16:58:33 CEST] <jamrial> in any case, it also wasn't working right in x86_64, where it's all sse, so maybe i fucked up somewhere else [16:59:13 CEST] <jamrial> it's in the ml, the float_dsp port patch. some tests fail if you run it as part of the entire suite but don't if you run them individually [16:59:32 CEST] <jamrial> they also doesn't make checkasm report the run as failed. it's weird [17:00:16 CEST] <jamrial> so i said fuck it, i'll stick with ints where everything is cozy :p [17:00:52 CEST] <ubitux> :) [17:04:05 CEST] <jamrial> https://pastebin.com/DWhMcwX3 this is the output on x86_64 [17:04:55 CEST] <jamrial> same seed, fails (but doesn't report it as failed at the end) if run as part of the whole suite, succeeds if ran alone [17:06:14 CEST] <jamrial> one of the "failing" tests is the one that retuns a float, something no other test currently in checkasm does [17:07:05 CEST] <jamrial> it's a direct port of the libavutil/tests file, which works fine, so i'm guessing something in the generic checkasm code is not playing nice with it [17:08:22 CEST] <mateo`> ubitux: only one function returns something (a float) in the sbrdsp [17:08:39 CEST] <ubitux> ah that's actually a float, ok [17:09:49 CEST] <mateo`> as ubitux said, i'm working on porting the sbrdsp to aarch64 (something like 40% of the work is done) [17:13:49 CEST] <BBB> ubitux: wanna test a patch for me? [17:13:57 CEST] <BBB> ubitux: for the hevc frame-thread issue [17:14:00 CEST] <ubitux> sure [17:14:12 CEST] <BBB> its totally hacky but maybe it works [17:14:31 CEST] <BBB> (itll take me a while to recompile everything with tsan and Im doing something else also :-/ ) [17:14:41 CEST] <BBB> https://pastebin.com/V2YW31Qm [17:14:50 CEST] <ubitux> well, i need to resetup a local build as well [17:14:55 CEST] <jamrial> from sbrdsp i wrote autocorrelate some time ago. i remember it was fun trying to figure out the first and last pass [17:15:00 CEST] <ubitux> it will take some time but i'll do it [17:15:06 CEST] <jamrial> looking at it now, i have no idea how to read all those "comments" i left next to some instructions :P [17:21:30 CEST] <BBB> hm no that doesnt do anything useful at all [17:23:53 CEST] <BBB> aha I made a mistake between eos/last_Eos [17:24:07 CEST] <BBB> ok a slight change makes it work and is probably fundamentally more correct than what it currently does [17:24:11 CEST] <BBB> lemme see if this fixes the tsan report [17:24:21 CEST] <BBB> oh I dont see it [17:24:22 CEST] <BBB> tsk tsk [17:24:25 CEST] <BBB> why not [17:25:48 CEST] <BBB> ubitux: ok so I cant reproduce the issue, it seems [17:25:53 CEST] <BBB> but I think I know how to fix it [17:25:54 CEST] <BBB> so... [17:25:56 CEST] <ubitux> which test is it supposed to fix, hevc-conformance-NoOutPrior_A_Qualcomm_1? [17:26:20 CEST] <BBB> yes [17:26:25 CEST] <BBB> but not the patch I just gave [17:26:26 CEST] <BBB> one sec [17:26:30 CEST] <ubitux> i can reproduce more easily increasing the number of threads [17:26:46 CEST] <ubitux> anyway, i have a tsan build for which i can reproduce [17:27:04 CEST] <ubitux> THREADS=8 fails almost all the time [17:27:29 CEST] <ubitux> btw, when it's all green i'll increase the number of threads [17:27:38 CEST] <BBB> https://pastebin.com/us4Rs1qp [17:27:39 CEST] <ubitux> (currently the fate instance is at 2) [17:27:56 CEST] <BBB> it reproduces here very sparringly [17:27:59 CEST] <BBB> I cant reproduce with that patch [17:28:11 CEST] <BBB> but itd be nice if you can reproduce more consistently to see if the patch fixes it [17:28:50 CEST] <ubitux> it seems to fix the issue indeed [17:28:52 CEST] <BBB> hey that would actually make hevc green [17:28:54 CEST] <BBB> nice [17:28:57 CEST] <BBB> on to h264 again then [17:29:19 CEST] <J_Darnley> I hope those Typhoons and Galaxies flying overhead are on their way to "bomb the shit" out of Brussels. [17:30:35 CEST] <ubitux> BBB: hevc-conformance-CONFWIN_A_Sony_1 failed thoughj [17:30:43 CEST] <ubitux> hevc-conformance-CAINIT_A_SHARP_4 as well [17:31:35 CEST] <ubitux> ah fuck i forgot to disable the stripping [17:32:13 CEST] <BBB> ohwell [17:32:21 CEST] <J_Darnley> :( pretty coloured smoke but no fire [17:35:17 CEST] <BBB> ubitux: patch sent [17:35:58 CEST] <ubitux> BBB: http://b.pkh.me/hevc-conformance-OPFLAG_C_Qualcomm_1.err [17:36:38 CEST] <BBB> oh that one again :( [17:36:42 CEST] <ubitux> (hevc-conformance-OPFLAG_C_Qualcomm_1) [17:36:42 CEST] <BBB> Ive seen that before [17:36:52 CEST] <BBB> I have not yet been able to determine what causes it [17:37:53 CEST] <BBB> it suggests a frame is decoded (multi-slice) in multiple threads, causing owner[] to be wrong [17:37:58 CEST] <ubitux> BBB: http://b.pkh.me/hevc-conformance-CAINIT_A_SHARP_4.err [17:38:02 CEST] <ubitux> how about this one? [17:38:02 CEST] <BBB> thats certainly possible and I havent dont a lot of work on that yet [17:38:08 CEST] <BBB> same [17:38:13 CEST] <BBB> better backtrace though [17:38:43 CEST] <BBB> its possible that owner[] is not set correctly in hevc for interlaced pictures split in two frame decoding threads [17:38:57 CEST] <BBB> I only fixed that for h264 [17:39:18 CEST] <nevcairiel> hevc interlaced is not something we support anyway [17:39:38 CEST] <BBB> maybe its something else [17:39:38 CEST] <ubitux> got it again with another test [17:39:39 CEST] <BBB> anyway [17:39:48 CEST] <BBB> I need to look deeply into one of these failing instances [17:39:51 CEST] <BBB> at some point I will [17:40:10 CEST] <BBB> its not very high priority because it has no practical implications [17:40:12 CEST] <BBB> afaik [17:40:16 CEST] <BBB> but it needs fixing obviously [17:40:41 CEST] <ubitux> it happens a lot with a bunch of threads [17:41:01 CEST] <ubitux> http://sprunge.us/fRQM [17:41:10 CEST] <ubitux> i haven't checked all but it does seem to be always the same [17:41:13 CEST] <BBB> ok, good to know [17:41:19 CEST] <BBB> Ill dig deeper at some point [17:41:26 CEST] <BBB> I sort of know what it means but not exactly ;) [17:41:34 CEST] <BBB> and hevc is & bla, vp9 much more important [17:41:37 CEST] <BBB> and at least vp9 is clean [17:42:24 CEST] <BBB> the h264 fail one is also the same btw [17:42:29 CEST] <BBB> so I guess I have no choice but to dig into it [17:42:31 CEST] <BBB> brrrrr [17:42:36 CEST] <BBB> I dont like that at all [17:42:39 CEST] <ubitux> :( [17:42:52 CEST] <ubitux> don't worry, once fixed, forever green [17:43:09 CEST] <BBB> ;) [17:43:17 CEST] <iive> until next merge ;) [17:43:57 CEST] <ubitux> speaking of merges, i got no answer wrt dashenc [17:44:05 CEST] <ubitux> (from the guy who did the bitrate thing) [17:44:15 CEST] <ubitux> no one wants to give it a try? [17:44:50 CEST] <RiCON> dashenc or dashdec? [17:44:54 CEST] <ubitux> enc [17:45:10 CEST] <RiCON> oh, libav merges [17:45:10 CEST] <RiCON> nvm [17:45:13 CEST] <ubitux> i feel like dropping our changes in favor of the work from the original author since we don't have a maintainer [17:45:42 CEST] <BBB> that may be reasonable, yes [17:45:44 CEST] <ubitux> (BBB: "Tthis") [17:46:06 CEST] <jamrial> ubitux: i'd say do that if the guy that wrote out bitrate thing isn't answering. there are a bunch of other dashenc commits from libav, so keeping our version may create conflicts in the future [17:46:10 CEST] <BBB> fixed [17:46:40 CEST] <ubitux> jamrial: yeah, it's going to be a pain [17:46:53 CEST] <ubitux> i don't want to do it :( [17:47:15 CEST] <thardin> has anyone gotten afl to work with ffmpeg? [17:47:25 CEST] <thardin> feeling slightly inspired to give it a go again [17:47:49 CEST] <BBB> is afl the complicated fuzzer? [17:47:57 CEST] <BBB> didnt kierank try that? [17:48:16 CEST] <kierank> thardin: yes it's pretty straightforward [17:48:28 CEST] <thardin> CC=afl-gcc or so? [17:48:46 CEST] <kierank> yes [17:48:49 CEST] <kierank> configure as well [17:48:53 CEST] <kierank> this might help [17:48:53 CEST] <kierank> https://github.com/openBroadcastSystems/fffuzz [17:49:00 CEST] <kierank> unless you specifically want to use ffmpeg.c [17:49:27 CEST] <kierank> tbh there's low hanging fruit with zzuf first [17:50:02 CEST] <thardin> lavf always seems more suspect to me, so fuzzing ffprobe kinda seems most relevant [17:50:05 CEST] <BBB> ubitux: oh, btw, I did at some point briefly look into fixing tsan for mpeg* [17:50:16 CEST] <BBB> ubitux: and its going to be a bloody and deadly and torturous pain [17:50:20 CEST] <BBB> ubitux: just so you know :-p [17:51:00 CEST] <BBB> hi gh0st__ [17:51:11 CEST] <ubitux> because of the famous and beloved MpegEncContext? [17:51:17 CEST] <thardin> but a minimal decoder CLI should be useful too, thanks [17:53:23 CEST] <BBB> ubitux: as part of update_context, it does update_context(MpegEncContext *dst, MpegEncContext *src) { [..] *dst = *src; [..] } [17:53:31 CEST] <kierank> thardin: fuzzing containers i think is a nontrivial problem [17:53:41 CEST] <BBB> ubitux: and lets just say that will take incredibly long to unwind in a way that can be proven to be correct [17:54:08 CEST] <BBB> ubitux: especially for someone like me who is superficially aware of mpeg bitstream syntax things but has no detailed knowledge of any of it [17:54:32 CEST] <ubitux> BBB: your face when you will realize it's for 5-6 sparse fields you'll have to figure out in a struct of 1000s of fields [17:54:35 CEST] <ubitux> :3 [17:54:42 CEST] <thardin> kierank: no doubt [17:54:57 CEST] <BBB> ubitux: you will hear me scream [17:55:11 CEST] <BBB> but yes [18:04:20 CEST] <gh0st__> BBB:hey! [18:05:23 CEST] <BBB> one more week! [18:05:56 CEST] <gh0st__> BBB: Yeah, I am ready! =) [19:14:27 CEST] <ubitux> [aarch64] so i have v0=AXXX v1=BXXX, i want f(v0,v1) = AABB; i'm going for zip1.2(zip1.4(v0,v0) zip1.4(v1,v1)) [19:14:32 CEST] <ubitux> anything else to suggest? [19:14:56 CEST] <ubitux> 3 zip are overkill, given that i have to do the same for all the other X [19:15:51 CEST] <ubitux> v0=ABCD v1=EFGH {AAEE, BBFF, CCGG, DDHH} [19:16:02 CEST] <ubitux> (4 different regs) [19:17:23 CEST] <atomnuker> <pshufb in aarch64-ise> [19:31:02 CEST] <iive> ubitux: no (un)pack equivalents? [19:31:22 CEST] <ubitux> that's what zip/uzp do [19:31:44 CEST] <ubitux> basically the tools are zip/uzp and the pshuf joker is "ext" [19:31:52 CEST] <ubitux> which i probably want to avoid [19:32:33 CEST] <iive> do you have a url to description of the (un)zip ? [19:32:48 CEST] <ubitux> with 4 zip i end up with {AACC, BBDD, EEGG, FFHH}, now i technically need 2 more instruction to swap CC with EE and swap DD with FF [19:32:55 CEST] <ubitux> iive: the arm doc is shit [19:33:23 CEST] <ubitux> https://developer.arm.com/docs/100069/latest/a64-simd-vector-instructions/u… [19:33:59 CEST] <ubitux> they still haven't dare to draw freaking graphics like in the intel doc [19:34:23 CEST] <ubitux> there are also the trn (transpose) ones that may help [19:35:02 CEST] <ubitux> so zip[12], uzp[12], trn[12] and ext joker [19:37:34 CEST] <iive> wow, that's bad... [19:37:42 CEST] <iive> what is arangement specifier? [19:39:23 CEST] <ubitux> it refers to the "type", treat it like 8, 16, 32 or 64 bits entries [19:39:33 CEST] <ubitux> and the variants refering to half of the ref [19:39:37 CEST] <ubitux> s/ref/reg/ [19:39:42 CEST] <iive> int or float too? [19:39:49 CEST] <ubitux> yeah it doesn't matter [19:39:52 CEST] <ubitux> it's the size [19:40:29 CEST] <ubitux> 8B is the lower-only version of 16B, 4H is the lower-only version of 8H, etc [19:40:37 CEST] <iive> so is 16b 16bits or 16 integer cells? [19:41:12 CEST] <ubitux> 16B refers to the "8-bit" layout of the whole SIMD register [19:41:18 CEST] <ubitux> "16 bytes" [19:41:36 CEST] <ubitux> 8H refers to the "16-bit" layout of the whole SIMD register [19:41:41 CEST] <ubitux> "8 half words" [19:42:33 CEST] <ubitux> the asm is pretty cool tbh [19:42:37 CEST] <mateo`> if you want the fancy version (armv7 only): https://community.arm.com/processors/b/blog/posts/coding-for-neon---part-5-… [19:42:38 CEST] <ubitux> the doc just sucks [19:42:47 CEST] <ubitux> yep i got that page [19:43:06 CEST] <ubitux> i'm interested in the rev, unfortunately it's in place and i don't want that [19:43:17 CEST] <ubitux> i want to swap the 2D layout of 2 given registers [19:45:43 CEST] <ubitux> v0.D={A,B} v1.D={C,D} and i want v0.D={A,C} v1.D={B,D} [19:45:53 CEST] <ubitux> i can probably do that with trn* [19:46:00 CEST] <ubitux> anyway. [19:46:01 CEST] <mateo`> trn if what you look for [19:46:22 CEST] <mateo`> if->is [19:47:12 CEST] <ubitux> yeah i need to figure out again what's it's doing [19:47:29 CEST] <ubitux> i really need to draw all of them at some point [19:47:32 CEST] <iive> ubitux: vzip as first step, give you {ABCD}x{EFGH} -> {AEBF}x{CGDH} , doesn't it? [19:48:08 CEST] <ubitux> this is arm, aarch64 is simpler, you have 1 dst for 2 src [19:48:32 CEST] <ubitux> you can ignore the inplace manipulations, it's much easier [19:48:39 CEST] <iive> so no vzip for you [19:48:46 CEST] <iive> ? [19:48:46 CEST] <ubitux> yeah but zip1() and zip2() [19:49:00 CEST] <iive> would they give you the same result? [19:49:50 CEST] <iive> {AEBF}x{AEBF}->{AAEE}x{BBFF} [19:50:03 CEST] <ubitux> v0=ABCD v1=EFGH {zip1(v0,v0), zip2(v0,v0), zip1(v1,v1), zip2(v1,v1)} {AACC, BBDD, EEGG, FFHH} [19:50:33 CEST] <ubitux> now i need 2 more instr to swap into {AAEE, BBFF, CCGG, DDHH} [19:50:39 CEST] <mateo`> my bad, trn is not going to work for that [19:50:54 CEST] <ubitux> mateo`: i think it might do it [19:51:02 CEST] <iive> ubitux: see what i've written [19:51:55 CEST] <ubitux> assuming i understood your notation it should be possible with 2 zip [19:52:58 CEST] <ubitux> (i'm understanding zip1(AEBF,AEBF)=AAEE and zip2(AEBF,AEBF)=BBFF) [19:53:05 CEST] <iive> 3 vzips or 6 zip?s [19:53:12 CEST] <ubitux> just 2 [19:53:21 CEST] <iive> to make all 4 results [19:53:45 CEST] <iive> also you need 2 register copy [19:53:47 CEST] <ubitux> 4 zip for all 4 results before the last 2 swap [19:56:01 CEST] <ubitux> in the end, i want v0=ABCD v1=EFGH {AACC, BBDD, EEGG, FFHH} [19:56:14 CEST] <ubitux> right now i have v0=ABCD v1=EFGH {AACC, BBDD, EEGG, FFHH} with 4 zip [19:56:27 CEST] <ubitux> and i need just 2 swap if i figure out the proper instruct [19:56:29 CEST] <iive> hum, no. no register copy required in zip1/2 [19:57:03 CEST] <iive> ubitux: you have to mix the 2 different vectors first [19:57:12 CEST] <iive> then mix the results with themselves [19:57:55 CEST] <ubitux> (sorry right in the end i want {AAEE, BBFF, CCGG, DDHH}) [19:58:10 CEST] <iive> yes [19:58:12 CEST] <ubitux> anyway, i'll go back in my brain instead of talking [19:58:21 CEST] <iive> and you have to get AEBF , CGDH first [19:58:37 CEST] <iive> let me write it down and i'll paste it [20:00:41 CEST] <iive> i think i'm done, should i paste it here or in pastebin? [20:00:45 CEST] <iive> or private? [20:01:02 CEST] <ubitux> doesn't matter [20:01:29 CEST] <iive> v1=ABCD [20:01:29 CEST] <iive> v2=EFGH [20:01:29 CEST] <iive> v3=zip1(v1,v2) ; {ABCD}x{EFGH}->{AEBF} [20:01:29 CEST] <iive> v4=zip2(v1,v2) ; {ABCD}x{EFGH}->{CGDH} [20:01:29 CEST] <iive> v5=zip1(v3,v3) ; {AEBF}x{AEBF}->{AAEE} [20:01:29 CEST] <iive> v6=zip2(v3,v3) ; {AEBF}x{AEBF}->{BBFF} [20:01:31 CEST] <iive> v7=zip1(v4,v4) ; {CGDH}x{CGDH}->{CCGG} [20:01:33 CEST] <iive> v8=zip2(v4,v4) ; {CGDH}x{CGDH}->{DDHH} [20:04:35 CEST] <ubitux> oh, so only with zips [20:05:18 CEST] <ubitux> let me check if it matches why original expectation [20:05:22 CEST] <ubitux> thanks [20:05:29 CEST] <iive> you are welcome :D [20:06:38 CEST] <iive> or course, v5/6/78 could reuse v1/2/3/4 [20:14:29 CEST] <BBB> ubitux: I have a hard time reproducing the update_context_from_user cases :-( [20:14:40 CEST] <thardin> running afl with a small test program. it was pretty good at finding obvious buffer overflow problems [21:01:41 CEST] <thardin> there we go, got afl-fuzz running with ffprobe on a small mxf file [21:19:29 CEST] <llogan> Compn, michaelni: I'll be unavailable to clear the ML queue from May 30-June 15 (mistakenly mentioned this in #ffmpeg). [21:20:24 CEST] <michaelni> llogan, ok, understood [21:35:06 CEST] <atomnuker> thardin: mxf? have you found something yet? [21:35:55 CEST] <thardin> only some "hangs" which only seem to be that it took a few more milliseconds than the default [21:36:38 CEST] <thardin> one case perhaps interesting, takes about 100 ms [21:37:22 CEST] <wm4> you should fuzz subtitle parsers lol [21:37:28 CEST] <wm4> I heard it's all the rage these days [21:38:54 CEST] <atomnuker> interesting, I thought mxf was pretty bad [21:39:18 CEST] <wm4> it just means the code isn't too bad, not the format [21:41:00 CEST] <thardin> lol yes [21:41:21 CEST] <cone-977> ffmpeg 03Michael Niedermayer 07master:89325417e7b3: avcodec/jpeg2000dec: Check tile offsets [21:41:22 CEST] <cone-977> ffmpeg 03Michael Niedermayer 07master:bce362d36ccf: avcodec/wavpack: Fix runtime error: left shift of negative value -1 [21:41:23 CEST] <cone-977> ffmpeg 03Michael Niedermayer 07master:cfec0d647525: avcodec/cllc: Check bitstream end before decoding pixels [21:41:35 CEST] <thardin> the format is fairly nasty, but at the same time maybe that encourages defensive coding [21:42:14 CEST] <thardin> don't think I have any file with subs [21:42:28 CEST] <thardin> I guess mkv would be a candidate [21:42:36 CEST] <thardin> I'll have to.. acquire some [21:44:05 CEST] <atomnuker> ubitux did some subtitle fuzzing not long ago btw [21:44:40 CEST] <atomnuker> he also has the entire opensubtitles database before the site went down [21:48:24 CEST] <thardin> I suppose the thing to do is convert a file with subs to rawvideo [21:49:01 CEST] <Compn> llogan : k buddy [21:49:48 CEST] <Compn> wm4 : so was the problem with opensubtitle plugin or everybodys glue code ? [21:50:01 CEST] <Compn> wm4 : also why not opensubtitle just sanitize these subtitle files [21:50:25 CEST] <wm4> 4 different bugs or something [21:50:27 CEST] <Compn> someone came into #mplayer asking about it , we had no idea haha [21:50:31 CEST] <wm4> in a sensationalistic package [21:51:00 CEST] <wm4> atomnuker: I'm pretty sure the site still exists? [21:51:38 CEST] <Compn> opensubtitles.org still up [21:51:51 CEST] Action: Compn wants kloofy to come back tho [21:58:51 CEST] <atomnuker> huh, I was sure it was "down" a few weeks ago because it had several million dmca requests and was useless because it complied [22:09:34 CEST] <RiCON> atomnuker: could you rebase your psyacoustics patch? i wanted to do a windows build to someone asking to test it in hydrogenaudio [22:10:28 CEST] <atomnuker> ok, hold on, I made some improvements to transient detection [22:12:28 CEST] <thardin> "sub2video: non-bitmap subtitle" [22:12:36 CEST] <thardin> do I need to configure with freetype? [22:15:14 CEST] <thardin> giving --enable-libfontconfig --enable-libfreetype a go [22:18:59 CEST] <thardin> nope [22:20:13 CEST] <wm4> lol [22:20:23 CEST] <wm4> welcome to the funny world of ffmpeg.c [22:20:37 CEST] <wm4> you need to pass the subtitle file to vf_sub I think [22:21:09 CEST] <thardin> trac wiki claims -vf subtitles=video.mkv [22:22:48 CEST] <wm4> yes [22:23:11 CEST] <thardin> no "subtitles" filter [22:23:16 CEST] <atomnuker> RiCON: https://pars.ee/temp/opuspsy_v3.patch [22:23:46 CEST] <RiCON> thanks [22:23:47 CEST] <atomnuker> only does transient detection, dual stereo search (useless for anything below 400kbps) and intensity stereo search (pretty good) [22:25:18 CEST] <atomnuker> it'll be much better once I complete my evil plan to have a lambda value which maps to band boosts [22:26:42 CEST] <atomnuker> RiCON: is that for Lumitopia? I hadn't checked the thread since I posted last month [22:26:48 CEST] <RiCON> atomnuker: yeah [22:28:18 CEST] <pfelt> is the person that is the moderator for the -devel mailing list on here? [22:28:33 CEST] <RiCON> it's llogan, i think? [22:30:44 CEST] <pfelt> oh yah know what, i just looked at the archives and figured out what happend. [22:30:47 CEST] <pfelt> nm. [22:44:56 CEST] <thardin> --enable-filter=subtitles [22:48:13 CEST] <thardin> didn't work :[ [22:49:30 CEST] <thardin> aha --enable-libass needed. it should complain about that [23:11:23 CEST] <wm4> does anyone know about the libopus encoder? someone claims it's very slow on lesser hardware [23:14:36 CEST] <atomnuker> wm4: not possible, the thing runs on mmu-less microprocessors [23:15:30 CEST] <TD-Linux> wm4, there's #opus (or me or atomnuker) [23:15:46 CEST] <wm4> I mean encoder, not decoder [23:16:12 CEST] <TD-Linux> no, the encoder is pretty fast too. you can run 1000 of them on a desktop in realtime [23:17:08 CEST] <wm4> hm I guess I'll have to take a closer look then [23:19:54 CEST] <TD-Linux> (note it does have speed levels, and it defaults the slowest which is about 100x realtime on my laptop, but you can set it faster if you want) [23:20:31 CEST] <thardin> managed to get a 4.4k mkv with subs to fuzz [00:00:00 CEST] --- Thu May 25 2017

1 0

ffmpeg.log.20170524
by burek 25 May '17

25 May '17

[00:09:52 CEST] <asmand> And finally I find some information, it seems that there is an issue in VLC actually: https://trac.videolan.org/vlc/ticket/14577 [01:18:42 CEST] <x86iac> any one know a bit more of this subtitle vulnerability - how does a text file (srt) cause anything? [01:29:02 CEST] <x86iac> to embarrassed? [01:59:59 CEST] <marquisor> hi [07:50:06 CEST] <buu> So, uh, how do I send commands to a running ffmpeg that is streaming output? [07:50:14 CEST] <buu> Specifically stop/start etc [12:42:57 CEST] <anchi> Hello, I want to know how can I hold the last frame of the video for 10 more seconds? [12:58:39 CEST] <SolidusAbi> Hi, can someone help me with smooth streaming? Im trying to use it but when i try to play the player say me "cannot find prueba.ism/Manifest" [13:00:45 CEST] <SolidusAbi> i used the ismindex tool [13:36:23 CEST] <chriss4349347293> Hi everyone, I'm searching for help with ffmpeg: [13:37:24 CEST] <chriss4349347293> I have a url to an jpeg image and I want ffmpeg to query it every second and make video out of it. [13:37:41 CEST] <chriss4349347293> what I tried: [13:38:01 CEST] <chriss4349347293> ffmpeg -hide_banner -loop 1 -analyzeduration 0 -r 1 -i "{url}" -an -copyts -f mp4 capture.mp4 [13:38:16 CEST] <colde> I doubt you will get ffmpeg to query it [13:38:23 CEST] <colde> you would likely have to query and download it yourself [13:38:24 CEST] <chriss4349347293> ah, ok [13:38:28 CEST] <colde> and then pipe that sequence into ffmpeg [13:38:29 CEST] <colde> iirc [13:38:32 CEST] <BtbN> while true; do curl ...; sleep 1; done | ffmpeg -f image2pipe -i - ... [13:38:41 CEST] <colde> or something like that [13:39:07 CEST] <chriss4349347293> ok, that would have been my next idea. just wanted to know, if ffmpeg can do by itself [13:39:17 CEST] <chriss4349347293> will give it a try with pipe [13:39:19 CEST] <colde> Not that I am aware of [13:49:32 CEST] <chriss4349347293> @colde, @BrbN : piping works!! thx for help [15:48:19 CEST] <mindw0rk> Hi everyone, I'm starting to use libav with the goal of grabbing a single frame from a live hls feed. Thought I ran into a problem that I can't get out of. [15:48:53 CEST] <mindw0rk> using avformat_open_input comes back with a "Invalid data found when processing input [15:54:17 CEST] <mindw0rk> "ffmpeg -i (same m3u8 url) -f image2 -vframes 1 snapshot.png" work great on the other hand. [15:57:59 CEST] <marquisor> i'm looking for a solution to just forward a DVB HD .ts (MP4, AC3) to a server (relay) and leave it as is (copy) to provide it for multiple clients. i read in bugreports ffserver is obsolete? and the copy (passthrough) is not supported at all? so what can i do in my scenario? [17:07:37 CEST] <mindw0rk> derp, nevermind. I forgot to call av_register_all [17:07:49 CEST] <SolidusAbi> Can someone help me with ismv example in ffmpeg? [17:08:02 CEST] <SolidusAbi> im trying to create a ism but it always fail [19:10:54 CEST] <Thomas> Hello :) [19:11:12 CEST] <Guest22458> I have a quick question about ffmpeg and hardware acceleration [19:11:59 CEST] <Guest22458> I compiled ffmpeg with options : --enable-gpl --enable-libx264 --enable-mmal --enable-omx-rpi --enable-omx --enable-nonfree --enable-libxcb [19:12:22 CEST] <Guest22458> So I can use h264_omx for hardware accelerated encoding [19:13:15 CEST] <Guest22458> There is no problem during compilation or execution, but I cannot see any difference in terms of performance between using standard x264 encoding (software) and supposedly-hardware-accelerated h264_omx. [19:14:15 CEST] <Guest22458> As my final objective is to use hardware acceleration to lighten CPU usage, can you help me please ? I use a raspberry pi zero [19:15:00 CEST] <furq> what's the command you're running [19:16:00 CEST] <Guest22458> ffmpeg -thread_queue_size 512 -f x11grab -vsync 0 -s 640x360 -r 2 -i :0.0+0,0 -thread_queue_size 512 -f alsa -async 1 -thread_queue_size 512 -ac 2 -channel_layout stereo -i pulse -ab 180k -c:v h264_omx -b:v 500k -bufsize 4000k -c:a aac -strict experimental -b:a 32k -ar 44100 -ac 2 -f flv rtmp://live.twitch.tv/app/ [19:16:09 CEST] <Guest22458> (Of course I deleted my Twitch key) [19:16:22 CEST] <furq> yeah you're probably cpu bound from doing the pixel format conversion and encoding aac [19:16:47 CEST] <Guest22458> oh :( [19:16:47 CEST] <JEEB> yup, armv6 and aac encoding is slow as beep [19:16:55 CEST] <JEEB> like, almost anything CPU intensive [19:16:56 CEST] <JEEB> will die [19:17:17 CEST] <JEEB> also what in that config requires nonfree? [19:17:38 CEST] <JEEB> omx-rpi? [19:17:40 CEST] <furq> nothing [19:17:43 CEST] <JEEB> yea [19:17:46 CEST] <JEEB> kind of wondered [19:17:49 CEST] <Guest22458> just added it in-case [19:17:52 CEST] <JEEB> lol [19:17:56 CEST] <Guest22458> :p [19:18:05 CEST] <JEEB> it's only needed for very specific libraries [19:18:10 CEST] <JEEB> not 1st party code [19:18:18 CEST] <Guest22458> Maybe I can disable sound and make an other test [19:18:19 CEST] <furq> fwiw you should see a massive difference between omx and x264 for that particular command [19:18:56 CEST] <kepstin> particularly on the pi zero, which is only signle core [19:18:57 CEST] <JEEB> also don't scale [19:19:10 CEST] <JEEB> also x11grab? [19:19:13 CEST] <Guest22458> ok [19:19:15 CEST] <JEEB> is that fast enough by itself? [19:19:24 CEST] <furq> he's not scaling, that's an input option [19:19:30 CEST] <JEEB> ... [19:19:31 CEST] <JEEB> fuck [19:19:34 CEST] <Guest22458> I need to record a window, and don't know any other input method [19:19:37 CEST] <furq> actually what does that do with x11grab [19:19:46 CEST] <JEEB> I mean, I know that it means size [19:19:53 CEST] <JEEB> but it is linked to vf scale [19:19:56 CEST] <JEEB> afetr -i [19:20:22 CEST] <kepstin> I think ffmpeg cli has some magic mapping so -s goes to the -video_size option on inputs? [19:20:31 CEST] <kepstin> i don't recall exactly. [19:20:36 CEST] <furq> well yeah but what does that do with x11grab [19:20:36 CEST] <JEEB> Guest22458: well first of all see how fast just the input and after -i you just put -f null afterwards [19:20:49 CEST] <JEEB> input options -i input -f null - [19:20:54 CEST] <furq> i assume something is having to scale it [19:20:58 CEST] <kepstin> well, x11grab accepts a size for the capture area with the -video_size avoption [19:21:08 CEST] <furq> oh right [19:21:09 CEST] <furq> nvm then [19:21:17 CEST] <marquisor> anyone with -acodec copy -vcodec copy passthrough ffserver? [19:22:05 CEST] <furq> i don't think there's anyone with ffserver [19:22:15 CEST] <marquisor> :< [19:22:19 CEST] <Guest22458> JEEB : ok [19:22:44 CEST] <marquisor> furq: alternatives to ffserver accepting ffmpeg input? [19:22:57 CEST] <furq> for what use case [19:23:18 CEST] <marquisor> i repeat: [19:23:22 CEST] <marquisor> i'm looking for a solution to just forward a DVB HD .ts (MP4, AC3) to a server (relay) and leave it as is (copy) to provide it for multiple clients. i read in bugreports ffserver is obsolete? and the copy (passthrough) is not supported at all? so what can i do in my scenario? [19:23:45 CEST] <furq> oh [19:23:50 CEST] <furq> shame it's ac-3 or you could just use rtmp [19:24:05 CEST] <furq> that's the only relatively painless streaming setup i know about [19:24:14 CEST] <marquisor> something like ffmpeg -i http://ip:port/stream.ts -acodec copy -vcodec copy http://server:port/feed.ffm [19:24:33 CEST] <furq> if you don't mind encoding the audio then rtmp is easy [19:24:41 CEST] <furq> as long as by mp4 you mean h264 [19:24:46 CEST] <marquisor> yep h264 [19:24:53 CEST] <marquisor> DVB-S2 [19:25:02 CEST] <furq> yeah rtmp is h264/aac [19:25:02 CEST] <Guest22458> JEEB : This way correct ? [19:25:04 CEST] <Guest22458> ffmpeg -thread_queue_size 512 -f x11grab -vsync 0 -s 640x360 -r 2 -i :0.0+0,0 -f null -thread_queue_size 512 -c:v h264_omx -b:v 500k -bufsize 4000k -f flv rtmp://live.twitch.tv/app/ [19:25:09 CEST] <JEEB> marquisor: is this for an internal or external network? [19:25:11 CEST] <marquisor> direct stream from my settop box satellite receiver [19:25:14 CEST] <marquisor> external [19:25:24 CEST] <JEEB> ok, so no multicast with UDP :) [19:26:04 CEST] <furq> Guest22458: check your cpu usage [19:26:15 CEST] <furq> that'll still be doing a pixel format conversion, which there's not really any way around with x11grab [19:26:46 CEST] <JEEB> marquisor: so what are the clients of this thing? [19:26:53 CEST] <JEEB> what would be playing these streams? [19:27:17 CEST] <marquisor> some different users, for private viewing [19:27:31 CEST] <JEEB> well the thing I was pointing is do you need to support browsers [19:27:43 CEST] <JEEB> or can the clients utilize some OSS player [19:27:45 CEST] <marquisor> they all connect to it via mediaplayer f.e. vlc [19:27:48 CEST] <JEEB> ok [19:28:31 CEST] <marquisor> it's just to save bandwidth and keep the source "secret" [19:28:34 CEST] <JEEB> marquisor: so what you need is something to give out the input stream as-is through HTTP or something [19:28:40 CEST] <Guest22458> I may wait a few minutes since I need to recompile ffmpeg (I recompile without the nonfree option) [19:28:41 CEST] <marquisor> like on this picture https://trac.ffmpeg.org/attachment/wiki/ffserver/ffserver_map.png [19:28:48 CEST] <JEEB> uhh, you won't save bandwidth [19:28:54 CEST] <JEEB> if you don't re-encode the stream [19:28:57 CEST] <furq> he will if the server is remote [19:29:04 CEST] <marquisor> yes server is remote [19:29:06 CEST] <JEEB> well yea it depends *where* he means [19:29:25 CEST] <furq> i assume he doesn't want all these people connecting to his house [19:29:29 CEST] <marquisor> yep [19:29:32 CEST] <marquisor> correct, furq [19:29:37 CEST] <JEEB> anyways, vlc is one of those things that can serve stuff over http [19:29:46 CEST] <marquisor> ok [19:29:48 CEST] <JEEB> so you could pass the stream to a vlc on the server [19:29:53 CEST] <marquisor> i considered vlc next. [19:29:55 CEST] <JEEB> and then you figure out the sout stuff [19:30:02 CEST] <JEEB> to output as-is without re-encoding [19:30:06 CEST] <JEEB> as http [19:30:09 CEST] <furq> i was really hoping you were going to name something better than vlc and http [19:30:19 CEST] <JEEB> well if he wanst the full DVB stream as-is [19:30:20 CEST] <marquisor> :) [19:30:25 CEST] <furq> i'd love an rtsp server or something which is as easy to work with as nginx-rtmp [19:30:42 CEST] <JEEB> and has no requirement for web browser spport [19:30:50 CEST] <JEEB> also man, nginx-rtmp brings out some weirdos [19:30:51 CEST] <marquisor> JEEB: sounds good [19:31:00 CEST] <furq> well yeah [19:31:16 CEST] <furq> marquisor: https://github.com/arut/nginx-rtmp-module [19:31:18 CEST] <JEEB> marquisor: you can hack the input through an ssh tunnel or something else [19:31:23 CEST] <furq> if you don't mind reencoding the audio then that's pretty good [19:31:29 CEST] <marquisor> JEEB: aaah read about that, yes [19:31:33 CEST] <furq> and meets all your other requirements [19:31:44 CEST] <marquisor> furq: i can spare cpu power for reencoding audio [19:31:45 CEST] <JEEB> ok, the main point being does he want to change the stream at all [19:31:50 CEST] <JEEB> *wants [19:31:54 CEST] <marquisor> but for video it'd be critical [19:31:58 CEST] <furq> it'd pretty much just be ffmpeg -i http://foo -c:v copy -c:a aac -b:a 128k rtmp://... [19:32:37 CEST] <JEEB> furq: there was this chinese spammer who herped a derp when I told him to make things take something else (not owned by adobe) in when he asked FFmpeg to hack HEVC into FLV unofficially [19:32:48 CEST] <furq> sounds like fun [19:32:52 CEST] <marquisor> you two got two reasonable solutions :) [19:32:54 CEST] <JEEB> because of nginx-rtmp most likely [19:33:00 CEST] <furq> yeah [19:33:01 CEST] <marquisor> gonna try it like that [19:33:11 CEST] <furq> even if ffmpeg had done that, what does he think clients will make of it [19:33:36 CEST] <furq> or servers for that matter [19:33:43 CEST] <marquisor> i have to ask, that rtmp:// doesn't need a running ffserver then? [19:33:45 CEST] <JEEB> yea, not sure how much the servers actually parse it [19:33:47 CEST] <marquisor> both solutions via ssh tunnel? [19:33:48 CEST] <furq> no [19:33:52 CEST] <marquisor> ah ok [19:34:02 CEST] <JEEB> marquisor: it needs a TCP connection to nginx with the nginx-rtmp module ocnfigured [19:34:08 CEST] <marquisor> kk [19:34:10 CEST] <marquisor> i see [19:34:11 CEST] <JEEB> the nginx will then serve whatever you configured [19:34:14 CEST] <furq> yeah you don't need an ssh tunnel for nginx [19:34:29 CEST] <kepstin> obviously you can run it through ssh or vpn or whatever if you want to tho. [19:34:32 CEST] <JEEB> well, you can still do it to further secure the pushing [19:34:40 CEST] <JEEB> vpn or port forward [19:34:41 CEST] <marquisor> yep [19:34:41 CEST] <furq> you could [19:34:43 CEST] <marquisor> ok [19:34:52 CEST] <marquisor> but won't need. it's not a permanent setup [19:34:54 CEST] <furq> although you can just deny publish all and whitelist your ip [19:35:08 CEST] <furq> or have some hacky password auth like i have [19:35:26 CEST] <marquisor> just for watching 24h Race on this weekend ;) [19:35:42 CEST] <furq> is it the vln this weekend [19:35:50 CEST] <marquisor> Nürburgring [19:35:52 CEST] <furq> yeah [19:35:54 CEST] <marquisor> Germany [19:36:53 CEST] <furq> oh it's a different 24 hour race [19:37:01 CEST] <furq> i forgot they have more than one [19:37:11 CEST] <furq> i'll be busy with monaco and indy anyway ;_; [19:37:24 CEST] <marquisor> http://www.nuerburgring.de/en/events-tickets/automobile/adac-zurich-24-hour… [19:38:06 CEST] <marquisor> tomorrow qualifying, first general rehearsal for the stream planned... [19:38:15 CEST] <furq> http://www.nuerburgring.de/en/events-tickets/automobile/vln.html [19:38:17 CEST] <furq> i was thinking of that one [19:39:41 CEST] <furq> sounds better than the wec race that only goes round the go-kart track anyway [19:41:28 CEST] <Guest22458> (still compiling) [19:42:11 CEST] <marquisor> hehe it's full track [19:42:24 CEST] <marquisor> and always a (the) big event in germany [20:08:47 CEST] <Guest22458> (still compiling) [20:13:20 CEST] <lystra> Anyone know if ffmpeg can hard burn more than one subtitle, one on top of the other to display multiple languages? [20:15:32 CEST] <Guest22458> I'm back [20:15:55 CEST] <Guest22458> I was wondering : What about using MP3 and not AAC encoder ? Maybe it will be ligther [20:16:12 CEST] <Guest22458> Lystra : I don't know [20:35:10 CEST] <furq> Guest22458: probably [20:35:24 CEST] <furq> you might want to try with fdk-aac or libshine, those are both optimised for mobile [20:35:41 CEST] <furq> idk if they'll be faster than lame on a pi zero though [21:02:23 CEST] <ThugAim> Hey guys, been trying to concat several files but it would usually show up 1000 frames duplicated error or DTS out of order error. Converting and joining .mpeg and .mxf files with libx264 and a watermark [21:03:07 CEST] <BtbN> with -f concat you can't concat arbitrary files [21:04:18 CEST] <marquisor> can you encode them to "one format" first and after then concat? [21:07:35 CEST] <llogan> michaelni: i'll be unable to clear the ML queue from May 30-June 15. [21:08:53 CEST] <BtbN> marquisor, yes [21:11:58 CEST] <ThugAim> :( [21:12:27 CEST] <ThugAim> thanks BtbN and marquisor ... is there hope for this function in the future? [21:12:59 CEST] <ThugAim> also, how can I specify the watermark to come in at specific times? [21:13:02 CEST] <BtbN> no, the concat demuxer is litterally just cat filea.mpeg fileb.mxf > inputfile.something [21:13:03 CEST] <marquisor> you could write a quite simple shell script for that [21:13:18 CEST] <BtbN> Use the concat filter if you want to concat arbitrary stuff [21:14:04 CEST] <ThugAim> I'll do some reading up and testing, you'll know if I've failed enough times. :D [21:14:59 CEST] <furq> ThugAim: if you're reencoding anyway then -i foo.mxf -i bar.mpg -filter_complex concat=n=2:v=1:a=1 [21:15:36 CEST] <furq> https://ffmpeg.org/ffmpeg-filters.html#Timeline-editing [21:15:39 CEST] <furq> and you can use that with overlay [21:19:05 CEST] <ThugAim> good info. will report back much later. [21:25:06 CEST] <CesarS> Hello, I've started having an issue where a call to avformat_write_header would rewrite my stream time_base. I think this is now by design, but unfortunately it broke the code I had before. Now my videos seem to have a crazy framerate like 1180 and I am not sure how to fix this... Would anyone try to help? [21:25:43 CEST] <CesarS> I think I have to update the way I was computing the pts for each frame, but I am not sure how [22:16:10 CEST] <Guest22458> I'm back [22:16:27 CEST] <Guest22458> So I'm on raspberry pi zero, I use hardware accelerated h264_omx [22:16:29 CEST] <Guest22458> ffmpeg -f x11grab -vsync 0 -i :0.0+0,0 -f null -c:v h264_omx -b:v 500k -bufsize 4000k -f flv rtmp://live.twitch.tv/app/live_157451802_XSl9Z3Z9wfoHstXOwBvoco3YU1eyaP [22:18:56 CEST] <Guest22458> (Twitch key is fake of course) [22:19:44 CEST] <llogan> Guest22458: no -maxrate? what's the "-f null" for? [22:19:52 CEST] <ChocolateArmpits> Guest22458, are you sure the encoder initializes vbv without maxrate specified ? [22:20:01 CEST] <furq> do -maxrate and -bufsize even work with h264_omx [22:20:07 CEST] <furq> also is there a question [22:20:08 CEST] <Guest22458> It"s just a command because CPU is overloaded, even with this test [22:20:22 CEST] <llogan> furq: add -g to that list [22:20:29 CEST] <furq> i think -g does [22:20:49 CEST] <Guest22458> How can I optimize this command ? [22:21:58 CEST] <furq> i'm not sure if there's much you can do [22:22:05 CEST] <Guest22458> :( [22:22:12 CEST] <furq> the only thing there that hits the cpu is the pixel format conversion [22:22:19 CEST] <furq> and there's no way around that with x11grab afaik [22:23:45 CEST] <Guest22458> ./ffmpeg -f x11grab -s 320x240 -r 5 -vsync 0 -i :0.0+0,0 -f null -c:v h264_omx -b:v 200k -bufsize 4000k -f flv test.flv [22:23:56 CEST] <Guest22458> With this command I get a really low CPU usage [22:24:12 CEST] <furq> yeah that seems right [22:24:27 CEST] <Guest22458> The parameter -g may be to 0 for a lower CPU conso ? [22:24:32 CEST] <furq> no [22:25:12 CEST] <furq> frame size and framerate is about all you can do afaik [22:25:39 CEST] <furq> also that "-f null" is doing nothing [22:26:15 CEST] <Guest22458> WTF There is no MP3 codec in audio [22:26:35 CEST] <furq> there's no built-in mp3 encoder [22:26:39 CEST] <furq> you need to build with libmp3lame or libshine [22:27:19 CEST] <Guest22458> ok [22:27:36 CEST] <Guest22458> I suppose the option is --enable-libmp3lame ? [22:27:54 CEST] <furq> yeah [22:27:58 CEST] <furq> libshine might be better for a weak cpu [22:28:04 CEST] <furq> no harm in doing both [22:32:12 CEST] <Ozgr> hi, which metadata keys are supported for webm container by ffmpeg? i have found this information: https://wiki.multimedia.cx/index.php/FFmpeg_Metadata but webm is not listed. i thought that those supported by matrsoka could also be supported by webm, since it is derived from matroska but only title key is written on the file, not the description. thanks [22:39:55 CEST] <Guest22458> Thanks [23:06:53 CEST] <Mavrik> hmm, dafuq, new x264 is all over the place when it comes to output bitrate [23:10:43 CEST] <BtbN> in cbr mode you mean? [23:11:17 CEST] <kepstin> it had better not be all over the place in cbr mode. or vbv mode. or even vbr mode when averaged over longer times. [23:11:44 CEST] <kepstin> now, crf mode, isn't the intent that the bitrate will be all over the place? :) [23:13:10 CEST] <BtbN> even in cbr/vbv mode, it can be all over the place, as long as it does not use more than it should. x264 does not support a minrate [23:13:14 CEST] <BtbN> at best you can turn on padding [00:00:00 CEST] --- Thu May 25 2017

1 0

ffmpeg-devel.log.20170523
by burek 24 May '17

24 May '17

[03:12:46 CEST] <rcombs> does anyone have up-to-date listening test results of AppleAAC vs ffAAC? [03:26:37 CEST] <Compn> rcombs : did apple change their encoder ? [03:27:01 CEST] <Compn> or you mean a new test after atomnuker's changes [03:27:09 CEST] <Compn> hmm i am not sure if double blinds have been done after that [03:27:17 CEST] <Compn> does hydrogenaudio handle those? [03:27:53 CEST] <Compn> http://wiki.hydrogenaud.io/index.php?title=Hydrogenaudio_Listening_Tests#AA… [03:28:00 CEST] <Compn> doesnt look like they've done any tests in a while [03:28:06 CEST] <Compn> you may want to ask them to do an updated one [03:29:03 CEST] <Compn> https://hydrogenaud.io/index.php/topic,111085.0.html [03:29:10 CEST] <Compn> oh they are more active on the forum than the wiki [03:29:14 CEST] <Compn> maybe atomnuker knows too [03:32:14 CEST] <Compn> that thread is just one guys' audio tests on his own files [03:32:30 CEST] <Compn> vs fdk aac [03:35:17 CEST] <JEEB> kamedo was the guy who also posted tests on the trac [03:36:09 CEST] <JEEB> while the patch set was still in development. but that's just fdk-aac and ffaac. and fdk-aac can't be distributed as binary [03:38:19 CEST] <rcombs> Compn: yeah I meant after atomnuker's changes [03:38:35 CEST] <rcombs> the only stuff I've seen is from partway through development [03:39:22 CEST] <rcombs> also, has anyone ever compared with Microsoft's encoder? [04:06:55 CEST] <cone-541> ffmpeg 03Michael Niedermayer 07master:cfd1ecdc0bb0: avcodec/asvdec: Check buf_size [04:06:55 CEST] <cone-541> ffmpeg 03Michael Niedermayer 07master:2002436b0c91: avcodec/xsubdec: Check that RLE coded image and colors fit in the buffer [04:06:55 CEST] <cone-541> ffmpeg 03Michael Niedermayer 07master:9726e9f80934: avcodec/ffv1dec_template: Fix runtime error: signed integer overflow: 202 + 2147483615 cannot be represented in type 'int' [04:06:55 CEST] <cone-541> ffmpeg 03Michael Niedermayer 07master:504d5804ac53: avcodec/g723_1: Fix runtime error: signed integer overflow: -1013481472 + -1139123755 cannot be represented in type 'int' [04:40:28 CEST] Action: Compn just hopes aac goes away one day [04:40:35 CEST] <Compn> like ogg and ra [07:44:37 CEST] <thebombzen> fdk-aac isn't nonfree, right? [07:44:44 CEST] <thebombzen> I thought its license was just GPL-incompatible [07:58:23 CEST] <rcombs> depends if you define "nonfree" as "GPL-incompatible" or not [09:20:55 CEST] <thebombzen> well I mean you can distribute binaries of fdk-aac [09:21:01 CEST] <thebombzen> just not ffmpeg binaries linked to it [09:21:22 CEST] <thebombzen> which makes it basically the same if you want to use ffmpeg, but slightly different in the US where laws are weird and bad [09:41:25 CEST] <thebombzen> Also, a bit overdue, but I made the SSH crash ticket. I listed is as "normal priority" because I'm unsure what the standards are for higher or lower priority [09:43:29 CEST] <thebombzen> are they published anywhere or do they play it by ear [10:14:05 CEST] <ubitux> so i got no answer for the dashenc thing [10:14:18 CEST] <ubitux> who maintains our dash stack? [10:15:33 CEST] <JEEB> wbs I think created it originally [10:17:03 CEST] <BtbN> Is there a hw pixel format that's an OpenGL texture? [10:17:30 CEST] <wm4> BtbN: no [11:12:12 CEST] <BtbN> What happened to the libav effort to make cuvid a classic hwaccel? [11:12:25 CEST] <nevcairiel> it still exists somewhere [11:12:56 CEST] <BtbN> Cause I'm thinking about doing that myself otherwise [11:15:18 CEST] <nevcairiel> its mostly done just has a few open questions [11:15:28 CEST] <nevcairiel> like, how to properly implement the delay so its not crawling slow [11:15:49 CEST] <wm4> BtbN: you should probably get in touch with elenril [11:48:00 CEST] <cone-588> ffmpeg 03Timo Rothenpieler 07master:a1652aca7e89: avcodec/nvenc: remove unnecessary alignment [11:58:46 CEST] <cone-588> ffmpeg 03Timo Rothenpieler 07release/3.3:3bc5e427e4da: avcodec/nvenc: remove unnecessary alignment [11:58:53 CEST] <wm4> do transport streams really not store a framerate? [12:00:25 CEST] <nevcairiel> of course not, they store no metadata [12:00:34 CEST] <nevcairiel> the codec might, of course [12:00:56 CEST] <cone-588> ffmpeg 03Timo Rothenpieler 07release/3.2:1f76235dd450: avcodec/nvenc: remove unnecessary alignment [12:01:22 CEST] <wm4> trying to make utils.c detect the framerate of a .ts file correctly (mediainfo gets it right, of course) [12:01:59 CEST] <nevcairiel> there is the timestamp-based heuristic which often works ok-ish, but of course TS and its 90kHz timestamps distort that as well to some degree [12:04:40 CEST] <cone-588> ffmpeg 03Max Justicz 07master:3766aa7343c4: avcodec/fmvc: Fix use of uninitialized memory when the first frame is not a keyframe [12:04:54 CEST] <cone-588> ffmpeg 03Timo Rothenpieler 07release/3.1:8c021166d194: avcodec/nvenc: remove unnecessary alignment [12:06:01 CEST] <wm4> who pushed 3766aa7343c4? [12:06:04 CEST] <wm4> it was not on the ML [12:06:44 CEST] <nevcairiel> git stores the commiter in the commit info [12:07:29 CEST] <wm4> something similar was reported to ffmpeg-security, but no patch [12:07:40 CEST] <wm4> so much for "patches should be submitted to the mailing list" [12:07:52 CEST] Action: wm4 adds another entry to his list of ffmpeg development bullshit [12:13:49 CEST] <durandal_1707> wm4: ffmpeg development is da best [12:19:01 CEST] <atomnuker> trivial commits like that don't need to go through the ml [12:19:56 CEST] <atomnuker> I too have been sent random one-line patches before [12:49:21 CEST] <kierank> wm4: only reliable way is from the codec [12:49:50 CEST] <wm4> is that generally signaled in stuff muxed into ts? [12:50:06 CEST] <kierank> in the main yes [12:52:59 CEST] <kierank> I've seen evil ateme files set the time_scale (note not framerate) to 1/27000000 [12:53:04 CEST] <kierank> but that was many years ago [12:53:32 CEST] <wm4> well this has to work with the cesspool that is fate+av_find_streaminfdo [13:20:41 CEST] <wm4> I can't comprehend this - so it checks against a list of "standard framerates", and considers 24.5 fps "better" than 25 fps? [13:22:06 CEST] <wm4> ah no, I made some mistake here [13:28:46 CEST] <cone-588> ffmpeg 03Matthieu Bouron 07master:224bb46fb857: lavc/mediacodec_wrapper: fix local reference leaks [13:28:46 CEST] <cone-588> ffmpeg 03Matthieu Bouron 07master:fb3228bee8fe: lavc/mediacodec_wrapper: do not declare JNIAMedia{Codec,CodecList,Format}Fields on the stack [13:28:48 CEST] <cone-588> ffmpeg 03Matthieu Bouron 07master:37de7f71758b: lavc/ffjni: add missing '\n' [13:34:33 CEST] <ubitux> git.videolan.org down? [13:35:03 CEST] <mateo`> did I just break it ? [13:35:41 CEST] <wm4> fate passes on this pig [13:51:26 CEST] <cone-588> ffmpeg 03Matthieu Bouron 07release/3.3:2fb25e2dd6ff: lavc/mediacodec_wrapper: fix local reference leaks [13:51:27 CEST] <cone-588> ffmpeg 03Matthieu Bouron 07release/3.3:cbae648eb8b2: lavc/mediacodec_wrapper: do not declare JNIAMedia{Codec,CodecList,Format}Fields on the stack [13:51:28 CEST] <cone-588> ffmpeg 03Matthieu Bouron 07release/3.3:3e38bf95c537: lavc/ffjni: add missing '\n' [13:54:35 CEST] <cone-588> ffmpeg 03Matthieu Bouron 07release/3.3:6ee4b20f4ae3: lavf/mov: make invalid m{d,v}hd time_scale default to 1 instead of erroring out [15:43:34 CEST] <BBB> J_Darnley: happy to assist with idct if you need help :) [15:44:11 CEST] <J_Darnley> Can you help right now? [16:06:42 CEST] <J_Darnley> Well, maybe later. See ya'll later [16:25:40 CEST] <cone-588> ffmpeg 03Michael Niedermayer 07master:620b452a118a: avcodec/aacdec_fixed: Fix multiple runtime error: shift exponent 127 is too large for 32-bit type 'int' [16:25:41 CEST] <cone-588> ffmpeg 03Michael Niedermayer 07master:948b54763b6c: avcodec/lagarith: Fix runtime error: left shift of 1 by 31 places cannot be represented in type 'int' [16:25:42 CEST] <cone-588> ffmpeg 03Michael Niedermayer 07master:fb75ad79cb8a: avcodec/celp_filters: Fix runtime error: signed integer overflow: 1892453989 + 381702783 cannot be represented in type 'int' [17:04:12 CEST] <BBB> J_Darnley: Im here, sorry [17:04:13 CEST] <BBB> ask [17:13:51 CEST] <BBB> J_Darnley: also feel free to email to the list and I can respond there [17:30:15 CEST] <ubitux> in the ps aac function, is it ok to assume n > 4 all the time? [17:30:47 CEST] <ubitux> it seems currently optimized ASM functions do assume that [17:31:27 CEST] <ubitux> (note: i'm writing some aarch64 code for this stuff and i'm wondering of the assumptions i can make) [17:35:06 CEST] <KGB> [13FFV1] 15dericed opened pull request #66: Definitions (06master...06definitions) 02https://git.io/vHt47 [17:35:06 CEST] <jamrial> ubitux: afaik, only ps_stereo_interpolate needs to check n [17:36:55 CEST] <ubitux> ok [17:36:58 CEST] <jamrial> stereo_interpolate, rather [17:40:35 CEST] <jamrial> btw, if you find a sample that sets ps->enable_ipdopd, send it my way [17:47:01 CEST] <ubitux> yeah i saw that it's optimized nowhere [17:51:26 CEST] <michaelni> wm4, the change 3766aa7343c4 was on ffmpeg-security in the form of english text posted by its author [19:31:39 CEST] <peloverde> jamrial: al_sbr_ps_04_new.mp4 [19:36:13 CEST] <jamrial> peloverde: cool, thanks! [19:37:45 CEST] <jamrial> ubitux: ^ in case you're interested [19:38:46 CEST] <peloverde> available at ftp://mpaudconf:adif2mp4@ftp.iis.fhg.de/mpeg4audio-conformance/compressedMp4/ [19:43:24 CEST] <BBB> I think I finally fixed my disconnect problem [19:43:40 CEST] <BBB> as in, I believe my computer no longer disconnects from IRC after a few minutes of inactivity [19:43:51 CEST] <BBB> \o/ [19:44:17 CEST] <BtbN> You added support for PING/PONG to your client? [19:45:48 CEST] <BBB> BtbN: no, a mac update added a new power savings feature to put my HD to sleep when the display goes to sleep (I specifically have it set up to not put network to sleep) [19:45:54 CEST] <BBB> and that disconnects the client, apparently [19:46:25 CEST] <BtbN> sounds more like it puts the whole machine into suspend? [19:46:35 CEST] <BBB> probably [19:46:43 CEST] <BBB> I dont know how to translate apple marketing speak into reality [19:46:49 CEST] <BBB> theres no literal dictionary for that conversion [19:47:09 CEST] <BtbN> just need some bouncer somewhere, so your client can freely disconnect [19:47:20 CEST] <BBB> I used to have one, but it got donged [19:47:30 CEST] <BBB> and too lazy to set up a new one [20:06:41 CEST] <jamrial> heh, love when you write an entire asm function and it works the first time you try it [20:07:22 CEST] <nevcairiel> i think thats always suspicious [20:07:53 CEST] <jamrial> yes, which is why i added a bullshit instruction so it would crash :p [20:08:01 CEST] <jamrial> to confirm it was actually running the code [20:09:25 CEST] <jamrial> anyway, 2x speed up for stereo_interpolate_ipdopd with sse3 [20:10:01 CEST] <jamrial> anyone wanting to play that sample will love the saved cycles, because i don't know if there's any other aac file out there that will [20:10:58 CEST] <wm4> adding an instruction that outputs corrupted data is better than making it crash [20:11:09 CEST] <wm4> because a crash means it's run, not necessarily that the result is used [20:11:25 CEST] <nevcairiel> well for dsp functions it kinda is [20:20:12 CEST] <ubitux> jamrial: thx, stolen :) [21:03:08 CEST] <pfelt> afternoon all. i'm looking at cleaning up some warnings in declink_dec and am having issues with av_dup_packet() and changing to av_packet_ref(). i'm positive that i'm not understanding the code, but i'm not finding any good help via google on understanding how this new call is supposed to work. [21:03:27 CEST] <pfelt> what's the best way to come up to speed on this referencing model? [21:40:46 CEST] <BBB> pfelt: whats the specific issue youre having? [21:41:38 CEST] <BBB> pfelt: its possible that declink actually requires you to copy the data to an external buffer and that using ref (which doesnt copy) would therefore change behaviour/ [21:41:39 CEST] <BBB> ? [21:50:08 CEST] <pfelt> bbb: decklink input (decklink_dec.cpp) pulls frames off the card for ingestion into ffmpeg. i'm converting the call to av_dup_packet() to av_packet_ref() per documentation. i was segfaulting in static int avpacket_queue_put(AVPacketQueue *q, AVPacket *pkt) , but i think i figured the right syntax out. [21:50:57 CEST] <pfelt> i find it a bit difficult to understand the code here to ensure that i'm soing the right thing, but it is running now without segfaulting. only question is, am i leaking memory now since i didn't add an unref() anywhere [21:51:36 CEST] <pfelt> overall the patch is pretty small [21:54:43 CEST] <BBB> if you can run it for an hour and memory usage is table [21:54:46 CEST] <BBB> youre probably fine [21:54:47 CEST] <BBB> ;) [21:54:54 CEST] <pfelt> hehe [21:55:09 CEST] <pfelt> yeah. i've got a stream pulling from one DL card and outputting on another [21:55:55 CEST] <pfelt> with av_dup_packet() who was supposed to do a free on the packet? [21:56:13 CEST] <pfelt> or i guess, what funcitons could be used to free a packet that has been "duped" [22:00:04 CEST] <BBB> I think the input device is supposed to return a reference [22:00:09 CEST] <BBB> and then whatever called will unref [22:00:14 CEST] <BBB> so if you return a reference thats fine [22:00:32 CEST] <BBB> so av_read_frame()s caller (e.g. ffmpeg.c, or a custom application using lavf) will do the unref [22:01:25 CEST] <pfelt> ok. then i *might* be good with this patch [22:02:07 CEST] <pfelt> how did dup work before? it seems like it took a single arg of the packet and returned an int [22:14:13 CEST] <BBB> pfelt: I believe it simply re-allocated the packet in-place [22:14:19 CEST] <BBB> not 100% sure though [22:14:26 CEST] <BBB> maybe realloc if n_refs > 1? [22:56:50 CEST] <atomnuker> nice to see some simd where you can be sure there will be mod 4 vector sizes [22:57:23 CEST] <atomnuker> me and iive have been getting many gray hairs and slower-than-c simd because its a pain to handle otherwise [22:59:31 CEST] <rcombs> who should I bother for listening tests? Kamedo? [23:01:36 CEST] <atomnuker> yes [23:02:02 CEST] <rcombs> is he usually on IRC? (under that nick?) [23:02:56 CEST] <atomnuker> nope, just ask him on twitter or hydrogenaudio [23:05:32 CEST] <rcombs> @kamedo2? [23:13:50 CEST] <nevcairiel> what are you improving that needs tests? [23:15:36 CEST] <atomnuker> rcombs: yes [23:16:15 CEST] <rcombs> nevcairiel: afaik there haven't been tests since atomnuker's AAC improvements were completed, and I'm not aware of any tests comparing the Windows AAC encoder against anything else [23:16:44 CEST] <nevcairiel> the last tests were in january 2016, since then not that much changed anymore [23:17:37 CEST] <atomnuker> nope, not much has since then [23:18:03 CEST] <rcombs> are they not linked in that thread? [23:19:27 CEST] <rcombs> last comments I saw in there mentioned that the tests linked were on an out-of-date patch [00:00:00 CEST] --- Wed May 24 2017

1 0

ffmpeg.log.20170523
by burek 24 May '17

24 May '17

[00:01:15 CEST] <lystra> Hi. I am trying to burn in Arabic subtitles to a video I created. I can do this for the English subtitles but the Arabic subtitles are not displaying correctly. I used Aegisub to create the subtitles, am using the Scheherazade Arabic font, and have the encoding in Aegisub set to "178 - Arabic". The subtitles are appearing as https://imagebin.ca/v/3NO5qw3AH8j9. My ffmpeg command and output is at https://pastebin.com/iwaJJy07. Any ideas? [00:01:15 CEST] <lystra> I tried playing around with -metadata:s:s:0 language=ara,-metadata:s:s:0 language=arb, and -sub_charenc ARABIC as well. [00:01:48 CEST] <Tatsh> the guy says 'shshshshshshshs' [00:01:52 CEST] <Tatsh> :) [00:03:32 CEST] <furq> lystra: it's probably an issue with libass [00:03:48 CEST] <furq> afaik it needs to have harfbuzz and fribidi support [00:03:55 CEST] <Tatsh> yes [00:04:03 CEST] <Tatsh> i have that stuff built on my machine everywhere [00:04:09 CEST] <Tatsh> fribidi global USE flag on gentoo [00:05:21 CEST] <lystra> On my system, ffmpeg links to /usr/lib/x86_64-linux-gnu/libass.so.5 which is linked to harfbuzz and fribidi. [00:05:39 CEST] <lystra> I am running Mint 18. [00:06:11 CEST] <Tatsh> o [00:06:14 CEST] <Tatsh> i'm getting the same output [00:06:30 CEST] <Tatsh> all shin and and the vowel marks show [00:06:36 CEST] <Tatsh> the ! is on the left side though for me [00:06:58 CEST] <Tatsh> could it be the font? [00:07:11 CEST] <lystra> I am using this font in LaTeX to display Arabic text fine. [00:07:58 CEST] <Tatsh> amiri font works [00:08:45 CEST] <Tatsh> it shows correctly at 3.3: marhaban baina [00:08:57 CEST] <lystra> I could easily switch the font force_style but I doubt it would matter. [00:08:59 CEST] <Tatsh> but it's like !marhaban baina :P [00:09:10 CEST] <Tatsh> make sure you get those order markers in however you can [00:09:14 CEST] <Tatsh> it's confusing as hell [00:09:35 CEST] <Tatsh> braina* [00:09:56 CEST] <Tatsh> lystra, try the amiri font if that is an acceptable font [00:09:56 CEST] <lystra> Tatsh: I'll try Amiri. No idea why it would work. [00:10:12 CEST] <Tatsh> http://www.amirifont.org/ [00:10:17 CEST] <Tatsh> it's a bit old school [00:10:28 CEST] <Tatsh> and i don't see any of those straight fonts available for free much [00:11:07 CEST] <lystra> Debian provides that font in the fonts-hosny-amiri package. [00:11:28 CEST] <Tatsh> i don't see why amiri works and not the other [00:11:49 CEST] <Tatsh> with my fontconfig, amiri is my default arabic-script font [00:12:17 CEST] <lystra> Tatsh: Amiri works. Odd. Trying the new Dubai font. [00:14:19 CEST] <lystra> Well, the new Dubai font mostly works though some characters aren't being displayed. Something else seems amiss here. [00:16:44 CEST] <lystra> Amiri Quran doesn't work. I get boxes like with Scheherazade. [00:21:51 CEST] <Tatsh> hmm very weird [00:24:40 CEST] <lystra> Tatsh: Thanks. [00:25:18 CEST] <Tatsh> fontconfig is involved with libass [00:25:23 CEST] <Tatsh> so try and adjust mine [00:25:54 CEST] <Tatsh> https://gist.github.com/Tatsh/ca2b0cd61cf6ffb5c0e4 [00:26:26 CEST] <Tatsh> 219 covers arabic [00:50:07 CEST] <lystra> Tatsh: Ok, thanks. [01:41:09 CEST] <pqatsi> Hello! Can I get help here about GPAC? If not, where can I ask a question? [02:36:50 CEST] <hellfar_> Hi everyone :) ! [02:36:50 CEST] <hellfar_> I came here because I have a log that proposes me to expose my problem here on this chanel. [02:36:50 CEST] <hellfar_> I am trying to install ffmpeg, but can't `configure` it. [02:36:50 CEST] <hellfar_> My command: [02:36:50 CEST] <hellfar_> `./configure --enable-gpl --enable-postproc --enable-swscale --enable-avfilter --enable-libmp3lame --enable-libvorbis --enable-libtheora --enable-libx264 --enable-libspeex --enable-shared --enable-pthreads` [02:37:11 CEST] <hellfar_> I was following this instructions: [02:37:11 CEST] <hellfar_> https://www.assetbank.co.uk/support/documentation/install/ffmpeg/ [02:37:57 CEST] <hellfar_> And the configure script return me the following message: [02:37:57 CEST] <hellfar_> "ERROR: speex not found using pkg-config" [02:38:08 CEST] <furq> what does config.log say [02:39:03 CEST] <hellfar_> ``` [02:39:04 CEST] <hellfar_> check_pkg_config speex speex/speex.h speex_decoder_init -lspeex [02:39:04 CEST] <hellfar_> pkg-config --exists --print-errors speex [02:39:04 CEST] <hellfar_> Package speex was not found in the pkg-config search path. [02:39:04 CEST] <hellfar_> Perhaps you should add the directory containing `speex.pc' [02:39:05 CEST] <hellfar_> to the PKG_CONFIG_PATH environment variable [02:39:07 CEST] <hellfar_> No package 'speex' found [02:39:10 CEST] <hellfar_> ERROR: speex not found using pkg-config [02:39:12 CEST] <hellfar_> ``` [02:40:29 CEST] <hellfar_> I do not really know if I installed the libspeex properly. [02:40:29 CEST] <hellfar_> I made `sudo apt-get install speex` and it told me "speex is already the newest version.". [02:40:42 CEST] <furq> install libspeex-dev [03:14:12 CEST] <hellfar_> It worked :) ! [03:21:12 CEST] <hellfar_> But then after trying to `configure` again, this is now with `libx264` I having the same problem, `x264` passed. [03:21:52 CEST] <DHE> same thing? x264-devel ? [03:21:57 CEST] <DHE> err, dev [03:22:00 CEST] <furq> libx264-dev [03:22:35 CEST] <furq> or you know [03:22:36 CEST] <furq> https://www.johnvansickle.com/ffmpeg/ [03:22:39 CEST] <furq> just install that [04:40:20 CEST] <hellfar_> Thanks, I was having another issue finally compiling ffmpeg ^^ [04:40:20 CEST] <hellfar_> But with the builds you showed me the gem I was trying to use works now fine :) [04:40:25 CEST] <hellfar_> Thanks a lot ! [05:22:57 CEST] <fullstack> is there a way to run ffmpeg to record a 720 IP camera that supports RTSP ONVIF? [05:23:14 CEST] <fullstack> record rtsp to disk. [05:34:20 CEST] <teratorn> fullstack: sure, ffmpeg -i rtsp://host:port/blah save.mp4, for example [05:39:31 CEST] <npnth> Is there a way to strip all EXIF data from a jpg? I'm looking for something like what `jhead -purejpg' does. [05:40:08 CEST] <npnth> I've been trying things like `ffmpeg -i JF4_021962.jpg -map_metadata -1 -c:v copy JF4_021962_maybe_stripped.jpg', but this still leaves lots of information. [05:41:22 CEST] <npnth> (Perhaps "EXIF" is the wrong term - maybe I should say "I want all removable metadata removed" or something.) [08:23:36 CEST] <k_sze[work]> If I have a streaming source that is just a bunch of jpeg packets, is there an easy way to pipe it into ffmpeg and mux it as mjpeg? [08:23:44 CEST] <k_sze[work]> (losslessly, of course) [11:02:21 CEST] <Guest69723> Hello, I am trying to use ffmpeg in a Unity project, but I keep getting a segfault on sws_scale. Frames are decoded correctly (I saved them in a file and was able to print them) but sws_scale keeps segfaulting. Here is my code : https://pastebin.com/5fJFBCXy [11:03:37 CEST] <Guest69723> If you can help me, I spent all my day yesterday trying to find this segfault, i'm using ffmpeg 3.3 and an C# wrapper [12:39:09 CEST] <DHE> Guest69723: sws_ctx is global? second execution sws_ctx wont' be null and you won't allocate the output buffer [12:39:16 CEST] <DHE> (not a C# programmer, may be missing more) [12:40:41 CEST] <Guest69723> no, DecodeFrame is a method of a class, sws_ctx is a member of the class [12:41:10 CEST] <Guest69723> I suspect the wrapper to have someproblems [12:41:27 CEST] <Guest69723> i an currently switching to an older release, see if it fixes the problem [12:42:40 CEST] <Guest69723> thanks for your answer x) [12:43:44 CEST] <DHE> line 23 should not be inside that if(){} block [12:43:53 CEST] <DHE> that's what I'm seeing [13:24:20 CEST] <thebombzen> fullstack: if you're going to do that, I recommend using -c copy [13:24:45 CEST] <thebombzen> as in ffmpeg -i rtsp://whatever -c copy save.mkv [15:07:32 CEST] <rofl____> more cpu cores or cpu with intel quicksync for encoding/decoding with ffmpeg? [15:07:38 CEST] <rofl____> intel quicksync cpus have 4 cores at most [15:08:35 CEST] <Mavrik> If you want good quality you don't use quicksync anyway [15:10:00 CEST] <rofl____> oh, right [15:25:40 CEST] <OT2_> Hello, I an Unity app using ffmpeg, it has to work on android and windows. I got the avcodec-57.dll and so on prebuild and they are working great. I finaly managed to build the android lib.so equivalents. but i noticed that the android libs are looking for each other using symbolic links : for example libavcodec-57.so looks for libavutil-55.so by using the symbolic link libavutil.so. Is there a way to compile ffmpegs libs so that they look directly for libs [15:25:40 CEST] <OT2_> without using symbolic links ? [15:27:04 CEST] <JEEB> OT2_: I'm pretty sure if you configure FFmpeg so that it notices it's android, then it disables SONAMEs [15:27:22 CEST] <JEEB> might only happen with install target so set --prefix according to where your linker is looking [15:27:34 CEST] <JEEB> and then `make insta` [15:27:37 CEST] <JEEB> *install [15:28:28 CEST] <OT2_> i build it for android with --target-os=android and --arch=arm [15:29:13 CEST] <OT2_> but I have to include my libsavcodec-57.so and a copy of them without the version for it to work [15:29:26 CEST] <JEEB> yes, so you are not running the install target :P [15:29:36 CEST] <JEEB> as that should disable SONAMEs at that point [15:29:40 CEST] <OT2_> what do you mean ? [15:30:03 CEST] <JEEB> it removes the SONAMEs when it installs the libraries into your prefer [15:30:04 CEST] <OT2_> should i past you my ./configure command ? [15:30:04 CEST] <JEEB> *prefix [15:30:33 CEST] <JEEB> for example I would do it --prefix=/home/user/armv7_prefix [15:30:45 CEST] <JEEB> and then make install would put the headers and libraries there under include/lib [15:31:10 CEST] <JEEB> you're not supposed to use the files from the build tree as-is [15:31:42 CEST] <OT2_> you mean i am not supposed to get the libs from ffmpegfolder/libavcodec/libavcodec-57.so [15:31:58 CEST] <OT2_> wait a moment I check what I did, thanks [15:31:59 CEST] <JEEB> yes. [15:32:11 CEST] <JEEB> you set a prefix and do make install [15:32:27 CEST] <OT2_> I do have set a prefix [15:32:28 CEST] <JEEB> that way you get things and a pkg-config file set correctly :P [15:32:41 CEST] <OT2_> but the folder is empty after compile [15:32:43 CEST] <JEEB> OT2_: hopefully not the default /usr/local [15:33:14 CEST] <OT2_> no its /mnt/e/ffmpegbuild (im using ubuntu bash for windows) [15:33:15 CEST] <JEEB> OT2_: well make install is the step does the final changes putting things into the configured prefix [15:33:21 CEST] <OT2_> oh ahah [15:33:25 CEST] <OT2_> i did not make install [15:33:26 CEST] <OT2_> my bad [15:33:45 CEST] <OT2_> idk I always assumed it was to imstall ffmpeg executable file [15:34:02 CEST] <JEEB> it also installs that under bin/ :P [15:34:19 CEST] <JEEB> if you build it that is [15:35:02 CEST] <OT2_> nice [15:35:11 CEST] <OT2_> seems to do what i expect [15:35:31 CEST] <OT2_> but is it possible to have them called libavcodec-57.so instead of libavcodec.so [15:35:50 CEST] <JEEB> android doesn't support SONAMEs [15:35:53 CEST] <OT2_> as my windows .dll have the version in it [15:35:56 CEST] <JEEB> that's why they get removed [15:36:24 CEST] <OT2_> i dont mean libavcodec.so.57 i mean libavcodec-57.so [15:36:48 CEST] <OT2_> and that they look for each others with this syntax [15:37:22 CEST] <JEEB> anyways, no. [15:38:13 CEST] <OT2_> Okay i will figure a way to get it working in the C# wrapper. thank you, helped a lot [15:45:57 CEST] <OT2_> one more question : when I compile ffmpeg it say pkg-config not found (seems like not included in android-ndk) and WARNING : using x264 without pkg-config [15:46:18 CEST] <OT2_> can i ignore those warnings ? [15:48:11 CEST] <JEEB> yes as long as your build works OK. and yea, NDK doesn't have pkg-config, you can usually utilize the pkg-config that comes with your distro. [15:48:56 CEST] <JEEB> just make a shell script with the NDK cross-prefix that sets PKG_CONFIG_LIBDIR to /path/to/prefix/lib/pkgconfig [15:49:28 CEST] <JEEB> and then the script just runs `exec pkg-config "$@"` after setting PKG_CONFIG_LIBDIR [15:50:00 CEST] <JEEB> in my case the script is called arm-linux-androideabi-pkg-config [15:50:09 CEST] <JEEB> as I build for armv7 [15:54:57 CEST] <OT2_> Build worked so I will not worry about, for the moment x) [16:58:07 CEST] <james999> NDK = Nvidia Development Kit? [16:59:06 CEST] <Mihaylov> Hello [16:59:16 CEST] <Mihaylov> "ffmpeg -i in.mkv -f segment -segment_time 10 -segment_format_options movflags=+faststart out%03d.mp4" [16:59:27 CEST] <Mihaylov> -segment_time X is in seconds? [16:59:48 CEST] <Mihaylov> -segment_time 600 splits the video in 20 minutes which I see no relation [17:00:03 CEST] <sfan5> james999: it's referring to the Android NDK (Native Development Kit) in that case [17:01:20 CEST] <Mihaylov> i tought it was seconds so i put 1200 for 20 minutes and it splited the video in 40 minutes, its logical that if 1200 40 min, then 600 = 20 min but that value makes no sense [17:01:37 CEST] <Mihaylov> what unit is the -segment_time parameter? [17:14:09 CEST] <kepstin> Mihaylov: it is in seconds, but whether or not it can split at the desired times depends on keyframe locations. That said, it being almost exactly double what you set is really strange& is this a really low fps video? [17:14:31 CEST] <furq> do the segments play back correctly [17:15:01 CEST] <dystopia_> 50fps played back at 25fps for maybe? [17:15:26 CEST] <Mihaylov> They do play correctly [17:16:09 CEST] <furq> well yeah that should work [17:16:26 CEST] <furq> could you pastebin the ffprobe output for the input and one of the segments [17:21:30 CEST] <Mihaylov> input: https://pastebin.com/TAyPAcND [17:22:23 CEST] <Mihaylov> segment: https://pastebin.com/aM4qXspp [17:35:48 CEST] <dystopia_> Duration: 00:10:01.28, start: 0.000000, bitrate: 4543 kb/s [17:36:03 CEST] <dystopia_> it's 10min's long or 600sec's [17:36:22 CEST] <dystopia_> also it is 50fps, maybe your media player was trying to play it back at 25fps [17:36:55 CEST] <Mihaylov> oops i sent the last segment [17:37:00 CEST] <Mihaylov> wich is 10 because no more video [17:37:45 CEST] <Mihaylov> im using vlc [17:38:08 CEST] <Mihaylov> ah no my bad, i sent you the right fragment [17:39:27 CEST] <Mihaylov> You are right, its 10 min and vlc is playing it at 20 min [17:40:31 CEST] <Mihaylov> weird the audio is perfect [17:42:41 CEST] <Mihaylov> Tools >>> Codec information says 50 fps [17:43:53 CEST] <dystopia_> i am unsure [17:44:05 CEST] <dystopia_> it's an issue with vlc and not ffmpeg though [17:44:12 CEST] <dystopia_> if you are on windows, try mpc [17:54:30 CEST] <Mihaylov> dystopia_ I just made a test, played the video for 10 min, even when it shows that the total duration is 20 min [17:54:50 CEST] <Mihaylov> it finished at 10 min [18:04:43 CEST] <Mihaylov> got to go, thanks for the enlightment [18:44:50 CEST] <alexpigment> hey guys. i'm trying to build a LGPL version of ffmpeg on Mac, and I've historically built with Homebrew [18:45:25 CEST] <alexpigment> I need to access the configure file prior to building/installing, but I have no idea where this is saved [18:45:34 CEST] <alexpigment> does it save a git somewhere before installing? [18:47:02 CEST] <DHE> configure will set up everything for building based on what you want. if you want lgpl, do NOT specify --enable-gpl to it [18:47:54 CEST] <alexpigment> am i right in thinking that what you're suggesting is incompatible with Homebrew? [18:48:23 CEST] <sfan5> why do you need access to the configure file? [18:48:35 CEST] <sfan5> also this sounds like something you should be asking the homebrew people instead [18:49:00 CEST] <alexpigment> because i'm going to be commercially licensing x264, and so I need to include x264 but still create an LGPL build of ffmpeg [18:49:05 CEST] <DHE> I don't know the specifics or homebrew. if you don't redistribute what you build you're allowed to do whatever you want with it [18:49:11 CEST] <DHE> oh... well then... [18:49:55 CEST] <DHE> this is a known issue where libx264 support doesn't properly handle the situation of where you buy an x264 license. you would have to hack it to override the GPL requirement [18:50:01 CEST] <alexpigment> i'm just a real novice with compilation, and I've got a build script for compiling on Windows via Linux, and I've got that fully configured as necessary. i just don't know how to replicate that for an OS X build [18:50:24 CEST] <sfan5> you will need to take a look at the instructions brew runs to build ffmpeg and edit those [18:50:28 CEST] <alexpigment> DHE: done and done with my Windows builds. I just have no idea what I'm doing on macOS [18:50:39 CEST] <alexpigment> k [18:50:48 CEST] Action: alexpigment digs into things he clearly doesn't understand ;) [18:54:31 CEST] <Fare> Weird. I'm trying to use ffmpeg to record myself doing a presentation. I can record either the sound or the video just fine, but trying to do both results in garbled sound. [18:57:33 CEST] <furq> alexpigment: it looks like you can run preinstall commands with brew edit [18:57:47 CEST] <furq> so i guess just call sed before configure [19:04:50 CEST] <alexpigment> furq: thanks for the info. i'm having to google things to actually process what you're saying, but i'll get back to you once i have a basic understanding of sed [19:05:55 CEST] <furq> sed -i '1504d' configure [19:06:27 CEST] <furq> obviously you'll need to change the line every time configure is updated [19:06:53 CEST] <alexpigment> is 1504d the current build? [19:07:05 CEST] <furq> line 1504 [19:07:07 CEST] <furq> the one you want to remove [19:07:51 CEST] <alexpigment> ok [19:07:56 CEST] <alexpigment> apologies, i'm just really lost here [19:08:14 CEST] <alexpigment> a) i can't find a configure file on my computer - so that may be problem #1 [19:08:34 CEST] <alexpigment> b) if i try to use brew edit in the termal, i get this interface where i have to choose a directory [19:08:41 CEST] <furq> if brew builds it on your machine then i assume it downloads that when you install it [19:08:51 CEST] <alexpigment> i assume so too [19:09:08 CEST] <alexpigment> i just don't know how to get it to download, let me edit, then finish the build process [19:09:32 CEST] <furq> are you doing `brew edit ffmpeg` [19:10:21 CEST] <alexpigment> no, i wasn't, but i'm doing that now [19:10:33 CEST] <alexpigment> and when i do it, i get the ffmpeg.rb file which i already was able to manually edit [19:10:55 CEST] <alexpigment> in this file, i can delete --enable-gpl , but then i get an error that x264 is gpl only [19:11:08 CEST] <furq> right, that's why you need to remove a line from configure [19:11:19 CEST] <alexpigment> so i really need to get access to the configure file to edit its list of GPL packages [19:11:27 CEST] <furq> https://github.com/FFmpeg/FFmpeg/blob/master/configure#L1504 [19:11:28 CEST] <alexpigment> right, and that's what i'm not sure of how to do [19:12:25 CEST] <furq> i assume you just add system("sed", "-i" "1504d", "configure") just before configure is invoked [19:13:05 CEST] <alexpigment> ok, does this make sense to you? [19:13:13 CEST] <alexpigment> i typed in sed -i '1504d' configure [19:13:38 CEST] <alexpigment> and terminal says "sed: 1: "configure": command c expects \ followed by text [19:14:17 CEST] <furq> oh right i forgot bsd sed requires an argument to -i [19:14:24 CEST] <furq> sed -i '' '1504d' configure [19:15:05 CEST] <alexpigment> no such file or directory [19:15:11 CEST] <furq> tbh you should probably do that using regex given that this is a ruby script [19:15:19 CEST] <furq> but i'd rather not remember how to use ruby [19:15:26 CEST] <alexpigment> ok [19:15:31 CEST] <furq> where are you running that [19:15:42 CEST] <alexpigment> i'm just in terminal [19:15:47 CEST] <furq> oh [19:15:56 CEST] <furq> well yeah it won't exist until brew downloads it [19:16:00 CEST] <alexpigment> right [19:16:01 CEST] <alexpigment> :) [19:16:18 CEST] <alexpigment> sorry, i tried to make that clear earlier, but i didn't do a good enough job [19:16:27 CEST] <alexpigment> configure doesn't exist on my computer yet [19:16:58 CEST] <alexpigment> so i'm really looking for a way for homebrew to start its thing, download git, PAUSE so i can edit the configure file, then go along its merry way [19:17:14 CEST] <furq> https://github.com/Homebrew/brew/blob/master/docs/Formula-Cookbook.md#grab-… [19:17:36 CEST] <furq> add `system("sed", "-i" "1504d", "configure")` at the top of `def install` [19:18:03 CEST] <alexpigment> got it [19:18:04 CEST] <alexpigment> 1 sec [19:18:05 CEST] <alexpigment> btw [19:18:19 CEST] <alexpigment> what am i doing with line 1504 with this command? [19:18:21 CEST] <alexpigment> just deleting it? [19:18:23 CEST] <furq> yeah [19:18:25 CEST] <alexpigment> because i think i need to be moving it [19:18:34 CEST] <alexpigment> to the non-gpl list [19:18:47 CEST] <alexpigment> is there a move command? [19:19:51 CEST] <furq> oh right [19:20:15 CEST] <furq> 1504d,1550ilibx264 [19:20:32 CEST] <furq> this isn't a particularly good solution anyway because it breaks if the line numbers change [19:20:38 CEST] <alexpigment> ok, so that's literally inserting the line text [19:20:41 CEST] <alexpigment> right [19:20:41 CEST] <furq> yeah [19:20:58 CEST] <alexpigment> let's hope the order of this is chronological [19:21:08 CEST] <alexpigment> because otherwise we might tear the fabric of space and time [19:21:12 CEST] <alexpigment> ;) [19:23:27 CEST] <alexpigment> just to clarify, should it be "1504d,1550ilibx264", "configure" [19:23:33 CEST] <furq> yeah [19:23:40 CEST] <alexpigment> or should each command be in its own quote? [19:29:49 CEST] <alexpigment> ok, the syntax fails, and i've edited it several times to no avail. is it easier to put a line in there that manually replaces the configure file with one either on my system or on a webserver? [19:30:05 CEST] <furq> maybe [19:30:37 CEST] <alexpigment> ok, i'm going to see what i can do here [19:34:24 CEST] <alexpigment> man, wish i just knew where the hell the files were being downloaded to. macOS is so bad about obscuring things like this [19:34:44 CEST] <alexpigment> i can't even watch the date modified changes to follow the folders [19:53:24 CEST] <alexpigment> furq: well, it turns out it's easier to just create my own tarball and reference that as the download [19:53:39 CEST] <alexpigment> thanks for your help on this, as it ultimately led me to a solution [19:53:50 CEST] <alexpigment> even if i was confused as hell there for a bit :) [20:05:59 CEST] <fullstack> Active: failed (Result: exit-code) since Tue 2017-05-23 08:50:01 PDT; 2h 15min ago [20:06:02 CEST] <fullstack> Process: 22132 ExecStop=/usr/bin/zmpkg.pl stop (code=exited, status=1/FAILURE) [20:10:49 CEST] <fullstack> [Note] Aborted connection 642142 to db: 'zm' user: 'zmuser' host: 'localhost' (Got an error reading communication packets) [20:11:01 CEST] <fullstack> that's in the mysql log like 10x per second continously [20:18:03 CEST] <blue_misfit> hey guys! I'm doing something like -ss 5000 -i file.mov -to 30 to extract 30 second chunks from a big ProRes MOV file and _sometimes_ the output is out of sync [20:18:08 CEST] <blue_misfit> any suggestions on how to debug / troubleshoot? [20:18:52 CEST] <c_14> fullstack: doesn't seem related to ffmpeg [20:19:05 CEST] <fullstack> c_14 thanks [20:20:31 CEST] <blue_misfit> this looks kinda suspicious: [mov @ 000000000262a740] timecode: tbc=24000/1 invalid, fallback on 24000/1001 [20:36:47 CEST] <ChocolateArmpits> blue_misfit, is the output out of sync when you seek after the input? [20:37:27 CEST] <blue_misfit> like, if I take the output file and play it, and seek around inside, does it stay out of sync? [20:37:39 CEST] <ChocolateArmpits> I mean place -ss after the input [20:37:45 CEST] <blue_misfit> oh [20:38:29 CEST] <ChocolateArmpits> it'll take longer to seek, but solves quite a few common problems [20:38:46 CEST] <blue_misfit> interesting [20:39:02 CEST] <blue_misfit> problem is it's a 200 GB file which is in cloud storage which we're addressing via https so it will have to read the whole file, huh? [20:39:12 CEST] <blue_misfit> I'm pulling it down locally to test anyway [20:40:49 CEST] <ChocolateArmpits> -ss when used as an input option skips the stream, when used as an output option decodes everything prior to the point [20:41:34 CEST] <blue_misfit> right [20:42:10 CEST] <blue_misfit> seems like ss for the input option is ideal for our case, and it usually works fine but some files are problematic like this [20:42:21 CEST] <blue_misfit> any words of wisdom for identifying the root cause of that? [20:43:07 CEST] <ChocolateArmpits> first thing make sure you are using latest ffmpeg [20:43:22 CEST] <blue_misfit> k [20:43:34 CEST] <blue_misfit> will start with latest zeranoe build on windows [20:44:39 CEST] <ChocolateArmpits> Does the sync issue happen for specific files or is it random? If for example you skipped a problematic file at a different point does the sync issue still occur ? [20:45:01 CEST] <blue_misfit> it does, I pulled a chunk from earlier in the file and it's still out of sync, just less so [20:45:51 CEST] <ChocolateArmpits> are you doing stream copy or reencoding too ? [20:46:33 CEST] <blue_misfit> for this test it's all stream copy [20:46:42 CEST] <blue_misfit> but in the real workflow we transcode with libx264 [20:46:55 CEST] <blue_misfit> split and stitch scale-out cloud transcoding workflow, pretty standard [20:54:57 CEST] <ChocolateArmpits> blue_misfit, pretty hard to say, but there are a few things you can play with. try using -r as an input option, try transcoding the audio while the video stream is left to be copied, check out -copyts, -copytb. Try rewriting timestamps using asetpts/setpts [20:55:10 CEST] <ChocolateArmpits> try any of them [21:16:26 CEST] <blue_misfit> ChocolateArmpits, thank you :) I will do some testing [21:23:01 CEST] <blue_misfit> ChocolateArmpits, turns out this was just a bad source file LOL but thanks regardless [21:23:11 CEST] <ChocolateArmpits> np [21:39:26 CEST] <OzgrK> Hi, the following command gives "error marking filters as finished" error filling the ram and putting the system in unstable position for version N-86175-g64ea4d1 however it works fine for one of the previous versions, version N-82324-g872b358. Any ideas? [21:39:33 CEST] <OzgrK> ffmpeg -f lavfi -i aevalsrc="sin((1000-t*0.0086805)*2*PI*t):s=48000:d=9" -i https://upload.wikimedia.org/wikipedia/commons/c/c4/PM5544_with_non-PAL_sig… -vf loop=size=1:loop=-1,hue=h=t*360/57600 -t 9 -pixel_format yuv420p -c:v libvpx-vp9 -lossless 1 -c:a libvorbis -qscale:a 10 test1.webm [22:56:04 CEST] <faLUCE> is there any wrapper library for ffplay ? [23:03:51 CEST] <kepstin> no, ffplay is basically a demo implementation of a player using the ffmpeg libraries [23:04:46 CEST] <faLUCE> kepstin: I'm searching for a shorter code than ffplay.c, in order to see a http mpegts stream [23:05:14 CEST] <teratorn> OzgrK: you're saying that command causes errors to be printed, and it leaks memory until your system becomes unstable? [23:05:41 CEST] <kepstin> faLUCE: why can't you just use an existing player, e.g. mpv? [23:05:56 CEST] <faLUCE> kepstin: because I have to optimize the latency [23:06:04 CEST] <faLUCE> (when receiving the http stream) [23:06:48 CEST] <teratorn> OzgrK: if I wanted to track it down to the commit that introducted the problem I would probably use git bisect [23:06:53 CEST] <kepstin> hmm, if you can't do it with configuring mpv, then I guess you are gonna have to write your own player :/ [23:08:11 CEST] <faLUCE> kepstin: I tried the dranger tutorial: http://dranger.com/ffmpeg/tutorial04.html. It works for video only, but doesn't work, with the 3.x version of ffmpeg, for audio+video [23:21:20 CEST] <james999> dranger tutorial? you mean your tutorial faLUCE? [23:24:17 CEST] <ChocolateArmpits> faLUCE, have you tried suggested settings for mpv here (wm4's answer)? https://github.com/mpv-player/mpv/issues/4213 [23:25:44 CEST] <faLUCE> ChocolateArmpits: no, mpv has bad support for http-mpegts [23:25:52 CEST] <ChocolateArmpits> so that's another issue [23:25:56 CEST] <faLUCE> james999: I mean dranger tutorial [23:26:17 CEST] <OzgrK> teratorn: yes, it prints "error marking filters as finished" and leaks memory until i quit ffmpeg and sometimes causes the system crash totally and reboot. but for version N-86175-g64ea4d1, not for version N-82324-g872b358, for example. i had no problem until i upgraded ffmpeg [23:28:08 CEST] <teratorn> OzgrK: well if it is still broken on git master, I would try to git bisect to find the commit that introduced the regression [23:28:32 CEST] <teratorn> OzgrK: then use the mailing list to try and get it fixed [23:31:07 CEST] <OzgrK> i will report it to user e-mail, i am not technically capable, thanx [23:51:50 CEST] <asmand> Hello there, I have a problem with dvb_subtitle stream copying to mkv. The problem manifests itself as unreadable subtitle stream (reported by VLC). Input file is mpegts, a straight copy of the streams I want, to output format mpegts works as expected. Output https://pastebin.com/Mfdqw2JN [23:53:00 CEST] <asmand> If I change the output container to mkv, however, the subtitle track is unrecognized, and we see something odd in the output: https://pastebin.com/V7hwnmyr [23:54:21 CEST] <asmand> The key differences is that with mpegts, the output metadata says "Stream #0:2(fin): Subtitle: dvb_subtitle ([6][0][0][0] / 0x0006)", while the same line for matroska says "Stream #0:2(fin): Subtitle: dvb_subtitle ([255][255][255][255] / 0xFFFFFFFF)" [23:57:36 CEST] <asmand> overlaying/burning in the subtitles works (-filter_complex "[0:0][0:2]overlay"), so that would suggest that the input file is ok, and that the subtitles can be read properly [00:00:00 CEST] --- Wed May 24 2017

1 0

ffmpeg-devel.log.20170522
by burek 23 May '17

23 May '17

[00:32:28 CEST] <cone-270> ffmpeg 03Carl Eugen Hoyos 07master:017a68e8fb8b: lavf/riff: Support more vlc fourcc's for 12 and 16 bit yuv4xx. [01:12:03 CEST] <cone-270> ffmpeg 03Carl Eugen Hoyos 07master:7f60dc03a0b8: lavf/img2: Add svg and svgz to allow reading image sequences. [02:18:01 CEST] <cone-270> ffmpeg 03Michael Niedermayer 07master:40fa6a2fa2c2: avcodec/mjpegdec: Fix runtime error: signed integer overflow: -32767 * 130560 cannot be represented in type 'int' [02:18:02 CEST] <cone-270> ffmpeg 03Michael Niedermayer 07master:a1cbf53c566b: avcodec/dct32_template: Fix runtime error: signed integer overflow: -1071326067 - 1088238847 cannot be represented in type 'int' [02:18:03 CEST] <cone-270> ffmpeg 03Michael Niedermayer 07master:8fb00b3e858b: avcodec/ivi_dsp: Fix multiple runtime error: left shift of negative value -71 [03:07:03 CEST] <BBB> j-b: can I ask you aom questions now that vlc is in it? [03:07:10 CEST] <BBB> j-b: or should I ask someone else in videolan? [09:13:49 CEST] <j-b> BBB: sure [09:45:48 CEST] <ubitux> can anyone confirm that isobase/mp4/mov do not have any timezone/time offset information? [09:46:16 CEST] <ubitux> time seems to be utc only, but except the potential gps info, i see no way of deducting the local time [11:09:35 CEST] <cone-089> ffmpeg 03Paul B Mahol 07master:49d0678181af: avfilter/avf_showspectrum: properly initialize pts [12:39:47 CEST] <cone-089> ffmpeg 03Hendrik Leppkes 07master:9fb293cfd8b7: Use AVOnce as a static variable consistently [12:39:56 CEST] <cone-089> ffmpeg 03Hendrik Leppkes 07release/3.3:8640339dbbb5: Use AVOnce as a static variable consistently [14:04:06 CEST] <J_Darnley> BBB or michaelni: Are either of you around? I need to ask you about simple_idct. [14:09:13 CEST] <J_Darnley> Can the horizontal and vertical transforms be reversed? [14:10:25 CEST] <J_Darnley> In the mmx code what do the upper and lower case comments mean? Is there a difference between "A0" nd "a0"? [16:14:14 CEST] <ubitux> INSTALLinstall-progs-yes [16:14:16 CEST] <ubitux> heh.. [16:17:05 CEST] <ubitux> btw, i think we have a leak with custom IO [16:17:17 CEST] <ubitux> and it may be a regression [16:25:12 CEST] <tdjones> atomnuker: You said that you were writing a new psychoacoustic model, what does the interface look like for me to write against in the vorbis encoder? [16:32:04 CEST] <atomnuker> tdjones: apply thins patch on top of git master and look at opusenc_psy.h [16:32:06 CEST] <atomnuker> https://0x0.st/ENB.patch [16:32:42 CEST] <tdjones> Will look at it, thanks [16:33:07 CEST] <atomnuker> for now only the _init, _end and ff_opus_psy_process would be needed [16:34:08 CEST] <atomnuker> the process function returns an int to tell you whether you need more buffers in the queue [16:35:14 CEST] <atomnuker> so replace it with some placeholder function which always returns 1 (e.g. each frame would be 64 samples large) [17:04:30 CEST] <ubitux> wm4: please check "[PATCH] lavf/concatdec: do not transfer custom IO flag" when you have a moment [17:06:31 CEST] <wm4> ubitux: check for what and where? [17:07:08 CEST] <wm4> ah on the ml [17:07:17 CEST] <michaelni> J_Darnley, "Can the horizontal and vertical transforms be reversed?" <-- no [17:07:52 CEST] <michaelni> it would change rounding and it would very likely lead to artifacts [17:10:26 CEST] <michaelni> or i misunderstand what you mean by "reversed" [17:21:16 CEST] <ubitux> wm4: this one? [17:21:17 CEST] <ubitux> - pls->ctx->flags |= s->flags; [17:21:19 CEST] <ubitux> + pls->ctx->flags |= s->flags & ~AVFMT_FLAG_CUSTOM_IO; [17:21:22 CEST] <ubitux> (lavf/hls.c) [17:23:24 CEST] <ubitux> we don't have other playlist formats with that? [17:24:02 CEST] <jamrial> tee? [17:24:20 CEST] <wm4> no, and tee wasn't affected [17:24:28 CEST] <wm4> I did only as much as to make fate pass [17:24:55 CEST] <wm4> once the side data merging is removed those changes can be reverted [17:31:20 CEST] <ubitux> ah? okay [17:31:31 CEST] <ubitux> well, i'll push anyway both fixes [17:33:00 CEST] <wm4> what does that make leak anyway? [17:33:42 CEST] <ubitux> the avio buffers [17:34:54 CEST] <wm4> oh [17:35:16 CEST] <wm4> (why does that depend on the demuxer flags instead of avio flags) [17:36:43 CEST] <ubitux> as a user you set the pb field, and internally it adds the flag [17:36:56 CEST] <ubitux> then it knows it must not free it [17:50:52 CEST] <cone-489> ffmpeg 03Clément BSsch 07master:0dcac9c3f0f8: lavf/concatdec: do not transfer custom IO flag [17:50:52 CEST] <cone-489> ffmpeg 03Clément BSsch 07master:ef0106122508: lavf/hls: do not transfer custom IO flag [19:49:11 CEST] <cone-489> ffmpeg 03Michael Niedermayer 07master:da617408c80a: avcodec/escape124: Check depth against num_superblocks [19:49:12 CEST] <cone-489> ffmpeg 03Michael Niedermayer 07master:42e42af76cff: avcodec/mlpdec: Fix runtime error: shift exponent -5 is negative [20:42:00 CEST] <J_Darnley> michaelni: Thanks. I did mean do vertical first then horizontal. I guess "transposed" would be a more accurate term? [20:46:39 CEST] <michaelni> yes [22:41:33 CEST] <iive> J_Darnley: btw, you do know that idct are permutated for easier simd? [22:41:47 CEST] <cone-489> ffmpeg 03Michael Niedermayer 07master:64d0dad93c18: avcodec/takdec: Fix multiple runtime error: signed integer overflow: -512 * 4563386 cannot be represented in type 'int' [22:41:48 CEST] <cone-489> ffmpeg 03Michael Niedermayer 07master:0ce7cf0c86a2: avcodec/takdec: Fix runtime error: left shift of negative value -360 [22:41:50 CEST] <cone-489> ffmpeg 03Michael Niedermayer 07master:f95f9f975598: avcodec/takdec: Fix runtime error: signed integer overflow: 2146548196 + 2156738 cannot be represented in type 'int' [23:10:43 CEST] <J_Darnley> iive: yes, vaguely. Did you mean something specific? [23:12:31 CEST] <iive> i mean that you can use this permutation to transpose the inital coefficient, do the horizontal tranformation (that is, do it vertically but using the rounding of the horizontal pass) [23:12:58 CEST] <iive> then transpose do the vertical transform and the result is in place. [00:00:00 CEST] --- Tue May 23 2017

1 0

ffmpeg.log.20170522
by burek 23 May '17

23 May '17

[11:27:36 CEST] <k_sze> ffmpeg doesn't use libjpeg at all? [11:27:47 CEST] <k_sze> (even for jpeg file or mjpeg support)? [11:34:10 CEST] <BtbN> It can optionally use libopenjpeg [11:35:36 CEST] <BtbN> Which seems to be jpeg2k though [12:01:32 CEST] <Nacht> Hi, I'm having some issues with the exec plugin. I'm running a .sh script located in /usr/local/bin which outputs in JSON. When I run telegraf with -test it just says: Plugin: inputs.exec, Collection 1. I even tried just making a simple .sh script which echos the example as given in README.md. But none of it shows any output [12:02:10 CEST] <Nacht> I have data_format = "json" [13:43:55 CEST] <BtbN> And that's related to ffmpeg in what way? [14:43:26 CEST] <SolidusAbi> Hi to everyone, can someone help with a smooth streaming? Im tringin to generate the .ism and .ismc but i have not found the ismindex operator. [15:08:24 CEST] <DHE> SolidusAbi: you're best off asked a specific question and waiting patiently [15:16:50 CEST] <zerodefect> Does the H.264 (libx264) support YUV422P? I could swear that on Friday I had it working with YUV422P, but now in avcodec_open2() it errors with "Specified pixel format YUV422P is invalid or not supported". [15:19:49 CEST] <zerodefect> I've tried changing the profile HIGH_422 but still. It makes me think I've changed something in the code which has broken it. [15:20:09 CEST] <zerodefect> "the code" being my code :) [15:20:54 CEST] <zerodefect> Other thing I'm wondering is if my package manager (Ubuntu Zesty) has pulled down a new version of the API [15:21:28 CEST] <BtbN> Debian/Ubuntu system packages stick to a stable version [15:21:58 CEST] <DHE> 422 is supported in the current version of ffmpeg and x264. did you actually set a profile that accepts 422 input though? [15:22:04 CEST] <zerodefect> Yeah, that's what I thought [15:23:33 CEST] <zerodefect> DHE: Yeah, I tried 'FF_PROFILE_H264_HIGH_422' but I'm only encoding PAL SD. Not sure if there are constrains on resolution with that profile? [15:23:54 CEST] <DHE> no resolution constraints on profiles. but there are resolution constaints if you set a 'level' [15:24:05 CEST] <zerodefect> Presumably, I only need to set that in the AVCodecContext? [15:24:08 CEST] <zerodefect> Ah ok [15:25:27 CEST] <microchip_> can't build ffmpeg with latest libx265 - this is what I get in config.log https://teambelgium.net/h5ai/files/text/ [15:25:34 CEST] <microchip_> sorry, libx264 [15:26:09 CEST] <microchip_> peeps of x264 added AVX512 support and that is causing issues with ffmpeg git [15:26:16 CEST] <DHE> wow, something's wrong with your copy of x264 [15:26:33 CEST] <DHE> that's not ffmpeg's fault. you're missing a symbol in x264 itself. that normally wouldn't happen... [15:26:47 CEST] <microchip_> ok, i'll try again [15:27:41 CEST] <DHE> oh, that's new in x264. I'm gonna try this [15:28:01 CEST] <microchip_> same thing [15:29:45 CEST] <BtbN> That's probably the worst paste-site possible. It puts a scroll bar at the bottom, and then opens something on top of it when you hover, so you can't reach it [15:30:11 CEST] <BtbN> And also does not allow me to select text to copy it, or to scroll it that way [15:30:25 CEST] <bencoh> :] [15:30:34 CEST] <microchip_> i'll put it elsewhere [15:30:50 CEST] <DHE> noscript + open in browser makes it tolerable. [15:31:31 CEST] <zerodefect> Found something interesting. I call avformat_alloc_output_context2 with an output format of "rtp_mpegts". It seems to be using MPEG4 part 2 (AV_CODEC_ID_MPEG4). That isn't part 10, is it? [15:31:32 CEST] <microchip_> https://pastebin.com/QrTSMus8 [15:32:19 CEST] <DHE> zerodefect: no, part 2 is also known as "DivX", or XviD compatbile [15:32:25 CEST] <DHE> part 10 is H264 [15:32:42 CEST] <BtbN> microchip_, yeah, that looks to me like that x264 is miscompiled in some way [15:33:01 CEST] <BtbN> it's missing symbols from itself [15:33:06 CEST] <microchip_> hmm [15:34:16 CEST] <DHE> kinda odd because x264.exe wouldn't be built successfully either if that happened. [15:34:32 CEST] <zerodefect> DHE: does AV_CODEC_ID_MPEG4 refer to part 10 then? [15:35:59 CEST] <DHE> zerodefect: no, you want AV_CODEC_ID_H264 [15:39:28 CEST] <JEEB> zerodefect: "mpeg4" in FFmpeg vocabulary when talking of video formats (not containers) is mpeg-4 part 2 [15:39:51 CEST] <zerodefect> Thanks for the clarification [15:40:51 CEST] <DHE> at this point in history mpeg4 [part 2] is a footnote nobody really talks about anymore. h264 has replaced it and not much in terms of hardware plays mpeg4 but not h264 [15:43:26 CEST] <zerodefect> When I call 'avformat_alloc_output_context2' should I specify an output context (currently the paramters is NULL) and force the video codec to H.264? What is the correct way? [16:00:38 CEST] <zerodefect> Using the format of "mpegts" gives a video codec of H.264. No RTP though; that will do for now. Thanks for the help. [16:04:47 CEST] <BtbN> if you do not specify a codec, ffmpeg will use a default based on the container. Which for mpegts happens to be mpeg4 [16:04:53 CEST] <BtbN> feel free to override [16:07:34 CEST] <bencoh> which is kinda odd, but anyway :) [16:08:26 CEST] <furq> it's mpeg2video here [16:08:31 CEST] <bencoh> (I guess it gives better results than mpeg2 while still being builtin - library-independant, but ...) [16:08:40 CEST] <furq> at least in ffmpeg it is [16:08:53 CEST] <bencoh> furq: that's what I had in mind back then as well [16:09:08 CEST] <furq> neither of which are as weird as nut using mpeg4 by default [16:10:50 CEST] <zerodefect> BtbN: Where/how exactly would I override the codec? [16:12:57 CEST] <BtbN> When creating the encoder [16:13:07 CEST] <BtbN> Or just with -c as an output option, when using ffmpeg.c [16:13:58 CEST] <zerodefect> Ok. Thanks. I'll give that a try. [17:14:39 CEST] <momomo> hello, any here. need urgent help [17:14:49 CEST] <momomo> I have this ffmpeg command ... it's fully working [17:15:39 CEST] <momomo> it generates hls files ... but it also transcode, scales and compresses using crf 23 ... however the source is already compressed so basically I only need to split it into segment files [17:15:50 CEST] <momomo> * and scale [17:16:10 CEST] <momomo> ooh forgot to post the link: https://hastebin.com/ojovacumim.swift [17:16:45 CEST] <momomo> what should I remove to enable less cpu power being used since I don't need to change anything [17:17:51 CEST] <momomo> here is a better formatted one: https://hastebin.com/ramikaragu.diff [17:19:00 CEST] <momomo> also, do I need to use preset .. and if I have to ... what would be the difference compared to ultrafast since there is no manipulation ... would the hls segment files still end up bigger? [17:37:08 CEST] <alexpigment> i'm trying to build an LGPL build of ffmpeg, but i'm having trouble telling if it actually worked [17:37:32 CEST] <alexpigment> if i type ffmpeg -L , should that give me a type of license? [17:39:54 CEST] <c_14> It lists license text here [17:43:40 CEST] <alexpigment> ok, i'm seeing it now. for some reason it was saying --enable-gpl before [17:43:58 CEST] <c_14> probably executed the wrong binary? [17:57:52 CEST] <alexpigment> possibly [18:35:13 CEST] <cmpscabral> hi everyone. I have this stream https://pastebin.com/gzpM9fEZ playing fine using vlc or vlc-plugin (firefox). my question is, is it possible to downscale and convert this stream with ffmpeg and output something a <video> element (native) or flash plugin would play in a browser? [18:36:58 CEST] <teratorn> cmpscabral: yes [18:41:02 CEST] <cmpscabral> @teratorn tks for your reply. do I need anything besides ffmpeg? [18:41:14 CEST] <cmpscabral> (to do the stream part) [18:41:48 CEST] <teratorn> cmpscabral: probably not [18:44:21 CEST] <cmpscabral> ok, thanks [18:55:24 CEST] <momomo> guys I am connecting using ffmpeg to stream ... but then I get this message mid through: [18:55:30 CEST] <momomo> 217247 : 1002352814 : [https @ 0x4b679e0] Stream ends prematurely at 29071192 [18:57:38 CEST] <momomo> Then it tries 10 times and gives up. However, i am not sure it is actually retrying correctly. [18:58:01 CEST] <momomo> and the process dies [18:58:46 CEST] <momomo> when I had -reconnect_stream 1 then it didnt die, but it kept looping the error ... it jumped back to the stream, and then it got into the same error it appears, then jumps back ... [18:58:56 CEST] <momomo> so using -reconnect_stream 1 is not an option [18:59:27 CEST] <momomo> but is there a way to have a higher error tolerance ... it continues until its ready again [19:16:31 CEST] <ChocolateArmpits> momomo, how are you executing your stream ? [19:22:45 CEST] <Tatsh> i have been unable to use my GPU's MJPEG decoding capabilities [19:22:56 CEST] <Tatsh> i get an error [19:24:16 CEST] <Tatsh> crap, what are the ffmpeg hot keys? [19:24:24 CEST] <Tatsh> besides q; i hit h and i don't know how to go back [19:24:33 CEST] <Tatsh> pressing h again shows a hex dump [19:34:01 CEST] <ChocolateArmpits> Tatsh, just click the same hotkey again [19:38:51 CEST] <Tatsh> where are the hotkeys documented? [19:47:23 CEST] <nyuszika7h> when I do `ffmpeg -i in.mkv -i in.mka -filter_complex '[0:v:0][1:s:0]overlay[vid]' -map '0:a:0' -c:v h264_qsv -c:a aac -b:a 128k out.mkv -y`, I get this error despite having not specified any rate control mode explicitly: [19:47:34 CEST] <nyuszika7h> [h264_qsv @ 0000000002749a20] Selected ratecontrol mode is not supported by the QSV runtime. Choose a different m [19:47:35 CEST] <nyuszika7h> Error initializing output stream 0:0 -- Error while opening encoder for output stream #0:0 - maybe incorrect para [19:47:36 CEST] <nyuszika7h> such as bit_rate, rate, width or height [19:47:42 CEST] <nyuszika7h> (also how do I use CRF or something similar with h264_qsv?) [19:48:49 CEST] <nyuszika7h> er, add a -map '[vid]' before -map '0:a:0' [19:48:54 CEST] <nyuszika7h> is my actual command [19:59:51 CEST] <OzgrK> hi, i'd like to report that native aac encoder produces glitches for low frequency sine waves, at least in my case. is this known? you may reproduce it with... [19:59:53 CEST] <OzgrK> ffmpeg -f lavfi -i aevalsrc="sin(32*2*PI*t):s=48000:d=1800" -c:a aac -b:a 384 -f matroska pipe:1 | ffplay - [20:10:00 CEST] <Tatsh> nyuszika7h, the options are never the same as libx264 [20:10:15 CEST] <durandal_170> atomnuker: ^^ [20:10:18 CEST] <Tatsh> so you have to trial and error with qmin/qmax [20:10:22 CEST] <nyuszika7h> I didn't specify any options related to that though [20:10:25 CEST] <Tatsh> if that's available [20:10:37 CEST] <nyuszika7h> except the filter which is not specific to the video codec I'd think [20:17:06 CEST] <Tatsh> nyuszika7h, filters are usually limited [20:17:08 CEST] <Tatsh> for hardware [20:17:55 CEST] <Tatsh> ffmpeg -h encoder=h264_qsv [20:17:59 CEST] <Tatsh> will show all options [20:18:36 CEST] <Tatsh> for me with CUVID there's no -crf, you have to use -cq instead which is a different scale and you have to trial and error with these sort of things [20:18:44 CEST] <Tatsh> see what quality/bitrate is acceptable to you [20:18:51 CEST] <atomnuker> nyuszika7h: the bitrate is insanely low [20:19:02 CEST] <atomnuker> 384 bytes per second [20:19:21 CEST] <furq> you mean OzgrK [20:19:59 CEST] <Tatsh> btw is h264_qsv for decoding or encoding or both? [20:20:06 CEST] <Tatsh> i don't have it on my CPU afaik [20:20:09 CEST] <furq> qsv is for encoding [20:20:14 CEST] <atomnuker> set it to 384k and it'll be right [20:20:52 CEST] <furq> quicksync itself has hardware decoding features but it's accessed through vaapi iirc [20:21:28 CEST] <Tatsh> well i must not have it built in [20:21:36 CEST] <furq> which cpu [20:21:51 CEST] <Tatsh> Intel(R) Core(TM) i7-5930K CPU @ 3.50GHz GenuineIntel [20:21:58 CEST] <furq> that should definitely have quicksync [20:22:28 CEST] <nyuszika7h> Tatsh: only used for encoding because it threw errors when trying to decide Hi10P [20:22:29 CEST] <Tatsh> yea but i don't have it in my build of ffmpeg on gentoo [20:22:33 CEST] <Tatsh> and there's no USE flag for it [20:22:40 CEST] <furq> you should be able to use h264_vaapi for encoding now [20:22:41 CEST] <Tatsh> unless that USE flag is vaapi [20:22:53 CEST] <furq> qsv isn't recommended any more afaik [20:23:14 CEST] <nyuszika7h> I just want it to accelerate temporary transcodes to burn in PGS subtitles [20:23:50 CEST] <nyuszika7h> And it seems I can't use CUVID on my laptop with dual graphics, even if I tell it to run ffmpeg.exe on the NVIDIA GPU [20:24:13 CEST] <nyuszika7h> h264_cuvid just gives me the exact same fps as libx264 [20:24:35 CEST] <furq> i don't think any hwdec stuff will work with hi10p [20:24:53 CEST] <Tatsh> 1080 has it [20:24:57 CEST] <Tatsh> i think [20:25:02 CEST] <furq> for avc? [20:25:06 CEST] <Tatsh> i only have a 980 so i lose [20:25:11 CEST] <furq> newer stuff has hevc main10 support [20:25:15 CEST] <furq> but that's not the same [20:25:16 CEST] <Tatsh> yes hevc [20:25:26 CEST] <Tatsh> my GPU lacks decoding support for main10 [20:25:28 CEST] <Tatsh> but 1080 has it [20:26:59 CEST] <Tatsh> i'm only finding the hardware acceleration stuff useful for lack of time [20:27:25 CEST] <Tatsh> although for 4K it's almost unreasonable to use CPU for anything longer than a few minutes [20:27:37 CEST] <furq> if your build has vaapi then use that instead [20:35:34 CEST] <ArsenArsen> Hello! Is there any examples/documentation on how to use FFMpeg in C/C++ code in order to do screen recording on Linux, Windows, and Mac OS X? Thanks in advance :) [20:37:27 CEST] <c_14> probably not, you'd have to use different input devices on each anyway (xcbgrab on linux, avfoundation on linux and one of the options on windows) [20:38:07 CEST] <ArsenArsen> I've done some reading up on it, I knew about x11grab already, and GDI, but I have no idea what's actually used on OS X. [20:38:43 CEST] <c_14> I think it's avfoundation on osx, I accidentally wrote linux [20:38:46 CEST] <BtbN> Just use OBS [20:39:06 CEST] <ArsenArsen> Can I even do that in my program? [20:39:24 CEST] <BtbN> It has a library, so I'd assume so [20:39:44 CEST] <BtbN> The ffmpeg screen capture methods are rather slow and not optimal for anything but a simple screencast [20:39:51 CEST] <ArsenArsen> Oh, it does? But, still, I think it'd be an overkill to be honest [20:40:02 CEST] <ArsenArsen> Wait doesn't OBS use ffmpeg? [20:41:08 CEST] <BtbN> for encoding and decoding, yes. [20:41:32 CEST] <ArsenArsen> Ahhhh I see [20:41:32 CEST] <OzgrK> atomnuker: it was k, 384 when i tried, i have missed "k" while writing here, sorry, even for 384k and even for 800k (not sure it is valid though:)) it produces glitches. you may reproduce it with... [20:41:41 CEST] <OzgrK> ffmpeg -f lavfi -i aevalsrc="sin(32*2*PI*t):s=48000:d=1800" -c:a aac -b:a 800k -f matroska pipe:1 | ffplay - [20:41:54 CEST] <Tatsh> furq, cool [20:42:00 CEST] <Tatsh> seems i could use vaapi for a faster deinterlace [20:42:05 CEST] <Tatsh> and scale [20:43:56 CEST] <Tatsh> me hating the blurriness of my phone's camera is making me want to buy a pro camera [20:44:00 CEST] <Tatsh> for both video and photo [20:45:08 CEST] <atomnuker> OzgrK: if it bothers you and you really want to encode just sine waves you can use -aac_coder anmr [20:45:30 CEST] <furq> OzgrK: it's not valid but it just clamps to the maximum valid bitrate [20:45:35 CEST] <ArsenArsen> At first I wanted to use Qt to capture the screen but came across a few QObject related limitations (such as the non thread safety) of QPixmaps and similar. I could achieve ~100FPS with conversion to RGBA8 though. [20:45:44 CEST] <atomnuker> else the defaults (not just for aac but really encoders in general) are good enough for music and speech, not sine waves [21:05:49 CEST] <OzgrK> atomnuker: i confirm that -aac_coder anmr works fine, even with down to -ab 1k :) however i just wanted to report the issue with native aac encoder, since it is not declared experimental anymore but just requires fine tuning for some sound characteristics [22:07:56 CEST] <grren_flag> evening, I need some help with live freetoair dvb-t streams going out of sync converted from astra->ffmpeg. I have 6 streams running and they can play fine for hours, then one wor two will drop out of sync. any help appreciated [22:10:41 CEST] <grren_flag> ffmpeg -i http://192.168.1.200:2345/stream1 -vcodec libx264 -preset superfast -s 640x360 -b:v 950k -b:a 96k -ar 44100 -bufsize 1900k -f flv rtmp://192.168.1.200:1935/live/stream1 [22:12:17 CEST] <grren_flag> i have noticed watching the raw stream that with the signal in our area it will get the odd glitch and resume to normal but not sure if that is the cause of the stream going out of sync [22:14:18 CEST] <dystopia_> dvb-t and astra ? [22:14:29 CEST] <dystopia_> astra runs satellites [22:14:29 CEST] <grren_flag> cesbo astra [22:14:33 CEST] <dystopia_> ahh [22:15:20 CEST] <dystopia_> the transmission glitching is probably what throws it out of sync yes [22:15:40 CEST] <dystopia_> can happen on a normal encode too [22:15:46 CEST] <grren_flag> sorry i have a TBS dvbt/c tuner and use cesbo astra to convert to http and then run through ffmpeg [22:16:40 CEST] <grren_flag> I was wondering if there was something i could put in to either use timestamp or similar to put the stream back into sync [22:17:04 CEST] <dystopia_> probably best to restart the steam periodically [22:17:11 CEST] <dystopia_> perhaps at the start of every hour [22:18:19 CEST] <grren_flag> that is how i originally did things, made a crontab to kill and start ffmpeg every few hours, but it can happen anytime [22:18:45 CEST] <grren_flag> that and became annoying the stream stopping for a short time every hour [22:18:57 CEST] <dystopia_> if you run it from a script you could do this https://paste.ofcode.org/R89n9C9Pv9tTLEnze7Pvmx [22:19:15 CEST] <dystopia_> should only be a couple of seconds [22:20:12 CEST] <grren_flag> i went the fancy route and installed monit and ran a script to check every 10 seconds if stream died, but that wasnt really a fix as the stream itself stayed solid [22:21:34 CEST] <grren_flag> do you think changing output from rtmp to mpegts would help or is that not relevant ? [22:22:04 CEST] <dystopia_> you can try but i don't think it would matter [22:22:32 CEST] <dystopia_> you could also try not encoding audio and just doing -acodec copy [22:22:42 CEST] <dystopia_> i doubt it would inflate stream size much [22:22:48 CEST] <dystopia_> 192k vs 96k [22:22:56 CEST] <grren_flag> good idea [22:23:34 CEST] <grren_flag> at least thats one thing to try [22:24:39 CEST] <furq> i doubt that'll work if you're going from dvb-t to rtmp [22:24:50 CEST] <furq> dvb-t audio is usually either mp2 or ac3 [22:24:54 CEST] <furq> neither of which are supported in flv iirc [22:26:05 CEST] <grren_flag> ok so mpegts supported ? [22:26:14 CEST] <furq> not by rtmp [22:26:29 CEST] <grren_flag> source is ac3 [22:26:41 CEST] <furq> mpegts supports ac3, but rtmp doesn't support mpegts [22:26:44 CEST] <furq> it's flv only [22:27:44 CEST] <grren_flag> well i only ouput to rtmp as thats all i have tried [22:28:01 CEST] <grren_flag> its not a requirement [22:28:17 CEST] <furq> i take it you have a good reason for not just watching the source stream [22:28:45 CEST] <grren_flag> yes mainly for watching over 3/4g [22:29:42 CEST] <grren_flag> and we have a holiday apartment we are planning to use [22:31:26 CEST] <grren_flag> just very frustrating seeing ffmpeg working really well for our needs and the sync issue ruining the whole experience [22:31:41 CEST] <furq> well yeah rtsp supports ac3 audio but i couldn't tell you a worthwhile rtsp server [22:31:46 CEST] <furq> and also i doubt it'll fix the issue [22:31:57 CEST] <furq> in my experience ffmpeg tends to have worse sync issues when copying streams [22:33:23 CEST] <grren_flag> what i can say is thata while back I replicated the whole process from a fta sat stream using an old hauppage card I had and it worked a charm, ran for about 2 weeks with no sync issues [22:33:43 CEST] <grren_flag> using the eact same ffmpeg comamnd [22:34:03 CEST] <furq> yeah it wouldn't surprise me if transmission errors were throwing it out of sync [22:34:30 CEST] <furq> ffmpeg isn't very robust when things go wrong [22:34:49 CEST] <grren_flag> just my luck that the 6 channels we watch arenet available on sat [22:35:13 CEST] <grren_flag> yes ive seen that, but on other hand when things are right its also very robust [22:35:56 CEST] <furq> you could maybe try adding -xerror to the command [22:36:19 CEST] <furq> that'll quit as soon as an error happens [22:36:24 CEST] <furq> i'm not sure if a decode error will trigger it though [22:36:36 CEST] <grren_flag> ok reading up now on xerror [22:37:10 CEST] <grren_flag> i agree i think that if ffmpeg saw the error id expect the stream to drop and monit to restart the stream [22:37:23 CEST] <grren_flag> but from monit los the streams never drop [22:37:25 CEST] <furq> well you could just run it in a shell loop if that works [22:37:35 CEST] <furq> decoder errors normally show up as warnings though [22:38:19 CEST] <grren_flag> how do you mean shell loop ? [22:38:25 CEST] <furq> while 1; do ffmpeg ...; done [22:38:40 CEST] <grren_flag> right [22:39:59 CEST] <grren_flag> i'll give xerror a go anyway, do I just put that at the end of the commad ? [22:40:07 CEST] <grren_flag> after the rtmp output ? [22:40:13 CEST] <furq> probably before the output [22:40:50 CEST] <grren_flag> ffmpeg -i http://192.168.1.200:2345/stream1 -vcodec libx264 -preset superfast -s 640x360 -b:v 950k -b:a 96k -ar 44100 -bufsize 1900k -xerror -f flv rtmp://192.168.1.200:1935/live/stream1 [22:40:53 CEST] <grren_flag> look right ? [22:40:56 CEST] <furq> yeah [22:41:00 CEST] <grren_flag> ok thanks [22:41:06 CEST] <furq> it's a global option so it might work after the output, i forget [22:41:09 CEST] <furq> it'll definitely work before though [22:41:32 CEST] <grren_flag> ok ill give it a go and report later on findings. really appreciate the help [22:41:55 CEST] <grren_flag> dystopia also, thanks [22:43:24 CEST] <OzgrK> I have a question about vp9 decoding in ff*. i made a video using -i https://upload.wikimedia.org/wikipedia/commons/c/c4/PM5544_with_non-PAL_sig… -vf loop=size=1:loop=-1,hue=h=t*360/57600 -t 57600 [22:43:28 CEST] <OzgrK> and encoded it as lossless vp9 using -pixel_format yuv420p -c:v libvpx-vp9 -lossless 1 [22:43:36 CEST] <OzgrK> the resulting videos decodes "wrong?" with ffplay. I'll explain what i mean by "wrong?" but vlc, and even firefox and chrome decodes "right". since the video is decoded "wrong" in ff*, ffmpeg encodes what it decodes and my transcodes to other encoders are also "wrong?". [22:43:56 CEST] <OzgrK> by saying "wrong" i mean that the colors in the second columns from each side of the test chart in input https://upload.wikimedia.org/wikipedia/commons/c/c4/PM5544_with_non-PAL_sig… are gone and those columns are just gray when ff* decodes. [22:44:07 CEST] <OzgrK> There are tiny vertical color lines in the original image as you can see in the png image in the link above. vlc, firefox and chrome decodes as the input png image, but ff* plays "wrong?" and also mplayer(smplayer) decodes the same as ff*. Any idea? Is there a problem with decoding vp9 in ff*? thanks [00:00:00 CEST] --- Tue May 23 2017

1 0

ffmpeg-devel.log.20170521
by burek 22 May '17

22 May '17

[02:12:29 CEST] <jamrial> michaelni: does https://pastebin.com/G06uardD fixes it for you? [02:13:19 CEST] <jamrial> in any case, perhaps this tempprefix feature should be removed, seeing that this meged commit in question gets rid of mktemp -u [02:25:31 CEST] <jamrial> ah, i see now you added it to make configure run faster with ccache. in that case better not remove it if possible [02:27:06 CEST] <michaelni> jamrial, the pastebin patch fixes it [02:27:10 CEST] <michaelni> thx [02:27:21 CEST] <jamrial> no prob [02:30:51 CEST] <cone-388> ffmpeg 03James Almer 07master:8794cfbbc5dd: configure: fix --tempprefix option [02:35:52 CEST] <cone-388> ffmpeg 03Michael Niedermayer 07master:6310fc714de3: avcodec/aacsbr_fixed: Fix multiple runtime error: shift exponent 170 is too large for 32-bit type 'int' [02:35:53 CEST] <cone-388> ffmpeg 03Michael Niedermayer 07master:384508b2ff69: avcodec/sbrdsp_fixed: fix runtime error: left shift of 1 by 31 places cannot be represented in type 'int' [04:45:21 CEST] <cone-388> ffmpeg 03Michael Niedermayer 07master:1d04fc94e102: avcodec/mlpdsp: Fix runtime error: signed integer overflow: -24419392 * 128 cannot be represented in type 'int' [04:45:22 CEST] <cone-388> ffmpeg 03Michael Niedermayer 07master:d66193252b40: avcodec/takdec: Fix runtime error: left shift of negative value -63 [04:45:23 CEST] <cone-388> ffmpeg 03Michael Niedermayer 07master:c3547dcbc326: avcodec/aac_defines: Fix: runtime error: left shift of negative value -2 [06:20:10 CEST] <thebombzen> did any of the SSH code change recently? I'm getting a segfault when I attempt to play a file over sftp:// [06:20:28 CEST] <thebombzen> to localhost that is. ffmpeg -i as well. and also I can ffplay the local file [06:31:07 CEST] <thebombzen> ah, okay, found the bug. so it appears if you try to use ffmpeg's sftp:// to read from a file where you don't have public key access, it'll segfault. [06:31:11 CEST] <thebombzen> With pubkey, it works fine. [06:31:49 CEST] <thebombzen> it should ask your for the password on the command line for password authentication, rather than segfault. or at least, exit gracefully. [06:33:29 CEST] <thebombzen> Here's my log by the way: http://sprunge.us/CIcd [06:34:06 CEST] <thebombzen> Should I put this on trac, or is this a sufficient bug report? [06:49:13 CEST] <thebombzen> update: ran valgrind on it: http://sprunge.us/RDHd [11:06:24 CEST] <atomnuker> rcombs: ping [11:06:29 CEST] <rcombs> PONG [11:07:13 CEST] <rcombs> asking for those patches again? [11:07:33 CEST] <atomnuker> yep, I think its time to push them [11:07:37 CEST] <atomnuker> lgtm'd by 2 people [11:08:02 CEST] <atomnuker> (the cover art attaching one at least) [13:36:32 CEST] <ubitux> michaelni: ah my bad, sorry i missed it. [13:36:50 CEST] <ubitux> i see that jamrial fixed it quickly [14:51:32 CEST] <cone-986> ffmpeg 03Michael Niedermayer 07master:3b09d7c264f3: avcodec/wavpack: Fix multiple runtime error: signed integer overflow: 548 * -2147483648 cannot be represented in type 'int' [14:51:32 CEST] <cone-986> ffmpeg 03Michael Niedermayer 07master:955db411929a: avcodec/takdec: Fix runtime error: signed integer overflow: 8192 * 524308 cannot be represented in type 'int' [14:51:32 CEST] <cone-986> ffmpeg 03Michael Niedermayer 07master:ec2b76aab44f: avcodec/vmnc: Check location before use [15:26:10 CEST] <jamrial> atomnuker: no, it can't be pushed as is. did you miss my review pointing out a couple issues? [16:04:44 CEST] <atomnuker> jamrial: sorry, I didn't notice you replied as well [16:20:17 CEST] <cone-986> ffmpeg 03Michael Niedermayer 07master:d4ee76780869: avcodec/vp9block: fix runtime error: signed integer overflow: 196675 * 20670 cannot be represented in type 'int' [16:20:18 CEST] <cone-986> ffmpeg 03Michael Niedermayer 07master:efeb47fd5d5c: avcodec/mpeg4videodec: Check for multiple VOL headers [20:34:16 CEST] <thebombzen> guys do you want me to submit a trac about the above bug [20:47:21 CEST] <cone-270> ffmpeg 03Michael Niedermayer 07master:b5228e44c7f3: avcodec/aacdec_fixed: Fix runtime error: shift exponent 34 is too large for 32-bit type 'int' [20:55:58 CEST] <jamrial> thebombzen: yes [00:00:00 CEST] --- Mon May 22 2017

1 0

ffmpeg.log.20170521
by burek 22 May '17

22 May '17

[00:02:26 CEST] <arpu> hmm the rtmpproto input timeout is only for rtmp listen server [00:02:50 CEST] <arpu> any idea how i can set the read timeout without using librtmp [00:13:48 CEST] <arpu> looks like this is a bug https://trac.ffmpeg.org/ticket/6372 [00:13:58 CEST] <arpu> anyone around help me to fix this [00:14:07 CEST] <arpu> should be set the tcp socket timeout [10:07:44 CEST] <Pandela> Aye, I'm having some trouble. Tired of trying to figure it out, but when I try to compile ffmpeg 3.3 in ubuntu with --enable-shared I get this [10:07:46 CEST] <Pandela> https://pastebin.com/raw/iJMQtxTk [10:09:47 CEST] <Pandela> pls send help [10:51:13 CEST] <c_14> have you tried recompiling with -fPIC? [14:45:38 CEST] <exutux> hi all [14:46:53 CEST] <exutux> I'm in trouble converting an .mkv video to .mp4 or .avi I need to keep subtitle but I cannot find how... this is my output paste https://pastebin.com/e0qBGGsa [14:47:01 CEST] <exutux> How can I keep subs? [14:50:15 CEST] <dystopia_> your log does not show what command you issued [14:51:39 CEST] <dystopia_> but what ever [14:52:10 CEST] <dystopia_> just make sure your have "-c:s copy" somewhere in your line and not "-sn" [14:52:18 CEST] <dystopia_> without the " doodar's of course [14:52:32 CEST] <exutux> ffmpeg -i "file.mkv" -codec:v mpeg4 -codec:a libmp3lame -c:s copy file.mp4 [14:54:13 CEST] <exutux> dystopia_: -c:s copy does't work [14:56:45 CEST] <exutux> Stream mapping: Stream #0:0 -> #0:0 (h264 -> mpeg4) Stream #0:1 -> #0:1 (aac -> libmp3lame) [14:56:48 CEST] <exutux> Press [q] to stop, [?] for help [14:57:23 CEST] <exutux> Still keeping only map 0:0 and map 0:1 [15:00:10 CEST] <exutux> trying this too https://pastebin.com/mb8AJU8X [15:17:53 CEST] <furq> -c:s mov_text for mp4 [15:17:58 CEST] <furq> although it should have just done that automatically [15:18:02 CEST] <furq> avi can't have embedded subtitles [15:18:12 CEST] <furq> also are you really sure you want to reencode this [15:18:37 CEST] <furq> mpeg4 and mp3 are much worse than h264 and aac, and mp4 is quite happy with h264 and aac [15:23:22 CEST] <exutux> furq: I need reencode because my tv can't recognize h264 and ac ( old TV ) :P [15:23:55 CEST] <exutux> I will try -c:s move_text [15:25:39 CEST] <exutux> ffmpeg -i Tredici.Oltre.i.Perchè.2017.ENG.SUB.ITA.mkv -codec:v mpeg4 -standard PAL -q:v 3 -codec:a libmp3lame -c:s mov_text -ac 2 -ab 128k -t 00:02:00 Tredici.Oltre.i.Perchè.2017.ENG.SUB.ITA.mp4 [15:26:15 CEST] <exutux> Still map only 0 and 1 stream :| [15:26:30 CEST] <exutux> Stream mapping: Stream #0:0 -> #0:0 (h264 -> mpeg4) Stream #0:1 -> #0:1 (aac -> libmp3lame) [15:32:05 CEST] <exutux> furq: subtitles are embed into mkv file and it is mapped on stream 0:3 [15:39:35 CEST] <exutux> furq: there is some way to get embed subtitles from mkv? [15:42:51 CEST] <exutux> furq: can I send you some little part of that video so you can try something about it? [16:11:37 CEST] <exutux> :| anyone have some suggest for me? [16:15:53 CEST] <dystopia_> dump subs with mkvmergegui? [16:15:54 CEST] <dystopia_> mmg [16:21:38 CEST] <exutux> dystopia_: I'll try [16:33:11 CEST] <exutux> dystopia_: grrrrrrrrrrrrrrrr -scodec mov_text "unknown encoder" [16:33:19 CEST] <exutux> getting me crazy [16:36:18 CEST] <exutux> ffmmeg -i file.mp4 -i file.srt file2.mp4 doesn't embed subs too [16:45:51 CEST] <zyme> exutux: what are you using to play the stream on your "old TV" ? [16:47:12 CEST] <exutux> zyme: is a program inside tv I can play video inside "media" and it's not compatible with some audio and video format [16:48:24 CEST] <zyme> so it's networked not usb or sd-storage only? (newer then any tv my family's ever owned =) [16:48:34 CEST] <exutux> usb [16:49:05 CEST] <exutux> I insert usb [16:50:42 CEST] <exutux> zyme: but the problem is how can I keep that subtitles when I encode video [16:50:50 CEST] <exutux> grrrrrr [16:51:44 CEST] <zyme> probably depends on the tv, try to search for a manual pdf and check if theres any firmware updates while your at it.. [16:54:29 CEST] <exutux> yeah the problem is the TV for sure...but theresn't way to see some not supprted format on that tv, for this reason I have to convert always all videos [16:55:04 CEST] <exutux> and no way to upgrade [16:55:54 CEST] <zyme> as a poor person I got a family member to buy me a $30 ChromeCast2 1.5 yrs ago, it's about the cheapest streaming decoder around and theres free software for every smartphone and os. Plays just about anything for me is it's <=1080p + the bitrate is under 300mbit/s~ish [16:56:00 CEST] <exutux> maybe there is no way to keep subtitles on mp4 video [16:56:09 CEST] <zyme> is the tv pre-HDMI? [16:56:16 CEST] <exutux> no zyme [16:57:03 CEST] <exutux> so your solution is to buy a new tv? :D It is good suggest but not a solution for now [16:59:36 CEST] <zyme> My suggestion was a better device hooked up to the TV, there's a large market out there I merely named one of the cheapest and cross-platform compatible options... [17:00:09 CEST] <zyme> Personally I would be converting everything to HEVC except the quality is terrible with reencodes. [17:01:20 CEST] <exutux> well [17:05:13 CEST] <zyme> I'm not that great with encoding as it is, but unless it's something avalible in bluray transcoding a file is almost always larger ig you don't wan't to loose a lot of detail. Even going from xvid avi's to HEVC usually breaks even although I'm sure close freeze-frame inspection would show quite a bit of detail lost. [17:11:07 CEST] <exutux> zyme: there are no format that I had tried that keeps subtitles [17:11:15 CEST] <exutux> I haven't got any ideas [17:25:15 CEST] <zyme> I've been wondering recently if there's any way to avoid decoding similar codec's and converting+add newer lossless methods (like what would be the effect of taking the image groupings in a x264 and combining some into the larger 265 formats with the previously used tracking data, increasing color bit debt and Layers), maybe some form of identifying color [17:25:15 CEST] <zyme> gradients and detecting banding and graphing a projected original gradient for automatic color remapping.. (a similar process is commonly done live w/audio processors), - That last idea is just something I wonder if it's been done or is too complex to automate yet - since recodes seem to show color banding more than any other distortion by far) [17:26:04 CEST] <zyme> exutux: I think handbreak lets you hardcode subtitles, surely ffmpeg has an option or filter to do the same? [17:27:11 CEST] <klaxa> it does [17:27:33 CEST] <klaxa> https://ffmpeg.org/ffmpeg-filters.html#subtitles-1 [17:28:19 CEST] <zyme> it not working or you just want to keep your subs soft-coded? [17:28:49 CEST] <klaxa> for potato-grade TVs i would also suggest hardsubbing if you definitely want subtitles [17:30:39 CEST] <klaxa> or you just hook up a raspberry pi :I [17:30:52 CEST] <klaxa> like, even the pi zero should be able to handle it? [17:30:56 CEST] <zyme> I'm not sure why, but..btw, do you use windows on any of your PC's? [17:31:01 CEST] <klaxa> the new one even comes with wifi [17:31:15 CEST] <zyme> I'm curious if you've looked at these files with MediaInfo [17:32:53 CEST] <zyme> it should point out exactly what codecs containers subs etc and their versions, and you might be able to use it as a guide with videos you know to work instead of all this trial/error. [17:38:53 CEST] <zyme> While I'm a fan of SoC's for media players they're a little isolated compared to other products that have a nice integration across different devices - it depends on what you have, need, +want though. or if your a developer with free time =) [17:45:35 CEST] <zyme> there was an android app for Chromecast emulating, kinda like those airplay servers out there, I once thought about using an adrino with that loaded and mod-ing it to accept more than just 264, it could probably play the role of a PC at the same time and have the power to conference in HEVC for tons less than a new atv.. [17:54:02 CEST] <knowsnothing> is there a filter that can be used to crossfade video and audio? [19:21:21 CEST] <ppw> why would a freshly cross compiled static ffmpeg.exe throw out a DLL missing error? [19:21:40 CEST] <ppw> libcrypto-1_1-x64.dll [19:22:39 CEST] <furq> what's your configure string [19:23:48 CEST] <ppw> https://pastebin.com/r6PyXLUL [19:24:56 CEST] <furq> yeah that's not fully static [19:25:16 CEST] <furq> that'll still use dynamic external libs in preference to static libs [19:25:21 CEST] <furq> you need --extra-ldflags=-static [19:25:59 CEST] <furq> --pkg-config='pkg-config --static' --extra-ldflags='-static -static-libgcc -static-libstdc++' [19:26:05 CEST] <furq> you probably want that for mingw [19:33:57 CEST] <ppw> well, it seems to be working. thanks furq! [20:24:23 CEST] <ppw> how would I preserve the original filename attributes after transcoding a video? [20:24:44 CEST] <ppw> I'm not talking about metadata, just a simple file changed/created attribute preservation [20:32:37 CEST] <ChocolateArmpits> that's OS specific, not something ffmpeg can do about [20:34:08 CEST] <ppw> ok. [20:35:15 CEST] <ppw> using yadif and mcdeint results in very verbose logging output which severely hampers encoding speed. what can I do to bypass this aside from completely disabling output with -loglevel 0? [20:36:26 CEST] <ChocolateArmpits> ppw, did you try out bwdif for deinterlacing ? [20:36:39 CEST] <ppw> no, only yadif and mcdeint so far. [20:36:47 CEST] <ChocolateArmpits> try it, you may like it [20:36:55 CEST] <ppw> ok [20:37:14 CEST] <ChocolateArmpits> combines yadif and w3fdif [21:07:53 CEST] <asie> Is there a way to make force_original_aspect_ratio in "-vf scale=320:240:force_original_aspect_ratio=decrease" respect codec width/height divisibility requirements (such as "height not divisible by 2 (320x213)")? [21:10:29 CEST] <DHE> what are you trying to accomplish? the SAR will be recalculated so that the player will render it properly. if you want to rescale to approximate resolution ratios use -2 for one of width or height to calculate that value as a multiple of 2 [21:11:02 CEST] <asie> i'm working on an automated transcoder for a large amount of videos, and i was expecting to be able to make a video fit in a maximum width, height *and* keep its aspect ratio [21:11:27 CEST] <asie> i can potentially run the necessary calculations outside of ffmpeg, was just curious if ffmpeg already has a solution for that [21:13:05 CEST] <DHE> well for example h264 can store an aspect ratio in the codec itself in case the resulting image should have non-square pixels. using this you can set any "scale=xx:yy" numbers you want and the video will have the same shape on the player [21:30:13 CEST] <ChocolateArmpits> I do size calcs outside off fmpeg myself, ffprobe not always gives aspect ratio info so I figure it may not be available to filters either [21:31:40 CEST] <ChocolateArmpits> Other times the aspects are wonky so I just have exceptions, like 16/11 sar [21:44:53 CEST] <DHE> but if you don't specify anything, the image aspect ratio will be preserved (by recalculating the sample aspect ratio) [21:44:58 CEST] <DHE> right? [21:45:43 CEST] <DHE> so usually you can do -vf scale=-2:240 to get a closest available width as a multiple of 2 and the player will rescale to retain the original aspect ratio [00:00:00 CEST] --- Mon May 22 2017

1 0