#ffmpeg-devel on 2023-12-15 — irc logs at libera.irclog.whitequark.org

2023-11-11 01:05 michaelni changed the topic of #ffmpeg-devel to: Welcome to the FFmpeg development channel | Questions about using FFmpeg or developing with libav* libs should be asked in #ffmpeg | This channel is publicly logged | FFmpeg 6.1 has been released! | Please read ffmpeg.org/developer.html#Code-of-conduct

00:09 elastic_dog has joined #ffmpeg-devel

00:53 navi has quit [Quit: WeeChat 4.0.4]

01:28 philipl has quit [Quit: leaving]

01:29 iive has quit [Quit: They came for me...]

01:32 philipl has joined #ffmpeg-devel

01:52 thilo has quit [Ping timeout: 246 seconds]

01:54 thilo has joined #ffmpeg-devel

01:54 thilo has quit [Changing host]

01:54 thilo has joined #ffmpeg-devel

02:19 cone-125 has quit [Quit: transmission timeout]

02:24 dellas has quit [Remote host closed the connection]

03:00 lemourin has quit [Read error: Connection reset by peer]

03:01 lemourin has joined #ffmpeg-devel

03:17 jamrial has quit []

04:13 AbleBacon has quit [Read error: Connection reset by peer]

04:34 Sk0tik has quit [Ping timeout: 245 seconds]

04:51 mkver has joined #ffmpeg-devel

07:06 mkver has quit [Remote host closed the connection]

07:06 mkver has joined #ffmpeg-devel

07:09 j45 has quit [Ping timeout: 255 seconds]

07:10 j45 has joined #ffmpeg-devel

07:10 j45 has quit [Changing host]

07:10 j45 has joined #ffmpeg-devel

07:36 Krowl has joined #ffmpeg-devel

08:15 Krowl has quit [Read error: Connection reset by peer]

08:24 Krowl has joined #ffmpeg-devel

08:27 kurosu has joined #ffmpeg-devel

09:05 qeed has quit [Quit: Leaving]

09:27 Krowl has quit [Read error: Connection reset by peer]

10:00 Krowl has joined #ffmpeg-devel

10:17 dellas has joined #ffmpeg-devel

10:18 <thilo> ePirat: I knew I didn't like the MAX_ versions

10:35 cone-538 has joined #ffmpeg-devel

10:35 <cone-538> ffmpeg Martin Storsjö master:081d69b78df8: avfoundation: Fix version checks

10:35 <ePirat> thilo, I told you they need the @available

10:37 <ePirat> the MAX_* ones allow the API being used if the SDK building with is recent enough, needing runtime checks. The MIN_* ones allow the API being used if the deployment target is recent enough, with no way they getting used even if at runtime the OS is new enough.

10:39 <thilo> ePirat: next time use a bigger stick for me 0:-)

10:39 <ePirat> anyway wbs fix looks good, feel free to push

10:40 <thilo> ^

10:40 <ePirat> oh nevermind :D

10:45 rvalue has quit [Read error: Connection reset by peer]

10:45 rvalue has joined #ffmpeg-devel

11:05 <Lynne> what advantages would a potential avpacket+avframe unification have for us?

11:05 <Lynne> bitstream filters in lavfi? no wrapped avframes?

11:22 odrling has quit [Remote host closed the connection]

11:24 odrling has joined #ffmpeg-devel

11:39 jamrial has joined #ffmpeg-devel

11:42 dellas has quit [Remote host closed the connection]

11:47 navi has joined #ffmpeg-devel

11:51 <kierank> Why not just use avbufferref

11:54 <Lynne> it's not exactly pleasant to work with, txproto uses avbufferrefs for everything from objects to events to contexts and I regret it

11:58 mkver has quit [Ping timeout: 252 seconds]

12:17 odrling has quit [Remote host closed the connection]

12:17 odrling has joined #ffmpeg-devel

12:27 lemourin has quit [Ping timeout: 240 seconds]

12:32 lemourin has joined #ffmpeg-devel

12:34 odrling has quit [Remote host closed the connection]

12:37 ngaullier has joined #ffmpeg-devel

12:43 Traneptora has quit [Remote host closed the connection]

12:43 Traneptora has joined #ffmpeg-devel

12:44 <cone-538> ffmpeg Wenbin Chen master:e01afa1c6d53: libavfilter/vf_dnn_detect: Fix an incorrect expression

12:44 <cone-538> ffmpeg Wenbin Chen master:833722343aa5: libavfilter/vf_dnn_detect: Add NULL pointer check

12:44 <cone-538> ffmpeg Wenbin Chen master:a1e8892a8a2e: libavfilter/vf_dnn_detect: Fix uninitialized variables error

12:44 <cone-538> ffmpeg Wenbin Chen master:22bebfa5e6cf: libavfilter/vf_dnn_detect: Set used pointer to NULL

12:45 odrling has joined #ffmpeg-devel

12:45 dellas has joined #ffmpeg-devel

12:55 Krowl has quit [Read error: Connection reset by peer]

13:12 NuoMi has joined #ffmpeg-devel

13:14 ccawley2011 has joined #ffmpeg-devel

13:20 Krowl has joined #ffmpeg-devel

13:25 <kierank> Lynne: macros

13:26 dykai has joined #ffmpeg-devel

13:26 <Lynne> I do that too, but you lose all type checking benefits

13:36 NuoMi has quit [Read error: Connection reset by peer]

13:46 mkver has joined #ffmpeg-devel

13:54 dykai has quit [Ping timeout: 276 seconds]

13:58 <JEEB> mkver: so was this enough of a change for you with regards to the linecount for the tests? https://patchwork.ffmpeg.org/project/ffmpeg/cover/20231212073242.10967-1-jeebjp@gmail.com/

13:58 <JEEB> or should I just drop the content checking

13:58 <JEEB> (personally I think it's useful to see what changed in the packets if the test starts failing)

14:03 iive has joined #ffmpeg-devel

14:05 * JEEB looks at trac and sees that there are no real reports on the new threading work

14:05 <JEEB> time to try :3

14:05 dellas has quit [Remote host closed the connection]

14:15 Raz- has joined #ffmpeg-devel

14:34 dellas has joined #ffmpeg-devel

14:52 <JEEB> always fun to find out you've utilized a static index in a loop

14:52 <JEEB> :3

15:00 darkapex has quit [Read error: Connection reset by peer]

15:00 darkapex has joined #ffmpeg-devel

15:32 Krowl has quit [Read error: Connection reset by peer]

15:35 ccawley2011 has quit [Ping timeout: 268 seconds]

15:37 ccawley2011 has joined #ffmpeg-devel

15:44 cone-538 has quit [Quit: transmission timeout]

15:44 <kurosu> int rand() { return 4; /* Randomized through a dice roll */ }

15:44 <kurosu> (obligatory xkcd etc)

15:59 Krowl has joined #ffmpeg-devel

16:53 cone-746 has joined #ffmpeg-devel

16:53 <cone-746> ffmpeg Leo Izen master:1a0495953200: fate/jpegxl: add multiframe permuted TOC image parser test

16:56 AbleBacon has joined #ffmpeg-devel

17:16 Krowl has quit [Read error: Connection reset by peer]

17:21 <kierank> are there any hlg prores samples?

17:32 ngaullier has quit [Ping timeout: 256 seconds]

18:08 Traneptora has quit [Quit: Quit]

18:18 mkver has quit [Ping timeout: 276 seconds]

18:34 <thilo> :w

18:55 <michaelni> haasn, "git checkout haasn/yuvj_removal_v10_p1 ; make distclean ; ./configure && make -j32 fate-filter-pan-stereo2" infinite loops here

18:56 <michaelni> i must be doing something wrong :)

19:17 Traneptora has joined #ffmpeg-devel

19:20 qeed has joined #ffmpeg-devel

19:34 <courmisch> is it me or rnd() gives even values more often than odd?

19:35 <Traneptora> courmisch: wdym by "more"

19:35 <Traneptora> how much more, and how did you test it

19:40 <courmisch> more than 50%

19:41 <courmisch> it just seems that the LPC autocorr length isote

19:41 <courmisch> is more often even than ldd

19:41 <courmisch> odd*

19:43 <Gramner> it uses AVLFG, no idea if anyone ever tested how good of a prng that actually is

19:50 <Gramner> (have fun yak shaving)

19:53 cone-746 has quit [Quit: transmission timeout]

20:08 Krowl has joined #ffmpeg-devel

20:42 zsoltiv has joined #ffmpeg-devel

20:58 <zsoltiv> how should an encoder calculate PTS if an input frame has AV_NOPTS_VALUE?

21:01 Krowl has quit [Read error: Connection reset by peer]

21:03 <BtbN> an encoder should not calculate PTS at all. It gets it as input

21:06 <zsoltiv> oh, that makes my easier then

21:06 <zsoltiv> so just av_rescale_q'ing it to the AVCodecContext's time_base is all that's needed?

21:07 <BtbN> Limited exceptions apply, if more frames come out than packages, but that should be simple math

21:07 <BtbN> rescale them to the timebase the encoder needs, or use the provided one

21:08 <zsoltiv> alright, thanks

21:17 s55 has quit [Ping timeout: 260 seconds]

21:17 sr55 has joined #ffmpeg-devel

21:42 sr55 is now known as s55

21:42 s55 has quit [Changing host]

21:42 s55 has joined #ffmpeg-devel

21:56 Marth64 has quit [Quit: Leaving]

21:59 Marth64 has joined #ffmpeg-devel

22:35 ccawley2011 has quit [Quit: Leaving]

23:01 <ubitux> my ffmpeg-fu is a bit rusty; do we have a way to dump a packet data to a file?

23:02 <JEEB> I (ab)used segment muxer with the rawvideo format

23:02 <ubitux> isn't rawvideo implying decode? 🤔

23:02 <JEEB> setting the segment duration to something like 0.0000001 and then not sure if I had to allow for non-keyframe things

23:02 <JEEB> I think it doesn't care? :D

23:02 <JEEB> similar to the data format

23:03 <JEEB> it just spews out what it gets fed

23:03 <JEEB> basically -c copy should keep the packet data as-is

23:03 lemourin has quit [Quit: The Lounge - https://thelounge.chat]

23:04 <JEEB> (you'd think "rawvideo" was limited to video streams, but it properly didn't care :))

23:06 <ubitux> mmh i'm seeing the data muxer

23:06 <JEEB> I might have utilized the rawvideo muxer because it had .rgb extension for it :D

23:06 lemourin has joined #ffmpeg-devel

23:06 <JEEB> (since I was abusing the segment muxer to split the packets into separate files)

23:07 <JEEB> while the data muxer has no extensions defined

23:07 <JEEB> and yes I did think of writing a small tool for dumping packets from input, but I never got to it

23:08 <ubitux> i'm currently using gstreamer for that :p

23:09 zsoltiv has quit [Remote host closed the connection]

23:09 <JEEB> break_non_keyframes and min_segment_duration or so in segment muxer

23:09 <JEEB> the first I recall but the latter I am not sure if that was the timing setting that you just set really low

23:15 <ubitux> i found a way without segment

23:15 <ubitux> -map v:0 -f data -c:v copy

23:16 lemourin has quit [Quit: The Lounge - https://thelounge.chat]

23:18 lemourin has joined #ffmpeg-devel

23:21 <JEEB> yea that will dump all things into a single file

23:21 <JEEB> I just needed to split packets which is why I went for segment -> rawvideo (or data if it could be forced without an extension)

23:35 lemourin has quit [Quit: The Lounge - https://thelounge.chat]

23:38 lemourin has joined #ffmpeg-devel

23:50 iive has quit [Quit: They came for me...]