01:07 <cone-369> ffmpeg James Almer master:76049d1c4564: avformat/iamf_writer: fix setting num_samples_per_frame for OPUS

01:08 <cone-369> ffmpeg James Almer master:4e2b9df48c6a: avformat/isom: use more of the existing channel layout bitmap defines

01:11 <fflogger> [editedticket] lingyikong: Ticket #11360 ([avcodec] H.264 Decoder Issue with Deblocking in Multi-Slice Field Video) updated https://trac.ffmpeg.org/ticket/11360#comment:1

01:15 <cone-369> ffmpeg Michael Niedermayer release/7.1:88126fc4ecff: avformat/iamf_parse: reject ambisonics mode > 1

01:15 <cone-369> ffmpeg James Almer release/7.1:888a6b54bcc8: avformat/iamf_parse: add checks to parameter definition durations

01:15 <cone-369> ffmpeg James Almer release/7.1:871c69a28bf4: avformat/iamf_parse: fix setting duration for the last subblock in a parameter definition

01:15 <cone-369> ffmpeg James Almer release/7.1:b168ed9b1463: avformat/iamf_writer: fix setting num_samples_per_frame for OPUS

06:07 <compnn> https://lists.ffmpeg.org/pipermail/ffmpeg-devel/2024-December/337757.html

06:52 <Marth64> CBS_UNIT_TYPE_POD - what is a POD vs. INTERNAL_REF?

08:33 <fflogger> [newticket] compn: Ticket #11361 ([avformat] avi demux Invalid sample rate) created https://trac.ffmpeg.org/ticket/11361

09:20 <compnn> ^ been a while since i found an avi file that ffmpeg could not handle

09:23 <JEEB> Marth64: nb_offsets == 0 I presume

09:53 <cone-914> ffmpeg Peter Ross master:c90e0777da97: avformat/iff: SndAnim decoding

09:55 <fflogger> [editedticket] pross: Ticket #5553 ([avformat] IFF ANIM: support anims with sound) updated https://trac.ffmpeg.org/ticket/5553#comment:5

10:22 <fflogger> [newticket] 0x20z: Ticket #11362 ([avcodec] memory leaks error at __interceptor_malloc) created https://trac.ffmpeg.org/ticket/11362

11:27 <fflogger> [newticket] bubbleguuum: Ticket #11363 ([avcodec] [Android] MediaCodec decoders/encoders do not work on Pixel 8 Pro (No output buffer available)) created https://trac.ffmpeg.org/ticket/11363

11:49 <Traneptora> dafuq

11:49 <Traneptora> "brave" sorry holdup

11:49 <Traneptora> how is a CC candidate censoring ML discussion about the CC vote "brave" exactly

11:50 <Traneptora> also why is it anyone's responsibility to stop flames before they occur, the whole point of moderation is after-the-fact response

11:50 <another|> because there is backlash to be expected?

11:51 <Traneptora> sure, but they just blanket blocked discussion

11:51 <Traneptora> there's an election going on, you can't just block discussion about the election

11:52 <Traneptora> because you think it might get to some point in the future

11:54 <another|> yeah, idk. I getting closer to unsubscribing with every mail

12:38 <haasn> unsubscribing won't help improve the situation

12:39 <haasn> if everybody leaves an online platform due to toxicity you get a kind of evaporative cooling effect where the toxicity gets reinforced

12:55 <fflogger> [editedticket] jamrial: Ticket #11362 ([avcodec] memory leaks error at __interceptor_malloc) updated https://trac.ffmpeg.org/ticket/11362#comment:2

13:11 <another|> haasn: sure, it's the selfish escape.

13:45 <fflogger> [editedticket] haasn: Ticket #2151 ([swscale] Warning: data is not aligned! This can lead to a speedloss) updated https://trac.ffmpeg.org/ticket/2151#comment:3

13:58 <JEEB> I think this can be used as a pretty good example of how early exit makes things simpler https://git.videolan.org/?p=ffmpeg.git;a=blobdiff;f=libavcodec/amfenc_av1.c;h=bc6ad3d7fbed09338482b238d0a2240af7a31fb1;hp=7646d38d8a1cb5ed216b144889bb7c382022884b;hb=95217872adff358af6ed4d06777e4625648904cd;hpb=c037eb8424399687712a9b44510bf3b3030b323a

14:07 <beastd> JEEB: as it's in the init function if I have grasped it correctly, I guess it would be best to factor all those longish if-bodies out into individual functions. not sure how feasible it is. did just skim over and didn't check for inter-if-body-relations.

14:07 <JEEB> yea, exactly :)

14:07 <JEEB> at least looking at that little pyramid it seems like it would be possible to extract

14:08 <beastd> sounds plausible

14:09 <JEEB> I was tempted to respond to that patch on ML regarding it, but then other things happened.

14:10 <beastd> maybe respond after the fact. dmitrii is listed as maintainer. he might be interested to do the refactor.

14:42 <compnn> Traneptora, feel free to ask me questions about ml moderating if you want

14:52 <Sean_McG> Hi folks. Just dropping by quick to wish you all the very best of the holiday season.

14:52 <JEEB> same to thee :)

14:53 <Sean_McG> Additionally I am taking a leave of absence for the forseeable(sp?) future and will be unsubscribing from the -devel ML for awhile. If you need anything from me or just want to talk, do send me an email.

14:58 <JEEB> pross: good catches with https://patchwork.ffmpeg.org/project/ffmpeg/list/?series=13561

16:18 <compnn> michaelni, are you on some kind of intel nuc or rpi ?

16:26 <michaelni> no

17:01 <fflogger> [editedticket] bubbleguuum: Ticket #11363 ([avcodec] [Android] MediaCodec decoders/encoders do not work on Pixel 8 Pro (No output buffer available)) updated https://trac.ffmpeg.org/ticket/11363#comment:1

20:14 <BtbN> michaelni: re enhanced flv, how strongly do you feel about the fourcc list? I don't particularily care, but I do prefer the switch-case. Specially because to implement your idea, it'd have to be two more dynamic allocations of arrays to store the fourcc for each stream, while just having a switchcase for each direction avoids that.

20:15 <BtbN> And with enhanced flv, really every packet carries the fourcc for every stream it seems

20:36 <michaelni> BtbN, i would have thought it would be similar to int64_t last_ts[FLV_STREAM_TYPE_NB]; or that all the stream related things like last_ts would be in a AVStream->priv_data

20:37 <michaelni> but its not my intend to complicate the code so if that doesnt work, sure ignore my suggestion, i thought it would be simple

20:37 <BtbN> I don't think there is a limit to the number of possible tracks, outside of integer type limits

20:38 <BtbN> It wouldn't be complicated to just iterate the mapping list every time, but I'd just rather not do that potentially multiple dozen times per packet

20:38 <BtbN> And caching it seems more complicated than just two switch cases in each direction

20:47 <Marth64> JEEB: it's a start but I have a CBS working that transforms between DVD CC and A53 CC, as well as the other way around