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 7.1 has been released! | Please read ffmpeg.org/developer.html#Code-of-conduct
cone-143 has joined #ffmpeg-devel
<cone-143> ffmpeg Sean McGovern master:4322be512b2b: avcodec/svq1enc: restrict Altivec acceleration to big-endian POWER configurations
<cone-143> ffmpeg Manuel Lauss master:d68d311bcd76: avcodec/sanm: codec37 buffers are private
<cone-143> ffmpeg Michael Niedermayer master:c733e2b5ed9c: Revert "avformat/mpegts: Add standard extension so hls can check in extension_picky mode"
<cone-143> ffmpeg Michael Niedermayer master:91d96dc8ddae: avformat/hls: Be more picky on extensions
<fflogger> [newticket] naruh: Ticket #11431 ([ffmpeg] FFmpeg no set languages metadata on subtitles files) created https://trac.ffmpeg.org/ticket/11431
<cone-143> ffmpeg James Almer master:ef1cb1c9c810: avfilter/avfilter: add a side_data field to AVFilterLink
<cone-143> ffmpeg James Almer master:7a025e1cb5fe: avfilter/buffersrc: add a side_data field
<cone-143> ffmpeg James Almer master:6707d970c041: avfilter/buffersink: add av_buffersink_get_side_data()
<cone-143> ffmpeg James Almer master:e61b9d4094d7: fftools/ffmpeg: propagate decoded_side_data from decoded streams to the filterchain
<cone-143> ffmpeg James Almer master:0bed1b895300: avutil/downmix_info: zero the allocated buffer
<cone-143> ffmpeg James Almer master:ea3c3b42dff5: avutil/frame: add a side data prop to signal channel layout dependent types
<cone-143> ffmpeg James Almer master:6a282cdf7e51: avfilter/af_aresample: propagate downmix metadata to swresample if present
<cone-143> ffmpeg James Almer master:4f9afbb1b203: fftools/ffmpeg_filter: reinitialize the filterchain if downmix metadata changed between frames
<cone-143> ffmpeg James Almer master:712140be75d1: doc/APIChanges: fix library name for a recent entry
thilo has quit [Ping timeout: 272 seconds]
thilo has joined #ffmpeg-devel
thilo has quit [Changing host]
thilo has joined #ffmpeg-devel
jamrial has quit []
mkver has quit [Ping timeout: 245 seconds]
cone-143 has quit [Quit: transmission timeout]
^Neo has quit [Ping timeout: 244 seconds]
Martchus has joined #ffmpeg-devel
Martchus_ has quit [Ping timeout: 252 seconds]
<Traneptora> Marth64: not to be too weird but can you please not top-post on the ML? it's confusing
<Traneptora> it's when the reply is above the thing it's replying to. It's common for email clients to do that but it makes following conversations much more difficult
<Traneptora> not to be like an ML police, just like, a friendly reminder I mean
<Marth64> no worries. yes I'll clear before I send
tantei3 has joined #ffmpeg-devel
<tantei3> Hello, I submitted a patch for x86 ASM and was asked to improve indentation https://patchwork.ffmpeg.org/project/ffmpeg/patch/20250125142546.1244665-1-16567adigashreesh@gmail.com/ and I was wondering if someone could clarify ASM indentation guidelines. Is there any auto formatting like clang-format to be run?
<Lynne> no, no tools like that exist
<Lynne> look at other files, in general the operands should be on the same column
<tantei3> Okay thanks for the clarification, will try to change and resubmit.
<Compn> i uh. thinking we should delete ff twitter account
<Compn> things are going from bad to worse here
<Marth64> is it getting spammed or something?
<Marth64> (I don't have an account so I just see what is visible to public)
<fflogger> [newticket] necros: Ticket #11432 ([undetermined] Empty output file after conversion if source is 5.1 audio (ac3>opus)) created https://trac.ffmpeg.org/ticket/11432
<Compn> Marth64, well this happened https://imgur.com/z0vQy3F
<Marth64> got it
<Compn> your opinion of getting rid of ff twitter yes/ no ?
System_Error has quit [Remote host closed the connection]
<Marth64> null. i don't support that gesture but i am also not into social media platforms. if having a profile page adds positive publicity and does not promote bad values, it could go either way. kind of like separating the actor/actress from the character
<Marth64> seems like many OSS projects have one still but I don't know the platform community enough to sense if the overall temperature is abandon ship
<Marth64> at least what I can see of the FFmpeg twitter itself it looks fine
System_Error has joined #ffmpeg-devel
<Compn> right but the guy doing that in the gif owns twitter and i'm wondering by using twitter we are supporting him
<Marth64> you are correct its a negative association. the angle i see is, >51.7K Followers is an asset for marketing, if lost it could be hard to rebuild
<Marth64> by marketing i mean the civil / ffmpeg is cool / new release update type posts
<Marth64> (this is a stretch--it may not be a realistic hypothesis at all) but if being on that social media account leads to eventual clicks to the public donation page for even 1% of that 51.7k, thats 517 donations lost
<Marth64> again total stretch and extrapolation
<Marth64> i'm just alluding to its a multi-angle decision
<Marth64> its also a case of being "grandfathered" into the current situation of what the platform is, I imagine the account has been around for a very long time predating whatever twitter is today
<fflogger> [editedticket] sean_mcg: Ticket #10955 ([undetermined] 7.0: checkasm failures on ppc64el) updated https://trac.ffmpeg.org/ticket/10955#comment:4
ElderLeader has joined #ffmpeg-devel
<ElderLeader> jamrial: your commit is introducing unused variable warning from compiler, check your work twice!
MisterMinister has quit [Remote host closed the connection]
MisterMinister has joined #ffmpeg-devel
abdu has joined #ffmpeg-devel
ElderLeader has quit [Quit: Client closed]
DauntlessOne4 has quit [Quit: Ping timeout (120 seconds)]
DauntlessOne4 has joined #ffmpeg-devel
kepstin has quit [Remote host closed the connection]
kepstin has joined #ffmpeg-devel
^Neo has joined #ffmpeg-devel
^Neo has quit [Changing host]
^Neo has joined #ffmpeg-devel
^Neo has quit [Ping timeout: 252 seconds]
ccawley2011 has joined #ffmpeg-devel
abdu has quit [Quit: Client closed]
abdu has joined #ffmpeg-devel
rvalue- has joined #ffmpeg-devel
rvalue has quit [Ping timeout: 252 seconds]
rvalue- is now known as rvalue
tantei3 has quit [Ping timeout: 276 seconds]
^Neo has joined #ffmpeg-devel
^Neo has quit [Changing host]
^Neo has joined #ffmpeg-devel
^Neo has quit [Ping timeout: 244 seconds]
<courmisch> any OSS project, this one included, should just state that it won't post on X anymore and provide links to their new comm channels
<courmisch> leaving altogether and closing the account is a bad idea, because it invites squatting
DauntlessOne4 has quit [Ping timeout: 252 seconds]
<courmisch> that being said, in FFmpeg's case, there are far worse problems in terms of external image to deal with
mkver has joined #ffmpeg-devel
<kierank> 21:38:06 <microchip_> abdu67: ChatGPT :P
<kierank> quite the opposite, chatgpt knows nothing about simd
<kierank> it produces the dumbest stuff
<sfan5> does ffmpeg even have a bluesky/fediverse presence?
<kierank> someone squatted the name
<kierank> and I personally will not post on the fediverse/mastodon/bluesky whatevefr
<sfan5> I do agree handling multiple platforms is annoying
ElderLeader has joined #ffmpeg-devel
<ElderLeader> make new social meeting platform and call it meta
<kierank> irc is social media
ccawley2011 has quit [Ping timeout: 260 seconds]
<ElderLeader> kierank: https://imgur.com/a/E244jNW
<Lynne> its not a social media platform unless its got influencers
ccawley2011 has joined #ffmpeg-devel
<ElderLeader> i failed to influence Lynne to finish lavu/tx so i did it myself
<ElderLeader> how to RE hardware IPU video mpeg1/mpeg2 decoder
<ElderLeader> ?
<kierank> ElderLeader: mmiotool
Mirarora has quit [Quit: Mirarora encountered a fatal error and needs to close]
Mirarora has joined #ffmpeg-devel
<kierank> who was it who had a standalone checkasm
<Lynne> unlord?
<ElderLeader> Lynne: write pixdesc for v210 pixel format, now!
<Lynne> you can't, afaik?
<ElderLeader> via BITSTREAM flag ?
uau has quit [Quit: ZNC 1.9.1+deb2+b1 - https://znc.in]
<kierank> won't work
<Lynne> didnt that flag not do what you think it did?
<kierank> All values of a component are bit-wise packed end to end.
<kierank> v210 has holes
uau has joined #ffmpeg-devel
<ElderLeader> so do other pixels
<kierank> definition of bitstream is packed end to end
mkver has quit [Remote host closed the connection]
mkver has joined #ffmpeg-devel
ccawley2011 has quit [Ping timeout: 265 seconds]
<microchip_> kierank: it was a joke, hence the :P
ccawley2011 has joined #ffmpeg-devel
Mirarora has quit [Quit: Mirarora encountered a fatal error and needs to close]
Mirarora has joined #ffmpeg-devel
<ElderLeader> kierank: holes in v210 are not the problem, but offsets are not constant for components, they change in fixed pattern
<kierank> the holes cause the offsets
<kierank> same thing imo
<kierank> ElderLeader: btw imo future 10-bit pixel format is 8-bit values, then 2-bit lsbs
<kierank> allows processing on MSBs if needed
<kierank> unpack if true 10-bit is needed
<kierank> svt-av1 used this
<ElderLeader> ugly hack
<kierank> it's pareto optimal
<kierank> if you want 10-bit you have to unpack anyway
<kierank> if you want 8-bit you just extend stride
<kierank> with v210 and others need to unpack anyhow
<ElderLeader> v210 is legacy and should be abandoned
<kierank> ElderLeader: how to transport 10-bit video then?
<kierank> without wasting 6 LSBs
<kierank> 37.5% wastage
<kierank> 8-bit with lsbs at the end is the optimal format
<kierank> you can do so much
<ElderLeader> for 422? just make it work with archaic pixdesc syntax
<ElderLeader> what other projects/programs/apps use for representing all pixel formats?
<kierank> any method of transporting over network
<kierank> ElderLeader: upipe has something like pixdesc, more advanced
<ElderLeader> link?
<ElderLeader> steal it, make vf_pf2pf.c use it.
<ElderLeader> haasn: we should probably invent better thing than pixdesc.c struct descriptor, and convert everything to it
<haasn> ElderLeader: I already invented this for libplacebo
<haasn> And I fully intend to copy paste the design
<haasn> Since it works
<nevcairiel> didnt the libplacebo design have issues with some packed subsampled formats, i faintly remember not being able to represent some of them with it
<ElderLeader> in libplacebo i see no way how to uniquely define v210 and have generic processor using such specs
<ElderLeader> and nothing  in upipe of even related struct
zsoltiv_ has joined #ffmpeg-devel
zsoltiv has joined #ffmpeg-devel
Mirarora has quit [Quit: Mirarora encountered a fatal error and needs to close]
Mirarora has joined #ffmpeg-devel
<haasn> Well I think some changes are needed to handle bitpacked formats etc
<haasn> And to bring it to a design closer to what our new read functions will need
<haasn> I’ll think about it
<haasn> nevcairiel: I think that’s less of an issue on CPU
<haasn> Because we don’t need to map to GPU formats as an intermediate
<haasn> And can directly sample from the same pointer with two different pixel strides
<haasn> Without needing to go through hoops to get the correct sampling swizzle for the GPU native format
<haasn> I think in a rewrite of libplacebo I’d make compute shaders a hard requirement and solve this problem by using direct memory access
tufei has joined #ffmpeg-devel
MisterMinister has quit [Ping timeout: 260 seconds]
HarshK23 has quit [Quit: Connection closed for inactivity]
lexano has quit [Remote host closed the connection]
tufei_ has joined #ffmpeg-devel
tufei has quit [Ping timeout: 264 seconds]
lexano has joined #ffmpeg-devel
witchymary has quit [Remote host closed the connection]
witchymary has joined #ffmpeg-devel
^Neo has joined #ffmpeg-devel
^Neo has quit [Changing host]
^Neo has joined #ffmpeg-devel
witchymary has quit [Remote host closed the connection]
<fflogger> [newticket] idest: Ticket #11433 ([undetermined] "Error parsing Opus packet header" when extracting an Opus audio stream from an MKV container) created https://trac.ffmpeg.org/ticket/11433
witchymary has joined #ffmpeg-devel
ElderLeader has quit [Quit: Client closed]
tufei__ has joined #ffmpeg-devel
tufei_ has quit [Ping timeout: 264 seconds]
tufei__ has quit [Ping timeout: 264 seconds]
HarshK23 has joined #ffmpeg-devel
abdu93 has joined #ffmpeg-devel
abdu has quit [Ping timeout: 240 seconds]
<fflogger> [editedticket] Balling: Ticket #8515 ([avcodec] vc1_parser erroneously flagged every packet as keyframe) updated https://trac.ffmpeg.org/ticket/8515#comment:7
ccawley2011 has quit [Read error: Connection reset by peer]
jamrial has joined #ffmpeg-devel
cone-083 has joined #ffmpeg-devel
<cone-083> ffmpeg James Almer master:f632ab53d981: fftools/ffmpeg_filter: remove accidental variable shadowing