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 7.1.1 has been released! | Please read ffmpeg.org/developer.html#Code-of-conduct
iive has quit [Quit: They came for me...]
Gramner has quit [Remote host closed the connection]
lemourin has joined #ffmpeg-devel
lemourin has quit [Client Quit]
lemourin has joined #ffmpeg-devel
Gramner has joined #ffmpeg-devel
lemourin has quit [Client Quit]
lemourin has joined #ffmpeg-devel
Traneptora has quit [Quit: Quit]
minimal has quit [Quit: Leaving]
System_Error has quit [Remote host closed the connection]
jamrial_ has joined #ffmpeg-devel
jamrial has quit [Read error: Connection reset by peer]
System_Error has joined #ffmpeg-devel
rodgort has quit [Ping timeout: 276 seconds]
rodgort has joined #ffmpeg-devel
Traneptora has joined #ffmpeg-devel
jamrial_ has quit []
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
indecisiveturtle has quit [Ping timeout: 252 seconds]
BradleyS has quit [Ping timeout: 276 seconds]
BradleyS has joined #ffmpeg-devel
TheVibeCoder has joined #ffmpeg-devel
MisterMinister has quit [Ping timeout: 245 seconds]
realies9 has quit [Quit: Ping timeout (120 seconds)]
realies9 has joined #ffmpeg-devel
TheVibeCoder has quit [Changing host]
TheVibeCoder has joined #ffmpeg-devel
indecisiveturtle has joined #ffmpeg-devel
ngaullier has joined #ffmpeg-devel
<fflogger>
[editedticket] guest: Ticket #11662 ([undetermined] ffplay does not detect when the remote side closes the connection) updated https://trac.ffmpeg.org/ticket/11662#comment:2
kode543 has joined #ffmpeg-devel
kode54 has quit [Read error: Connection reset by peer]
kode543 is now known as kode54
<fflogger>
[editedticket] guest: Ticket #11662 ([undetermined] ffplay does not detect when the remote side closes the connection) updated https://trac.ffmpeg.org/ticket/11662#comment:3
<TheVibeCoder>
adding WebP Animation support is rocket science!
<aaabbb>
no it's easy. just take the webp animation encoder, turn the code upside-down, and now it'll start decoding!
mkver has joined #ffmpeg-devel
<pross>
Compn: do u think its worth highlighting the twoccs that ffmpeg currently supports. e.g. colorise the table background green ?
cone-097 has joined #ffmpeg-devel
<cone-097>
ffmpeg Peter Ross master:853ad9e77e1d: avcodec/adpcm: squelch uninitialized variable warnings
<linkmauve>
TheVibeCoder, I once wrote a hardware lossy WebP decoder using V4L2, but animations require me to not just hand out dmabufs, which was the main point of the decoder, so I never wrote that part yet.
<linkmauve>
Unlike e.g. MJPEG or animated AVIF, which have no notion of keeping the previous frame and modifying it outside of the codec.
<cone-097>
ffmpeg Marvin Scholz master:82aeee3c19f4: avformat/rtsp: fix leak of options dict on error
<fflogger>
[editedticket] guest: Ticket #11662 ([undetermined] ffplay does not detect when the remote side closes the connection) updated https://trac.ffmpeg.org/ticket/11662#comment:4
<ePirat>
of course one needs a special flag to make openssl verify the hostname in a standard way...
jamrial has joined #ffmpeg-devel
microlappy has joined #ffmpeg-devel
<TheVibeCoder>
michaelni: i demand you to make almpeg work/code publicily available
<TheVibeCoder>
also proresdec_raw is weird name for files
mkver has joined #ffmpeg-devel
<TheVibeCoder>
Lynne: also, the your CPU decoder is 10x times slower than mine
<TheVibeCoder>
the banding waveform issue is not 100% resolved
Traneptora has quit [Quit: Quit]
<Compn>
pross, if it helps you keep track of whats supported or not, go for it. the wiki used to have a page that said how many people would access a given page. said another way, i'm not sure if anyone is using the wiki at all besides a few codec developers
Everythi1g has quit [Quit: leaving]
<TheVibeCoder>
Lynne: tested static ffmpeg build with your changes vs librempeg build
zsoltiv_ has quit [Read error: Connection reset by peer]
<TheVibeCoder>
Lynne: without any scaling conversions
<TheVibeCoder>
Lynne: with librempeg rgb0 is slower than gbrp12 if you want to watch it, because with mpv and bayer i get most of time blue frame output and mpv red errors when not using scaler
<beastd>
jamrial: `(!!is_unreliable * DECODER_FLAG_TS_UNRELIABLE)` the !! in your new code shoudn't be needed and is a bit irritating to read. anyway correct and you already backported. so fine to leave as is for me. sorry i didn't notice it earlier.
<TheVibeCoder>
Lynne: ah, its because on ffmpeg, frame threading is used and not slice threads by default
<TheVibeCoder>
but with librempeg both frame and slice are similary fast
<TheVibeCoder>
strange
<Lynne>
so banding issues are still not solved
<Lynne>
does librempeg do better with respect to banding?
<TheVibeCoder>
same
<TheVibeCoder>
if you mult DCs by 2 banding goes away, but now you get blocks
<TheVibeCoder>
perhaps with vulkan decoder you can experiment more and find batter way to resolve this
<Lynne>
you use --vf=format=gbrp12,waveform=c=7:d=overlay to view it, right?
<TheVibeCoder>
yes
<TheVibeCoder>
its most obvious in v0 for hq
ngaullier has quit [Remote host closed the connection]
<Lynne>
hmm... I have an idea to figure out for sure if qscale is 2 fields for ac/dc
ngaullier has joined #ffmpeg-devel
ngaullier has quit [Remote host closed the connection]
cone-097 has quit [Quit: transmission timeout]
zsoltiv has quit [Read error: Connection reset by peer]
<thardin>
and the results look kinda worse than motion compensation used in video codecs
<thardin>
I can't imagine how someone would think 4x temporal upconversion would ever look decent. it doesn't look decent on TV:s, and there you can delay the video by a few frames without the user noticing. not so with games
<Lynne>
TheVibeCoder: qscale definitely is not 2 fields
<Lynne>
same input, HQ vs non-HQ, qscale top input remains the same, but both DC and AC take less bits than before
<fflogger>
[newticket] fluoriteByte: Ticket #11664 ([undetermined] ffmpeg can't allocate memory when using any vulkan hwaccel) created https://trac.ffmpeg.org/ticket/11664
<fflogger>
[editedticket] fluoriteByte: Ticket #11664 ([undetermined] ffmpeg can't allocate memory when using any vulkan hwaccel) updated https://trac.ffmpeg.org/ticket/11664#comment:2
indecisiveturtle has joined #ffmpeg-devel
<fflogger>
[editedticket] fluoriteByte: Ticket #11664 ([undetermined] ffmpeg can't allocate memory when using any vulkan hwaccel) updated https://trac.ffmpeg.org/ticket/11664#comment:3
<beastd>
thardin: i got the impression it serves the market of gamers that want to play with RT at 4k resolutions with reasonably fast framerates, which hardware and games cannot really deliver.
<beastd>
not sure if RT is really worth it. personally i'm not so fond of ray tracing in games. sure it looks nice in tech demos. but in games i didn't feel it adds much value.
<jamrial>
beastd: i added it for consistency (see the line immediately above), but yes, it doesn't look pretty