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...]
Mirarora has joined #ffmpeg-devel
uau_ has joined #ffmpeg-devel
Mirarora has quit [Quit: Mirarora encountered a fatal error and needs to close]
uau has quit [Ping timeout: 276 seconds]
uau_ has quit [Quit: ZNC 1.9.1+deb2+b3 - https://znc.in]
uau has joined #ffmpeg-devel
Mirarora has joined #ffmpeg-devel
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
minimal has quit [Quit: Leaving]
mark4o has joined #ffmpeg-devel
markh has quit [Ping timeout: 272 seconds]
mark4o is now known as markh
bwu25 has joined #ffmpeg-devel
jamrial has quit []
bwu25 has quit [Ping timeout: 248 seconds]
_whitelogger has joined #ffmpeg-devel
_whitelogger has joined #ffmpeg-devel
TheVibeCoder has joined #ffmpeg-devel
jarthur_ has joined #ffmpeg-devel
jarthur_ is now known as jarthur
jarthur has quit [Quit: jarthur]
jarthur_ has joined #ffmpeg-devel
jarthur_ is now known as jarthur
MisterMinister has quit [Ping timeout: 260 seconds]
jarthur has quit [Quit: jarthur]
TheVibeCoder has quit [Read error: Connection reset by peer]
compn has quit [Read error: Connection reset by peer]
compn has joined #ffmpeg-devel
bsFFFFFF has joined #ffmpeg-devel
cone-105 has joined #ffmpeg-devel
<cone-105>
ffmpeg Niklas Haas master:0a5ae743ef56: avfilter/vf_thumbnail: switch to query_func2
<cone-105>
ffmpeg Niklas Haas master:cf18b280f0f4: avfilter/vf_thumbnail: support more planar formats
<haasn>
Is there a good way of vectorizing a histogram? The only approach I can think of is to have a separate histogram for each element of the vector, e.g. for 128-bit vectors we store 16 separate 256-entry histograms, then use a gather/scatter to update the corresponding histogram for each element
<haasn>
but I'm not sure that will even be faster
<haasn>
and apparently AVX2 does not even have good instructions for this
<thardin>
seems like AVX-512 at least has enough space in the registers to do some neat tricks for htat
<thardin>
if it has a good enough barrel shifter than you could put single bits into at least three registers, then add them up (two bits per count), then add that to memory
<thardin>
but then you need a loop to skep pairs of bits that are zeros, so you get lots of branches
<thardin>
maybe there's a clever way to use masking to count up to 63 using all the available register bits
<thardin>
or 64
<Lynne>
ePirat: ping
rvalue has quit [Ping timeout: 248 seconds]
rvalue has joined #ffmpeg-devel
TheVibeCoder has joined #ffmpeg-devel
cone-105 has quit [Quit: transmission timeout]
minimal has joined #ffmpeg-devel
bsFFFFFF has joined #ffmpeg-devel
<Lynne>
haasn: ping on the lut vf_libplacebo patch, do you think its worth making the option static?
<Lynne>
surely parsing a small text file upon settings being reloaded isn't that bad
<kasper93>
it's not compile time, because it declares function prototype thats not compatible with asm one
<kasper93>
or _c one, but still we don't carry prototypes in headers apparently and declare it inline
<JEEB>
yea, that would be where it would be catchable as well. but if we don't carry the prototypes then... alas
<JEEB>
uhh, the HTTPS clone url sure is not feeling good
<JEEB>
wanted to just fetch latest
mateo`_ has joined #ffmpeg-devel
mateo` has quit [Ping timeout: 248 seconds]
<JEEB>
ok, managed to fetch
MisterMinister has joined #ffmpeg-devel
elvis_a_presley has quit [Quit: smoke-bomb ; grapple-hook]
elvis_a_presley has joined #ffmpeg-devel
<JEEB>
kasper93: btw I'm essentially giving you a LGTM, I'm just waiting for my x86_64 setup to first fail in FATE for those things, and then apply the patch and see it work
<JEEB>
(since this seemed to touch C file function declarations)
<BtbN>
Yahoo apparently is blocking all ffmpeg mails, due to too many user complaints. Fun.
<BtbN>
Guess if you're with Yahoo, you're just out of luck
Traneptora_ has joined #ffmpeg-devel
Traneptora has quit [Ping timeout: 248 seconds]
wyatt8740 has quit [Ping timeout: 265 seconds]
wyatt8740 has joined #ffmpeg-devel
<fflogger>
[newticket] mms: Ticket #11669 ([undetermined] Compatibility issue with RTMP push from Dahua cameras) created https://trac.ffmpeg.org/ticket/11669
<kasper93>
"too many complaints" from whom?
bsFFFFFF has quit [Ping timeout: 272 seconds]
bsFFFFFF has joined #ffmpeg-devel
<haasn>
Lynne: I would just make the option static
<haasn>
I think you are mistaken in assuming all LUTs are small
<haasn>
and yes, libplacebo already logs errors, so just return AVERROR_EXTERNAL or EINVAL or something
<Lynne>
ok
jarthur has joined #ffmpeg-devel
Traneptora_ has quit [Quit: Quit]
Traneptora has joined #ffmpeg-devel
cone-403 has joined #ffmpeg-devel
<cone-403>
ffmpeg Jiasheng Jiang master:35a6de137a39: libavcodec/alsdec.c: Add check for av_malloc_array() and av_calloc()
mkver has joined #ffmpeg-devel
bsFFFFFF has quit [Quit: bsFFFFFF]
iive has joined #ffmpeg-devel
mkver has quit [Ping timeout: 244 seconds]
TheVibeCoder has quit [Ping timeout: 272 seconds]
cone-403 has quit [Quit: transmission timeout]
<BtbN>
ePirat: I got a bit further, the client succeeds in the first handshake step now (mtu was set to 0, which openssl did not like)
<BtbN>
But now the same obscure error happens when the server side tries to continue the handshake...