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 7.0.2 has been released! | Please read ffmpeg.org/developer.html#Code-of-conduct
yigithanyigit has quit [Remote host closed the connection]
feiw2 has joined #ffmpeg-devel
feiw1 has quit [Remote host closed the connection]
thilo has quit [Ping timeout: 258 seconds]
yigithanyigit has joined #ffmpeg-devel
yigithanyigit has quit [Remote host closed the connection]
thilo has joined #ffmpeg-devel
thilo has quit [Changing host]
thilo has joined #ffmpeg-devel
jarthur has quit [Ping timeout: 260 seconds]
IndecisiveTurtle has quit [Ping timeout: 248 seconds]
arch1t3cht4 has joined #ffmpeg-devel
arch1t3cht has quit [Ping timeout: 248 seconds]
arch1t3cht4 is now known as arch1t3cht
jarthur has joined #ffmpeg-devel
feiw1 has joined #ffmpeg-devel
feiw2 has quit [Ping timeout: 252 seconds]
feiw2 has joined #ffmpeg-devel
feiw1 has quit [Remote host closed the connection]
witchymary has quit [Read error: Connection reset by peer]
witchymary has joined #ffmpeg-devel
feiw1 has joined #ffmpeg-devel
feiw2 has quit [Read error: Connection reset by peer]
jarthur has quit [Quit: jarthur]
mkver has quit [Ping timeout: 248 seconds]
jamrial has quit []
Martchus_ has joined #ffmpeg-devel
Martchus has quit [Ping timeout: 260 seconds]
feiw2 has joined #ffmpeg-devel
feiw1 has quit [Remote host closed the connection]
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
ramiro has quit [Ping timeout: 276 seconds]
ramiro has joined #ffmpeg-devel
feiw1 has joined #ffmpeg-devel
feiw2 has quit [Remote host closed the connection]
unturned3 has quit [Quit: Connection closed for inactivity]
feiw2 has joined #ffmpeg-devel
feiw1 has quit [Remote host closed the connection]
System_Error has quit [Remote host closed the connection]
Krowl has joined #ffmpeg-devel
System_Error has joined #ffmpeg-devel
ccawley2011 has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
lexano has quit [Ping timeout: 258 seconds]
rodgort has quit [Quit: Leaving]
lexano has joined #ffmpeg-devel
rodgort has joined #ffmpeg-devel
IndecisiveTurtle has joined #ffmpeg-devel
feiw1 has joined #ffmpeg-devel
feiw2 has quit [Ping timeout: 248 seconds]
Krowl has joined #ffmpeg-devel
feiw2 has joined #ffmpeg-devel
feiw1 has quit [Read error: Connection reset by peer]
bencoh has quit [Remote host closed the connection]
bencoh has joined #ffmpeg-devel
IndecisiveTurtle has quit [Ping timeout: 260 seconds]
j45_ has joined #ffmpeg-devel
j45 has quit [Ping timeout: 276 seconds]
j45_ is now known as j45
j45 has quit [Changing host]
j45 has joined #ffmpeg-devel
IndecisiveTurtle has joined #ffmpeg-devel
<rajivharlalka>
is there a filter where EOTF/OOTF conversion has been implemented? wanted to have one as a reference but grep doesnt provide a result
<JEEB>
zscale filter utilizes zimg library, and then libplacebo filter utilizes libplacebo library. both handle transfer functions
<JEEB>
zscale being CPU based, libplacebo being GPU (vulkan) based
<rajivharlalka>
had seen zscale but I wish that the input to my filter is linearized if it isn't already without the need to pass another filter( zscale in my case) as a predecessor
IndecisiveTurtle has quit [Ping timeout: 264 seconds]
IndecisiveTurtle has joined #ffmpeg-devel
s55 has quit [Server closed connection]
Krowl has quit [Read error: Connection reset by peer]
s55 has joined #ffmpeg-devel
System_Error has quit [Remote host closed the connection]
jamrial has joined #ffmpeg-devel
zip6como has quit [Server closed connection]
zip6como has joined #ffmpeg-devel
mkver has joined #ffmpeg-devel
cone-825 has joined #ffmpeg-devel
<cone-825>
ffmpeg James Almer master:a754ee084409: avcodec/h2645_parse: replace three bool arguments in ff_h2645_packet_split with a single flags one
<cone-825>
ffmpeg James Almer master:80606442377a: avcodec/shorten: Fix discard of ‘const’ qualifier
signalhunter has quit [Server closed connection]
signalhunter has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
ramiro has quit [Ping timeout: 258 seconds]
ramiro has joined #ffmpeg-devel
System_Error has joined #ffmpeg-devel
ramiro has quit [Ping timeout: 252 seconds]
ramiro has joined #ffmpeg-devel
Kei_N_ has quit [Ping timeout: 265 seconds]
ramiro has quit [Ping timeout: 272 seconds]
ramiro has joined #ffmpeg-devel
ramiro has quit [Ping timeout: 252 seconds]
ramiro has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
Kei_N has joined #ffmpeg-devel
rvalue has quit [Read error: Connection reset by peer]
rvalue has joined #ffmpeg-devel
Kei_N_ has joined #ffmpeg-devel
Kei_N has quit [Ping timeout: 248 seconds]
Krowl has joined #ffmpeg-devel
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
<ramiro>
courmisch: will the rvv assembly need to be updated if we change the strides in mpegvideoencdsp to ptrdiff_t?
<courmisch>
ramiro__: but you need to patch the init.c ofc
Krowl has quit [Read error: Connection reset by peer]
kasper93 has quit [*.net *.split]
blb has quit [*.net *.split]
hbbs has quit [*.net *.split]
zsoltiv_ has quit [*.net *.split]
nitroxis has quit [*.net *.split]
lemourin has quit [*.net *.split]
bilboed has quit [*.net *.split]
wyatt8740 has quit [*.net *.split]
englishm has quit [*.net *.split]
APic has quit [*.net *.split]
nitroxis has joined #ffmpeg-devel
englishm has joined #ffmpeg-devel
lemourin has joined #ffmpeg-devel
APic has joined #ffmpeg-devel
bilboed has joined #ffmpeg-devel
kasper93 has joined #ffmpeg-devel
zsoltiv_ has joined #ffmpeg-devel
blb has joined #ffmpeg-devel
hbbs has joined #ffmpeg-devel
hbbs has joined #ffmpeg-devel
hbbs has quit [Changing host]
wyatt8740 has joined #ffmpeg-devel
ramiro__ is now known as ramiro
<ramiro>
would it make sense to have clear_block_c() use AV_ZERO128() instead of memset()?
<ramiro>
if AV_ZERO128 has been optimized for an architecture, it assumes the pointer is aligned and issues better code for clear_block_c()
jarthur has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
<jamrial>
ramiro: the c version is boilerplate. no point trying to optimize it
<jamrial>
and a good compiler will probably make that memset into something similar to AV_ZERO128
aljazmc has joined #ffmpeg-devel
feiw1 has joined #ffmpeg-devel
feiw2 has quit [Remote host closed the connection]
ccawley2011 has quit [Ping timeout: 244 seconds]
<JEEB>
Lynne: just to verify, AVVulkanDeviceQueueFamily is something that external API clients are meant to utilize? or is it only meant to be utilized between avcodec vulkan video and the avutil hwcontext?
j45_ has joined #ffmpeg-devel
j45 has quit [Ping timeout: 276 seconds]
j45_ is now known as j45
j45 has quit [Changing host]
j45 has joined #ffmpeg-devel
j45_ has joined #ffmpeg-devel
j45 has quit [Ping timeout: 248 seconds]
j45_ is now known as j45
j45 has quit [Changing host]
j45 has joined #ffmpeg-devel
ccawley2011 has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
cone-291 has joined #ffmpeg-devel
<cone-291>
ffmpeg James Almer master:9d15fe77e33b: avcodec/container_fifo: add missing stddef.h include
vipyne has joined #ffmpeg-devel
feiw2 has joined #ffmpeg-devel
feiw1 has quit [Read error: Connection reset by peer]
thilo has quit [Ping timeout: 245 seconds]
thilo has joined #ffmpeg-devel
thilo has quit [Changing host]
thilo has joined #ffmpeg-devel
dionisis has quit [Quit: WeeChat 3.8]
Sean_McG has quit [Remote host closed the connection]
dionisis has joined #ffmpeg-devel
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
<Lynne>
JEEB: they're meant to utilize, since it also indicates which queues are supported and initialized
<Lynne>
if you let lavu init the hwcontext, its the only way you'll know what you are able to use
<JEEB>
alright. thanks.
dionisis has quit [Quit: WeeChat 3.8]
dionisis has joined #ffmpeg-devel
<JEEB>
just trying to figure out how and what/where to adjust so that libraries which do not themselves utilize hwcontext_vulkan, but then provide helper headers that include hwcontext_vulkan could better check whether vulkan is new enough. I guess a configure check which attempts to include libavutil/hwcontext_vulkan.h is the only way to make sure.
<JEEB>
because 3rd parties adding AVUTIL version checks each time hwcontext_vulkan.h changes its minimal required version for utilized types doesn't sound too feasible
<Lynne>
I didn't bump configure version though, old pre-encoding vulkan headers should work
<Lynne>
they do require decoding, but that was merged in 2022 december, surely that's old enough to universally support?
<JEEB>
VkVideoCodecOperationFlagBitsKHR is what is now causing failures a few things downstream
<JEEB>
so just trying to wonder how to make it so that projects can figure this out at configure time. and it seems like the only way is to handle it so that if you have something like config.h in a project, you make sure you don't include hwcontext_vulkan if it cannot be included
<JEEB>
and for the record, this is for builds where FFmpeg and the end-of-the-line project have vulkan stuff disabled. another library just has a helper header which then as long as avutil is new enough and the library's own vulkan usage flags are enabled includes hwcontext_vulkan
<JEEB>
so I guess instead of having that avutil version check it's just better to check whether hwcontext_vulkan could be included at build time (as it is IIRC always installed into include/), and block its inclusion from helper headers used by downstream if it was not possible (unless overridden by defines or so)
<JEEB>
and just to put names onto things, libplacebo currently enables vulkan stuff in its FFmpeg helper header if libplacebo's own vulkan requirements are met, and the avutil version is new enough. which breaks down if hwcontext_vulkan as a header requires newer vulkan than libplacebo :)
<JEEB>
and since with just the lib build the library itself doesn't utilize that helper header, that breakage unfortunately only is noticed then in downstream projects during build time :D (even if in that downstream project vulkan is disabled)
<JEEB>
so yea, tl;dr it seems like as hwcontext_vulkan can change its requirements it's better to just check whether it can be included with the current build env vulkan headers
<JEEB>
(as the current method only works if libplacebo's vulkan functionality requirements are >= hwcontext_vulkan's requirements for vulkan)
<JEEB>
thanks, I think I've rubber duckie'd this somewhat :)
Gramner has quit [Ping timeout: 252 seconds]
vipyne has quit [Quit: Leaving.]
Gramner has joined #ffmpeg-devel
<haasn>
I will once again request that ffmpeg vulkan not unconditionally pull in bleeding edge vulkan header features
Krowl has joined #ffmpeg-devel
<JEEB>
on FFmpeg's side it could be improved - for headers which are installed whether or not vulkan is actually enabled in FFmpeg - by adding the stuff that has bumped the requirement into its own header, thus those API users which require that stuff would not suddenly have the requirements of that existing public header get bumped.
<Lynne>
bleeding edge? the video queue stuff was merged 2 years ago
<Lynne>
2022-09-29
<JEEB>
13489c8a2154a2e0e8fd3c3c45f7856b4c3110b0
<JEEB>
AVVulkanDeviceQueueFamily utilizing VkVideoCodecOperationFlagBitsKHR bumped the requirement for that public header
<JEEB>
which is installed irrespective of whether vulkan is enabled in FFmpeg IIRC
<Lynne>
yes, that enum was merged into the vulkan spec in 2022
<JEEB>
ah
<wbs>
JEEB: btw, do you want to try to pick up and finish the videotoolbox AV1 patch that was posted a couple months ago?
<JEEB>
wbs: yea, currently I'm stuck trying to build an iOS app to test it since that's the only thing which I have that supports AV1 :D
<JEEB>
since apple in their great wisdom did not add swdec
<wbs>
JEEB: ah - I have an M3 where I can test it, but haven't had time to pick up the patch and push it across the line
<JEEB>
going to verkkokauppa.com just to curl a binary and hope not to look too suspicious at it
<JEEB>
is a bit tiring >_>
<JEEB>
(yes, I did that once)
<wbs>
lol
<JEEB>
("all I got was this segfault")
<JEEB>
apparently the guy's additions made it not segfault for him
<wbs>
yep, that form of the patches seems to run fine for me
<wbs>
the main thing iirc was to prepend sequence headers to the frame data, or something like that
<JEEB>
but yea, access to an M3 would make things simpler since my current iOS test app is just able to call avutil version string function
<JEEB>
Lynne: anyways the build failures are mostly because API users got used to "oh if I find vulkan well enough for my project and avutil is new enough to contain the vulkan hwcontext I can just include its header". so FFmpeg itself probably had vulkan as a feature disabled, and the public header is always installed. and an old assumption was found to no longer hold true.
<JEEB>
and the only reason I think this is noticed with vulkan is because most of the other hwcontexts' exposed external API types etc haven't changed in quite a while
Traneptora has joined #ffmpeg-devel
Krowl has quit [Quit: Krowl]
Krowl has joined #ffmpeg-devel
witchymary has quit [Remote host closed the connection]
witchymary has joined #ffmpeg-devel
vipyne has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
witchymary has quit [Remote host closed the connection]
witchymary has joined #ffmpeg-devel
cone-291 has quit [Quit: transmission timeout]
IndecisiveTurtle has quit [Ping timeout: 244 seconds]
jarthur has quit [Quit: jarthur]
jarthur has joined #ffmpeg-devel
aljazmc has quit [Remote host closed the connection]
aljazmc_ has joined #ffmpeg-devel
IndecisiveTurtle has joined #ffmpeg-devel
ccawley2011 has quit [Read error: Connection reset by peer]
feiw1 has joined #ffmpeg-devel
feiw2 has quit [Remote host closed the connection]
witchymary has quit [Remote host closed the connection]
witchymary has joined #ffmpeg-devel
aljazmc_ has quit [Quit: Leaving]
IndecisiveTurtle has quit [Ping timeout: 245 seconds]