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 6.1.1 has been released! | Please read ffmpeg.org/developer.html#Code-of-conduct
HarshK23 has quit [Quit: Connection closed for inactivity]
hamzah has quit [Ping timeout: 250 seconds]
MrZeus has quit [Read error: Connection reset by peer]
MrZeus has joined #ffmpeg-devel
Traneptora has quit [Quit: Quit]
Marth64 has joined #ffmpeg-devel
kurosu has quit [Quit: Connection closed for inactivity]
mkver has quit [Ping timeout: 272 seconds]
Marth64 has quit [Ping timeout: 264 seconds]
Marth64 has joined #ffmpeg-devel
clarkh_ has quit [Quit: Connection closed for inactivity]
Traneptora has joined #ffmpeg-devel
rvalue has quit [Ping timeout: 268 seconds]
rvalue has joined #ffmpeg-devel
thilo has quit [Ping timeout: 268 seconds]
thilo has joined #ffmpeg-devel
Marth64 has quit [Ping timeout: 255 seconds]
Marth64 has joined #ffmpeg-devel
<kierank>
BBB: John carmack liked your x86inc ASM blog post
lexano has quit [Ping timeout: 272 seconds]
<BBB>
tnx for posting that on twitter :)
MrZeus has quit [Ping timeout: 272 seconds]
<kierank>
He retweeted it to his 1m followes
<kierank>
Followers
mateo` has quit [Ping timeout: 264 seconds]
Marth64 has quit [Ping timeout: 256 seconds]
Marth64 has joined #ffmpeg-devel
mateo` has joined #ffmpeg-devel
Marth64 has quit [Ping timeout: 256 seconds]
Marth64 has joined #ffmpeg-devel
tmm1 has quit [Ping timeout: 256 seconds]
tmm1 has joined #ffmpeg-devel
jarthur has quit [Quit: jarthur]
<Lynne>
I wanna write an xhe-aac encoder now
<Lynne>
it is a total bitstream rewrite that fixes the biggest barrier to writing a good encoder, and lets you just write a much simpler one
<Lynne>
it scraps the fucking scalefactor delta coding with range limits for just using RLE
AbleBacon has quit [Read error: Connection reset by peer]
jamrial has quit []
HarshK23 has joined #ffmpeg-devel
Sean_McG has quit [Quit: leaving]
Teukka has quit [Quit: Not to know is bad; not to wish to know is worse. -- African Proverb]
Teukka has joined #ffmpeg-devel
Teukka has quit [Changing host]
Teukka has joined #ffmpeg-devel
hamzah has joined #ffmpeg-devel
MisterMinister has joined #ffmpeg-devel
Martchus_ has joined #ffmpeg-devel
Martchus has quit [Ping timeout: 260 seconds]
hamzah has quit [Ping timeout: 250 seconds]
Labnan has joined #ffmpeg-devel
t0001 has joined #ffmpeg-devel
<t0001>
Hello.
hamzah has joined #ffmpeg-devel
Marth64 has quit [Ping timeout: 252 seconds]
Marth64 has joined #ffmpeg-devel
t0001 has quit [Ping timeout: 250 seconds]
hamzah has quit [Ping timeout: 250 seconds]
t0001 has joined #ffmpeg-devel
t0001 has quit [Client Quit]
<Marth64>
bunch of closed captions patches coming
<Marth64>
waiting for fate to pass one last time
pzy has quit [Read error: Connection reset by peer]
pzy has joined #ffmpeg-devel
pzy_ has joined #ffmpeg-devel
pzy has quit [Ping timeout: 260 seconds]
pzy has joined #ffmpeg-devel
pzy_ has quit [Ping timeout: 260 seconds]
t0001 has joined #ffmpeg-devel
t24 has joined #ffmpeg-devel
t0001 has quit [Ping timeout: 250 seconds]
pzy has quit [Ping timeout: 252 seconds]
t24 has quit [Quit: Client closed]
t0001 has joined #ffmpeg-devel
t0001 has quit [Client Quit]
kurosu has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
MisterMinister has quit [Ping timeout: 240 seconds]
elastic_dog has quit [Ping timeout: 255 seconds]
Krowl has quit [Read error: Connection reset by peer]
elastic_dog has joined #ffmpeg-devel
ngaullier has joined #ffmpeg-devel
<JEEB>
TIL av_dict things are deprecated since 2011 :D
<JEEB>
yet we have plenty of code using it
<JEEB>
although we don't have #if guards for it so it doesn't seem to be "really deprecated"
cone-302 has joined #ffmpeg-devel
<cone-302>
ffmpeg Anton Khirnov master:639208cc6ada: fftools/ffmpeg_filter: don't forward EOF in close_output()
<cone-302>
ffmpeg Anton Khirnov master:ce178be8b0ca: fftools/ffmpeg_filter: do not assume av_buffersrc_get_nb_failed_requests()>0
<cone-302>
ffmpeg Anton Khirnov master:f344277a880b: fftools/ffmpeg_filter: merge block with the enclosing one
mkver has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
pzy has joined #ffmpeg-devel
rajivharlalka has joined #ffmpeg-devel
<rajivharlalka>
Hi, Is there a document on writing fate tests? Couldn't find one on the docs and the tests directory in the source code is getting quite overwhelming.
<elenril>
not really
<rajivharlalka>
I wish to write some tests for an avfilter, av_atempo.c
<elenril>
look at tests/fate/filter-audio.mak?
klaxa has quit [Remote host closed the connection]
klaxa has joined #ffmpeg-devel
Marth64 has quit [Ping timeout: 268 seconds]
Marth64 has joined #ffmpeg-devel
agrosant has joined #ffmpeg-devel
pzy has quit [Ping timeout: 255 seconds]
pzy has joined #ffmpeg-devel
pzy has quit [Ping timeout: 240 seconds]
<cone-302>
ffmpeg Stefano Sabatini master:5828aaa2b5da: lavf/fifo: sort options by name
Krowl has quit [Read error: Connection reset by peer]
agrosant has joined #ffmpeg-devel
mkver has quit [Remote host closed the connection]
mkver has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
lexano has joined #ffmpeg-devel
omegatron has joined #ffmpeg-devel
rajivharlalka has quit [Quit: Connection closed for inactivity]
jamrial has joined #ffmpeg-devel
<cone-302>
ffmpeg Fei Wang master:c252d2b84a35: lavc/vvcdec: Add missed chroma sampling factor for crop offset
<cone-302>
ffmpeg Fei Wang master:4027c2e1d039: lavc/vvc_ps: Correct NoOutputBeforeRecoveryFlag of IDR
Krowl has quit [Read error: Connection reset by peer]
<elenril>
why does the fifo muxer exist anyway
<JEEB>
probably to not block ffmpeg looping
<JEEB>
the whole main loop which is now split into multiple threads, I think
* JEEB
has used fifo+tee for async muxing and outputting same packets into multiple end points
<mkver>
elenril: As a workaround for the non-threading in the ffmpeg cli, I presume?
<mkver>
I would like to see it nuked.
<elenril>
so should we remove it then?
<mkver>
It is the only part in the codebase that reuses AVFormatContexts after av_write_trailer().
<mkver>
This is the reason that av_write_trailer() contains a bunch of freeing code.
Krowl has joined #ffmpeg-devel
aaabbb has joined #ffmpeg-devel
agrosant has quit [Ping timeout: 264 seconds]
MrZeus has joined #ffmpeg-devel
<JEEB>
fun, dict.h points at tree.h and then I check usages: avutil/tree.h => 6 hits, avutil/dict.h => 100 hits
<jamrial>
yeah, the idea apparently was for tree to replace dict, but it never happened
<JEEB>
I only learned of that earlier today when someone who happened to actually open the dict.h header and then told me that it was deprecated :D
Marth64 has quit [Ping timeout: 260 seconds]
Marth64 has joined #ffmpeg-devel
agrosant has joined #ffmpeg-devel
agrosant has quit [Ping timeout: 264 seconds]
<cone-302>
ffmpeg Andreas Rheinhardt master:9b9b57419197: avcodec/tiff: Fix handling of av_strdup() failures
<cone-302>
ffmpeg Andreas Rheinhardt master:957053b52a2d: avcodec/tiff: Avoid duplicating strings
<cone-302>
ffmpeg Andreas Rheinhardt master:f6f94c335765: avcodec/tiff: Don't check before av_freep()
<cone-302>
ffmpeg Andreas Rheinhardt master:c14dc99f44c6: avcodec/tiff: Improve inclusions
<cone-302>
ffmpeg Marth64 master:9b981e311214: avcodec/rcwtenc: canonize name and refresh documentation
<cone-302>
ffmpeg Marth64 master:b3d87fb0c4aa: avformat/sccdec: remove unused bprint.h include
agrosant has joined #ffmpeg-devel
<Marth64>
will send patch for compiler warning in dvd shortly
<Marth64>
but yay, colors and menus
<jamrial>
Marth64: can it handle Shrek's DVD menus? :p
Marth128 has joined #ffmpeg-devel
Marth64 has quit [Killed (NickServ (GHOST command used by Marth128!~Marth64@98.159.224.162))]
Marth128 is now known as Marth64
<Marth64>
jamrial: I do have that one, I'll try it out this week :P
<Marth64>
it's somewhere in the bin of discs
<Marth64>
1 DVD drive died during this journey, RIP
<Marth64>
it literally started scratching discs in circular patterns on its last days
agrosant has quit [Ping timeout: 264 seconds]
AbleBacon has joined #ffmpeg-devel
<Marth64>
are there any decent linters folks use? I'd like a computer to catch me making silly mistakes
<Marth64>
cppcheck is usually quiet for me
<Marth64>
and clang-tidy I need to tune
<Lynne>
how did you get a drive killed? I wrote a CD ripper without killing one, and it does seeks at every frame
agrosant has joined #ffmpeg-devel
<Marth64>
Lynne: a lot of obsessive testing over the course of a year, but i think what pushed it is a phase i was testing with rotted/scratched/purposely bad sector discs
<Marth64>
and letting ddrescue run for hours at a time
agrosant has quit [Ping timeout: 264 seconds]
<Marth64>
also, CD ripper? :o I have been wanting to let go of EAC for some time
<JEEB>
cyanrip IIRC
<Marth64>
oh. neat.
Krowl has quit [Read error: Connection reset by peer]
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
<gnafu>
cyanrip is amazing.
<gnafu>
I am tempted to rerip my entire collection when 1.0 is cut >_>.
<Marth64>
Lynne: if you need any drives for testing, I have at least 20 in a box of all kind of varieties
<Marth64>
(or feel free to delgate any tests to me, I can do here)
justache is now known as fotastache
<Marth64>
I know things get funky with scsi commands
Krowl has joined #ffmpeg-devel
<kepstin>
looks like it uses libcdio-paranoia for interfacing with the drives, so i wouldn't expect any issues there (and if they are, probably should be reported to that project)
<Marth64>
libcdio is nice. I have mostly used aaru and commend the developer for their patience doing direct operations with different drives
<Marth64>
at one point I added drive speed settings to dvd demuxer, but things started to get bloated, I want to stop adding to it except for the 2 essentials left (seeking/listing of what's on the disc)
stevenliu has joined #ffmpeg-devel
stevenliu_ has quit [Ping timeout: 272 seconds]
MrZeus_ has joined #ffmpeg-devel
MrZeus has quit [Ping timeout: 255 seconds]
<Oneric>
Marth64: thx for the review!
<Oneric>
(btw, you replied to the v3 instead of v4 thread, but as there were no changes inbetween it doesn’t matter much; v4 was just rebased due to FATE being generally broken during v3)
Sean_McG has joined #ffmpeg-devel
<Marth64>
Oneric: NP, yes I had realized it later. I started to test the BiDi marks but didn't finish .. will pick on it more this week
<Marth64>
mkver: PRId64 ?
<mkver>
yes
<Marth64>
✔
<Sean_McG>
I haven't heard from anybody on either of my patches and it has been quite a few days now, both the one for aacenc_pred and sub2video
<mkver>
Sean_McG: I remember that I replied to your aacenc_pred patch.
<mkver>
But I have forgotten what I said.
pzy has joined #ffmpeg-devel
<Marth64>
mkver: since you have taught me this, I will fix the other %ld to be PRId64 (int64_t) and PRIlu64 (uint64_t) - log statements in same file
<mkver>
PRIu64, not PRIlu64
<Marth64>
yes, typo, sry.
<Marth64>
Sean_McG: I migt have a big endian somewhere to validate sub2video test
<Sean_McG>
re: the sub2video patch -- I wondered if it is considered correct to force a pix_fmt in FATE... if not, we might need to generate big-endian refs and switch appropriately in the Makefiles
<Sean_McG>
(probably for all the tests that output as raw_video)
<Sean_McG>
errr rawvideo, no _
<mkver>
Sean_McG: We already force pix_fmt in FATE in these cases.
<Sean_McG>
OK cool, so this one was just missed
<Sean_McG>
probably because FATE had no BE coverage for quite some time
jarthur has joined #ffmpeg-devel
pzy has quit [Quit: Konversation terminated!]
pzy has joined #ffmpeg-devel
<cone-302>
ffmpeg James Almer master:7cf4e305a2d2: fate/h264: add missing swscale dependency to h264-bsf-mp4toannexb-new-extradata
<cone-302>
ffmpeg Andreas Rheinhardt master:c9eb36e0a0b0: avcodec/tiff: Use ff_tget_long() where appropriate
fotastache has left #ffmpeg-devel [bye?]
Krowl has quit [Read error: Connection reset by peer]
omegatron has quit [Quit: Power is a curious thing. It can be contained, hidden, locked away, and yet it always breaks free.]
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
averne has quit [Ping timeout: 264 seconds]
hamzah has quit [Quit: Client closed]
Traneptora has quit [Quit: Quit]
Traneptora has joined #ffmpeg-devel
hamzah has joined #ffmpeg-devel
gnafu has quit [Quit: Rebooting for kernel update.]
gnafu has joined #ffmpeg-devel
cone-302 has quit [Quit: transmission timeout]
Livio has joined #ffmpeg-devel
averne has joined #ffmpeg-devel
kurosu has quit [Quit: Connection closed for inactivity]
jarthur has quit [Ping timeout: 268 seconds]
jarthur has joined #ffmpeg-devel
johnmcnuggets has joined #ffmpeg-devel
<elenril>
michaelni: any further comments on the ffmpeg CLI set?
johnmcnuggets has left #ffmpeg-devel [Leaving]
<Marth64>
Is there a golden example of ffio_ensure_seekback()?
<Traneptora>
apngdec has a few calls
<Marth64>
ty
<Traneptora>
as does mpegts if that's insufficient
<Marth64>
apng looks tidy. I'll look there. thx
cone-636 has joined #ffmpeg-devel
<cone-636>
ffmpeg Leo Izen master:a356815a3192: avcodec/libjxlenc: add option to disable xyb encoding
<Traneptora>
btw elenril were you still looking to review the exif patch I sent as-is or should I make a v2 addressing andreas's points before review
<Traneptora>
I remember you mentioned reviewing it was on the todo list but guessing it got overshadowed by busy stuffs
iive has joined #ffmpeg-devel
<cone-636>
ffmpeg Andreas Rheinhardt master:41ab25cb2d58: avformat/fifo_test: Move into tests/fifo_muxer.c
MrZeus_ has quit [Read error: Connection reset by peer]
<frankplow>
I've started getting a lot of -fsanitize=function errors from all over the codebase when compiling with clang-usan recently, never used to get them. Any ideas what might have caused this?
<mkver>
frankplow: Clang-UBSan gained the capability to check the function pointer types.
<mkver>
You get these errors from resample.c, didn't you?
<mkver>
More exactly, Clang-usan gained this ability for C.
<frankplow>
mkver: Ah okay, so they're not necessarily spurious?
<mkver>
No, they are real.
<Sean_McG>
oh, cute
<mkver>
Sean_McG: Sent a patch for your ppc-be box to test.
<Sean_McG>
cheers, it's very slow to build so I'll start that now
<cone-636>
ffmpeg Kristoffer Brånemyr master:51d303e20cbb: avformat/webvttdec: Skip more parts of header to let parsing continue
<cone-636>
ffmpeg Nicolas Gaullier master:275add93287f: fate: fix generating references when sh=dash
<cone-636>
ffmpeg Christian Lee Seibold master:384cc270d25e: avformat/gopher: Add audio and video itemtypes
<frankplow>
Most recently, I've gotten them from: avfilter.c in three locations, hscale.c and parser.c
<mkver>
Sean_McG: Actually, you should only need to rebuild one file. But I guess running the fate suite on an emulator with ubsan is indeed taxing.
MrZeus has joined #ffmpeg-devel
<Sean_McG>
hey um, are we absolutely sure that the sh=dash fix is still compatible with bash?
<mkver>
frankplow: I will rerun fate with clang ubsan.
<mkver>
frankplow: It was way worse before f705b8b5b4f and ab2173c0a5306.
<Sean_McG>
wow Trac is suddenly so slowwwwww
<frankplow>
mkver: Yeah the other day I had dozens of errors, had to add -fno-sanitize=function to find the UB I was looking for. Thanks for sorting that.
MrZeus has quit [Read error: Connection reset by peer]
<Traneptora>
trac is suddenly so slowwwww?
<Traneptora>
it's been slow as all heck for over 10 years
<Traneptora>
Sean_McG: dash supports a strict subset of bash so what do you mean?
<Marth64>
needs janitor to purge old tickets and optimize db
<Traneptora>
dash is essentially posix-only, plus iirc it also supports the local keyword
<Traneptora>
bash is posix-compliant and local works the same way in bash and in dash
<Traneptora>
officially dash and bash are "incompatible" but that's because Bash is not strictly posix-compliant. anything that is posix-compliant and works in bash should also work in dash
<Sean_McG>
OK, if you're all comfortable then that's fine
Livio has quit [Ping timeout: 256 seconds]
Livio has joined #ffmpeg-devel
<Marth64>
mkver: sending a small dead code deletion patch for opt_common.c. I am curious your opinion. I spent ~45 minutes trying to justify it's existence, and could not.
<Sean_McG>
OK this build will probably take like > 2 hours, especially since it is also doing a FATE run already. I should go eat.
<Sean_McG>
just for reference, building gcc 13.2.0 on this emulator took like 2.5+ days -- but I think it was worth it for the newer UBsan stuff
<nevcairiel>
in that time you could setup cross-compiling :D
<Sean_McG>
yes but stress level would have been through the roof
<mkver>
Sean_McG: Found the avi issue. Patch on ML.
<Sean_McG>
nice. I'm heading out for food so I'll have a peek later.
<Marth64>
memory management is hard coming from java background
<Marth64>
you get lazy and used to not dealing with pointers
* Marth64
facepalm
<Traneptora>
yea, that was me at first, but honestly it's nicer that you can do it yourself
<Traneptora>
large array allocs in java are expensive so you often have to pool them and then you feel like the automatic memory management is pointless
MrZeus has joined #ffmpeg-devel
<compn>
i wonder if there are cross compiling vm you can just download and boom
MrZeus has quit [Read error: Connection reset by peer]
<compn>
make things easier
<compn>
everythings a virtual machine now
<Traneptora>
I'm sure they exist, but windows is not a free operating system, so I'm not really sure how you'd just *do* that
<Traneptora>
without paying or having some kind of license