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
<Traneptora>
haasn: film_grain_params.c, line 77
<Traneptora>
there's a continue; but it's inside a do { } while (0);
<Traneptora>
so does it actually do anything?
<Traneptora>
wouldn't that just break out of the do loop
<Traneptora>
you might need a label
<Traneptora>
or to just remove the do/while
<iive>
do {} while() executes at least once
<iive>
do {} while(0) executes exactly once
<iive>
and the continue is probably used as a way to avoid using goto.
<Lynne>
IndecisiveTurtle: two reasons, blocking artifacts and signalling overhead
<Lynne>
haar by itself is a lossless transform, as are all wavelet transforms
<Lynne>
but once you start to quantize the coefficients, the block edges quickly start to diverge, so you get discontinuities
<Lynne>
signalling overhead because with smaller blocks, you'll have more blocks to encode
any1 has quit [Read error: Connection reset by peer]
any1 has joined #ffmpeg-devel
lexano has quit [Ping timeout: 245 seconds]
<Traneptora>
iive: no, the issue in this case is that inside the do while(0) is a continue statement guarded by if and *nothing else*
<Traneptora>
the intent is likely to continue the outer loop, but it won't work since it's inside a dowhile0
<iive>
yeh, in this case continue and break do the same thing.
<Traneptora>
ye, but it looks like an oversight, hence my patch
<iive>
oh, there is nothing after the continue, that's the whole thing.
<iive>
you might want to keep the {} brackets.
<iive>
hum... in this case I think it would be better if "check" is just the if() part, so the usage is something like "CHECK(fgp->luma, bit_luma, 0) continue;" or even
<Traneptora>
dvdvideo_nav_header is a static const uint8_t[4]
<Traneptora>
would it make more sense to define the dvdvideo_nav_header as a uint32_t with MKTAG, and then use AV_RL32(&read_buf[foo]) == dvdvideo_nav_header?
rvalue has quit [Ping timeout: 255 seconds]
<mkver>
It will probably not make a difference. Compilers know memcmp (and memcpy).
<Traneptora>
that is fair
<mkver>
Anyway: This format is BE, so it should be AV_RB32(buf) == 0x000001BF.
<Traneptora>
sure, tho if you were to MKTAG it the endianness would swap anyway
tufei has quit [Remote host closed the connection]
tufei has joined #ffmpeg-devel
<jkqxz>
Lynne: I thought that was linked up in the generic code, but it isn't. Ignore that, then.
cone-174 has joined #ffmpeg-devel
<cone-174>
ffmpeg Andreas Rheinhardt master:07064f9bdac3: avformat/demux: Restore pkt->stream_index assert check
<cone-174>
ffmpeg Andreas Rheinhardt master:ccd2b7f858d2: avformat/demux: Combine "Packet corrupt" logmessages
<cone-174>
ffmpeg Andreas Rheinhardt master:0d43adcbef9a: configure: Explicitly check for static_assert, _Static_assert
IndecisiveTurtle has quit [Ping timeout: 260 seconds]
<mkver>
Traneptora: A start code is not really a tag. And even if it were: If the endianness of MKTAG differs from the endianness of the format, then the former should be avoided.
<Traneptora>
true
<mkver>
bye MSVC 19.27.
<Lynne>
jkqxz: thanks for looking at it
rvalue has joined #ffmpeg-devel
<BtbN>
Anyone able to look at the vf_stack patch? Depending on input resolution, it straight up heap overflows right now.
<BtbN>
The patch is most definitely not the correct solution, so I'd like to figure out how to fix it correctly.
Krowl has quit [Read error: Connection reset by peer]
tufei has quit [Remote host closed the connection]
tufei has joined #ffmpeg-devel
rvalue- has joined #ffmpeg-devel
rvalue has quit [Ping timeout: 268 seconds]
rvalue- is now known as rvalue
jamrial has quit []
jamrial has joined #ffmpeg-devel
Livio has joined #ffmpeg-devel
Marth64 has joined #ffmpeg-devel
<Marth64>
seems cc_dec silently skips text sometimes when switching normal->italic->normal
<Marth64>
will look into it sometime
Guest52 has joined #ffmpeg-devel
IndecisiveTurtle has joined #ffmpeg-devel
marcj has quit [Ping timeout: 240 seconds]
tufei has quit [Quit: Leaving]
<cone-174>
ffmpeg Marton Balint master:e6c2c8703732: avformat/mov_chan: respect channel order when parsing and creating chan atom
tufei has joined #ffmpeg-devel
<cone-174>
ffmpeg Marton Balint master:9a5627ea9a21: avfilter/af_channelmap: fix error message if FL source channel was missing
<cone-174>
ffmpeg Marton Balint master:1bea3e9ee2f3: avfilter/af_channelmap: fix mapping if in_channel was a string but out_channel was not specified
<cone-174>
ffmpeg Marton Balint master:2f754a96bd4a: avfilter/af_channelmap: disallow channel index 64
<cone-174>
ffmpeg Marton Balint master:bba6dd391f4e: avfilter/af_channelmap: factorize checking indexes against a channel layout
<cone-174>
ffmpeg Marton Balint master:eaca78eff832: avfilter/af_channelmap: add some additional checks for the mappings