ChanServ changed the topic of #ffmpeg to: Welcome to the FFmpeg USER support channel | Development channel: #ffmpeg-devel | Bug reports: https://ffmpeg.org/bugreports.html | Wiki: https://trac.ffmpeg.org/ | This channel is publically logged | FFmpeg 7.1.1 is released
Kruppt has quit [Quit: Leaving]
Shine_ has joined #ffmpeg
Shine__ has joined #ffmpeg
Shine_ has quit [Ping timeout: 244 seconds]
dallemon has quit [Ping timeout: 244 seconds]
iive has quit [Quit: They came for me...]
YUiNA_ has quit [Remote host closed the connection]
jtgd has quit [Quit: WeeChat 4.6.3]
jtgd has joined #ffmpeg
jtgd has quit [Client Quit]
jtgd has joined #ffmpeg
apteryx has quit [Quit: connection reset by purr]
Shine__ has quit [Read error: Connection reset by peer]
minimal has quit [Quit: Leaving]
damian101_ has joined #ffmpeg
damian101 has quit [Ping timeout: 276 seconds]
apteryx has joined #ffmpeg
xvaclav has quit [Remote host closed the connection]
javabean has quit [Quit: well, shoot]
javabean has joined #ffmpeg
acidbunny has quit [Read error: Connection reset by peer]
Shine_ has joined #ffmpeg
maxim_d33 has quit [Ping timeout: 265 seconds]
maxim_d33 has joined #ffmpeg
s55 has quit [Ping timeout: 276 seconds]
s55 has joined #ffmpeg
Guest58 has joined #ffmpeg
Guest58 has quit [Client Quit]
libra811 has joined #ffmpeg
Shine_ has quit [Read error: Connection reset by peer]
kasper93 has joined #ffmpeg
kasper93 has quit [Client Quit]
kasper93 has joined #ffmpeg
MisterMinister has quit [Ping timeout: 248 seconds]
lavaball has quit [Remote host closed the connection]
Richardcavell has quit [Ping timeout: 276 seconds]
cantelope has joined #ffmpeg
MisterMinister has joined #ffmpeg
qqe has quit [Quit: Lost terminal]
Richardcavell has joined #ffmpeg
rsx has quit [Quit: rsx]
kasper93_ has joined #ffmpeg
kasper93 is now known as Guest2629
Guest2629 has quit [Killed (silver.libera.chat (Nickname regained by services))]
kasper93_ is now known as kasper93
microchip_ has quit [Remote host closed the connection]
microchip_ has joined #ffmpeg
lavaball has joined #ffmpeg
Vonter has quit [Quit: WeeChat 4.6.3]
Richardcavell has quit [Ping timeout: 260 seconds]
minimal has joined #ffmpeg
YUiNA_ has joined #ffmpeg
YUiNA has quit [Ping timeout: 276 seconds]
five618480339176 has joined #ffmpeg
Shuriko has joined #ffmpeg
beaver has joined #ffmpeg
YUiNA has joined #ffmpeg
YUiNA_ has quit [Ping timeout: 248 seconds]
\\Mr_C\\ has joined #ffmpeg
System_Error has quit [Remote host closed the connection]
Luvveous has joined #ffmpeg
lavaball has quit [Quit: lavaball]
lavaball has joined #ffmpeg
NotWarcop has quit [Remote host closed the connection]
jmcantrell has joined #ffmpeg
vulpine has quit [Quit: Connection reset by purr]
hussein1 has joined #ffmpeg
vulpine has joined #ffmpeg
lusciouslover has quit [Quit: \]
mtoy has quit []
lusciouslover has joined #ffmpeg
iive has joined #ffmpeg
Richardcavell has joined #ffmpeg
jmcantrell has quit [Ping timeout: 248 seconds]
dallemon has joined #ffmpeg
coldfeet has joined #ffmpeg
jmcantrell has joined #ffmpeg
Traneptora has quit [Quit: Quit]
jmcantrell has quit [Ping timeout: 252 seconds]
bsFFFFFF has quit [Read error: Connection reset by peer]
bsFFFFFF has joined #ffmpeg
Shine_ has quit [Read error: Connection reset by peer]
bsFFFFFF has quit [Ping timeout: 244 seconds]
bsFFFFFF has joined #ffmpeg
bsFFFFFF has quit [Quit: bsFFFFFF]
coldfeet has quit [Remote host closed the connection]
Blacker47 has quit [Quit: Life is short. Get a V.90 modem fast!]
beaver has quit [Quit: don`t touch bout baby]
beaver has joined #ffmpeg
jmcantrell has joined #ffmpeg
Richardcavell has quit [Ping timeout: 252 seconds]
flyingtoaster has joined #ffmpeg
<flyingtoaster>
Hi everyone. I have a file that's exhibiting some strange behaviour. It's a video from a Google Pixel 6: `Stream #0:0[0x1](und): Video: vp9 (Profile 0) (vp09 / 0x39307076), yuv420p(tv, bt709), 1920x1080, 3158 kb/s, 30 fps, 30 tbr, 15360 tbn (default)`
<flyingtoaster>
I've found that when reading through the file, certain frames will cause `avcodec_receive_frame` to return `AVERROR_EOF`. Looking at the output of `ffprobe -show_frames`, I don't see anything different about these frames.
<flyingtoaster>
I did notice that FFmpeg required that I enable the wrapped_avframe decoder in order to read the file in the first place. Is there some difference in how these are intended to be read?
<another|>
Is this a custom build?
<flyingtoaster>
Of FFmpeg? Yes.
<BtbN>
Why would decoding a normal file yield you wrapped avframes?
<flyingtoaster>
I'm not entirely sure. This is what I got before enabling wrapped_avframe: https://pastebin.com/FCevidAj
<BtbN>
format null?
<BtbN>
are you missing the vp9 decoder?
<flyingtoaster>
This was run: `ffmpeg -i video.mp4 -vf vfrdet -an -f null -`
<flyingtoaster>
So it's probably a red-herring.
<BtbN>
Does it work on a bog standard build from latest master?
<another|>
well, yeah. If you're gonna mux to null you'll need wrapped_avframe
<flyingtoaster>
ffplay from master doesn't seem to have any issue with the file.
<flyingtoaster>
If I move past the bad frame in the timeline in my application, it continues to decode properly.
<another|>
unless you specify something elsse
<flyingtoaster>
ffplay from my package manger, 5.1.6, doesn't seem to have an issue with the file either.
<flyingtoaster>
In this file, pts = dst
<flyingtoaster>
*pts = dts
jmcantrell has quit [Ping timeout: 248 seconds]
bleb has quit [Ping timeout: 260 seconds]
bleb has joined #ffmpeg
beaver is now known as pong
Luvveous has quit [Quit: Going offline, see ya! (www.adiirc.com)]
<flyingtoaster>
I guess my best bet to figure out what's going wrong is to step through the FFmpeg source and see where it's raising the EOF.
tel has quit [Ping timeout: 252 seconds]
<flyingtoaster>
Ah, that time it gave me: `[mov,mp4,m4a,3gp,3g2,mj2 @ 0x7fa404ae0a00] Unknown dref type 0x206c7275 size 12`
bleb has quit [Ping timeout: 260 seconds]
bleb has joined #ffmpeg
pong has quit [Remote host closed the connection]
<flyingtoaster>
I'm not entirely sure what a "dref" is, so I have no idea if this is suspicious or not.
<JEEB>
dref is one of the box types in mp4-like container formats
<JEEB>
you can utilize something like l-smash's boxdumper or atomicparsley to dump the structure of a MP4 or QTFF (mov) style file :)
pong has joined #ffmpeg
<JEEB>
the specification has a name ISO Base Media File Format, and it used to be free. then the ISO spec mafia changed the rules on how specifications become free (against the wishes of the authors), so now you have to search for "14496-12 2015 pdf" just to find the last freely available edition
Muimi has joined #ffmpeg
<JEEB>
dref is a data reference box, and 0x206c7275 if you parse those hex bytes is 'lru ', which in the other endian is 'url ', which matches the url data reference type :)
<flyingtoaster>
Right, so this is just noise.
<JEEB>
so for whatever reason the file is referencing data reachable through a URL (which is not supported as randomly opening I/O is a nice avenue for attack), but I'm not sure if that has anything to do with that VP9 problem
<JEEB>
with one of those applications I mentioned you can probably show what URL it refers to load data
<flyingtoaster>
boxdumper doesn't unfortunately
<JEEB>
with the `--box` argument?
bleb has quit [Ping timeout: 248 seconds]
<flyingtoaster>
Yeah, it just says: `[url : Data Entry Url Box] position = 383 size = 12 version = 0 flags = 0x000001 location = in the same file`
<flyingtoaster>
Unless "in the same file" is the URL.
<JEEB>
yea
<JEEB>
the url data entry box literally just has the location there as a value :)
<flyingtoaster>
Trying AtomicParsley gave me: `AtomicParsley error: unsupported MPEG-4 file brand found 'iso6'`
<JEEB>
ahh
<JEEB>
first byte in flags is @same file@ >D
<JEEB>
whoops, off-by-one on the keyboard physically
<JEEB>
I guess using url for this is rare? because I don't recall seeing this message from FFmpeg's parser on pretty much any of the files I've written out
bleb has joined #ffmpeg
<JEEB>
and that matches the size since it says in the spec the structure shall end with the flags field, and no actual location url string is to be written
<technomancy>
so is anubis working OK to keep trac from getting DDoSed now?
<JEEB>
technomancy: at least I don't see BtbN herping a derp about the load
<BtbN>
Well, a large chunk of US AT&T residential IPs are still just flat out blocked
<BtbN>
Didn't check if they are still trying to query trac
Traneptora has joined #ffmpeg
<technomancy>
but if they have, it hasn't been to the degree that it's destabilizing
xx has quit [Ping timeout: 244 seconds]
<flyingtoaster>
So it looks like `avcodec_send_packet` is returning `AVERROR_EOF` because `dc->draining_started` is true.
<flyingtoaster>
Which is strange because I have an assertion to make sure that the codec context is non-null before passing into `avcodec_send_packet`.
<flyingtoaster>
*avpacket - not context.
lavaball has quit [Quit: lavaball]
Muimi has quit [Quit: Going offline, see ya! (www.adiirc.com)]
termos_ has joined #ffmpeg
Mirarora has joined #ffmpeg
termos has quit [Ping timeout: 248 seconds]
flyingtoaster has quit [Ping timeout: 272 seconds]
hackkitten has quit [Ping timeout: 244 seconds]
SakuraChan has joined #ffmpeg
jarthur has joined #ffmpeg
Sakura`Kinomoto has quit [Ping timeout: 252 seconds]