BtbN 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 7.0 is released
lexano has quit [Ping timeout: 244 seconds]
<aaabbb>
another|: no, v4l2 does not support audio on linux (ie /dev/videoN itself does not do audio)
FlorianBad has quit [Quit: Konversation terminated!]
rvalue has quit [Read error: Connection reset by peer]
rvalue has joined #ffmpeg
sugoi has quit [Ping timeout: 260 seconds]
jemius has quit [Quit: Leaving]
fling has joined #ffmpeg
sugoi has joined #ffmpeg
sugoi has quit [Ping timeout: 252 seconds]
HerbY_NL has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
xx has joined #ffmpeg
sugoi has joined #ffmpeg
sugoi has quit [Ping timeout: 260 seconds]
iconoclasthero has joined #ffmpeg
Suchiman has joined #ffmpeg
sugoi has joined #ffmpeg
fling has quit [Ping timeout: 260 seconds]
sugoi has quit [Ping timeout: 265 seconds]
rsx has joined #ffmpeg
sugoi has joined #ffmpeg
sugoi has quit [Ping timeout: 252 seconds]
HerbY_NL has joined #ffmpeg
sugoi has joined #ffmpeg
sugoi has quit [Ping timeout: 255 seconds]
HerbY_NL has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
sugoi has joined #ffmpeg
sugoi has quit [Ping timeout: 260 seconds]
Sciencentistguy has quit [Quit: o/]
sugoi has joined #ffmpeg
sugoi has quit [Ping timeout: 252 seconds]
sugoi has joined #ffmpeg
GingerGeek has quit [Ping timeout: 252 seconds]
welder has quit [Ping timeout: 252 seconds]
javabean_ has joined #ffmpeg
JavaBean has quit [Read error: Connection reset by peer]
mrelcee_ has joined #ffmpeg
GingerGeek has joined #ffmpeg
mrelcee has quit [Ping timeout: 252 seconds]
sugoi has quit [Ping timeout: 244 seconds]
mrelcee_ is now known as mrelcee
Dotz0cat has quit [Ping timeout: 252 seconds]
Arokh has quit [Ping timeout: 260 seconds]
YuGiOhJCJ has quit [Quit: YuGiOhJCJ]
Dotz0cat has joined #ffmpeg
sugoi has joined #ffmpeg
rv1sr has quit []
Arokh has joined #ffmpeg
rv1sr has joined #ffmpeg
sugoi has quit [Ping timeout: 246 seconds]
coldfeet has quit [Remote host closed the connection]
DauntlessOne has quit [Ping timeout: 255 seconds]
sugoi has joined #ffmpeg
jemius has joined #ffmpeg
sugoi has quit [Ping timeout: 252 seconds]
sugoi has joined #ffmpeg
coldfeet has joined #ffmpeg
luva88 has quit [Remote host closed the connection]
sugoi has quit [Ping timeout: 264 seconds]
sugoi has joined #ffmpeg
sugoi has quit [Ping timeout: 260 seconds]
MoC has joined #ffmpeg
sugoi has joined #ffmpeg
Sciencentistguy has joined #ffmpeg
sugoi has quit [Ping timeout: 265 seconds]
minimal has joined #ffmpeg
javabean_ is now known as JavaBean
Sciencentistguy4 has joined #ffmpeg
Sciencentistguy has quit [Ping timeout: 252 seconds]
Sciencentistguy4 is now known as Sciencentistguy
sugoi has joined #ffmpeg
sugoi has quit [Ping timeout: 252 seconds]
HerbY_NL has joined #ffmpeg
zmt00 has joined #ffmpeg
swamp_ has quit [Ping timeout: 252 seconds]
sugoi has joined #ffmpeg
sugoi has quit [Ping timeout: 246 seconds]
sugoi has joined #ffmpeg
HerbY_NL has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
sugoi has quit [Ping timeout: 252 seconds]
sugoi has joined #ffmpeg
sugoi has quit [Ping timeout: 260 seconds]
<jemius>
hmm. I encoded an old black-white movie with h.265. It became surprisingly large. I suspect that might be due to the codec not being able to compress the quite strong noise
<JEEB>
with any modern format you should be able to compress things quite a bit (anything sane like x264 for that format or newer), so if something didn't compress it means that you didn't make it compress hard enough rate control wise
<JEEB>
if you do not care about a specific bit rate, then with CRF you just raise the number until you get into ranges that you are OK with. then you can look at whether such compression looks good enough for you
<jemius>
I just kept the CRF default, which is 20
sugoi has joined #ffmpeg
<JEEB>
which is just some default set by the encoder. I do not expect you to know what it leads to result-wise for your content type (fps, what kind of content etc)
<JEEB>
so just bang it upwards if in your opinion it doesn't compress enough
<jemius>
sure. I might try that. But might be a good opportunity to fire up a denoiser, too
<JEEB>
yea I am just saying that without knowing what the expected quality level given by that quantizer range to which that CRF value leads to, you can't really say whether it's a case of "not being able to compress (enough)"
<JEEB>
with modern formats unless you're going deep into the weeds ("I'd like this 20 min animated thing with audio fit into 1.44MiB"), "not being able to compress" should not be a problem with any modern format with a sane encoder
sugoi has quit [Ping timeout: 260 seconds]
<JEEB>
then the question is whether that compression ratio that you'd like looks good enough. that should then be the trigger for thinking about additional filters.
<JEEB>
(if you really require the content within that range - even if modified)
<JEEB>
although I would first poke at "make the encoder try more things and look at more things" first, also looking at the GOP duration. 5-10 second GOPs are normal for final playback content that doesn't need to be quickly seeked through into any frame.
<jemius>
JEEB, what's a GOP?
<JEEB>
group of pictures
<JEEB>
how long the encoder may utilize until the next random access picture
<jemius>
that's not the same as a key frame, is it?
<JEEB>
no
<JEEB>
in earlier formats (before H.264 etc) keyframe was generally also a random access point
<JEEB>
keyframe = all intra, random access point = point from which a decoder can start decoding with the promise that following images can also be decoded
sugoi has joined #ffmpeg
<JEEB>
actually now that I think about it, didn't earlier formats like mpeg-2 video or mpeg-4 visual let you do intra only (= keyframe), but not promise that you can decode anything further starting from it
<JEEB>
but yes, for example due to Older Simpler Times the flag for random access points in FFmpeg is called FLAG_KEY :D
<JEEB>
but to my knowledge keyframe is just an intra picture that you can decode by itself (if you have the headers to initialize the decoder)
<aaabbb>
JEEB: iirc keyframe is any frame that can be independenly coded *and* which is an idr point, so i frames in an open gop are not considered keyfames by ffmeg even though they are all intra blocks
Sketch has joined #ffmpeg
Sketch has quit [Read error: Connection reset by peer]
<aaabbb>
so like ffv1 is an intra-only codec and all frames are i frames, but you can actually set the gop and only have one keyframe every 15 frames for example (much more doesn't make sense for ffv1 because it doesn't use motion compensation at all), it's just that the frames within a gop are compressed together so damage to one frame will damage the whole gop, and the first i frame is vital. in that case only
<aaabbb>
the first frame of the gop is a keyframe, even though they are all iframes
HarshK23 has quit [Quit: Connection closed for inactivity]
namazso has quit [Remote host closed the connection]
namazso has joined #ffmpeg
namazso has left #ffmpeg [#ffmpeg]
Sketch has joined #ffmpeg
Muimi has joined #ffmpeg
Warcop has quit [Remote host closed the connection]