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
OctopusET has quit [Ping timeout: 248 seconds]
OctopusET has joined #ffmpeg
Shine_ has joined #ffmpeg
iive has quit [Quit: They came for me...]
NaviTheFairy has quit [Ping timeout: 244 seconds]
Shine_ has quit [Read error: Connection reset by peer]
NaviTheFairy has joined #ffmpeg
UltraFuzzy has joined #ffmpeg
relue has quit [Ping timeout: 252 seconds]
mven9713 has quit [Ping timeout: 245 seconds]
mven9713 has joined #ffmpeg
YUiNA has joined #ffmpeg
YUiNA_ has quit [Ping timeout: 252 seconds]
relue has joined #ffmpeg
mort has quit [Quit: Ping timeout (120 seconds)]
mort has joined #ffmpeg
jmcantrell has quit [Ping timeout: 260 seconds]
maxim_d33 has quit [Ping timeout: 245 seconds]
maxim_d33 has joined #ffmpeg
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg
ja_02 has quit [Read error: Connection reset by peer]
ja_02 has joined #ffmpeg
Fiji has quit [Ping timeout: 248 seconds]
cantelope has quit [Quit: Connection closed for inactivity]
SakuraChan has quit [Remote host closed the connection]
<Wallboy>
I have some files that don't have their "field_order" set, and show up as "unknown" in ffprobe. I know these files are interlaced, so I can simply remux the file with "-field_order tt", which works fine when the output container is mkv. However, when I try to remux to mp4, the field_order stays as "unknown" afterwards. Is this a limitation of the mp4 container, or should I create a bug report?
lavaball has quit [Remote host closed the connection]
Richardcavell has joined #ffmpeg
<Richardcavell>
Hi everyone. I'm trying to use ffmpeg in my Makefile to automatically convert sounds. Every time I call ffmpeg, it gives me a message: "Guessed Channel Layout: mono
<Richardcavell>
"
<Richardcavell>
I want to suppress this. How do I do it?
<aaabbb>
do you want to suppress all log output?
<Richardcavell>
Not necessarily all
<Richardcavell>
It's just that it appears in bright yellow and there's no need to guess
<Richardcavell>
My command line is ffmpeg -channel_layout mono -i $< -v warning -af "dynaudnorm=p=0.9:s=5,lowpass=f=3750,aresample=ochl=mono:osf=u8:osr=8192:dither_method=triangular" -f u8 -c:a pcm_u8 $@
<aaabbb>
you know i've actually always wondered that too, ever since -channel_layout mono didn't do anything i just gave up and didn't really look it up
Mister_Magister has quit [Excess Flood]
Mister_Magister has joined #ffmpeg
<furq>
-v error will suppress it but also any other warnings
<aaabbb>
yeah that's what i was gonna suggest but that's not specific to channel layout
<aaabbb>
furq: there's no way to inhibit it by making it *not* guess?
<furq>
i actually don't get that warning
<furq>
what's $< pointing at
UltraFuzzy has joined #ffmpeg
realies9 has quit [Ping timeout: 272 seconds]
<tkzv>
Hello, is there a simple way to extract (copy, without recoding) audio track from a video file, without knowing the data format? "ffmpeg -i <infile> -acodec copy <outfile>", won't work when I incorrectly guess outfile extension. Or is there some universal container? Will .WAV work that way?
Blacker47 has joined #ffmpeg
<furq>
mka will work for most things
<tkzv>
furq: Thanks, I'll try it.
<tkzv>
furq: Nope, the program I need (whisper.cpp) hangs from AAC+MKA. And refuses to work with AAC+WAV.
<another|>
If the consumer only eats certains input, encode to that input?
<bpmedley>
another|: When you consumer, are you referencing a player? And, by input are you asking about container or codec or something else?
<bpmedley>
*say consumer
lusciouslover has quit [Quit: \]
<another|>
consumer is whatever is supposed to consume the output, in this case, apparently, whisper.cpp
<another|>
sorry, I realize now it was poorly formulated
<another|>
tkzv: If whisper only consumes certain inputs, why don't you encode to something it likes?
beastd has quit [Ping timeout: 245 seconds]
YUiNA_ has joined #ffmpeg
YUiNA has quit [Ping timeout: 245 seconds]
beastd has joined #ffmpeg
<tkzv>
another|: It consumes several inputs -- PCM, MP3, AAC, Opus and some more. (It should support everything through ffmpeg, but that's a work in progress.) Recoding takes time and affects timestamps. I can parse ffprobe output (or is there a command to simplify that?) and set the correct extension, but I asked if a switch for that already exists.
<another|>
mp4 should support all of that. But... Just decode to PCM?
psyb0t has quit [Remote host closed the connection]
<tkzv>
MP4 support is broken. Ditto MKV and MKA. Decoding to PCM works, but the subtitles may appear a few seconds off, if a file is large. Extracting AAC or Opus works best, but I need to know beforehand if it's AAC, Opus or something else.
rvalue has quit [Read error: Connection reset by peer]
rvalue has joined #ffmpeg
qqe has joined #ffmpeg
coldfeet has joined #ffmpeg
coldfeet has quit [Ping timeout: 265 seconds]
Richardcavell has quit [Ping timeout: 248 seconds]
Shine_ has quit [Read error: Connection reset by peer]
Ozzymandias has joined #ffmpeg
<Ozzymandias>
Hello ffmpeg experts. I have a giant video (9.5 hours, ~6000kbps, 1080p, ~26GB) that I want to convert to something more portable (for one thing, fit on fat32 drive so files <4GB).
<Ozzymandias>
So, the first thing I Tried was breaking it up into chunks, witht he following syntax:
<Ozzymandias>
I was expecting it to make chunks an hour or so each, but instead, they are often 1.5 hours each, so that is kinda weird....
<Ozzymandias>
Some came out about an hour, but some 90 minutes, not sure why. So I guess my first question is, is there a better way to split it up?
<LimeOn>
hello Ozzymandias, you are splitting it into pieces, not shrinking it. i guess you can do it from the first step using "-ss" and "-t"
<Ozzymandias>
Yeah, I know splitting it up is not shrinking it. I just wnated reasonable length files before I reencode it.
<LimeOn>
i guess that since you are using "-c copy" the pieces are bigger(so ffmpeg does not need to cut the frames or something like that)
<technomancy>
a while back I heard that LLM scrapers were getting around ffmpeg's anubis setup; is that still happening?
<Ozzymandias>
I feel like the "-segment_time 00:56:00" isnt doing what I Think it is. I'm looking up "-ss" and "-t" now.
<LimeOn>
related: -ss position (input/output): When used as an input option (before "-i"), seeks in this input file to position. Note that in most formats it is not possible to seek exactly, so ffmpeg will seek to the closest seek point before position. When transcoding and -accurate_seek is enabled (the default), this extra segment between the seek point and position will be decoded and discarded. When doing stream copy or when -
<LimeOn>
noaccurate_seek is used, it will be preserved.
<LimeOn>
--> When doing stream copy or when -noaccurate_seek is used, it will be preserved.
<furq>
Ozzymandias: -f segment -segment_time would be the recommended way
<furq>
i don't know why it would be so far out
<furq>
mkvmerge --split 2G might be better if you want to split by size
<LimeOn>
there is also : -fs limit_size (output)
<furq>
that only works for one segment
<LimeOn>
oh, he would need to run the command several times, i see
<furq>
well also you would have no idea where to start the second segment
<furq>
you can only seek into a file by timestamp
<furq>
i guess you could pull the duration from the first file but that seems very jank
<LimeOn>
yeah, its not a good idea, more trouble to implement
pastelowl has joined #ffmpeg
lavaball has quit [Remote host closed the connection]
pastelowl has quit [Ping timeout: 260 seconds]
pastelowl has joined #ffmpeg
<Ozzymandias>
ah okay I seem to be wrong. the durations are indeed about an hour when I check properties with file manager. However when I play them it says they're longer. I dont understnad what's going on.
<Ozzymandias>
it shows my file manager in the background with all the lengths what I expected.... but it shows me playing the second one and it says its 1:31 or something int he video player
<furq>
what duration do you get from ffprobe
<Ozzymandias>
I think somehow I broke something doing the split so video player confused about timing. I had hoped "-reset_timestamps 1" would take care of that
<Ozzymandias>
This is ffprobe of the .001 video (the second one) which is the same one I was playing in screenshot that video player thinks is 91 minutes. https://paste.debian.net/1385431/ It shows duration of ~56 minutes like file manager does (and like i targetted)
<Ozzymandias>
vlc also thinks its 91 minutes (not just the gnome media player)
<furq>
if you're transcoding it anyway then you can probably just ignore that as long as ffprobe shows the right duration
<furq>
maybe concat the split files and double check everything is ok around the split points before you start
<furq>
if you have the disk space
<Ozzymandias>
Okay. I'll combine split and transcode into one command, hopefully the output doesn't have any glitches. I don't have new/fast computer so this will take hours before I know if it works, so trying to get it right the first time.
<furq>
that would kind of defeat the purpose of splitting
<Ozzymandias>
My purpose for splitting is twofold, one is filesize under 4GB for fat32, but also reasonable legnth so i can find sections i want easier, etc. (chapters above my pay grade)
<Ozzymandias>
I'm going to prepare a command that I hope will split and reencode in one, as I am also short on disk space too so will delete my previous split version first.
<Ozzymandias>
And I'll run it by here before I start it.
<Ozzymandias>
I think that'll convert it to 720p, keep the audio, and break it into ~63 minute chunks.
<Ozzymandias>
Does that seem right? (The threads is so i can keep using the pc while it goes)
<Ozzymandias>
I guess not, got syntax errors trying to run it.
<Ozzymandias>
okay, fixed one problem. should be scale=1280:-2
<Ozzymandias>
and fixed the second problem, should be segment_time 01:03:28, duh
<Ozzymandias>
and now it can't find the moov atom. damnit, where did I put that. never gunna find it, atoms are tiny.
<Ozzymandias>
In all seriousness, I looked up that error and it suggests using "-movflags faststart"; so real question, should I abort and start over with that added? Or is it irrelevant....
Blacker47 has quit [Quit: Life is short. Get a V.90 modem fast!]
iive has joined #ffmpeg
enigma9o7_irc has joined #ffmpeg
Ozzymandias has quit [Ping timeout: 252 seconds]
derjanni_ has joined #ffmpeg
enigma9o7_irc has quit [Client Quit]
derjanni has quit [Ping timeout: 272 seconds]
Fiji has quit [Ping timeout: 252 seconds]
bsFFFFFF has quit [Quit: bsFFFFFF]
bsFFFFFF has joined #ffmpeg
bsFFFFFF has quit [Ping timeout: 276 seconds]
bsFFFFFF has joined #ffmpeg
E81l7HT8T7sF9JdA has joined #ffmpeg
Nw6mLRq1TuIuHsNd has joined #ffmpeg
E81l7HT8T7sF9JdA has quit [Ping timeout: 252 seconds]
Nw6mLRq1TuIuHsNd has quit [Ping timeout: 245 seconds]
E81l7HT8T7sF9JdA has joined #ffmpeg
Nw6mLRq1TuIuHsNd has joined #ffmpeg
E81l7HT8T7sF9JdA has quit [Ping timeout: 252 seconds]
Nw6mLRq1TuIuHsNd has quit [Ping timeout: 245 seconds]