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
realies has quit [Quit: Ping timeout (120 seconds)]
realies has joined #ffmpeg
<noobaroo> xx Without a plus it will be the only fflag, if you use a plus it adds to existing ones and can specify multiple +bitexact+fastseek+genpts+igndts etc
<noobaroo> Not sure if -bitexact is even a thing, if it works then its probably just a shortcut
crossby1004 has joined #ffmpeg
crossby1004 has quit [Client Quit]
crossby1004 has joined #ffmpeg
<ThePendulum> hm my Pi produces a different hash and yet another different filesize, but at least within the same ballpark as my workstation, not 10x the size as what they're expected to be
<furq> xx: -flags is for libavcodec, -fflags is for libavformat
<furq> i think -bitexact just enables both
<furq> the docs imply that but i've never tried it
<furq> ThePendulum: what's the exact command
<furq> if you're getting three different hashes for the same video then something weird is happening
<ThePendulum> direct output to png or jpg seems fine however
<xx> thanks
<xx> noobaroo: -bitexact is documented but the documentation is not explaining it well
<furq> ThePendulum: what happens without -c:v ppm
<ThePendulum> with png instead of ppm still different hashes, let me try without entirely
<furq> without -c:v it just tests the decoder
<furq> granted idk why -c:v ppm would break it because ppm doesn't have any metadata that could be different
<ThePendulum> without it still different hash
<ThePendulum> let me remove the other options too
<ThePendulum> so just ffmpeg -i "$1" -f hash -, still gives three different hashes
<furq> well that's no good
<ThePendulum> consistent between runs on each device locally though
<furq> does it happen with other videos
<ThePendulum> let me try, I'm already using a different video than I was when I first noticed the images that it was spitting out went from 3MB to 24MB between my PC and my NAS
<ThePendulum> ooo interesting those are the same, let me try now with the original video again
<ThePendulum> alright so the original videos also produce the same hash without any of the flags, let me see what happens if I put the flags back
<ThePendulum> video is longer so it takes a while to run
Keshl has quit [Read error: Connection reset by peer]
Keshl has joined #ffmpeg
wyatt8740 has quit [Ping timeout: 252 seconds]
iive has quit [Quit: They came for me...]
makidoll has quit [Remote host closed the connection]
makidoll has joined #ffmpeg
<ThePendulum> still the same hash now... alright, so I suppose the fact it was generating a different hash for the other video file is a different possibly unrelated issue
rv1sr has quit []
crossby1004 has quit [Quit: leaving]
gvg has quit [Ping timeout: 260 seconds]
gvg has joined #ffmpeg
Mirarora has quit [Quit: Mirarora encountered a fatal error and needs to close]
Mirarora has joined #ffmpeg
<ThePendulum> alright, that was indeed a red herring, still curious why that was happening, but when I use the imagemagick appimage on both machines the images output is the same, so evidently the main issue is with imagemagick
gchound has joined #ffmpeg
gchound has quit [Client Quit]
gchound has joined #ffmpeg
deetwelv- is now known as deetwelve
<furq> well at least you didn't find a serious bug in ffmpeg
<ThePendulum> there's still the question why it was outputting different hashes for the test file :P
moxie has quit [Quit: WeeChat 3.8]
moxie has joined #ffmpeg
xx has quit [Ping timeout: 260 seconds]
poorboy_ is now known as poorboy
Dagger has quit [Ping timeout: 260 seconds]
nasso has quit [Read error: Connection reset by peer]
vlm has joined #ffmpeg
Dagger has joined #ffmpeg
System_Error has quit [Ping timeout: 260 seconds]
Dagger has quit [Ping timeout: 265 seconds]
Dagger has joined #ffmpeg
System_Error has joined #ffmpeg
Suchiman has quit [Quit: Connection closed for inactivity]
Dagger has quit [Ping timeout: 265 seconds]
Dagger has joined #ffmpeg
sihloo has quit [Quit: ZNC 1.9.1 - https://znc.in]
talismanick has quit [Read error: Connection reset by peer]
talismanick has joined #ffmpeg
Cheetahze has joined #ffmpeg
StephenLynx has quit [Remote host closed the connection]
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg
gchound has quit [Quit: WeeChat 4.1.1]
lusciouslover has quit [Read error: Connection reset by peer]
lusciouslover has joined #ffmpeg
System_Error has quit [Ping timeout: 260 seconds]
rv1sr has joined #ffmpeg
System_Error has joined #ffmpeg
travisghansen has joined #ffmpeg
coldfeet has joined #ffmpeg
<Forza> Hi. I noticed the loudnorm filter upsample auto to 16bit 192kHz. Is it possible to change the bit depth to 32bit or float?
<furq> loudnorm resamples to 64-bit float
<furq> you can see with -v verbose
MoC has joined #ffmpeg
jemius has joined #ffmpeg
Sketch has quit [Remote host closed the connection]
Sketch has joined #ffmpeg
FH_thecat has quit [Read error: Connection reset by peer]
FH_thecat has joined #ffmpeg
jemius has quit [Quit: Leaving]
xx has joined #ffmpeg
rsx has joined #ffmpeg
nigetilly has joined #ffmpeg
kepstin has quit [Remote host closed the connection]
kepstin has joined #ffmpeg
Marth64[m] has quit [Ping timeout: 252 seconds]
kasper93_ has joined #ffmpeg
kasper93 has quit [Ping timeout: 252 seconds]
kasper93_ is now known as kasper93
kasper93_ has joined #ffmpeg
kasper93 has quit [Ping timeout: 255 seconds]
kasper93 has joined #ffmpeg
kasper93_ has quit [Ping timeout: 272 seconds]
kasper93_ has joined #ffmpeg
kasper93 has quit [Ping timeout: 255 seconds]
gvg has quit [Ping timeout: 260 seconds]
gvg has joined #ffmpeg
kasper93_ has quit [Remote host closed the connection]
faxmodem has quit [Ping timeout: 248 seconds]
faxmodem has joined #ffmpeg
<Forza> Ok thanks. I think rhe issue is i had null output which seems to do pcm 16bit: Stream #0:0 -> #0:0 (dts (dca) -> pcm_s16le (native))
<JEEB> that's the default audio "codec" for null
<JEEB> see `ffmpeg -h muxer=null`
kasper93 has joined #ffmpeg
<welder> When i record video with -f x11grab and audio with -f pulse, then pressing "q" to stop cuts last 2 seconds of audio. On the other hand, when recording with a given time limit (-t 10), then whoel audio is kept
<welder> Is it a known issue?
<JEEB> it should write out everything that's in the buffers of the tool
<JEEB> is this on which version of FFmpeg btw?
<welder> Debian stable, ffmpeg version 5.1.6-0+deb12u1
<welder> I think that happens only with pipewire-pulse though
<welder> Let me check on a laptop where i use pulseaudio globally
<JEEB> try out https://github.com/BtbN/FFmpeg-Builds/releases/tag/latest -master-gpl without shared. download, extract somewhere under homedir and then call the binary with full path
<JEEB> that seems to include pulse and x11
<JEEB> since the whole tool has changed a lot since 2021 or whenver 5.1.x was branched
lavaball has joined #ffmpeg
<Forza> Is it possible to speed up loudnorm some? It seems to be single threaded when I run it.
<welder> same thin on the laptop with pulseaudio and ffmpeg 7.1. I will check these builds
<welder> same thing.
<JEEB> this will spam a lot, but should show all reads and writes in the tool, `-v verbose -debug_ts`
<JEEB> in other words, does the ffmpeg tool even receive audio packets with such timestamps
<JEEB> might be worth doing `2> funky.log`
<JEEB> verbose log level by itself btw is the one I utilize, highest one that doesn't spam (and provides some additional useful info). debug_ts is the one that logs movements of packets and frames within ffmpeg the tool
<welder> will test. Btw before a moment i split the audio and video into separate files with -map 0:v video.mp4 and -map 1:a audio.mp4 and the audio stream is 9.18, video 11.11 in length
<welder> Video is 12.11, audio 9.89 in size. Here's the log: https://0x0.st/s/3tED-X4FkV7euXqWZ_tEdw/XdTA.log
<welder> The command: ./ffmpeg -v verbose -debug_ts -y -f x11grab -i "$DISPLAY" -f pulse -i alsa_output.pci-0000_05_00.6.analog-stereo.monitor -map 0:v /tmp/foo.mp4 -map 1:a /tmp/foo_audio.mp4 2> /tmp/funky.log
<JEEB> cheers
<JEEB> I would first look at how demuxer+tsfixup looks like, since that zeroes the start time in general (starts from zero)
<JEEB> demuxer is basically as ffmpeg the tool reads from the input module
<JEEB> and the +tsfixup just notes that those values are after ffmpeg makes adjustments to the received packet's fields
<JEEB> yea just by looking at the highlighting of lines matching ist_index:1 with demuxer+tsfixup, it seems like the pulse input is only receiving data in bursts
<JEEB> since all inputs are running in their own threads in this new of a version
<JEEB> it seems like either the pulse side or the pulse reader module are having a bit more than a second (1.2 or so?) there of buffer, so that plus the input thread getting killed at a specific time leads to the audio finishing up quicker
<JEEB> but this basically means that the command line tool is fine, just that it has no such input (audio module returns data in bursts)
sihloo has joined #ffmpeg
<JEEB> check `ffmpeg -h demuxer=pulse`
<JEEB> if there is a relevant option
wobbol has joined #ffmpeg
Mirarora has quit [Quit: Mirarora encountered a fatal error and needs to close]
Mirarora has joined #ffmpeg
<welder> there's frame_size and fragment_size
<welder> does not seem to help
ewomer has quit [Quit: WeeChat 4.4.3]
ewomer has joined #ffmpeg
<welder> but I wonder how come nobody bumped onto this. There's one eaxct same issue from 8 years ago
<Forza> Is there a filter like showfreqs that also displays the scale so that I can see what frequenzy is across the x axis?
duckworld has joined #ffmpeg
vampirefrog has quit [Ping timeout: 244 seconds]
nigetilly has quit [Quit: Konversation terminated!]
billchenchina has joined #ffmpeg
lavaball has quit [Remote host closed the connection]
vampirefrog has joined #ffmpeg
coldfeet has quit [Remote host closed the connection]
MisterMinister has quit [Ping timeout: 252 seconds]
Suchiman has joined #ffmpeg
StephenLynx has joined #ffmpeg
Dagger has quit [Ping timeout: 244 seconds]
Dagger has joined #ffmpeg
lavaball has joined #ffmpeg
EmleyMoor has quit [Ping timeout: 252 seconds]
EmleyMoor has joined #ffmpeg
Marth64 has joined #ffmpeg
EmleyMoor has quit [Ping timeout: 260 seconds]
lusciouslover has quit [Ping timeout: 246 seconds]
lusciouslover has joined #ffmpeg
EmleyMoor has joined #ffmpeg
Cheetahze has quit [Quit: Connection closed for inactivity]
phantomics_ has quit [Quit: Leaving]
phantomics has joined #ffmpeg
treefrob has quit [Quit: Leaving]
Sl4yer has quit [Remote host closed the connection]
Everything has joined #ffmpeg
treefrob has joined #ffmpeg
Everything has quit [Quit: leaving]
drew has left #ffmpeg [WeeChat 4.4.2]
Exa has quit [Quit: see ya!]
Exa has joined #ffmpeg
EmleyMoor has quit [Ping timeout: 255 seconds]
coldfeet has joined #ffmpeg
EmleyMoor has joined #ffmpeg
Vonter has quit [Quit: WeeChat 4.4.3]
stolen has joined #ffmpeg
grufwub has quit [Remote host closed the connection]
dlb76 has quit [Changing host]
dlb76 has joined #ffmpeg
grufwub has joined #ffmpeg
Suchiman has quit [Quit: Connection closed for inactivity]
Dotz0cat has quit [Ping timeout: 260 seconds]
zsoltiv_ has joined #ffmpeg
Mister_Magister has quit [Quit: bye]
Mister_Magister has joined #ffmpeg
theprof86 has joined #ffmpeg
<theprof86> I'm copying a MOV file into MP4 and using -filter_complex "[0:v:0]thumbnail,scale=320:-2,trim=start_frame=0:end_frame=1[thumb]" \ to generate and add a thumbnail as disposition in the same command. Everything works fine.
<theprof86> The playback is also fine
<theprof86> The problem is that if I examine it with ffprobe, I see a warning "stream 0, timescale not set" which I think is the thumbnail stream
<theprof86> how would I fix it in the simplest way?
rv1sr has quit [Ping timeout: 252 seconds]
rv1sr has joined #ffmpeg
theprof86 has quit [Quit: Client closed]
Marth64 has quit [Quit: Leaving]
coldfeet has quit [Remote host closed the connection]
Suchiman has joined #ffmpeg
hightower2 has joined #ffmpeg
lavaball has quit [Remote host closed the connection]
l4yer has joined #ffmpeg
coldfeet has joined #ffmpeg
triscal has quit [Ping timeout: 252 seconds]
lusciouslover has quit [Ping timeout: 252 seconds]
lusciouslover has joined #ffmpeg
microlappy has joined #ffmpeg
coldfeet has quit [Remote host closed the connection]
microlappy has quit [Client Quit]
billchenchina has quit [Ping timeout: 265 seconds]
triscal has joined #ffmpeg
rv1sr has quit []
DauntlessOne has quit [Read error: Connection reset by peer]
DauntlessOne has joined #ffmpeg
Everything has joined #ffmpeg
ZLima12 has quit []
ZLima12 has joined #ffmpeg
<znf> Can you "engineer" a HLS playlist so it loops a single segment multiple times?
ZLima12 has quit []
gvg has quit [Read error: Connection reset by peer]
ZLima12 has joined #ffmpeg
<znf> copying the same segment multiple times just results into DTS/PTS issues, it seems, obviously it wasn't meant to be that simple :)
gvg has joined #ffmpeg
ZLima12 has quit []
ZLima12 has joined #ffmpeg
Everything has quit [Quit: leaving]
lavaball has joined #ffmpeg
<noobaroo> There is no #sox channel on Libera nor OFTC, I'm starting to experiment with the program directly instead of using ffmpeg's version. What is the difference between the lowpass effect and the sinc effect ? From wikipedia page about "sinc filter" I can see it's a type of lowpass filter
<noobaroo> The sox lowpass effect seems to do the job either very poorly or not at all, according to audacity. The sinc filter works perfectly.
Galisma has joined #ffmpeg
<BtbN> gotta signal a discontinuity before every repeat
<BtbN> Though I think ffmpeg just ignores that tag anyway
Galisma has quit [Read error: Connection reset by peer]
ThePendulum has left #ffmpeg [#ffmpeg]
Mister_Magister has quit [Ping timeout: 248 seconds]
mw1144 has joined #ffmpeg
mw1144 is now known as darkgreen
Mister_Magister has joined #ffmpeg
darkgreen has quit [Quit: Client closed]
Dotz0cat has joined #ffmpeg
lucasta has joined #ffmpeg
vampirefrog has quit [Quit: Leaving]
vampirefrog has joined #ffmpeg
vampirefrog has quit [Remote host closed the connection]
lavaball has quit [Remote host closed the connection]
Everything has joined #ffmpeg
System_Error has quit [Ping timeout: 260 seconds]
MoC has quit [Quit: Konversation terminated!]
<znf> BtbN, would this look right? https://pastie.dev/K7OJ7S.m3u
<BtbN> should be. But should also make zero difference for ffmpeg
<BtbN> the tag is ignored completely
<BtbN> Other stuff might care about it though
<znf> ffplay chokes on it :(
<BtbN> Could very much see that happen, yeah. ffplay is not intended as a complete player
<znf> VLC stops after 1st segment, too
<BtbN> VLC should handle it
<BtbN> Try with the discontinuity
<znf> nope, it stops at 8s with
<znf> main error: Timestamp conversion failed for 7466667: no reference clock
<znf> main error: Could not convert timestamp 0 for FFmpeg
chandash has joined #ffmpeg
Galisma has joined #ffmpeg
Galisma has quit [Quit: Konversation terminated!]
SuicideShow has quit [Ping timeout: 272 seconds]
SuicideShow has joined #ffmpeg
theprof86 has joined #ffmpeg
Dagger has quit [Ping timeout: 265 seconds]
Dagger has joined #ffmpeg
<theprof86> -filter_complex "[0:v:0]thumbnail,scale=320:-2,trim=start_frame=0:end_frame=1[thumb]" is causing stream 0, timescale not set warning, is there a way to prevent this warning?
chandash has quit [Quit: WeeChat 4.4.2]
<furq> i think you can just ignore that warning