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
EmleyMoor has joined #ffmpeg
deus0ww has quit [Ping timeout: 246 seconds]
deus0ww has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
five618480338 has joined #ffmpeg
five61848033 has quit [Ping timeout: 256 seconds]
five618480338 is now known as five61848033
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
lucasta has joined #ffmpeg
kuba-orlik0397 has quit [Read error: Connection reset by peer]
kuba-orlik0397 has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
lemourin has quit [Quit: The Lounge - https://thelounge.chat]
lemourin has joined #ffmpeg
feiwan1 has joined #ffmpeg
Corroddity has quit [Quit: WeeChat 4.2.2]
Corroddity has joined #ffmpeg
Corroddity has quit [Max SendQ exceeded]
housemate has joined #ffmpeg
feiwan1 has quit [Read error: Connection reset by peer]
Corroddity has joined #ffmpeg
Corroddity has quit [Max SendQ exceeded]
Corroddity has joined #ffmpeg
Corroddity has quit [Max SendQ exceeded]
Corroddity has joined #ffmpeg
Corroddity has quit [Max SendQ exceeded]
Corroddity has joined #ffmpeg
Corroddity has quit [Max SendQ exceeded]
Corroddity has joined #ffmpeg
Corroddity has quit [Max SendQ exceeded]
Corroddity has joined #ffmpeg
Corroddity has quit [Max SendQ exceeded]
Corroddity has joined #ffmpeg
Corroddity has quit [Max SendQ exceeded]
Corroddity has joined #ffmpeg
Corroddity has quit [Max SendQ exceeded]
Corroddity has joined #ffmpeg
Corroddity has quit [Max SendQ exceeded]
housemate has quit [Remote host closed the connection]
housemate has joined #ffmpeg
housemate has quit [Remote host closed the connection]
Corroddity has joined #ffmpeg
Corroddity has quit [Max SendQ exceeded]
Suchiman has quit [Quit: Connection closed for inactivity]
Corroddity has joined #ffmpeg
lucasta has quit [Remote host closed the connection]
beaver has quit [Remote host closed the connection]
fling has quit [Remote host closed the connection]
fling has joined #ffmpeg
beaver has joined #ffmpeg
Hackerpcs has quit [Quit: Hackerpcs]
Hackerpcs has joined #ffmpeg
Hackerpcs has quit [Max SendQ exceeded]
AbleBacon has quit [Read error: Connection reset by peer]
HarshK23 has joined #ffmpeg
FH_thecat has quit [Quit: Leaving]
FH_thecat has joined #ffmpeg
xx has quit [Remote host closed the connection]
hussein1 has quit [Read error: Connection reset by peer]
xx has joined #ffmpeg
hussein1 has joined #ffmpeg
wallacer has quit [Quit: wallace dozed off]
lavaball has joined #ffmpeg
iliv has joined #ffmpeg
function1 has quit [Ping timeout: 264 seconds]
vlm has joined #ffmpeg
hussein1 has quit [Ping timeout: 260 seconds]
iliv has quit [Quit: "<paniq> you know when i walk out the door, there is plenty of stupid people. i open irc, there is plenty of intelligent people. so the choice comes easy."]
lavaball has quit [Remote host closed the connection]
nd has quit [Quit: bye]
nd has joined #ffmpeg
vlm has quit [Quit: Leaving]
hussein1 has joined #ffmpeg
Suchiman has joined #ffmpeg
function1 has joined #ffmpeg
rv1sr has joined #ffmpeg
farter has joined #ffmpeg
<farter> here's my next problem: what's the correct way of handling video input with color space yuvjxxxp (in other word, yuv with full-range values) in ffmpeg? searching the web doesn't find quite valid info, among many who asked about the warning "about yuvjxxxp is deprecated"... just "ffplay"ing it, i've experimented various combinations of the range, pixel format converting filter.
<farter> but get no correct result by just inspecting the black level. or even worse ones like magenta/green stripes. however, video players on the web, or the video player that comes with ubuntu 18.04 works well. fyi, the avi is generated by opencv VideoWriter MJPG encoder.
<farter> https://ibb.co/6J8MQh0 here's what the ffplay results look like
<JEEB> I think everything should be flagging the AVFrame's color range in addition to the hopefully-soon-gone separate J pixel formats
<JEEB> so you can base off of that, and treat otherwise the J and non-J variants the same
<JEEB> it also includes the reference to the BT.2100 spec that is the latest spec which defines the limited and full range representation
<JEEB> so the J pixel format just used to be the way to flag the color range before AVFrames had color range etc
<CounterPillow> Ubuntu 18.04 lmfao
<farter> i've done some programmatic call to api that encodes full range yuv to h264, and save mp4 or and sent it through rtp then watched from the web browser, they seem to work fine.
<JEEB> just read the color range definition and that should tell you what full range and limited range are
<JEEB> CounterPillow: for singular users ubuntu made long-term LTS stuff free for N machines so you get people utilizing old stuff
<farter> but when it turns to opening video, doing some experiments... both ffplay, or just convert the first frame to png, etc... didn't find a way to get it right
<farter> btw using the ffmpeg 7.0.1 compiled
<JEEB> well to my knowledge common full range cases should work, since the handling of full range content hasn't changed in a good while. also I recommend testing out BtbN's automated master builds that you can just extract and call with the full path to the binary
vincejv_ has joined #ffmpeg
<JEEB> and if you think you are getting something wrong, post the full command with `-v verbose` added (last log level that is not spammy) and the full terminal output to a pastebin service of chocie
<JEEB> then link here
vincejv has quit [Ping timeout: 255 seconds]
vincejv_ is now known as vincejv
HerbY_NL has joined #ffmpeg
<farter> yes i'm also believing that, according to docs and probe output etc. it all looks like well supported and should just work (despite the warning that yuvj is deprecated), but..
<farter> https://ibb.co/LJr1WJD converting to png
<farter> to avoid what might be wrong with ffplay
<farter> ffmpeg -i Breast_20231219102226436_buf.avi -c:v png -t 0.01 _out%d.png
vincejv has quit [Ping timeout: 256 seconds]
<farter> that commandline i don't think is very wrong..?
<JEEB> please re-read my note
<Curid> `ffmpeg -i input.avi -frames:v 1 out.png`
<JEEB> I did specifically tell him to bump verbosity to verbose, and then post both the command and the output
<JEEB> he specifically *hid* his log (and it's an image)
<JEEB> that is why I told him to re-read my sentence
<JEEB> also I do recommend testing with these automated ones https://github.com/BtbN/FFmpeg-Builds/releases/tag/latest
<JEEB> if that is what I think it is, thank you
<farter> cus too many repetition of lines of warning about that deprecation, so i hide them to show my original command
<JEEB> > disable-x86asm
<JEEB> that is why you're getting all the spam of warnings
<JEEB> (if you are building for x86 or x86_64, just install nasm and you're fine - you're losing a bunch of optimizations)
<farter> uhm i got this from some web tutorial about compiling ffmpeg, yes i've been in doubt what it does
<JEEB> yea, so check the Btbn's automated builds. it's all in the public and no secrets :P
<JEEB> grab a tarball, and then extract to homedir
<JEEB> then just call the binary with the full path
<JEEB> farter: if you are actually on x86(_64), you just made your FFmpeg a whole metric load slower
<CounterPillow> I can't believe a web tutorial lied to me
<JEEB> since most of the heavily utilized decoders for example all have hand-written SIMD optimizations
<farter> i think i've tried that before, though not on invesgating this specific problem. cus i prefer to use nvidia hwaccel, seems not supported by precompiled binaries...
<JEEB> Btbn's has ffnv headers included since he maintains those, too :P
<JEEB> anyways, the log in general shows that things are OK
<JEEB> look at the auto_scale_0 lines
<farter> i'll try deleting that config item, but is it related to color range handling?
<JEEB> it may be if most people are utilizing some code path, and others not :P but no, it was a general remark
<JEEB> also just for this issue, please test with BtbN's build
<JEEB> but as noted, the auto_scale_0 lines that log input and output configuration seem fine
<farter> k will do, thx
<farter> yes i also see logs that looks fine
<JEEB> range:pc YCbCr (yuvj) to range:pc RGB
<JEEB> so that would then mean that either there is a bug (which would also show up if you fed the program standard full range content instead of this specific sample), or (more likely) the content does not actually follow standard full range
<JEEB> there are a few full range samples in FATE suite
<JEEB> or heck, all normal JPEG stuff would be full range :P
function1 has quit [Ping timeout: 264 seconds]
<JEEB> so if you ffprobe a random JPEG file you could check if it is YCbCr and full range
<farter> but the output png just slaps me.. cusz they're medical images, ultrasound machine UI by convention has some parts that's definitely black, and we want to encode the image losslessly (to 8-bit, though i'a aware quantization of untrasonic signals is lossy).
<JEEB> then you should be able to test the same command with a known full range content file
<JEEB> if that one looks good, then your input does not follow full range value ranges
<JEEB> also if you want to look at the JPEG included in the AVI, `ffmpeg -v verbose -i input.avi -map 0:v -c copy -frames:v 1 out.jpg`
<JEEB> that should enable you to look at the frame by something else :P
<farter> https://ibb.co/grV8T3x tried -c copy output jpg, still had black non-black *facepalm*
<farter> i'm pretty sure in c++ the recording routine the RGB cv:Mat are fed with zeroes on those pixels.... might be opencv videowriter problem??
<JEEB> and codec copy copies the JPEG from the inside of that AVI, so you can see it with any other application handling JPEGs
<JEEB> so if you are seeing that the signal in that file is weird, then it's not normal full range content
<farter> *thonk*
<JEEB> and yea, in that case it's probably a good idea to check what is generating that data :P
<JEEB> and how it's doing it
darkapex has quit [Ping timeout: 268 seconds]
darkapex has joined #ffmpeg
<farter> i see that in jpg realm mostly uses full range, no wonder why it's named after j.. so it might be, these video are composed with jpeg inside marked as full-range, but actually encoded by opencv mjpg as limited-range, and those players all ignored the full-range directive, showing them as limited range, so wrong-on-wrong turns right?
<farter> but then... why not even one in those filter:v combinations get the right output? shoudn't "scale=in_range=mpeg,out_range=full" work?
peterLustig67 has joined #ffmpeg
<farter> ../ffmpeg-btbn/bin/ffmpeg: /lib/x86_64-linux-gnu/libpthread.so.0: version `GLIBC_2.28' not found (required by ../ffmpeg-btbn/bin/ffmpeg)
<farter> haha, libc... kinda breaks as expected..
<farter> i dont remember exactly but seems i also had this sort of issue with nasm..
<JEEB> nah, that's unrelated
<JEEB> that just means that BtbN's builds now require a newer base OS, as they're built against a newer glibc
<JEEB> I guess because general support for 18.04 ended a while ago
<JEEB> (and something needed something from newer glibc)
<farter> yes i just mean i remember encountering something in build chain requiring newer libc, then it fails. not sure if it's nasm...
<farter> fails to build something from source
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
hightower3 has quit [Remote host closed the connection]
function1 has joined #ffmpeg
Livio has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
Blacker47 has joined #ffmpeg
{{{{{}}}}} has quit [Quit: Client closed]
<farter> https://ibb.co/h7V0qwW ummmmm tried to roundtrip, seems my guessing makes some sense...?
{{{{{}}}}} has joined #ffmpeg
peterLustig67 has quit [Remote host closed the connection]
peterLustig67 has joined #ffmpeg
<farter> however without disable-x86-asm, magenta and green stripes never appear, just none of these conversions make any difference..
<BtbN> C11 needs glibc 2.28
<farter> confirmed that the video player in ubuntu 18.04 doesn't respect full-range color video... and also opencv mjpg outputs fake full-range frames...
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
Livio has quit [Ping timeout: 246 seconds]
farter has quit [Remote host closed the connection]
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
Tinos has quit [Remote host closed the connection]
Tinos has joined #ffmpeg
rv1sr has quit [Ping timeout: 256 seconds]
rv1sr has joined #ffmpeg
peterLustig67 has quit [Ping timeout: 250 seconds]
javabean_ is now known as Javabean
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
HerbY_NL has quit [Ping timeout: 246 seconds]
waleee has joined #ffmpeg
rsx has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
luva85 has joined #ffmpeg
kuba-orlik0397 has joined #ffmpeg
luva8 has quit [Ping timeout: 240 seconds]
luva85 is now known as luva8
theobjectivedad has quit [Ping timeout: 240 seconds]
theobjectivedad has joined #ffmpeg
theobjectivedad has quit [Read error: Connection reset by peer]
theobjectivedad has joined #ffmpeg
theobjectivedad has quit [Remote host closed the connection]
{{{{{}}}}} has quit [Quit: Client closed]
{{{{{}}}}} has joined #ffmpeg
{{{{{}}}}} has quit [Client Quit]
theobjectivedad has joined #ffmpeg
{{{{{}}}}} has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
waleee has quit [Ping timeout: 268 seconds]
vlm has joined #ffmpeg
HarshK23 has quit [Quit: Connection closed for inactivity]
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
utsweetyfish has quit [Ping timeout: 252 seconds]
utsweetyfish has joined #ffmpeg
lavaball has joined #ffmpeg
BUSY has quit [Ping timeout: 255 seconds]
function1 has quit [Ping timeout: 256 seconds]
minimal has joined #ffmpeg
function1 has joined #ffmpeg
BUSY has joined #ffmpeg
mven976 has joined #ffmpeg
mven97 has quit [Ping timeout: 268 seconds]
mven976 is now known as mven97
zmt01 has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
zmt00 has quit [Ping timeout: 260 seconds]
lucasta has joined #ffmpeg
coldfeet has joined #ffmpeg
rsx has quit [Quit: rsx]
kuba-orlik0397 has quit [Remote host closed the connection]
waleee has joined #ffmpeg
kuba-orlik0397 has joined #ffmpeg
vincejv has joined #ffmpeg
{{{{{}}}}} has quit [Quit: Client closed]
{{{{{}}}}} has joined #ffmpeg
lavaball has quit [Quit: lavaball]
waleee has quit [Ping timeout: 264 seconds]
choix has joined #ffmpeg
<choix> How do I list all the available audio formats for a given container?
<choix> Maybe codecs instead of formats is a better word.
ttys000 has quit [Quit: Textual IRC Client: www.textualapp.com]
lucasta has quit [Quit: Leaving]
lucasta has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
ttys000 has joined #ffmpeg
lucasta has quit [Remote host closed the connection]
beaver has quit [Remote host closed the connection]
beaver has joined #ffmpeg
rv1sr has quit []
iliv has joined #ffmpeg
j45 has quit [Ping timeout: 268 seconds]
j45 has joined #ffmpeg
<kepstin> choix: there's no real list anywhere; the way it works in ffmpeg is that the output format either accepts or gives an error when you try to put a certain codec into it :/
<choix> kepstin: Ok, thanks.
kuba-orlik0397 has quit [Remote host closed the connection]
tranzistor has quit [Ping timeout: 272 seconds]
kuba-orlik0397 has joined #ffmpeg
choix has quit [Quit: Leaving.]
HarshK23 has joined #ffmpeg
lucasta has joined #ffmpeg
tranzistor has joined #ffmpeg
minimal has quit [Quit: Leaving]
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
GenTooMan has quit [Quit: Leaving]
GenTooMan has joined #ffmpeg
HerbY_NL has joined #ffmpeg
Livio has joined #ffmpeg
dkh has quit [Quit: Ping timeout (120 seconds)]
dkh has joined #ffmpeg
Javabean has quit [Quit: well, shoot]
Javabean has joined #ffmpeg
HerbY_NL has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
rv1sr has joined #ffmpeg
Tinos has quit [Remote host closed the connection]
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
function1 has quit [Ping timeout: 240 seconds]
function1 has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
devinheitmueller has joined #ffmpeg
coldfeet has quit [Remote host closed the connection]
rv1sr has quit []
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
lucasta has quit [Remote host closed the connection]
HerbY_NL has joined #ffmpeg
odrling has quit [Remote host closed the connection]
odrling has joined #ffmpeg
odrling has quit [Remote host closed the connection]
odrling has joined #ffmpeg
HerbY_NL has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Livio has quit [Ping timeout: 246 seconds]
j45 has quit [Ping timeout: 240 seconds]
j45 has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
Livio has joined #ffmpeg
jarthur has joined #ffmpeg
lavaball has joined #ffmpeg
nurupo has quit [Quit: nurupo.ga]
nurupo has joined #ffmpeg
iliv has quit [Ping timeout: 240 seconds]
Livio has quit [Ping timeout: 255 seconds]
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
noumenon has joined #ffmpeg
<noumenon> <noumenon> I'm trying to use ffmpeg to cut a video, but experiencing some issues; first I tried the -t option with a duration, but in that case the beginning of the clip missed by some seconds, including time at the start before the indicated starting time, then I tried putting the input first and using -to, which yielded the right duration, but now there's a slight freeze in the video at the very beginning for a second or two
<noumenon> (but
<noumenon> <noumenon> not in the audio, which plays along just fine while the video is frozen); is there any way to do it right?
<noumenon> sorry for that getting split into three, should've anticipated that
<noumenon> I think I read a long time ago that it has something to do with certain keyframes or something
<CounterPillow> Don't do stream copy
<noumenon> I was about to try doing it with re-encoding, and someone suggested that in a different channel too, but it was so slow that I'd hoped there was some other way
<noumenon> but maybe there isn't
Muimi has joined #ffmpeg
<{{{{{}}}}}> Hi #ffmpeg o/
<{{{{{}}}}}> I finally got this working: ffmpeg -i /dev/video0 -c:a copy -c:v libx264 -vpre ultrafast  -f flv -listen 1   rtmp://127.0.0.1/live.mp4
<{{{{{}}}}}> but I needed help to add an additional output to a file
<{{{{{}}}}}> an exact duplicate of what is being streamed to rtmp
<{{{{{}}}}}> except i want to segment the stream  for the file output where the output file name has a timestamp and is segmented every hour
<BtbN> live.mp4 + rtmp makes very little sense
<{{{{{}}}}}> how can I do that, it feels like i'd have to spend many hours reading docs and testing it out
<{{{{{}}}}}> BtbN , would flv be better?
coldfeet has joined #ffmpeg
<BtbN> It's rtmp
<BtbN> there are no files
<BtbN> You can add a tee muxer if you also want a local file
Muimi has quit [Ping timeout: 268 seconds]
function1 has quit [Ping timeout: 268 seconds]
<BtbN> But you really should be using some streaming server, not ffmpeg listen mode. That's merely a simple proof of concept, that only supports one singular client
<{{{{{}}}}}> yeah, i basically want to tee it to a series of files
<{{{{{}}}}}> BtbN, any suggestions for linux?
function1 has joined #ffmpeg
<{{{{{}}}}}> vlc couldn't do the job
<BtbN> There is pretty much only nginx-rtmp
<BtbN> and some JavaScript written something, but I doubt you'd want to touch that
<{{{{{}}}}}> yeah, for now though i'd like to stick with rtmp, having to setup nginx on this box is a PITA
<BtbN> ?
<{{{{{}}}}}> i want to do nginx+jitsi embedded front end or something
<{{{{{}}}}}> just not right now
<BtbN> ffmpeg is not a suitable streaming server
<{{{{{}}}}}> but how would it work with nginx-rtmp, wouldn't it need ffmpeg to transcode the input anyways?
<CounterPillow> mediamtx
<BtbN> nginx-rtmp is pretty much universally what people use, and it can also produce HLS for you
<BtbN> Why would it need to transcode at all?
<BtbN> it can do simple remuxing itself
<{{{{{}}}}}> BtbN, it's a capture device, i'd think for compression?
<BtbN> what?
<BtbN> nginx-rtmp is not a capture device
<{{{{{}}}}}> the input is a capture device, something needs to transcodeit
<BtbN> yes, ffmpeg.
<{{{{{}}}}}> alright, but right now i don't have time for setting that up. any chance you can show me how to do the tee + segment output since i need to do that anyways even with nginx-rtmp
<BtbN> I don't have any examples for that, no. But the tee muxer examples should be simple enough to derive from
<BtbN> And I don't know what you plan to set up, but ffmpeg in rtmp listen mode as a server is not a suitable solution for most of anything
<BtbN> It will block until something connects, and will exit after it disconnects, and only handle exactly that one client
<{{{{{}}}}}> yes, i got that part. i was just reading up on mediamtx, i think i'll try that out instead of listen mode
<{{{{{}}}}}> i wish there was an example of tee muxer + segmented output, i'll keep trying later
minimal has joined #ffmpeg
<dostoyevsky2> why can't the -t and -ss options not always work on exact frames and instead just seem to refer to points where the codecs will receive and stop getting data from the stream, and thus then the video does not play the first frames, and may play longer than the -t specified
<dostoyevsky2> +?
<bpmedley> {{{{{}}}}}: You can change the .yml file for mediatmx and in like 5 - 10 minutes have what you want.
<bpmedley> Change record: no to yes. And set your duration. Might need to mkdir recordings, as well.
<noumenon> same question as dostoyevsky2 :D
<bpmedley> It's single binary, for me, on macOS.
<noumenon> still looking for a simple way to quickly cut a video at the exact indicated times, re-encoding or no
<noumenon> I did try looking a bit into filter_complex, but I don't quite understand it
<dostoyevsky2> noumenon: I asked it here before, but nobody seemed to be able to answer... :)
waleee has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
<CounterPillow> Guys why does video compression reuse previous frames to compress next frames therefore creating a data dependency? Is everyone just unable to answer?
<galad> it must be conspiracy
<CounterPillow> Thanks I totally didn't know the answer!
HerbY_NL has joined #ffmpeg
Livio has joined #ffmpeg
<Curid> Is player support for edit lists really that bad?
<psykose> CounterPillow: i think it has something to do with deez
Sketch has quit [Remote host closed the connection]
<kepstin> CounterPillow: to summarize really quickly: video frames that are close in time are very similar, so the biggest improvement people have figured out on how to compress video smaller is to re-use similar parts from previous frames instead of encoding them again separately.
dlb76 has quit [Ping timeout: 252 seconds]
<dostoyevsky2> kepstin: are there still video formats where the data is compressed frame by frame, or maybe even not compressed at all?
<CounterPillow> Yes thank you I know, I was being facetious in the face of the two people in here knowing what a stream copy does but not making the obvious connection as to why that won't allow precise cuts in every case
<CounterPillow> Yes there are intra-only formats and profiles
Sketch has joined #ffmpeg
<dostoyevsky2> CounterPillow: well, -ss and -t could simply start/stop decoding earlier/longer, and then you could still just acccess each frame, provided the video would be reencoded
<Curid> dostoyevsky2: what command did you use?
HerbY_NL has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<dostoyevsky2> noumenon: Do you have a ffmpeg commandline to show your problem? (It's noumenon's question actually)
<kepstin> anyways, _as long as you are re-encoding_, ffmpeg will use frame-exact start/stop points based on the -ss and -t options; you don't have to do anything special.
<dostoyevsky2> kepstin: Ah, good to know!
<kepstin> (there's options that can change that behaviour, but "accurate seek" mode is the default)
vincejv has quit [Remote host closed the connection]
<noumenon> dostoyevsky2: sure, the second and last thing I tried (where the video froze for the first 1-2 seconds) was: ffmpeg -i in.mkv -ss 00:46:50.00 -to 00:47:44.00 -c copy out.mkv
<noumenon> well, the third thing I tried was with reencoding, but that's when I realized it would take extremely long
<Curid> you lose quality too
Sketch has quit [Remote host closed the connection]
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
<noumenon> yeah, but what I'm wondering is if there's a way to get it precise without doing that
<noumenon> first I tried -ss and -t, in which case the output started before the indicated time
<Curid> which video players does it need to work in?
<noumenon> then -i first with -ss and -to, but then the first 1-2 seconds of the video froze
<noumenon> uh, doesn't the file format determine that?
<Curid> you assume video players actually follow the spec
<noumenon> well, ultimately the goal is uploading it to X
Sketch has joined #ffmpeg
<dostoyevsky2> noumenon: if you are not reencoding then you have to cut according to how the coded chunked your video... so you might be able to not have the freeze by repositioning your cut according to the chunks
<Curid> xitter will probably transcode it again
<noumenon> dostoyevsky2: well, yeah, I assume that's what happened when I did -ss and -t first, when it started before the indicated starting time, but that's of course not desirable either
<noumenon> or maybe you meant something else
<noumenon> I just want the cut times to be exactly what I specify xD
<Curid> it may go faster if you enable hardware acceleration
<noumenon> yeah, that's what someone in the other channel suggested
<noumenon> I briefly looked into that, maybe I'll end up doing that with reencoding if there's no other way
<noumenon> that's what I use to play YT videos with my shitty laptop as well
aphysically has quit [Ping timeout: 264 seconds]
<Curid> what if ffmpeg squished first few frames together to avoid reencoding?
AbleBacon has joined #ffmpeg
<Curid> i'm going to try that at some point
<noumenon> Curid: you mean like, keep track of where the indicated start time actually was vs. where it ended up being, and then calculate the frames between the two, and squish them?
<Curid> take all samples in the start time gop and set the duration of the ones before the start time to 0
HerbY_NL has joined #ffmpeg
Blacker47 has quit [Quit: Life is short. Get a V.90 modem fast!]
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
<noumenon> Curid: is there any way to do that programmatically?
HerbY_NL has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
iliv has joined #ffmpeg
Vonter has quit [Ping timeout: 252 seconds]
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
Vonter has joined #ffmpeg
vincejv has joined #ffmpeg
yans has joined #ffmpeg
mrelcee has quit [Ping timeout: 264 seconds]
coldfeet has quit [Remote host closed the connection]
mrelcee has joined #ffmpeg
function1 has quit [Ping timeout: 268 seconds]
function1 has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
Vonter has quit [Ping timeout: 264 seconds]
Vonter has joined #ffmpeg
rv1sr has joined #ffmpeg
iliv has quit [Quit: "<paniq> you know when i walk out the door, there is plenty of stupid people. i open irc, there is plenty of intelligent people. so the choice comes easy."]
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
noumenon has left #ffmpeg [Leaving]
Livio has quit [Ping timeout: 256 seconds]
vlm has quit [Quit: Leaving]
yans has quit [Remote host closed the connection]
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
yans has joined #ffmpeg
SuicideShow has quit [Ping timeout: 255 seconds]
SuicideShow has joined #ffmpeg
beaver has quit [Remote host closed the connection]
beaver has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
rv1sr has quit []
Haxxa has quit [Quit: Haxxa flies away.]
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
mrelcee has quit [Ping timeout: 264 seconds]
xx has quit [Ping timeout: 260 seconds]
mrelcee has joined #ffmpeg
kuba-orlik0397 has quit [Remote host closed the connection]
kuba-orlik0397 has joined #ffmpeg
Vonter has quit [Ping timeout: 268 seconds]