[Ffmpeg-devel-irc] ffmpeg.log.20190120
burek
burek021 at gmail.com
Mon Jan 21 03:05:01 EET 2019
[00:00:04 CET] <JEEB> the headers I think have that standardized and the header parsing code expects that
[00:00:20 CET] <JEEB> but I think the cookie option wants a simpler thing because it's supposed to be jsut some endline-delimited thing
[00:00:26 CET] <JEEB> and it internally handles the cookie part
[00:00:31 CET] <pink_mist> tdr: the \r is carriage return, the \n is linefeed ... aka newline
[00:00:32 CET] <JEEB> uhh, header creation part I mean
[00:00:32 CET] <tdr> ick. if thats what it uses, then stick with it tho
[00:00:39 CET] <JEEB> tdr: the header thing that is
[00:00:41 CET] <JEEB> which is raw headers
[00:00:44 CET] <JEEB> as far as I can see :P
[00:00:48 CET] <JEEB> there's a separate cookie option
[00:00:52 CET] <tdr> pink_mist, whatever the names, \n is unix-y and \r is windows
[00:00:56 CET] <JEEB> which is just endline delimited as far as I can tell
[00:01:07 CET] <pink_mist> tdr: windows needs both \r and \n
[00:01:18 CET] <JEEB> and it creates the header format from the cookie format
[00:02:30 CET] <nadermx> it's not in windows, but even if I try -cookies "GPS=1; PREF=f1=50000000&hl=en; VISITOR_INFO1_LIVE=YYVg5wZlZM4; YSC=nqFTzbjynHk; s_gl=1d69aac621b2f9c0a25dade722d6e24bcwIAAABVUw==;\n" it still shows me the no CRLF warning
[00:02:42 CET] <JEEB> nadermx: DELIMITED
[00:02:57 CET] <JEEB> the docs say it pretty clearly
[00:03:57 CET] <keglevich> JEEB: I found out if I set burst_bits parameter as well the output bitrate is OK....now the issue is I don't know what kind of value that burst_bits should really have?
[00:04:18 CET] <JEEB> keglevich: no idea, I've never had to tweak those things
[00:04:30 CET] <keglevich> ffmpeg -re -i 1.mp4 -r 25 -c:v libx264 -x264opts nal-hrd=cbr -b:v 8000k -minrate 8000k -maxrate 8000k -bufsize 320k -muxrate 8500k -pcr_period 30 -c:a mp2 -ac 2 -b:a 192k -ar 48000 -f mpegts "udp://239.10.10.10:10000?pkt_size=1316&burst_bits=1000000&bitrate=8500000"
[00:04:44 CET] <keglevich> that's what I have so far...it somehow works...not 100% ok, but at least it works
[00:05:15 CET] <keglevich> I tried to look for more info regarding this parameter, nowhere a single mention of the preferred values
[00:05:44 CET] <JEEB> the docs just say "When using bitrate this specifies the maximum number of bits in packet bursts. "
[00:05:57 CET] <JEEB> so it probably controls how much is sent out max
[00:06:00 CET] <JEEB> at a single send()
[00:06:09 CET] <JEEB> or no idea really
[00:07:41 CET] <keglevich> yes, that's the only thing I found as well...is there a way to find out the preferred values somehow, without jus try&error method?
[00:07:55 CET] <JEEB> reading the logic from the code
[00:08:03 CET] <JEEB> and matching that against your use case
[00:08:14 CET] <keglevich> and that's something I'm really good at ;)
[01:41:21 CET] <raytiley> Are there any parameters required for h264_qsv?
[01:41:48 CET] <raytiley> Trying a simple ffmpeg command and get tons of "incompatible video parameters" warnings
[01:42:10 CET] <raytiley> https://www.irccloud.com/pastebin/t8orRYyB/h264_qsv-log.txt
[01:44:13 CET] <BtbN> Is this Linux?
[01:44:24 CET] <BtbN> You really don't want to be using QSV on Linux, use libva instead
[01:45:00 CET] <raytiley> windows
[09:28:01 CET] <johnjay> question about mp4 files
[09:28:13 CET] <johnjay> is it possible to cut a portion from the start or middle without reencoding it?
[09:28:18 CET] <johnjay> or is that not possible in general
[09:44:55 CET] <fella> johnjay: depends on your INPUT. and even if it does work you might run into problems afterwards. from wrong timestamps to standalone players refusing to play those files at all, so please keep that in mind when you try something like: ffmpeg -ss 00:10:00.00 -i INPUT -t 00:01:00.00 -c copy OUTPUT
[09:46:38 CET] <fella> skip 10min from the beginning (-ss) and cut out 1min (-t)
[12:33:41 CET] <saucecode> Hello, I'm trying to compile ffmpeg 4.1 on ubuntu 16.04. configure is telling me "ERROR: libnpp not found". I compiled ffmpeg 3.4.1 with these configuration options, so I'm not sure what's going on. Am I missing something?
[12:33:54 CET] <saucecode> https://pastebin.com/raw/1WWnNWz5 here's what I'm passing to ./configure
[12:35:02 CET] <JEEB> ffbuild/config.log
[12:35:05 CET] <JEEB> is what you want to look at
[12:35:09 CET] <JEEB> it should contain the actual error :P
[12:40:06 CET] <saucecode> thanks! I think I see what I'm missing now
[14:04:21 CET] <Jeen_> Hey
[14:04:45 CET] <Guest31128> What is the webm equivalent of libx264?
[14:04:54 CET] <Guest31128> libwebm doesn't works.
[14:06:23 CET] <Mavrik> webm is a container
[14:06:35 CET] <Mavrik> And it's supported natively by ffmpeg
[14:06:59 CET] <Mavrik> You probably mean VP8 or VP9 which are usually put inside webm. The encoding library is called libvpx
[14:07:48 CET] <Guest31128> Mavrik: Thank you.
[18:57:10 CET] <shfil> hi, I'm getting a lot of `Invalid return value 0 for stream protocol`, I guess it's caused by this commit https://github.com/FFmpeg/FFmpeg/commit/a606f27f4c610708fa96e35eed7b7537d3d8f712#diff-fa223699156807771d5267b3b52f1f8f
[18:57:10 CET] <shfil> What can cause that read_packet returns 0?
[18:58:53 CET] <shfil> when there's no EOF?
[18:59:54 CET] <shfil> my code: https://github.com/rwengine/openrw/blob/master/rwengine/src/audio/SoundSource.cpp#L241
[19:00:30 CET] <shfil> everything goes in `void SoundSource::loadSfx(LoaderSDT& sdt, size_t index, bool asWave)`
[22:29:07 CET] <g1itch> I'm attempting to use ffmpeg to capture a clip from a live rtsp stream and i get the following output. it looks like it's connecting successfully but getting a bad response code? https://paste.w00t.cloud/hunayadoxe.bash
[00:00:00 CET] --- Mon Jan 21 2019
More information about the Ffmpeg-devel-irc
mailing list