Ah. I was going to point out that the issue was that this was in no way an AVI.
Looks as if you figured that out by the time I woke up. I will give the update
a try. With the header working correct I suspect things will fall into place.
@dhannyz
I figured out an ffmpeg visualization output la
I dug through my test scripts from last year and found this output that semi
worked. The issue here is that it's Video Only, no audio tracks. Even though I
did include the audio on the encoding side. From my testing it doesn't work
properly currently, I can't test more because I am having issu
I haven't tested this since v1.3.3 but this hasn't worked for quite some time
(I speculate it ever worked). From what I can remember, there's no way for
ffmpeg to know what the input is and synchronize with it (specifically the
synchronization). Use GStreamer instead. Vastly more complex but
Install it using opam and not the tar source package.
https://www.liquidsoap.info/doc-1.3.6/install.html
I recommend adding to the basic install fdkaac, flac and opus. So it'd be
something like;
opam depext taglib mad lame vorbis cry samplerate fdkaac flac opus liquidsoap
opam install taglib m
Do you have an example configuration? What version of the fdk-aac library are
you using? The latest version commonly used in Linux is v2.0. I am currently
using v1.6 and not having any obvious cut off issues using ffplay or VLC
players. I did have problems with older versions.
https://github.c
I believe you've over complicated it. But I understand. A lot of this program
doesn't make any sense or changes based on context. I believe based on your
sample code that there's no place in the source path that you've placed the
jingle to be added into the stream. All you need to do is create
I want to chime in here and say again. There is no such thing as a 44.1khz opus
file. The encoder does not support that sample rate. The DECODER will resample
to the best available samplerate that the hardware supports. The information
stored in the file is the SOURCE sample rate.
libsamplerate
This is a bit beyond the scope of what I can provide. It appears that you maybe
using the Docker version of AzuraCast. I'd recommend installing Liquidsoap via
opam independently on the VPS and test it with a simple script to see if you
get the same results. If you are using the Docker install v
>From searching google you can issue to see the package versions;
`$ apt-cache policy `
It appears for v18.04.2 that libopus is v1.1.2. and Liquidsoap package is
v1.1.1.
These are quite old versions. I highly recommend to install the latest
Liquidsoap via opam to start with.
--
You are receiv
Opus is way different from the encoders of the past. Opus outputs to 48khz
samplerate only. The max_bandwidth setting sets 'effective' sample rates for
the decoder. The samplerate configuration in LS is for the input audio source
and not the output. I'd recommend keeping the default (I think it
I've been continuously testing ALSA output and, it's been solid ever since I
restarted the process.
The only difference I can think of between when it initially stopped working
and now is; I believe Sunday evening, weekly cron jobs are activated.
Specifically pretty cpu/ram intensive ones such
Quick testing and got about 15 hours in before it stopped working for me. LS
completely stopped output inthe log files. I forgot to test the telnet harbor
before killing the process. Using Fedora, ALSA v1.1.8,, AMD Athlon X2 cpu and
ALC888 sound on motherboard.
--
You are receiving this beca
To further expand on the ladspa plugin feature. This isn't by default. In opam
you need to install the ladspa plugin feature for LS. You'll want to install
the system ladspa and ladspa-devel packages, then opam install ladspa. Opam
wlll install the ladspa plugin and prompt to re-compile LS. F
Don't forget ladspa plugin abilities. There appears to be several ladspa gate
plugins.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/savonet/liquidsoap/issues/751#issuecomment-481971020__
It's been over 24hrs and working great with the rpm version of fdk-aac. I
couldn't figure out how to tell opam to use the latest version in
/usr/local/lib other than removing the rpm one. Will test it out later on after
a week or so of using v1.6 I also had the old aacplus and vo-aac libs. Rem
Fedora's rpm of fdk-aac is v0.1.6 in the rpm-fusion-nonfree.
I just did an ldd on the liquidsoap binary. It appears to be using an
fdk-acc.so in /usr/local/lib/ with a 2016 date. I also found remnants of
v1.2.0 and v1.3 when I compiled it outside opam. I must have downloaded and
compiled the
I am not aware of LS to stream the playlist content as-is. As far as I
understand it, that is not how it works. It decodes the playlist files (which
can be various formats) into a raw format, then based on the output
configuration sends that raw data to encoders (internal or external) or
physic
I suspect your ezstream setup isn't re-encoding it's output. Liquidsoap is
decoding, then re-encoding based on your output settings. It can also encode to
multiple formats simultaneously quite easily. This is where a substantial
amount of CPU time is used. Memory usage is due to the playlists
**Describe the bug**
Setting the AAC afterburner to true will cause LS to drop the Icecast
connection after about 24-30 hours, it will stop logging any log file output.
The telnet harbor still functions and LS appears to respond to all the commands
as if it was working properly.
**To Reproduc
I've been using this code I got somewhere off the site or maybe some other,
can't remember. I don't think this is would be a proper place for the
simple_crossfade function.
`# Switch to Live Source when connected
# Transistion to Live source with Jingle
# Define a transition that fades out the
The PPM image file needs to be in RGB24.
$ffprobe -i logo.ppm
Stream #0:0: Video: ppm, rgb24, 2407x2022, 25 tbr, 25 tbn, 25 tbc
The article you refer to is from 2013. Consider that to be outdated and a 'it
maybe kinda supports those files but probably not'.
--
You are receiving this because
And this code works for me;
`lines = get_process_lines("youtube-dl --get-title --get-filename --
'https://www.youtube.com/watch?v=rmCA3qQkqso'")
log("lines = #{lines}")
line = list.hd(default="",lines)
log("line = #{line}")
`
```
2019/03/13 18:23:13 [lang:3] lines = ["Michael Jackson - Billie J
Do you get the same behavior using https:// trying to fetch a file from a web
server? I get the same error as well. I suspect this has something to do with
calling external processes.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on
You are correct. It is broken. I am not sure what Toots is rambling on about.
It is also a failable source which requires fallback.
Even with that, I don't know how this would work. The protocol.liq script
executes two passes of youtube-dl.
The first one is the youtube-dl --get-title --get-file
Convert the image file to a PPM image file and see if that works. Had same
issue and the ppm file worked.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/savonet/liquidsoap/issues/731#issuecomment-472255
What does the command do when you enter it from the shell from the server?
`$ youtube-dl --get-title --get-filename --
"https://www.youtube.com/watch?v=rmCA3qQkqso"`
This command just retrieves the title and filename. LS hasn't gotten to the
routine to download the actual file yet.
--
You ar
Install via Opam. I can't remember the last version that worked by compiling it
outside of Opam. Opam will tell you what modules is needed for it. They are
all there within Opam. You'll also need to using Opam v4.06 unless LS has been
updated for v4.07. This won't affect the system installed LS
> Ok I changed the first line in test.liq to source =
> single("/etc/liquidsoap/test.avi") and now I'm getting this message:
>
> At line 1, char 9-8: the variable test.liq used here has not been previously
> defined.
This error is essentially FILE NOT FOUND. You are attempting to launch
liquid
Look here;
http://www.liquidsoap.info/doc-1.3.3/cookbook.html
Under 'Dynamic Input With Harbor' and 'Switch Based Transitions'.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/savonet/liquidsoap/issues/61
This error ties in with the previous report #579 and most likely #578.
The reported error above is related to too many files open, which is also TCP
Error 24 in report #579.
I believe this is erroneously marked as 'can't reproduce' and should be marked
as 'Insufficient Information'.
What versi
Firefox v60,0.2 will play the stream but only after pausing the stream and
starting it again.
ffplay produces the same errors as above in initial report but will play the
stream.
VLC produces no errors, plays the stream.
Chrome v67.0.3396.99 is the most troublesome on if it will play the strea
ffmpeg does not produce these errors and the video works for all clients.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/savonet/liquidsoap/issues/549#issuecomment-401944951---
I am not entirely sure if the theora encoding is using gstreamer or it's own
built in code. Because gstreamer also produces this error I just made
assumption that LS is using gstreamer to create the stream.
Further testing resulted in the two major browsers Chrome and Firefox were not
able to p
[gking@geoffpc ~]$ clamscan -r -i ../Documents/liquidsoap-1.2.0-
full/liquidsoap-1.2.0/src/liquidsoap: Unix.Trojan.Mirai-5607459-1
FOUND[gking@geoffpc ~]$ clamscan --versionClamAV 0.99.2/22952/Thu Jan
26 23:52:36 2017[gking@geoffpc ~]$
--
34 matches
Mail list logo
