Le Sunday 07 December 2008 14:38:59 Romain Beauxis, vous avez écrit :
> However, I would need more informations for find the issue. They are:
> * More verbose logging: could you send a similar log but with the setting:
> set("log.level",5)
> * A dump of the vorbis stream: run the script with the option dumpfile
> = "/tmp/debug.ogg" added to the input.harbor source, and send it to me (in
> private if you care about copyright..)
>
> With these informations, the issue should be fixed quite quickly..
Thanks to wine, I have been able to run edcast on my machine and reproduce the
issue.
Indeed, the ogg/vorbis stream generated by edcast is completely broken.
It is composed of:
* An initial empty vorbis stream: the last header packet is followed by an
end of stream without any audio data
* Another stream with the same ID as the previous....
It is not suprising that we catched this issue and not icecast, since
ogg/vorbis streams sent by source encoders are mangled by icecast, and this
operation is only header-aware, doesn't care about duplicated IDs, and
renumerate the logical streams IDs..
On the contrary, when liquidsoap receives a source stream, it decodes the
audio (and soon the video), so any broken stream is more likely to get
catched.
However, many thanks polemon for this report, such streams are now correctly
parsed by liquidsoap: when receiving a stream without any data, we try at
least one to parse and get the next logical stream.
Romain
------------------------------------------------------------------------------
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you. Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
_______________________________________________
Savonet-users mailing list
Savonet-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/savonet-users
