Re: [Savonet-users] Liquidsoap HDSP 9632
On Wed, Mar 11, 2009 at 12:25 AM, wrote: > I've tried to change the frame.samplerate to 48000 and withdifferent > values for he direct input.alsa - but no changes - I've the same error log > : Error while setting rate: Alsa.Invalid_argument! Ah, that's annoying. I don't know what more to try... Just to make sure, I'll precise that 48000 was only an example, what I meant is that you should find what rate your card accepts, and use that. It might be 92k as well -- which actually seems more appropriate for that high-end card, from a quick googling. Good luck, -- David -- Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com ___ Savonet-users mailing list Savonet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/savonet-users
Re: [Savonet-users] Liquidsoap HDSP 9632
Hi David,
also many thanks for your quick answer and the explanations.
> Here, you should force liquidsoap's internal samplerate to match one
supported by your card. The default is 44100, perhaps your card only
takes 48k? In any case, you can change it using
> set("frame.samplerate",48000).
For the moment, the HDSP works with Liquidsoap only with Jackd and
Qjackctl (for the patchbay) and it's a little bit tricky and use lot of
CPU.
I've tried to change the frame.samplerate to 48000 and withdifferent
values for he direct input.alsa - but no changes - I've the same error log
: Error while setting rate: Alsa.Invalid_argument!
I'm wondering if this may come from a wrong .asoundrc file or channel
selection at a low level end ? Or something else ?
>> input= input.jack(id="input")
>>
>> this stop liquidsoap and log :
>> 2009/03/09 01:18:18 [stderr:3] ERR:
jack_wrapper.c::JACK_OpenDevice(838) jack server not running?
>
> It's good that you sent a separate mail about that, it's a different
issue. Probably a bug. But right now I don't have any idea about it.
Problem fixed, HDSP works with Jackd under liquidsoap user (see last post).
Best,
Lb
--
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
easily build your RIAs with Flex Builder, the Eclipse(TM)based development
software that enables intelligent coding and step-through debugging.
Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
___
Savonet-users mailing list
Savonet-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/savonet-users
Re: [Savonet-users] Liquidsoap HDSP 9632
Hi Romain, many thanks for your answer. Romain Beauxis a écrit : > Hi ! > >>> input= input.jack(id="input") >>> >>> this stop liquidsoap and log : >>> 2009/03/09 01:18:18 [stderr:3] ERR: jack_wrapper.c::JACK_OpenDevice(838) >>> jack server not running? >> It's good that you sent a separate mail about that, it's a different >> issue. Probably a bug. But right now I don't have any idea about it. > > Is the jack server running ? Sorry but I've found the problem : it was a newbie user:group problem between liquidsoap and Jack... When launching Jack under the same liquidsoap user:group, everything works well ! Best, Lb -- Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com ___ Savonet-users mailing list Savonet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/savonet-users
Re: [Savonet-users] Streaming theora video failed.
Le Tuesday 10 March 2009 15:43:26 Romain Beauxis, vous avez écrit : > mplayer is generaly not very good for decoding live Theora streams, while > vlc behaves quite good. Also, you can expect a commit to enhance this very > soon. Done ! Our ogg page muxing is the better possible one now :-D Romain -- ___ Savonet-users mailing list Savonet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/savonet-users
Re: [Savonet-users] periodic metadata rewriting
Hi, Gergely CZUCZY wrote: > However, i've read > through the 0.3.8 and 0.9.0 API reference, but I couldn't find any > functions that might give me the next song, or anything. I've also > checked the queues, if there might be something. Yes, we still haven't found a way to auto-generate the reference of all server commands (which is not the same as builtin script functions). It is likely that the server interface will become more structured and hence more easily documentable for 1.0.0. But you can get a list of all available commands by typing "help" in telnet. For playlists, there is a command ".next". > The closest i could find is playlist.parse, which returns a > wicked multidimensional associative array, somehow I will have to get > the current uri playing, look that up and go forth 1 iteration for the > next song. If this is what you've been thinking of, could you please > give me a hint how to handle a [([(string*string)]*string)] ? I think that it's unnecessary to go in that direction. The next command gives you URIs directly. > In the .metadata output in the telnet session i see a "rid" field, > incrementing. Might that be the nth item of the playlist? Using that > this might be a bit simpler. No, the RID is a global identifier for requests. The RID of a new request depends on which requests have been created and deleted before, which depends on which other sources are active, etc. Destroyed requests keep their RID as long as possible, but when no more RID is available their ID is given to newly created requests. In short: it's not reliable. Good luck, David -- ___ Savonet-users mailing list Savonet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/savonet-users
Re: [Savonet-users] Streaming theora video failed.
Hi ! Le Tuesday 10 March 2009 15:25:30 fruity, vous avez écrit : > # output to a icecast server > output.icecast.theora(port=8000, password="passwd", host="icecast", > mount="test.ogv", name="test.ogv", protocol="http", > genre="Experimental,Eclectic,Free pirate radio", description="Online > since 1997", public=true, final) I don't see anything ovbiously wrong here. You can check the maximun source limit in your icecast.xml config file, since the same error message is triggered when icecast refuses the source when there is already the maximun number of them... Ha, also, I am currently struggling to enhance the synchronization between audio and video in our streams. Clearly, this kind of videos, with a single image and audio, is very hard to stream: there can be a fairly important time without video keyframe (main image, in jpeg, the other ones are coded by difference with this one), and then the player is confused because it has nothing to display for the video. mplayer is generaly not very good for decoding live Theora streams, while vlc behaves quite good. Also, you can expect a commit to enhance this very soon. Romain -- ___ Savonet-users mailing list Savonet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/savonet-users
[Savonet-users] Streaming theora video failed.
Hi to all,
I'm doing something wrong here, I just copy paste the example from toots
blog on rastageeks to stream a theora video.
# Log to stdout
set("log.file",false)
set("log.stdout",true)
# Disable real-time processing, to process with the maximun speed
set("root.sync",false)
# Enable video
set("frame.video.channels",1)
# The video source using a static image
s = video.image(width=240,height=240,x=40,"/test.ppm")
# The audio song.
audio = single("/test.mp3")
# The special metadata function, that
# also triggers the shutdown..
def end_meta(_) =
shutdown ()
[("title","prout")]
end
# This is the blank source
bl = map_metadata(insert_missing=true,update=false,end_meta,blank())
# And the final source: the sequence song and blank added
# with the video of the static image..
final = add([sequence([audio,bl]),s])
# output to a icecast server
output.icecast.theora(port=8000, password="passwd", host="icecast",
mount="test.ogv", name="test.ogv", protocol="http",
genre="Experimental,Eclectic,Free pirate radio", description="Online
since 1997", public=true, final)
but i get:
notification: This doesn't look like a Speex file
notification: This doesn't look like a Speex file
2009/03/10 14:23:56 [rcyb1(dot)ogv:2] Shout socket error: timeout,
network failure, server shutdown? Restarting the output in 3 seconds.
what am I doing wrong?
thanks in advance
fruity
--
___
Savonet-users mailing list
Savonet-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/savonet-users
Re: [Savonet-users] periodic metadata rewriting
We've got a switched am/pm playlist, that makes our source, and that's put out in two formats. So we're using playlist(). However, i've read through the 0.3.8 and 0.9.0 API reference, but I couldn't find any functions that might give me the next song, or anything. I've also checked the queues, if there might be something. The closest i could find is playlist.parse, which returns a wicked multidimensional associative array, somehow I will have to get the current uri playing, look that up and go forth 1 iteration for the next song. If this is what you've been thinking of, could you please give me a hint how to handle a [([(string*string)]*string)] ? In the .metadata output in the telnet session i see a "rid" field, incrementing. Might that be the nth item of the playlist? Using that this might be a bit simpler. On Mon, 9 Mar 2009 16:18:43 +0100 David Baelde wrote: > > And there's one more thing we'd like to do. We'd like to display on > > the website the current and next song for each stream (artist, > > title and maybe album would be enough). I haven't seen functions > > for this functionality in the docs, nor ways how could another > > application recieve this information. Could you also help us please > > on this one? > > On dolebrai.net, we use a on_metadata() handler which calls a script, > which itself interacts with liquidsoap via telnet, to obtain the > current and last ten tracks, uptime and other info. This information > is put in an XML file, which can then be re-used in several ways; in > our case, it's directly processed by javascript using AJAX. > > The hard part is to obtain the next track. In liquidsoap there is no > builtin way of doing that: the script language makes it too > complicated, and it's even worse if you take into account possible > external events. But, there's a good chance that _you_ know how _your_ > script decides how to choose the next track. For example, if you have > a playlist and a request queue that overrides it, you can ask the > request queue what's next (via the server) and if there's nothing ask > the playlist (also via the server although I must admit that the > command is not very handy right now). -- Sincerely, Gergely CZUCZY Tel: +36-30-8414202 -- ___ Savonet-users mailing list Savonet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/savonet-users
