Hi guys,

Le 13 janvier 2012 10:00, Brandon Casci <[email protected]> a écrit :
> a long shot, but maybe you can hack something together using sox
>
> http://sox.sourceforge.net/sox.html
>
>
>
> 2012/1/13 Jean-Francois Mauguit <[email protected]>
>>
>> Hello,
>>
>> On 13 Jan 2012, at 00:45, Frédéric wrote:
>>
>>
>> Hi,
>>
>> As far as I understand LS, it would perhaps be possible to:
>> - duplicate your source (ie. source_A = source_B)
>> - use "filter" on source_B to enable a high pass filter that would cut
>> any frequency below 20 Khz
>> - use "on_blank" to do nothing when source_B is blank, else your
>> triggered action when not blank.
>>
>>
>> It could work. I'll have to set up a lab to test that :-)
>>
>>
>> Another way to do this would be to add a third track to your stereo
>> source.
>> You'd be able to use audible "tops" on the third track without needing
>> to duplicate and filter your original stereo source.
>> And remove the third track before outputting the source.
>>
>>
>> As it comes from an FM broadcaster, I don't think they could synchronize a
>> second source with the main stream to send the top in this second source.

I do not think that we have a direct support for this. I see, however,
several possible ways to hack this:
* Custom *-pass filter and on_blank (already mentioned): if that works
and is not too cpu intensive, then great!
* Plug through sox (already mentioned)
* Plug through jack
* Use a custom LADSPA plugin

All in all, I'd like to mention an important point, specially for you
radionomy guys: this might turn out to be quite CPU intensive. Thus,
you'd better pick-up the most efficient solution right away. Here, I
think it is the LADSPA plugin one, here's why:
* Sound processing within liquidsoap is not very efficient as of now
(sam is working on this, maybe he'd like to insert something here..)
* Sox and jack requires multiple processes, which can really be a mess
to monitor

On the contrary, the LADSPA API is relatively simple and integrates
smoothly into liquidsoap. See documentation there:
  http://liquidsoap.fm/doc-1.0.0/ladspa.html
LADSPA support in liquidap is also quite old so a single LADSPA plugin
should work also with old versions of liquidsoap. Finally, being pure
C and dynamically loaded into liquidsoap, LADSPA plugins are probably
the most efficient way to implement custom sound processing in
liquidsoap for now..

The only difficulty with this solution is that I do not think that a
LADSPA plugin can interact directly with a running instance, like
change a variable etc.. So, if you want to trigger a transition when
your LADSPA plugin detects some sound, you may have to think about how
this may be done. There are a lot of possibilities, though, such as
flipping a switch directly within your own software, calling a telnet
command, touching a file, etc.. You could even think of a plugin
transforming the source into a source with only blanks and bips, which
bips/blanks when something has to happen... Just use your imagination
:-)

Finally, please let us know if you find a nice solution that you are
able to share with us, I am sure this would be very interesting for
others as well :-)

Romain

------------------------------------------------------------------------------
RSA(R) Conference 2012
Mar 27 - Feb 2
Save $400 by Jan. 27
Register now!
http://p.sf.net/sfu/rsa-sfdev2dev2
_______________________________________________
Savonet-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/savonet-users

Reply via email to