Hi Majik,
majik wrote:
---Here is an email I sent a little while ago to Kai
Vehmanen:
I have been looking through the Soundtracker
0.6.7 code as I have been
wanting to improve the jack output. It would be excellent if Soundtracker
could output each of its channels to a Jack port, instead of (as well as?)
outputting the mix the two mono channels. Unfortunately, Im only a beginner
when it comes to audio coding and was wondering if someone could hack it
together for me?
---And his reponse:
that would be a cool feature to have, but unfortunately not a trivial
thing to add (though not impossible either). You could try to ask about
this on the soundtracker mailing list (or possible on linux-audio-dev)...
maybe someone else is also interested and has the time to help.
Unfortunately I don't have much time for FOSS-development atm, so I'll
have to pass. .(
---Will anyone help with this? I believe that the problem is that the mixer
code works in a monolithic way, thus needs a rewrite.
This issue's been discussed on the soundtracker mailing list, months ago.
Here's a quote from Yury Aliaev, a Soundtracker contributor :
<quote>
In the current state of ST such a thing (multichannel output) is almost
impossible because of the monolitic structure of the mixer code. The
optimal way to solve this is (I mean) rewriting the whole mixer in the
modular way. This also will make ST more flexible and universal and, in
particular, will make adding new effects (including LADSPA processing)
more easy.
Currently I have some ideas how to do this (but still have no free time
for this :( ), but there is another way: libremix written by Conrad
Parker (see
remix.sf.net) seems to be good for this purpose.
</quote>
A question I asked, a few weeks later :
<quote>
About per track JACK outputs : In your answer to
Emiliano Grilli, on
june 1st, you explain that this needs a big rewrite. But do you
believe a sort of hack is possible ? Like a small patch that lies
around for those (I like JACK :) who need this feature before
soundtracker engine gets rewritten... If yes, any advise ? Does it
imply playing with the mixer assembly routines ?
</quote>
And Yury's answer :
<quote>
Unfortunately, it will be a havy hack of assembler routines :( Because
they mix sounds from different channels directly after resampling,
rather then write to the separate buffers and then mix them. This is why
I decided to rewrite the mixer entirely rather then inventing kludges...
</quote>
Hope it helps... Yury may have started to rewrite the mixer.
Regards
--
og