Let's see if got the existing logic right.
After being created each part goes through its own group of up to three effects.
It can then pass though one of the 8 insertion effects (but only one part can
go though an insertion effect).
After this all parts are summed and then passed though the system effects and
from there, on the the master L & R outputs.
Now it seems to me that without changing any of this existing structure it
would be possible to take a copy of any part before the mix and send it to its
own pair of outputs.
I would agree that this clone of an individual part should *not* go though the
system effects, and arguably should not go though the insertion effects either.
Does yoshimi get any message from jack to say what ports are connected? Doesn't
jack tell its clients when it wants more data? If so, then surely here is the
mechanism for dynamically working out exactly which processes should be run and
minimise processor overhead - Yes, I know that's much easier to say than to do!
Thoughts welcome :)
--
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.
