[linux-audio-dev] ladspa diff

[Jesse Chappell]

Paul Davis wrote on Fri, 12-Sep-2003:

 > >Well, I'm *half* paying attention...
 > >
 > >So we're essentially saying that the momentary thing is a trigger that
 > >operates on 0->positive but not positive->0. I must admit I'm not 100% clear
 > >about the randomisable case from the text.
 > >
 > >Do we have example plugins for the two cases?
 > 
 > sooperlooper is the example for the momentary case. in that case, your
 > definition isn't quite correct. a momentary control is one that only
 > causes something to happen when its value transitions between any two
 > meaningful values. this means that it needs to be reset to a
 > meaningless value after being set to any meaningful value. this is
 > required so that "pressing the button twice" does what the user expects.
 > its actually slightly more complex than that at the code level.
 > 
 > the more i try to explain this, the less convinced i am that we should
 > add it. it requires the definition of a meaningless value, for a
 > start, which will require warping/overloading the meaning of some
 > other value (e.g. using the default as the meaningless value).

I agree, I do not think a momentary hint would really help in this specialized
case.  Sooperlooper will never be a ladspa plugin that any generic
host can use.  The better solution to the SL problem is audio-rate
control ports.  Then we get sample accurate control, and can easily
specify the "off" value in the plugin's custom semantics.

So the hint we *really* need is the audio-rate control port (or
the control-hinted audio port :)

 > however, jesse tried to come up with a way of modelling this kind of
 > control with existing LADSPA options and couldn't do it. the "requires
 > reset after every use" is hard to do without it.

The audio rate control port would fix this, but again, only a
custom host could use this plugin, which I'm fine with... apps
to use it are on the way.

jlc