[LAD] LADSPA dilemma

Fons Adriaensen fons at kokkinizita.net
Sat Jun 16 15:01:06 UTC 2007

On Sat, Jun 16, 2007 at 04:34:19PM +0200, Tim Goetze wrote:

> [Fons Adriaensen]
> >It would also be possible to define that calling run() with nframes
> >equal to zero means 'update your internal state to the current 
> >control values', but this not specified in the LADSPA specs.
> I really like this idea.  True, it might break some plugins, namely of 
> course those that divide by the number of frames without looking at 
> its value -- but these are likely to be the ones that do parameter 
> smoothing already, and would be updated first.

That would include almost all of mine :-(
It's not really backward compatible w.r.t. hosts - no host should
ever do this until _all_ plugins are updated...
> A simple redefinition of what is already in place that gains us quite 
> a lot of control over parameter smoothing.  Very elegant, I think.

Note that this is not really simpler than the flag I proposed earlier.
It requires exactly the same code to update the internals, and instead
of testing the flag you have to test nframes.


Follie! Follie! Delirio vano è questo !

More information about the Linux-audio-dev mailing list