[LAD] LADSPA plugins in ardour

[Fons Adriaensen]

On Sun, Aug 09, 2009 at 03:05:16PM -0400, David Robillard wrote:

> On Sun, 2009-08-09 at 13:52 -0400, Paul Davis wrote:
> > On Sun, Aug 9, 2009 at 11:22 AM, Fons Adriaensen<fons at kokkinizita.net> wrote:
> > > Hello all,
> > 
> > wrong maiing list, really. but continuing on ...
> > 
> > > I have an ardour mixer strip with 4 input and 2 outputs,
> > > and insert (post-fader) a plugin which has 4 inputs (A,B,C,D)
> > > and two outputs (X,Y).
> > >
> > > The signals at the output of the strip seem to be
> > > X + C + D and Y + C + D instead of X and Y.
> > >
> > > Is this 'documented behaviour', and if yes what
> > > purpose does it serve ?
> > 
> > its certainly not documented, but not much is. it doesn't serve any
> > purpose other than being one of the several approaches one could take
> > with this somewhat unusual (but certainly not wrong) i/o
> > configuration. basically, ardour has the concept of "data streams"
> > within the signal processing route. imagine the situation *without*
> > the plugin for a moment: what happens to the "extra" data streams C &
> > D that don't obviously map to the outputs? executive decision: add
> > them to the outputs (and do it badly, by adding C *and* D to both
> > outputs, just for good measure). add the plugin to the situation ....
> > presumably ardour should know that there are now really just two data
> > streams post-plugin and that nothing needs to be done. but it appears
> > that it doesn't discover this, and this may be intentional but more
> > likely is just a bug.
> 
> This behaviour will be manually configurable (and smarter by default) in
> 3.0, FWIW.  Particularly with (LV2) plugins that give more useful
> information about what ports represent.

Any workarounds today ?

AFAICS, even LV2 would need an 'extensions' to say if it makes sense
or not to replicate a plugin (yes for EQ or delay, very probably no
in most other cases).

Ciao,

-- 
FA

Io lo dico sempre: l'Italia è troppo stretta e lunga.