Subject: Re: [linux-audio-dev] LADSPA run() blocksize
From: Benno Senoner (sbenno_AT_gardena.net)
Date: Fri Nov 17 2000 - 20:04:33 EET
On Fri, 17 Nov 2000, David Benson wrote:
> >
> > Its an optimization yes, but then we need optimizations like a fish
> > needs water.
well said !
>
>>
> For this and the run_adding() argument, I'd love to see measurements
> (your intuition and mine differ on their importance)
I think run_adding() can make a a big difference in large mixer setups,
and if you consider the fact that we will not use 32bit float forever ,
it becomes even more important.
>
> Summary: I'm worried very much about adding/recommending plugin complexity
> that could be done (with say >99% efficiency) by the host.
99% would be ok, but I fear that this drops to 70% (certain cases of
run_adding) which is quite a performance hit.
And what's the deal of transforming a
out = .....
in
out += ....
in order to support run_adding() ?
And if you are lazy, could even use a macro in order to avoid code duplication.
But as David O. mentioned, run_adding() makes much more sense if provded with
a standarized gain control.
Benno.
This archive was generated by hypermail 2b28 : Fri Nov 17 2000 - 18:46:24 EET