Try sending very large values to "roomsize", which would make longer reverb tails, and very small or 0 value to "damping", which increases decay time. My guess is that [freeverb~] calculates the decay time in samples somehow, thus the very short reverb times at very fast sampling rates. Alternately, move [freeverb~] out of the upsampled subptach, it's probably eating tons of CPU by being there anyways. Try to concentrate only the objects which might produce aliasing in you upsampled section... oscillators, pretty much, although I don't know exactly what you are doing with waveshaping so I can't tell if that could produce aliasing as well. But anything else getting upsampled besides the potentially aliasing objects and the filters used to anti-alias them is really a huge waste of CPU.

best! d.

David F. Place wrote:

On Mon, 2008-10-20 at 20:08 +0200, Derek Holzer wrote:

...

Multiply the decay time by the same factor that you are upsampling?

There isn't a parameter that corresponds directly to decay time? Fooling with the parameters doesn't make much difference.