[PD] vanilla partitioned convolution abstraction

[Roman Haefeli]

On Wed, 2019-01-09 at 11:27 -0200, Alexandre Torres Porres wrote:
> yeah, I also get lots of glitches and artifacts with a 64 minimum
> window. I have to increase the delay up to 50ms so I get rid of them,
> which is kinda bad, even though it seems incredibly efficient. 

Just to clear: I'm not experiencing Pd glitches at any configured audio
setting whatsoever. What I mean is the abstractions produces 'wrong'
results when configured to 64. I don't understand enough to give any
hint why 128 is still fine and 64 sounds off.

> Brent is not using gardner's approach I assume. And what this does is
> that it increases the window over and over in the following
> partitions, this is why it gets so much lighter on the CPU

Yeah. From what I gather without reading William Brent's [convolve~]'s
source code, is that it uses n partitions with same size. At least the
printed number seems to suggest that. Thus, using a smaller delay leads
to a significantly increased CPU load. With Gardener's approach, the
penalty of a short delay is almost not noticeable.

Roman

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 488 bytes
Desc: This is a digitally signed message part
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20190109/b52d6d08/attachment-0001.sig>