I have a home rolled external which I ported from Linux to W2K. The
external is a physical model of a Tibetan bowl which employs 8
banded digital waveguides.
The port was done in Visual C++, and all the appropriate changes were made
to the external's code and compiled. pd for W2K v 0.35, appropriately
recognizes the external. My machine is a 800mhz P3 laptop.
On a Linux 500 Mhz desktop, on which the original external was developed
and tested, the audio and (massive) control patches worked fine with
minimal notice of processor overload.
Problem: On my windoze laptop, with the exact same patches, and the newly
compiled bowl external, the processor is max'd out at 100%, causing
significant control latency, lethargic graphics, and no sound output.
I was wondering if anyone has had similar problems. I somehow think there
is a problem, given the significant realtime buffer reading and writing,
with the way W2K dll is dealing with the buffers (as apposed to the lean
Linux way). I would use Linux for this project if I weren't having to
deal with a MIDI interface without a reliable Linux driver.
Thanks.
Carr