How many voices should I be able to get out of K2 on such a system, assuming proper optimizing?
-1.5 gigs of Kingston DDR2700
-Dedicated sample drive (standard issue parallel ATA 7200rpm, 8mb cache)
-M-Audio Delta 66, running at 5.8 ms latency, 128 sample buffers
I'm able to get up to about 100 voices before things choke - tested using the "forearm cluster on piano sound" test method. Is this typical? What are the recommended DFD settings for a comparable system? I'd hoped for more voices.
Also, is it best to have as many instruments as possible running in fewer instances of K2? I'm inclined to have an instance open for ww, brass, strings, and percussion (and sometimes further division), but I wonder if that is significantly less efficient.