19 Jun
2003
19 Jun
'03
3:41 p.m.
i jsut did a test with nqpoly4, using two grains that were identical except one had an extra throw-catch pair in it. the one with the throw catch pair started to max out my system (119% cpu) with 300 voices, but the straight connected one was still getting only about 68% cpu. a grain using an s~/r~ connection did only slightly better than throw/catch, using about 115% cpu.
pix.
On Thu, 19 Jun 2003 11:21:10 +0200 "Guilherme Carvalho" guilherme@wanadoo.fr wrote:
Hello list,
I was wondering: is using a lot of throw~/catch~ pairs (or s~/r~) more expensive in terms of CPU than connecting objetcts directly with cords? Things do look better without all the criss-crossing, but I must save all I can in terms of CPU.
Thanks in advance, Guilherme
--