Dietz wrote:One of the Beta testers had MIR running and working on Bootcamp and reported ok-ish results. (Ron, would you please recapitulate your findings ...? TIA!)
Dietz, my pleasure:
Folks, one of my test systems was an early 2008 Mac Pro octocore 3.2
ghz with 32 gigs of ram, running Windows 7 and an RME Fireface 800,
cubase 5.0.1 is running on the same computer. This system is certainly no slouch, but is a generation old.
My findings were as follows:
Buffer 128 is not really a viable option
for more than a hanful of instruments. At buffer 512, I can run
roughly 15-18 instruments at approximately 75% cpu consumption before clicks and pops begin (all perf-legato patches for these testing purposes). I can
basically double that at 1024. Buffer set at x2. Keep in mind, however, that at 1024 w/ x2 buffer, there is some latencty. It's actually a
fairly viable MIR solution for smallish-medium orchestrations, but don't expect to mockup Mahler ;-). Also, MIR's buffer should be set at x2 or above.
If
someone were planning to use this kind of rig for MIR, I would
definitely recommend having the Sequencer on another computer.
I'm happy to answer any questions about this kind of setup.
Best,
-Ron