Teiwaz
Saturday 30 November 2002 à 09:47
It's true that TDM doesn't degrade audio in the slightest, you just hit a brick wall and can't enable any more channels of I/O or plug-ins, when you run out of DSP on the PCI cards. Unfortunately, with TDM, I have discovered that I run out of DSP steam way faster than I do on the ASIO/VST system...

The situation is rather different on ASIO/VST/RTAS; experiencing a sort of "fizziness" and a softening of dynamic range when approaching the max CPU load on the host Mac's processor.

I find that in this situation I just bounce my virtual instruments (well, maybe just a few - the most CPU hungry ones like Battery, Absynth or Stylus) to hard disk and then run them as audio tracks. On a dual 1Ghz Mac this gives up to, in some cases, a good 60% of DSP power back to the CPU!

And then, a great deal of the compressor/reverb/eq/etc tasks are taken care of by the UAD-1 card's powered plug-ins - this configuration takes a great deal of stress off the Mac's processor too...
Of course, I do realise that I am kind of lucky to be the owner of a fast mac with a UAD-1 card, and that slower machines are more tricky when running large amounts of audio/instruments.
When bouncing an instrument to disk, I always solo the instrument that's being bounced, so that I have plenty of CPU reserve to ensure optimum audio quality during the bounce.
Also, I never actually record audio on the ASIO/VST mac. I record audio separately on a TDM based mac, so I can be sure of optimum record quality at all times. I feel that if I were forced to overdub & record additional audio tracks in a song that was already running hot on the CPU on the ASIO/VST system, that this would compromise the quality of the audio source actually being recorded.
I feel fortunate enough to be running dual systems...this provides flexibility and assurance of audio quality (both systems run Logic and are synchronized together, and I never have any major system problems (touch wood!)