# of Voices per DSP Problem

[SHIRK]

When I use more than 1 DSP for voices in the Playback Engine Setup, I hear little clicks and pops during playback (Sounds like a clock problem). If I just use 1 DSP, everyhting sounds fine. Once I use 2 DSP's (across two cards) at min/med/max voices I get these noises. Going further to 3 or 4 DSP's (across 3 cards) gives me the same problem.

I'm on a Powerbook G4 1.25Ghz / 1.5 Gig RAM / OS 10.3.5 / PT 6.4cs9 / HD3 (non Accel) in Magma 4-slot expansion chassis (w/ load resistor) / 192i/o / Sync i/o / external OWC Mercury Elite FW400 drives

Digitest says everything is OK

Playback engine settings are as follows:
512 samples / 85% CPU / 64 voices over 4 DSP's / ADC long Buffer / DAE buffer 2/

I've tried moving these settings around and heard no improvement. I've loaded up sessions that use up almost all of my DSP chips with plugins and everything sounds and behaves fine...so long as the the Voice Allocation is 1 DSP (core card) As I mentioned before, Once I use a DSP on the process cards for Voices, I get the noises. I tried switching order of the process cards and there was no improvement.

The loop master clock is the 192 internal. Tried using the Sync i/o's internal clock as loop master and saw no improvement. Because my setup works fine with only 1 DSP ( the one on the core card) that leads me to believe that somehow the proper clock signal isn't getting to the process cards ???

Could this be a problem with the flexcable that connects the Core and Process cards together.

Any ideas?

Steve

[SHIRK]

bump

[SHIRK]

Anyone? Digi?

[Seeee]

bad ribbon cable? just a guess.

[SHIRK]

Just tried a brand new set of flexcables between the core card and the 1st & 2nd process cards. Still no improvement. I had a bad DSP chip on my second process card so I had digi overnight a replacement process card. Today I got the replacement card, dropped it in slot 2 took out both of the original process cards, so that I was only dealing with a good working core card and a brand new process card and a brand new flexcable. What happens, same thing. 1 DSP for 48 voices....no problem, works great. 2 DSP's for 32, 48 or 64 voices....rice crispies...snap crackle and pop.

I just don't get it. This pretty much rules out that the problem could be in the flexcable. The core card works great when it's on it's own but when a process card's DSP is used for voice allocation, I get problems. Most times I can get away with working with 48 voices...but the minute I have to tell a client that my $20k rig can't do any more tracks becasue of %@#$$^%, I'm going to get look stupid.

Please digi, a little attention here. I'm stumped. Spoke to tech support yesterday, and they were like "uuuuugggghhh!"

[SessionKat]

Please digi, a little attention here. I'm stumped.

Hang on Shirk - I'm gonna get some answers for you ASAP.

-tv

[SessionKat]

Quickly Shirk, do you have I/O's on each card? And if so, is the "loop sync" properly connected? If not, this could certainly be the problem.
check it out if applicable.

-tv

[SHIRK]

Thanks SessionKat,

192i/o is connected to the core card. It's in Loop Sync with my Sync i/o and the 192 internal clock is the master. Cabling seems to be correct. as far as the loop goes. The Sync i/o is connected to the core card's digi serial port as well.

[SHIRK]

To answer your question, no additional interfaces are connected to my system. Just the 192 to the core card.

Thanks

[SHIRK]

For what it's worth, I just tried using 4 dsp's for 64 voices (2 dsp's on core and 2 dsp's on process) This still resulted in the same clicking problem....however, after 10 seconds of playback I got the spinning wheel then a (-9162) error. After the spinning wheel supsided, I hit play again and I got this message (DAE was unable to communicate with the audio hardware. Please save your session and restart your computer. (-6039)

Don't know what that tells us.