Offline Bounce Weirdness

[GeeMan]

OK I had a fairly quick search and came up with nothing particlarly revealing about this...

 

We have "upgraded" to L9.18 and 10.8.2. 6 dual core machine.

 

I have a not-very-complex track happening with VI's and audio. Much of the instrumentation is NI, inc, processing with GR5 etc.

 

After I added guitar rig to some busses (I like to be able to process stereo guits etc with one plug/bus) and started getting spikes during an otherwise <10% on each core average I tried to track the problem down. Yesterday the track bounced fine and I've only added processing to this track. I've been through (the usual) extensive elimination tests, and this time, nothing makes much sense.

 

I got the track behaving again, by bouncing effects (not really acceptable connotations for my future workflow) except: when I go to offline bounce, when the track hits the same position as I had the original problem, the bounce slows to a crawl. (Like the bounce would take <50 minutes or something). Realtime works.

 

So just to start with: Is anyone else having extremely weird stuff like this with NI plugins involved?

I've never seen LP act quite this stupid before. Not least, because it's stalling in a large gap where the bus/plugin isn't working on any input.

[GeeMan]

Right.

I'm pretty sure I have tracked the issue down. It bounced. SO unless more issues occur after more tweaking... I have a culprit.

 

Not a 3rd party after all.

 

Match EQ.

 

On a bus, stereo input. The stuff I had ADDED was NI, and up until last night, this track bounced fine, which is why I couldn't trace this more easily. It worked, then today it didn't.

 

The thing was the latency setting - "Zero Latency" in the single instance caused such a massive spike (understatement: four cores were spiking - bugged) it made realtime bounce fail and the offline work incorrectly.

 

One would think, in OFFLINE bounce mode the zero latency setting would be irrelevant. But, no. Not to Apple, somehow.