CPU spike when 'print track' put into input

[[email protected]]

Hey guys, I know similar things have been posted, but this is a head scratcher.
I have a mix session that's pretty complex, but sit's nicely around 50%-60%
Now, without change, it jumps to 90%-100% when I put the 'print track' into input. (or record)
Video attached:

https://www.dropbox.com/s/yh1xxno8vb...spike.mov?dl=0

[[email protected]]

Bump!

[[email protected]]

Nothing? No one else is seeing this?

[clancychris]

http://duc.avid.com/showthread.php?t=394178

I've had exactly the same issue. Had to totally change my workflow to try and get around it. Somebody has opened a case about it but if you read through this thread there's other people with the same issue. There's also a test session posted in there which I'm sure people would value your feedback on

[Carl Kolchak]

Could it be due to Pro Tools new(ish) dual buffer system?
When you're in playback, it's in the high latency large buffer, when you activate input monitoring, or record arm a track, it goes into the low latency smaller buffer - as set by the hardware buffer size.


Sent from my iPhone using Tapatalk

[Darryl Ramm]

So then a useful question is at what IO buffer size are you running? If that is small what happens when it is made very large?

Is this a problem? Why exactly? What AAE error do you get? Often looking at the CPU meters is a waste of time anyhow.

If there is a problem can freeze/commit or bounce to disk be used here instead of printing tracks?


Sent from my iPad using Tapatalk

[[email protected]]

Originally Posted by Darryl Ramm

So then a useful question is at what IO buffer size are you running? If that is small what happens when it is made very large?

Is this a problem? Why exactly? What AAE error do you get? Often looking at the CPU meters is a waste of time anyhow.

If there is a problem can freeze/commit or bounce to disk be used here instead of printing tracks?


Sent from my iPad using Tapatalk

at 1024. Pretty big.

[[email protected]]

Originally Posted by clancychris

http://duc.avid.com/showthread.php?t=394178

I've had exactly the same issue. Had to totally change my workflow to try and get around it. Somebody has opened a case about it but if you read through this thread there's other people with the same issue. There's also a test session posted in there which I'm sure people would value your feedback on

Thanks. I checked the other thread. Similar stuff. Hopefully Avid could chime in at some point.

[[email protected]]

Originally Posted by Darryl Ramm

So then a useful question is at what IO buffer size are you running? If that is small what happens when it is made very large?

Is this a problem? Why exactly? What AAE error do you get? Often looking at the CPU meters is a waste of time anyhow.

If there is a problem can freeze/commit or bounce to disk be used here instead of printing tracks?


Sent from my iPad using Tapatalk

I just realized you asked a few more questions:
Getting all kinds of errors- 9073, 72, 9093, and a bunch of others.
Freezing/committing/bouncing is not an option as it sounds different AND I have a hybrid setup that uses hardware inserts. (please- let's not get into the whole- sounds the same debate)

[Darryl Ramm]

Ouch. Do those errors happen specifically when printing the track when you see the CPU increase or also at other times?

What happens on sessions with no hardware inserts, are they better behaved? Regardless of anything else broken here, when you run stuff through hardware inserts you are working the system harder on the “low-latency” side of the house... but at 1024 samples it should handle a lot.

(Yes they sound the same but I’ll let you off because you mentioned hardware inserts )


Sent from my iPad using Tapatalk