Latency and performance optimization

As I’m on macOS I leave this here:

While going on in my GP evaluation last weekend I recognised some glitches and stuttering audio. Unfortunately the CPU load meters on both of my instances showed only 20%-30% and below 10% load. :thinking:

I was able to get rid of this with increasing the audio buffers which adds of course some latency, which I don’t care in my instrument rig instance at the moment.

BTW: it’s a cool feature that I can setup different audio buffer sizes in the audio options of each instance. So I can increase buffers in my instruments instance, and lower buffers for audio in processing in the 2nd instance :sunglasses:

But I’m curious: How do you manage your CPU load / latency?

When looking at Logic or MainStage I can readout plugin latency by hovering on the plugin slots and optimise or changing the plugin. I’ve not seen something like this in GP?

Bildschirmfoto 2021-04-28 um 17.44.03

We have a category for feature suggestions in the community forum. It is perhaps still time to propose something like that until GP4 comes out… :wink:

1 Like

GP4 does report the plugin latency in samples and ms. You’ll be able to get that functionality in June when GP4 comes out. See the image below:

Screen Shot 2021-04-28 at 12.05.35 PM

As for different buffer sizes - yes. This was done deliberately so that you can run one GP instance for recording at a high buffer size which will then consume minimal resources and your main instrument instance could be ran at a lower buffer size. Other applications may report the buffer sizes and/or overall latency differently. Some simply have some kind of inconspicuous box saying something like “safety buffer” which in practice could double your latency without your knowledge.

GP always reports the exact latency and even includes the latency measurement tool that will measure the actual latency of your hardware as well.


Perfect. I already like GP4 :slight_smile: