Hi Ross,
Does this relate to the number of cores as hypothesized below? Or have you landed on a different number of instances for the transform.go? Once the fourier is running the maximum instances of DeFeedback is it typically also tapped out for other plugins? Apologies if that’s a basic question. I’m just gearing up to install a couple of these systems at my venue.
Hey jcrocker,
No basic questions here!
Running at the minimum latency means running at the smallest buffer size possible, and this typically means Defeedback is using a large amount of your DSP budget. if you wished to use more plugins on top of that, you would need to increase the buffer size, which naturally increases the total latency.
Since the transform platform does core isolation, you can imagine where a transform.engine with 6 DSP cores can handle x instances, the transform.go has 2 cores, so there is some simple maths to run to get an aproximation. We don’t want to comit to a number of instances diretly just yet as we haven’t done the extensive testing that we did with the transform.engine.
Hope this helps,
Ross
Product Specialist
1 Like