First I would like to send my hopes for the wellbeing of you and your loved ones in these trying times.
I am always trying to get the most out of my computer: XPS 15 9550 i7 6700HQ. I ran LatencyMon and as usual the ACPI driver in Windows impacted DPC latency.
Please bear with me. I am aware that, in worrying about this driver, I may be chasing phantoms, and missing the point "if the music gets processed, stop whining".
I ran LatencyMon, but excluded core 0. ACPI.sys disappeared from the analysis. The analysis ran for 30 minutes and stayed green.
What I would like to know is, if I set the affinity on Studio One to exclude core 0, will that negatively affect performance? I could try to check myself, but anything I do would be subjective. I'm looking for advice from better heads than my own.
Thanks and stay safe,
That's a good question. Unfortunately I can't give you a definite black or white answer.
I would suggest the following -
1. Define what you mean by 'performance' and what your criteria are for it being affected either negatively or positively.
2. Set up a control Studio One song / project so that you can make some system resource utilisation observations - e.g. CPU, memory, disk activity, round trip latency etc.
3. Make the affinity change to exclude the core.
4. Repeat the observations made in step 2. Compare what you expect to happen with what actually happens.
5. Roll-back your change if necessary. If not observe system performance over a period of time during normal use.
MS Perfmon (shipped with Windows) is your friend here as well as LatencyMon. You could set up a basic Perfmon template to record CPU, Memory, Disk usage (read / writes per second etc) that you can run whilst carrying out step two.
The usual caveats of backing up stuff first (e.g. the contents of C:\users\<username>\AppData\Presonus) and making sure you can properly back out of the change apply.
Not lazy at all.
As I said it's a very good question. It makes good sense to ask if anyone has been down this particular rabbit-hole rather than re-inventing the wheel.
I'd be interested to see your findings.
Users browsing this forum: peterkellett and 15 guests