Hi, I am new here and read a lot of topics but didn’t find the right one:strong text
I want to move from MainStage to Gig Performer. Mainly driven by higher flexibility and lower CPU - usage.
The thing is: I rebuilt some of my MainStage patches in Gig Performer (1:1).
All parameters are the same (Buffer size, sample rate). The same plugins, the same presets. But in Gig Performer I have audio glitches and ongoing overload while in Mainstage the CPU - load is max, 70%.
I mean: Deskew makes a lot of advertising regarding the lower CPU - usage.
What might be the reason for the higher CPU usage in Gig Performer?
Ok, in principle it is not the best idea to rebuild exactly what you did in Mainstage in Gig Performer.
Because Gig Performer is not bus oriented as Mainstage is.
Gig Performer supports free routing - Mainstage does not.
But in Gig performer you can build like “bus oriented” - not the best idea.
Gig Performer does have rackspaces and variations, a concept Mainstage does not have.
So when you exactly rebuilt - you did not use all possibilities Gig performer offers.
Believe me, I was a long time user of Mainstage and switched to Gig Performer - a much better experience for my use cases. I do not know yours.
Any chance to upload your (smallest possible) gig where you face this issue?
I could take a look at that.
This can often be misleading as the two applications have a different set of settings for this.
Do you have the “I/O Safety buffer” checkbox checked in MS under the buffer size?
When’re is your “Driver Latency” slider positioned? More towards less latency or towards more safety?
Finally - what is the resulting latency in ms displayed in that dialog within MS?
If you have these “safety” options set up - it simply means that your buffer size is not what is displayed, but it is doubled.
Take a look at these two screenshots from MS and GP. Both set to use the same interface, same sample rate and in MS - I deliberately selected twice as small buffer size of 128 and in GP I selected the buffer size of 256.
The resulting latency is the SAME which means that 128 in MS is actually resulting in the same latency as 256 in GP (or other applications). The difference is that the actual latency will most likely bi higher in MS even with half the buffer size. If you look at the reported output latency you will see 10ms in Main Stage and GP says that if should produce 7.3ms output latency.
Exactly. I tried to rebuilt a patch I use in Mainstage.
The rackspaces / patches I programmed are the same.
A rackspace with 4 instances of Omnisphere (each one with a single patch, no stacked patches) and one instance of diva plus two effects (Synthorus 2, Eventide H3000), one on diva, the second on a an omnissphere patch leads to audio glitches in GP. In MS the CPU load is max 70% max in this case, no glitches at all.
