All blocks muted, but CPU maxed

As others have stated, the blocks are always "active", and if I remember correctly this is to ensure that engage/bypass functions don't produce artifacts/gaps/etc in the audio.

It's counter-intuitive to many, but it's the most reliable method of ensuring horsepower management by the end user. Performance-disrupting issues are far more likely if this wasn't the case.

That said, I suspect that Fractal's highly anticipated "gapless" update to the FM3 firmware will render such a discussion obsolete. If you need more CPU headroom in your current preset, consider using only a single standard-length IR for your cab block, deleting unnecessary shunts, reducing any effect/tone qualities to "Economy", or reduce the number of Control Switches by finding ways to program them with Scenes. Any combination of these may be enough to get you back in the green.
 
This has been a helpful thread for me. I never really thought about shunts. I went in and cleaned up my live presets that were around 79-80 and it brought them to 77-78. I'm much more comfortable with that in a live situation. No need to be pushing my luck when there's a host of other things I can do to screw our show up :oops:
 
If you happen to use Control Switches in your preset, I'd definitely try and get that usage another percent or two juuuust in case!
 
If you happen to use Control Switches in your preset, I'd definitely try and get that usage another percent or two juuuust in case!
I used to use CS per scene to lower the gap between clean and drive tones by staying on one amp/cab, but I settled on living with the gap so I could have a Fender clean and 5153 drive and lead. Many of the gain amps clean up pretty nicely, but not to a sparkly Fender clean.
 
Do away with any blocks that are empty and the shunts if you can go to a spring reverb and lower the quality of the reverb that will help some.
 
Also useing economy settings will lower your CPU on your blocks. For live settings, you won’t notice the difference.
 
Are you sure one IR uses less CPU than two?
I tried many times to leave second slot empty, never noticed any changes in CPU usage against channels with two IRs.
Standard resolution, of course.
I believe on the FM3 the IR processing is offloaded to accelerator processor... Maybe that's why?
 
Are you sure one IR uses less CPU than two?
I tried many times to leave second slot empty, never noticed any changes in CPU usage against channels with two IRs.
Standard resolution, of course.
Yes. On my unit, both Legacy and Dynacabs save roughly .5% CPU by using a single IR instead of two.

Dynacab, 2 IRs: 24%
Dynacab, 1IR: 23.6%
Legacy Cab, 2 IRs: 24%
Legacy Cab, 1 IR: 23.6%

Waited a couple minutes and then recorded the highest CPU spike for each at idle. Preset attached in case I didn't account for something!
 

Attachments

  • Do IRs Change CPU.syx
    24.1 KB · Views: 0
Back
Top Bottom