[Not a Bug] Error sound with Solo100 Rhy, IF it´s drive is controlled with Scene controllers

I Have a little problem with Solo 100 Rhy.. There is a VERY nasty error (Chemical) sound. -Sound like a broken synthesizer, WHEN I have the drive button controlled by the scene controller (scene controller 1).. It's the only amp of what I have heard until now, with that nasty error. My other preset with diffirent amps is okay, with the same scene controller connected. It seems like the MFC 101 is also effected, in this particular preset, that it is not responding very well to this preset.

To my it sound like a very strange bug..?
 
Seems like the amp model uses some serious extra cpu (+10%) when the scene controller is attached to drive. That is the cause of the crackle. Is this wai?
 
Aaah, YES you are in the right area.. this was serious stopper for the "noise".. But I can NOT understand it use so much cpu? Im right now on 93% and have only Volume > wah > drive > Amp > Cab > Chorus > Fx loop (1. row) > Multi delay > delay > Reverb. I this much..? I don't think so..
But thank you anyway..
 
Confirmed. It's not just the Solo 100 Rhy model. It does it on several other models as well (59 Bassguy, 1959 SLP Treble, Brit 800,...) Attaching a scene modifier to Input Drive causes about a 10% jump in CPU usage on the affected models.

To recreate, assign a scene controller to Input Drive, change scenes, and note the CPU usage in Axe Edit. Use the arrows to scroll through different amp models in the amp block and watch on some of them the CPU usage will jump up about 10%. If you then remove the scene modifier from Input Drive on those models, the CPU usage then drops back about 10% after you change scenes.

I did not hear any noise or odd distortion while testing with just an amp and cab block, so OP's issue there was probably the CPU jump is making it go over 90% CPU on a busy preset, making the audio glitch.

Axe II XL running Q9.02
 
Last edited:
On a quick run through all the models, these are the ones that gave me the ~10% CPU jump with the scene controller attached.

59 Bassguy
1959SLP Jump
1959SLP Treble
1987X Jump
1987X Treble
Bludojai Clean
Bludojai LD 2
Bludojai LD Pab
Bogfish Strato
Brit 800 #34
Brit 800 Mod
Brit 800
Brit JM45
Brit Super
CA OD-2
CA3+ Rhy
Cameron CCV 1A
Cameron CCV 1B
Capt Hook 2A
Capt Hook 2B
Capt Hook 3A
Capt Hook 3B
JS410 Crunch Or
JS410 Crunch Rd
JS410 Lead Or
JS410 Lead Rd
ODS-100 Clean
ODS-100 Ford 1
ODS-100 Ford 2
ODS-100 Ford Md
ODS-100 HRM Mid
ODS-100 HRM
Plexi 50W 6550
Plexi 50W Hi 1
Plexi 50W Hi 2
Plexi 50W Jump
Plexi 100W 1970
Plexi 100W High
Plexi 100W Jump
Solo 88 Rhythm
Solo 100 Rhy
Two Stone J35 1
Two Stone J35 2

All of the other models seem to add only around 1% or 2% with the controller attached.
 
Extra CPU use to modify a parameter isn't unexpected.

Is the preset over the tipping point with respect to CPU use with the controller attached to the parameter maybe?
 
Extra CPU use to modify a parameter isn't unexpected.

Sure I would expect some added CPU, but why the big discrepancy between models though, like say between 1959SLP Treble and 1959SLP Normal. Normal model adds about 3% with the scene controller while the Treble model adds 13%.
 
Last edited:
Takes a lot of math to calculate the Drive. Not a bug.

Wouldn't that math be the same for all the amp models though (controller value, not the internal audio calculations)? Seems like the scene controller provides the same Input Drive parameter values regardless of amp model chosen. Don't all the amp models have the same range of Input Drive parameter values (0.00 to 10.00). I guess I don't understand where the CPU load difference is happening with regard to the scene controller. With no controller attached, there's practically no difference in the CPU meter between amp models. I was under the impression that the dedicated amp block CPU handled any differences in amp model architecture.
 
Last edited:
Don't all the amp models have the same range of Input Drive parameter values (0.00 to 10.00). .
But the amp models don't all have the same input network. When you change the Input Drive, you're telling the system to generate a new way to process the signal, and that takes some math. In circuit analysis, a small increase in circuit complexity can make the math much hairier.
 
Last edited:
Yes but aren't those differences handled by the dedicated amp block CPU? If it's a difference in amp architecture, then why doesn't the CPU load change between models when there's no controller attached to Input Drive?

It's not the changing Input Drive value, it's the fact that there's a controller attached that is causing a different CPU jump between models.

For example, starting each from a blank preset with a row of shunts and a single amp block:

1959SLP Normal model, all defaults, CPU = 11%
1959SLP Treble model, all defaults, CPU = 11%
1959SLP Normal w/ scene controller attached to Input Drive, CPU = 14% (+3% for controller)
1959SLP Treble w/ scene controller attached to Input Drive, CPU = 22% (+11% for controller)

Why the big difference for the controller load when the base load between the two models is the same?
 
Last edited:
this is how it's worked for a while. wasn't the Drive on the AX8 not allowed to have a modifier because of this?
 
This has been discussed before.

Cliff: "Depending upon the amp model it can take a lot of CPU to calculate the Input Drive network. Some amps have simple networks that are rapidly solved. Others, like the Hook Lead and Rhythm models have complex networks that require more math. If you attach a modifier to the Input Drive it is constantly recalculating the network which increases CPU usage."

(wiki)
 
?? I'ts somehow an automated process, anyway.. From where the scene controller is set to only change input drive, to a specific value, when and only when you change scene? Nothing else.. The increase in cpu 10-13% for this "calculation" is compared to ad a reverb block?
 
Last edited:
Btw.. it would be an extremely handicap for the possibilities for have automated drive shift and in the same time, have some other good effects.. IMO
 
Sure I would expect some added CPU, but why the big discrepancy between models though, like say between 1959SLP Treble and 1959SLP Normal. Normal model adds about 3% with the scene controller while the Treble model adds 13%.
The math the unit has to do when the parameter changes varies from model to model.
 
Back
Top Bottom