Warning about big preset loads in the lower 90%

josephyballew

josephyballew

Inspired
I was working on a preset that was 92 percent when finished with Axe edit connected. All communication took forever to happen then froze. I rebooted both the Axe and the Axe edit. Now Axe edit and The Axe 3 will not communicate. If I rebooted my Axe to the same preset, the Axe would go to scene one but if I tried to go to scene 2 would freeze up. If I booted the Axe up and changed presets, the Axe was fine. I could not get Axe edit or Fractal bot to work. I had to uninstall the USB driver and reinstall it. All is good now. I am just guessing here, but I think the new firmware should consider 90% the new 100 percent max load. This was bad enough to destroy the old driver and lock down the Axe on that preset. I redid the preset and it is at 85% and all is good. The reinstalled fixed the driver. So before you decide to load up all you can, I would keep it under 90% until this is checked by the engineers. So I am thinking that bottle necking is happening at 90% and above, enough to destroy a driver.
 
I seriously doubt that a CPU-overloading preset can "damage" the USB driver for Windows.

Also, the manual mentions 90% as being the limit.
 
  • Like
Reactions: Ole
I didn't put 2 + 2 together until your thread ..., but I think that's exactly what happened to me and what I was trying to report here: https://forum.fractalaudio.com/threads/problem-w-axe-edit-iii-or-axe-fx-iii.139088/

The preset I'm building was in the low 80% ( IIRC ) and I tried to add a 2nd Amp block .... Axe-Edit wouldn't let me do that because the CPU load became too high. It locked everything up and I also had to re-boot ....

That's when the trouble started w/ Axe-Edit not seeing the Axe-III ....

Same fix for me too .... I uninstalled and reinstalled both Axe-Edit and the drivers.

All seems to be good now !
 
Did you reboot the computer and III before retrying Axe-Edit (before reinstalling the driver)?
 
yek said:
Did you reboot the computer and III before retrying Axe-Edit (before reinstalling the driver)?
Click to expand...

I did .... multiple times both just power cycling the Axe-III and rebooting my computer .... Every time I did that the first launch of Axe-Edit would time out .... Then the 2nd launch of Axe-Edit it would begin reading the User Cabs at an extremely slow rate ( approx. 1 Cab per second - SERIOUSLY ! ) .... Then when it reached User 1 100 ( or so ) it would accelerate to reading them at normal speed and eventually go live ....

Uninstalling and reinstalling both the USB Drivers and Axe-Edit seem to have solved the issue ....

It didn't dawn on me until this thread that it might have happened because I tried to max out an individual preset ?
 
I've always thought 90% was a weird number to not exceed. Seems like 100% would be more logical with fewer questions and concerns from the user base. The 90% to 100% dead range seems to confuse a lot of folks. If someone gets to 90% they naturally think they have 10 more percent to go. Just my 2 cents.
 
josephyballew said:
...but I think the new firmware should consider 90% the new 100 percent max load.
Click to expand...

I've always wondered why FAS has yet to change this. My brother has an FX8 and runs into this all the time as well.
Can FAS modify to show 100% cpu utilization when the system wide cpu utilization is only at 90%.

Think of it as the Spinal Tap "these go to 11" situation, only reverse. :)

Im sure this has been discussed before and Im sure there is a good reason for not changing...
 
Axe EDIT III runs super slowly once the CPU gets a high load on it. I havent tried using the front panel when this happens but im sure it isnt slow.
 
Same exact thing happened to me. All of a sudden the AXE FXIII would not connect with the PC. Rebooting did not help at all.
So I reinstalled the USB driver, rebooted the PC and they connected.
I'm assuming somehow the driver got corrupted, and it may have been because my laptop updated to the latest windows 10 the day before. So my suggestion is, if this happens to you, reinstall the computer's USB driver and reboot.
 
yek said:
I seriously doubt that a CPU-overloading preset can "damage" the USB driver for Windows.

Also, the manual mentions 90% as being the limit.
Click to expand...


All I know is the driver was working perfectly and I made that preset and the driver stopped working after the Axe and Axe edit froze. I rebooted the computer and the Axe too. Facts are facts. I won't make another preset that big. I was trying to make an all in one preset and learned a big lesson. Never go over 89 percent on a preset. Also Axe Edit went to a crawl and take forever to change anything before it quit. As soon as I realized the driver was trashed, I reloaded the driver, delete several modules in the preset and all is great again. I sequence a lot too and every once in a while my midi driver will get trashed. It happens when something happens to the driver that there is no code to cover.
 
Last edited:
guitarcal said:
I've always wondered why FAS has yet to change this. My brother has an FX8 and runs into this all the time as well.
Can FAS modify to show 100% cpu utilization when the system wide cpu utilization is only at 90%.

Think of it as the Spinal Tap "these go to 11" situation, only reverse. :)

Im sure this has been discussed before and Im sure there is a good reason for not changing...
Click to expand...


I had to take two programming classes in College for my degree and it would not take too much code to show 90% as 100 percent I would not think. Keeping everything linear would be more code. There may be more variables than we know to change this. Yek, said the manuals says 90% is the max. I don't remember reading it, but I am sure it is there. I don't read a whole manual, just the parts when there is something that I don't know how to do. You have some that will read every word. I am just not one of those guy. I got burned here because of it. It is no big deal. Just a driver replacement and to narrow the preset 10%.
 
It does seem more logical to call it 100%... then there wouldn't be all these posts and possible support calls. I also get that there's some leeway in there - some combinations of fx can probably handle a few more % than others just due to how it all threads in the processor; and for example we know that the Drive blocks vary their cpu usage depending on how hard they're being played, at least that was the case on the AFXII so I'd guess it's that way on the III as well? (I would suggest they just calculate it as if it's always at the upper limit). But yea, having a set limit # at 100 would probably cause less confusion.
 
@josephyballew Mac or Windows? That you had to reinstall the driver is troubling and warrants investigation. Should be easy enough to reproduce if we know what OS you're working with.
 
I highly doubt there's any driver damage occurring. That would be.... ahem.... quite a feat, which you could probably present at DEFCON.

The next time it happens try a power off/on cycle (not a reboot). I bet it's fine after that without re-installing any driver.
Your PC's USB controller is probably in a bad state - something which I've been occasionally experiencing on mine while the Axe-Fx III is on - in my case only a power cycle gets it unstuck (again not a reboot).
 
My car’s redline is 8000rpm, it can rev a little higher before it hits the limiter but it’s not recommended. I usually shift around 6-7k rpm.

I would not want them to relable my tachometer 1000rpm lower just so I can think I’m shifting at 8000rpm when it’s really only 7k.

I also don’t want FAS to “dumb down” the cpu meter 10% becasue they don’t think I can utilize it on my own. Give me an accurate account of what it’s running, and give me the “safe” values, so I as the operator am in control of how I create my patches.

If I want 93% and have been informed to keep it below 90 that’s on me.
 
AlbertA said:
I highly doubt there's any driver damage occurring. That would be.... ahem.... quite a feat, which you could probably present at DEFCON.

The next time it happens try a power off/on cycle (not a reboot). I bet it's fine after that without re-installing any driver.
Your PC's USB controller is probably in a bad state - something which I've been occasionally experiencing on mine while the Axe-Fx III is on - in my case only a power cycle gets it unstuck (again not a reboot).
Click to expand...
If you are referring to my post, you would lose that bet. I shut everything down entirely 3 times, waiting in between. Then powered back up with no luck. AXE Edit would not talk to the III. The driver was definitely damaged and had to be reinstalled. But being a computer tech since DOS 1.0, I blamed it on the recent Windows 10 upgrade.
 
lqdsnddist said:
My car’s redline is 8000rpm, it can rev a little higher before it hits the limiter but it’s not recommended. I usually shift around 6-7k rpm. I would not want them to relable my tachometer 1000rpm lower just so I can think I’m shifting at 8000rpm when it’s really only 7k.
Click to expand...

Interesting perspective! Most vehicles will have a rev limiter though which will prevent you from exceeding the red line, but I get what your driving at (pun intended! :)).

Maybe this would be the solution, programmatically add a "rev limiter". If you select an effect that would take you over the 90%, display an error message instead of setting the effect. Just a thought...

I believe its a poor user experience if you allow the effect to be chosen, which then in turn renders the axe fx unusable...
 
Last edited:
I'm interested to know what kind of presets on Axe III are running short on CPU?

Also, there must be a difference in architecture that causes AxeEdit operation to be affected by cpu level on the Axefx3 - this does not occur on Axe2 where Axe Edit will continue to operate normally when Axefx2 CPU is high (ie I can remove / adjust blocks in AxeEdit to get cpu back in line without AxeEdit taking a long time to respond). I'm interested to know what causes this effect on Axe3.
 
You must log in or register to reply here.
Back
Top Bottom