Mk1 FW:17 preset increased CPU usage

Oh, so that's why my kitchen sink preset all of a sudden has combinations of channels that go way over 80 %. I was just in the middle of redoing it and my MMGT22 setup to reduce the CPU usage because of this.
 
AXE-III Mk1 - after FW 17.0 upgrade with Axe Edit The Reading Cab, or blocks when connecting my MacMini and the AXE via USB is very slow ... the 16.05 FW would take about 15 to 20 seconds of reading... now it takes about 2-3 minutes :( , and if I make changes on the axe it self ... the computer displays it much slower then it did before..... the communication via usb seem to be affected.
 
Sorry for delay..I was downgrading to 16.05 and the preset was somehow replaced with one I had before on that slot some time ago, so I remake it again...anyway. After I make it again in 16.05, I save it, and then upgrade to 17 and saved the upgraded version also. Both versions attached. In FW17, CPU is between 82 to 85, with warning mgs appearing from time to time. I'm just afraid that the unit will lag during performance, and no one wants that. I know that this can be somehow fixed by reducing the IR length in cab block from max to lets say 1024, and the CPU usage will stabilize around 80% without the warning message, but that's beside the point...There is definitely an increase of CPU usage after upgrade.
Thanks
Not seeing this here. On 16.05 your preset runs at 81.3%. On 17.00 it runs at 81.5%.
 
I see a rise in CPU load as well.
Please find an example Preset with moderate CPU load attached.

On FW 16.5 it shows a - relatively constant - CPU load ~ 58,5 %.

On FW 17 it shows a varying CPU load ~ 60,5 % .... 64 %.
This preset has been saved for firmware 17.00 so I can't load it in 16.05.
 
Some of the increase is unavoidable due to the "Improved scene switching to prevent signal “leaking” into bypassed blocks." This required adding some code to the input of every block.

I'm working on some optimizations and have already improved things a bit.
 
I tried with my main preset that I use with my band.
Obviously I have turn on/off my Axe3 Mk1. The Axe-Edit version is the same for each firmware (Axe Edit 1.10.00).
DSP usage WITHOUT playing.
16.05 --> 70.5%
17.00 --> 73.2%
16.05.png17.00.png
 
Last edited:
Some of the increase is unavoidable due to the "Improved scene switching to prevent signal “leaking” into bypassed blocks." This required adding some code to the input of every block.

I'm working on some optimizations and have already improved things a bit.
We all enjoy your abilities and your relentless work.
So much fun with your products. :)

Thank you.
 
Firmware 17 is a disaster! Every time you turn the axe unit on it re-reads all the cabs and such like you are doing a refresh. Also, the cpu usage is off the scale on some presets now - full for no reason. Anyhow, must simply be a lemon firmware, it happens on occasion
You're describing an Axe-Edit behavior as a firmware issue... Seems like you are mixing things up.

By the way, the scanning of cabs and presets on startup is normal behavior for Axe-Edit, but you can disable it in Preferences.

Now, if you're saying that it's all the block definitions, that's a different story.

I believe Axe-Edit does that when it detects a new firmware version only...
 
Cliff stated that every block had a little bit of code added to deal with a scenes/bypass artifacts bug and that he's working on further optimizing it. The more blocks you have in the preset, the more increase you'll likely see.
 
I don't see it re-reading the cabs, only re-reading the presets in Axe III Edit, however, the preset re-reading seems to have slowed down quite a bit.
I just booted up Axe-Edit III 10.00.01 and it read presets and cabs, but holy sh*t it was FAST. I mean like 3-5 seconds. Mk I unit.
 
Back
Top Bottom