Axe-Fx III Firmware 23.00 Release

Status
Not open for further replies.
Tommy Tempest said:
With regard to cpu usage, here is a screenshot of one of my preset templates. I have gapless on as well.

I don't have any cpu issues om my FXIII Turbo. Not really sure what I'm doing different.

View attachment 129662
Click to expand...

I know you don't have CPU issues and this would affect both firmware versions, but getting rid of all those shunts and moving the INPUT blocks to the end of where the shunts are now may actually save you a couple %. That's a lot of shunts and each shunt does add a small amount of CPU load.

1699983217982.png
 
RockYoWorld said:
I know you don't have CPU issues and this would affect both firmware versions, but getting rid of all those shunts and moving the INPUT blocks to the end of where the shunts are now may actually save you a couple %. That's a lot of shunts and each shunt does add a small amount of CPU load.

View attachment 129669
Click to expand...
About .12% for each shunt, so approximately 3% loss, hardly enough to worry about, in this scenario. I compress the shunts when needed, but like a certain visual layout in Axe Edit, so I don't worry when I have enough CPU space.
But I see your point if I needed more memory.
I had noticed several claiming they have CPU issues. So my post was to show how CPU usage can be streamlined to achieve what's wanted and still not overload the CPU.
 
KyleHowdy said:
View attachment 129606

What a freaking update! We are truly blessed. The Vince McMahon meme doesn't even come close to having enough blank spots to add all the goodness in this update.
Click to expand...
Im with this guy.. Imsure there are those who will get excited abouted the spring reverb for those classic fender vibes but not really my thing. THe other effects, and gapless are huge to me though... Yet, Im here on the fm9 turbo waiting.....
 
No increase in CPU usage here, Axe III MKII (non Turbo).
Maybe it’s just related to specific blocks?
 
Tommy Tempest said:
About .12% for each shunt, so approximately 3% loss, hardly enough to worry about, in this scenario. I compress the shunts when needed, but like a certain visual layout in Axe Edit, so I don't worry when I have enough CPU space.
But I see your point if I needed more memory.
I had noticed several claiming they have CPU issues. So my post was to show how CPU usage can be streamlined to achieve what's wanted and still not overload the CPU.
Click to expand...

Do you need noise gates on inputs 2,3 and 4? Setting the input block noise gate level thresholds to off saves a few percent.
 
Potential bug; swapping reverb channels seems to cause some audio artefacts to appear.

Noticed it on this preset swapping between scenes 4 & 5

Tried with Spillover & gapless off but noise still seems to be there.
 

Attachments

  • Sadade JazzE.syx
    48.2 KB · Views: 6
While we are used to Cliff achieving impossible things, at some point improving the fidelity of amps and effects modules will inevitably have an impact on CPU consumption.
In that sense, I would prioritize the advancement of these quality improvements, to the detriment of the number of modules to be used simultaneously. That is, do we really need to use 8 different Multidelays and 8 plexydelays in the same preset?
It would be a shame if this need for so many modules was a brake on sound improvement.
 
I took note of CPU numbers of a dozen or so presets. From FW23 Release Candidate beta 6 to FW23 I've seen no increase at all. The fluctuation was 3-4% on both firmwares...
 
Last edited:
rcm78 said:
I took note of CPU numbers of a dozen or so presets. From FW23 beta 7 to FW23 I've seen no increase at all. The fluctuation was 3-4% on both firmwares...
Click to expand...

It looks to me like the increase (with gapless switching turned off) occurred between beta 6 and beta 7. This preset goes from 81% (beta 6) to 85% (beta 7). But I don't know, maybe even with gapless switching turned off, an increase in cpu usage is to be expected? Or at least some presets it does?
 

Attachments

  • Skylight.syx
    48.2 KB · Views: 10
I have gone through a few of my custom presets since upgrading an hour ago. So far, I am not seeing any appreciable increases in CPU demands. My presets are normally quite simple and have low CPU demands of around 35% to 50% with my Axe Fx III Turbo. So far, the greatest increase in CPU demand I have seen with gapless switching on has been around 3% over what it was previously. Tomorrow I will go through the rest of my custom presets and a number of the presets I have obtained from others which are more complicated and power hungry. The gapless switching works fantastically well! Thanks to Cliff and all the rest of the team that developed the new firmware and did all the testing!
 
I think it's pretty safe to assume at this point that while maybe something is misconfigured, it's more likely that the little tweaks that were performed between beta 7 and this current final release just increased the CPU usage a tad more, depending on what arrangement of blocks you have in the preset.

At some point the firmware improvements will most likely either have to stop in order to accommodate the Mk1's, or split, because at some point the continued advancements are going to exceed the older Axe3 capability.

Awesome update Fractal.. by far the best company out there.
 
With the new gapless switching, I do think it's a good idea at least to try spreading out kitchen sink presets between a few different presets that do different things, as long as you don't switch between the presets rapidly. Like, if there's a part of a song that requires some crazy and fast tap dancing, you can put on an Eric Johnson workout video and program those particular changes to a single preset. But for anything else, why not spread it around, make it easier on the CPU, and just incorporate preset switches where some of your scene, channel, and effect switches used to be. Just a thought.

I'm having a blast programming songs in ways that are so much easier than they were before gapless channels were introduced. I've said it before, but this was the secret wish I had that I was sure would never be possible in this generation of the hardware. Thank you so much Cliff!
 
Status
Not open for further replies.
Back
Top Bottom