Axe-Fx III Firmware 23.00 Public Beta #4

Status
Not open for further replies.
You FA guys are heroic.

I’m reading through all this thread. There are a huge number of variables, some with fractal gear and many others with aftermarket elements, and you guys just crunch through it.

All good over here.
 
Here is a clip of what I'm hearing...seems like I can hear it glitch right before it changes....maybe just me? biggrin.gif



 
Rock Chalk said:
Here is a clip of what I'm hearing...seems like I can hear it glitch right before it changes....maybe just me? View attachment 128939



Click to expand...

You could zoom in on the wave form to verify. But that sounds like what I get here. I don't hear the obvious stutter heard in a different example, that was likely caused from having the editor active (it's recommended to pause the editor).

But the gaps are larger with both Amp blocks and the Cab block all changing channels at the same time. If you manually change the channel of any one of them by itself, the change is much quicker. It's an interesting case study/stress test for this new function.

Personally, I am so used to avoiding channel changes on specific high-CPU blocks that I try to find work-arounds using Scene Controllers, Control Switches, and/or other low-CPU block places before or after the Amp block as needed to alter the gain, pre-EQ, and post-EQ. It seems that in some cases, this won't be needed going forward.
 
Last edited:
Moke said:
You could zoom in on the wave form to verify. But that sounds like what I get here. I don't hear the obvious stutter heard in a different example, that was likely caused from having the editor active (it's recommended to pause the editor).

But the gaps are larger with both Amp blocks and the Cab block all changing channels at the same time. If you manually change the channel of any one of them by itself, the change is much quicker. It's an interesting case study/stress test for this new function.

Personally, I am so used to avoiding channel changes on specific high-CPU blocks that I try to find work-arounds using Scene Controllers, Control Switches, and/or other low-CPU block places before or after the Amp block as needed to alter the gain, pre-EQ, and post-EQ. It seems that in some cases, this won't be needed going forward.
Click to expand...
Well that's cool, I just must be hearing the gap.....no big deal as it's not huge, but there none the less.....thanks for checking it out.
 
Rock Chalk said:
Well that's cool, I just must be hearing the gap.....no big deal as it's not huge, but there none the less.....thanks for checking it out.
Click to expand...
Most of the extra gap seems to be coming from the Cab block additionally changing channels. If this was my personal preset, or if a client really wanted this specific preset to switch scenes more smoothly, I would make four new IRs of the current Dyna-Cab and Legacy cab blends. Load them into the first two slots of two stereo two Cab blocks, pan everything to access each of them as needed without needing to change any Cab block channels. As well as some other possible changes depending on the clients specific preferences.
 
Rock Chalk said:
Here is a clip of what I'm hearing...seems like I can hear it glitch right before it changes....maybe just me? View attachment 128939



Click to expand...

Try copying your preset into four preset slots. Setup the four presets to load on consecutive scenes, first copy-scene 1, second copy-scene 2, etc...

Now make sure that 'Gapless Switching' is set to 'ON', and 'Spillover' is set to 'ALL'.

Note that switching between each preset (and consequently between the scenes) is actually faster than switching between the scenes within a single preset. You would think that switching scenes within a preset would be faster than completely reloading another preset. Maybe there is a clue here for Cliff @FractalAudio ??
 
Last edited:
Moke said:
Try copying your preset into four preset slots. Setup the four presets to load on consecutive scenes, first copy-scene 1, second copy-scene 2, etc...

Now make sure that 'Gapless Switching' is set to 'ON', and 'Spillover' is set to 'ALL'.

Note that switching between each preset (and consequently between the scenes) is actually faster than switching between the scenes within a single preset. You would think that switching scenes within a preset would be faster than completely reloading another preset. Maybe there is a clue here for Cliff @FractalAudio ??
Click to expand...
Wow....that cuts the gap way down.......great idea!
 
Rock Chalk said:
Wow....that cuts the gap way down.......great idea!
Click to expand...
I am probably one of the weird ones, but I've always used presets instead of scenes. I just found it easier for me, since I'm old school, and like setting them up like a regular amp. Clean, Crunch, Distortion, Heavy distortion, etc.. Then I setup all my effects accordingly. I am gapless or at least almost.
 
Just watched the Marco Fanton YouTube live and I now really really hope that this will come to the fm3…

The cpu usage used for the gapless preset switching is nothing compared to the advantages of it.
 
Just updated to this beta and think this is a great improvement - thanks FAS!

Having said that, it inspired me to export some of my FM9T presets back to the AXEIII. I am surprised to see that the CPU usage on the AXEIII for each preset is around 10% to 15% higher than the FM9T. I have checked all of my settings and cannot see any reason for the differential. Has anyone else experienced this as up until l now I had expected the opposite to happen i.e. lower CPU on AXEIII? (FM9T is running FW5.01)
 
CoffsterP said:
Just updated to this beta and think this is a great improvement - thanks FAS!

Having said that, it inspired me to export some of my FM9T presets back to the AXEIII. I am surprised to see that the CPU usage on the AXEIII for each preset is around 10% to 15% higher than the FM9T. I have checked all of my settings and cannot see any reason for the differential. Has anyone else experienced this as up until l now I had expected the opposite to happen i.e. lower CPU on AXEIII? (FM9T is running FW5.01)
Click to expand...
CPU allocation is different in FM9, and the dedicated processor for (iirc) reverb means a preset with a busy reverb block will run at a lower CPU percentage than expected....
 
Frodebro said:
Make a duplicate of your preset in each unit, then delete the reverb block from each and see what you have.
Click to expand...
Thanks for the suggestion - the key culprits for the differences are:

Reverb block - uses 8% more CPU on the AXEIII
Dyna Cab block - uses 4% more CPU on the AXEIII

That surprises me as I had assumed that the extra firepower in the AXEIII would always result in lower CPU usage overall.
 
CoffsterP said:
Thanks for the suggestion - the key culprits for the differences are:

Reverb block - uses 8% more CPU on the AXEIII
Dyna Cab block - uses 4% more CPU on the AXEIII

That surprises me as I had assumed that the extra firepower in the AXEIII would always result in lower CPU usage overall.
Click to expand...

You were running the beeta which has new reverb algorithms and other cool enhancements (triodes, gapless, etc) not on the FM9; no?
 
FractalAudio said:
https://www.fractalaudio.com/downlo...xe-fx-3/23p0/axefxiii_dsp_rel_23p00_beta4.zip

Fixes a bug where Output 4 isn't working.

Axe-Fx III Firmware Release Notes​


23.00


Added “Pre-Delay Tap” parameter to Reverb block. This selects the input to the reverb engine. When set to OUTPUT the behavior is as before. When set to INPUT the initial delay is absent. This allows more natural “Echo-Verb” sounds. This is now the default for these types of models and existing presets are automatically updated.
Click to expand...
Hi Cliff, i've just a little report: the pre-delay tap seems to be set to "output" by default when i reset the channel, instead to be set to "input" on your note.?
 
sebastos08 said:
the pre-delay tap seems to be set to "output" by default when i reset the channel, instead to be set to "input" on your note.?
Click to expand...
The release notes say "input" is the default for "these types" but doesn't specify which types. Resetting a channel selects Medium Room type. I assume "output" is the intended default for that.

The types that default to "input" are Echo Hall, Echo Pitch Plate, Echo Room, Fields of Pyrite, Large Deep Hall.
 
Status
Not open for further replies.
Back
Top Bottom