Axe-Fx III Firmware 31.00 Public Beta #2

[sebastos08]

Hello everyone. I've been building a complete cab library for several months now, with realistic cab names. I also use the "global block" function in "dynacab" mode to link a global block to several presets. Apparently, the cab names/numbers linked to my old libraries are no longer correct, and some cab names in slots cab1 and 2 (using the global block linked function) are incorrect too.
I don't know if you're experiencing this as well, and if it could be due to the latest firmware updates?

There's definitely something wrong with the "linked global blocks" cabs too; The cabs numbers aren't automatically and correctly recalled when previous presets are loaded. When I reload the global block manually via Axedit, it works...
It seems the FAS team needs to investigate this further.

 

[BROCKSTAR]

Most likely because it needs the new editor to match? This has been mentioned in here already that a new editor will come out eventually.

 

[sebastos08]

Most likely because it needs the new editor to match? This has been mentioned in here already that a new editor will come out eventually.

I don't think it's related to Axedit version since the behavior is identical from the front panel

 

[yek]

Sometimes a change a firmware can cause a Global block to lose the link. Unlink and link again to restore the link.

 

[biggness]

Is there a listing anywhere of all of them, or at least of the the makers?

(Personally I'm not that interested in using them due to workflow considerations, but I'm curious.)

https://shop.fractalaudio.com/collections/software-and-irs-dynacab-packs

Yes, that’s them. I was mainly thinking of the York Audio Mesa cab, the one I’m mainly interested in, and thought that there were more 3rd party options.

Would love to tinker with all those mic options with the new Auto-Delay feature, but like you, Dave, find the workflow a bit clunky with only being able to tweak them with Cab Lab.

Wish there was some solution to baking them in the firmware.

 

[FractalAudio]

There's definitely something wrong with the "linked global blocks" cabs too; The cabs numbers aren't automatically and correctly recalled when previous presets are loaded. When I reload the global block manually via Axedit, it works...
It seems the FAS team needs to investigate this further.

Global blocks are not automatically updated. You'll need to manually update them.

 

[sebastos08]

Global blocks are not automatically updated. You'll need to manually update them.

Thanks for the feedback, it's good to know. So I'll have to redo my entire Marshall presets series and more.
Is it the same for library blocks?

 

[deathbyguitar]

Yes, that’s them. I was mainly thinking of the York Audio Mesa cab, the one I’m mainly interested in, and thought that there were more 3rd party options.

Would love to tinker with all those mic options with the new Auto-Delay feature, but like you, Dave, find the workflow a bit clunky with only being able to tweak them with Cab Lab.

Wish there was some solution to baking them in the firmware.

The York Mesa one is the only thing I use.. It's glorious, especially with the Recto Straight SIC

 

[RockYoWorld]

This is a part of the discussion I'm still really not clear on. Afaiu, all the Dynacab selections are individual mic captures available for us to mix. So how can the phase differences between different types of microphones be already baked in? (it can't know what we want to mix with what). Unless this point is just related to individual mic captures having leading silence removed to make them more user mixable without phase issues? What am I missing?

What I'm about to say comes from a lot of assumptions in how the system is designed and how it was measured, so I could very be wrong, but I at least want to explain how I understand it.

I believe the DynaCabs are made by essentially capturing an IR of a cab with a specific microphone in a bunch of locations in 3D space. You have the radial distance from the center of the cone and the distance from the cab/cone. There must be some sort of interpolation between data points so we can simulate a specific position, even if there wasn't a capture in that exact location. Once again, that's my understanding, but I could be wrong.

Taking my assumption as truth, then in the process of capturing the dynacab, having a dynamic mic and a condenser mic in the same exact position will result in the differing phase conditions that you would get if you really used the real microphones in those positions because the model is derived from actual measurements. This is true IF there is no automatic alignment being used, which is the whole topic being discussed.

The phase differences are baked in because , while capturing the dynacab, as you move the microphone away from the cab, there is a delay as the sound waves get to the microphone. The model must capture some sort of transfer function of [Electrical signal coming from microphone / electrical signal going into the cab]. That signal coming out of the microphone has the delay/phase baked into what the microphone is seeing.

That can be corrected with phase/timing alignment, which used to be ON by default but will be OFF going forward, which must be a decision that was made to be align with the usual goal for Fractal being as accurate as possible by default while letting us tweak parameters to get "ideal" sounds that aren't (easily) possible with the "real thing."

 

[jayjay347912]

Just amazing how many changes, additions and improvements this device has undergone during my 1.5 years of ownership.

And I didn't need to "unlock" them or buy them through micro transactions, or enable telemetry or even give you access to my personal data in return...

Just plain effing awesome.

We purchased around the same time frame, and I get you on the micro- transactions. Games aren't the same anymore...I do play Battlefield 6...but they will never see any of my $$ for some dumb skin. My son knows better too The Axe Fx 3mk2 still blows by mind, from the feel, to the forums. We spent a lot of money on these units, and to have the buyers remorse disappear at first play, AND see the creator still improve and engage almost daily?! It's something that I've never seen before, but it should be studied and used as an example of excellence in business and product support

 

[Wired]

I’m getting some really pleasing results moving mics around with Auto Delay off. I quite like this realistic approach.
The amps sound and feel excellent.

 

[Bones86]

Could there be a better company around? I really really mean that. Thank you so much this unit just keeps refreshing itself with awesomeness and how much did I pay for it again??

Rock on good folks.

 

[DLC86]

Taking my assumption as truth, then in the process of capturing the dynacab, having a dynamic mic and a condenser mic in the same exact position will result in the differing phase conditions that you would get if you really used the real microphones in those positions because the model is derived from actual measurements. This is true IF there is no automatic alignment being used, which is the whole topic being discussed.

The phase differences are baked in because , while capturing the dynacab, as you move the microphone away from the cab, there is a delay as the sound waves get to the microphone. The model must capture some sort of transfer function of [Electrical signal coming from microphone / electrical signal going into the cab]. That signal coming out of the microphone has the delay/phase baked into what the microphone is seeing.

That can be corrected with phase/timing alignment, which used to be ON by default but will be OFF going forward, which must be a decision that was made to be align with the usual goal for Fractal being as accurate as possible by default while letting us tweak parameters to get "ideal" sounds that aren't (easily) possible with the "real thing."

I think you're making some confusion between phase and time. Phase information is stored in the IR but has nothing to do with the auto delay/alignment in the Axe FX. You don't even need to store the actual delay of each IR as it's pretty easy to calculate it based on mic distance and it would also be a waste of storage and processing power... or loss of resolution in the IR if it has a fixed length (like in the Axe FX)

 

[unix-guy]

Thanks for the feedback, it's good to know. So I'll have to redo my entire Marshall presets series and more.
Is it the same for library blocks?

Only presets are updated.

Global blocks are not stored in the preset and neither are library blocks.

 

[rsvette12]

I’m getting some really pleasing results moving mics around with Auto Delay off. I quite like this realistic approach.
The amps sound and feel excellent.

Auto delay switch is on the unit itself correct ? Axedit is not updated so I will have to wait - I dont use on unit adjustments if at all possible

 

[Wired]

Auto delay switch is on the unit itself correct ? Axedit is not updated so I will have to wait - I dont use on unit adjustments if at all possible

Correct. Front panel.

 

[RockYoWorld]

I think you're making some confusion between phase and time. Phase information is stored in the IR but has nothing to do with the auto delay/alignment in the Axe FX. You don't even need to store the actual delay of each IR as it's pretty easy to calculate it based on mic distance and it would also be a waste of storage and processing power... or loss of resolution in the IR if it has a fixed length (like in the Axe FX)

In this case, they're the same thing. The microphone's phase is also captured as a function of distance. That will be captured in the data points collected for the DynaCab.

 

[austinbuddy]

you can kind of do this align thing now in the cab block with legacy ir's in the align tab by adding a slight delay to one of the cabs in a pair. both have to be panned centre, of course. it can be very useful in taming bass and harsh frequencies. it was cliff's "secret sauce" for live use when it was introduced way back. it's a very useful tool for tonal manipulation.

Yep. I used this with some larger values (distance between two mics on cans) in teh Fall 2025 updated Brit Rock Royalty to get some of that Zep/style Page/style amp tone studio magic.

Especially if one of the IRs is a room
mic blended in. Pan that off hard to one side with other close mic speaker other side or middle. Presto. Love that can be replicated already in the III already. Looking forward to digging in to this new Fw once the editor catches up!

 

[DLC86]

In this case, they're the same thing. The microphone's phase is also captured as a function of distance. That will be captured in the data points collected for the DynaCab.

Again, nope. Phase rotation and delay can be considered "the same thing" only when the signal contains a single frequency (i.e. a steady sine wave). Example: you have a 100 Hz sine wave and delaying it by 5ms (half cycle) would be the equivalent of a 180° phase shift.
But for a 50 Hz sine wave, that same 5ms delay would be 1/4 of a cycle, thus the equivalent of a 90° phase shift.
And from this example I think you can already tell why, for IRs, phase and delay can't be considered the same thing... at all.
Delay is the only thing that changes linearly (and predictably) with the distance from the source, phase does not.

 

[Corinthian]

There must be some sort of interpolation between data points so we can simulate a specific position, even if there wasn't a capture in that exact location. Once again, that's my understanding, but I could be wrong.

This is how other brands like TwoNotes does it but, IIRC, Cliff stated when Dynacabs were launched that they are created by actually capturing each possible position (to what resolution I don’t know but one Dynacab file has 3,000+ IRs).