Fractal Audio Firmware Update Thread

Maybe I’m not getting it, but choosing a mic and moving it around been around for a decade (Two Notes Wall of Sound). Why is this such a big deal?
 
hw_impulse.jpg
 
Maybe I’m not getting it, but choosing a mic and moving it around been around for a decade (Two Notes Wall of Sound). Why is this such a big deal?

Not revolutionary, but a very nice implementation of new features for the Axe-Fx.
And even if you regard that as "meh", just realize that as an Axe-Fx user, you're getting loads of new IRs for free.
Somehow Cliff has found a way to fit this into the available memory.
 
Last edited:
Being able to mix together 4 cabs on the Axe-Fx 3 actually puts it ahead of the capabilities in other stuff, except ML Sound Lab MIKKO plugin (which can mix up to 9 mics/cabs).

I hope they bring in 4x10 Super Reverb and Bassman cabs for final release or a future update tho!
 
I've never found any need to have a higher spatial sampling frequency in the mic locations. That seems like a strange thing to chase at the expense of more mic options, which provide a lot more range of tones than being able to move the mic 0.05" instead of 0.1"
The spatial resolution of the product we're supposedly copying is around 1.7" for a 12" speaker. Everything in between is interpolated and my studies show this is insufficient to accurately reproduce those in-between points.
 
The spatial resolution of the product we're supposedly copying is around 1.7" for a 12" speaker. Everything in between is interpolated and my studies show this is insufficient to accurately reproduce those in-between points.

^^^ Genuinely not sure what you're getting at with this ?

I *presume* you are referring to the Helix FW 3.5 Release ?

If not .. please *ignore* the following :)

If yes ..... for what its worth ....D.I stated when 3.5 was released - and it was also included in the early drafts of the 3.5 Release Notes - that each individual "adjustment increment" of each Mic, Placement, On-Axis, Off-Axis, Distance-In-Out-Sideways etc... etc.... etc.... is its own single-unique IR shot .... the 3.5 IR shots don't have any "in-between" points ..... the 3.5 adjustment increments are not as detailed / micro-adjustable as the current Axe Beta, but they are each a single-actual-individual IR shot ..... not sure though if this is or isn't what you are referring to ?

Ben
 
^^^ Genuinely not sure what you're getting at with this ?

I *presume* you are referring to the Helix FW 3.5 Release ?

If not .. please *ignore* the following :)

If yes ..... for what its worth ....D.I stated when 3.5 was released - and it was also included in the early drafts of the 3.5 Release Notes - that each individual "adjustment increment" of each Mic, Placement, On-Axis, Off-Axis, Distance-In-Out-Sideways etc... etc.... etc.... is its own single-unique IR shot .... the 3.5 IR shots don't have any "in-between" points ..... the 3.5 adjustment increments are not as detailed / micro-adjustable as the current Axe Beta, but they are each a single-actual-individual IR shot ..... not sure though if this is or isn't what you are referring to ?

Ben
I don't know what was claimed. All I know is the math and the measurements.

Look at it this way:
Say a product allows position from 0 to 10 in 0.1 steps and distance from 1" to 12" in 1/4" steps. That's 100 positions times 44 distances = 4400 unique IRs for a single speaker and mic combo.

Now multiply that by 12 mics x 2 angles x 24 speakers = 4400 x 12 x 2 x 24 = 2.53 MILLION(!!!) IRs.

Now multiply that by 1024 samples x 2 bytes/sample = 5.2GB of data.

Do you really think said product is storing 5.2GB of data when it only has a 32MB memory chip?

Or do you think it's more likely that said product stores, say, two dozen IRs per speaker/mic combination which results in about 28MB of data.
 
Last edited:
Oh, and for all the people claiming we "copied" this. We started working on this almost a year ago. These things don't happen overnight. We had to write our own software to control the robot and acquire the data, Matlab processing routines, firmware modules, etc., etc.

That said, it's not a unique idea. I believe ML Sound Labs was the first to do this, or maybe Two Notes. But to give credit where credit is due, the enabling technology is the mic robot. Without that none of this is possible. So the real innovators are the people who created the mic robots.
 
Is there a discussion to be had about whether all of this control (and associated compromises) need to be 100% stored within a HW unit?

This strikes me as something that would make far more sense to run externally on a computer - you can download and audition appropriate cabs and mic positoons, and then from there decide which ones are loaded on to HW (and in what resolution for interpolation). Do all users require all cabs/mics/positions taking up space on their device?

It already feels as though several compromises have had to take place for it to exist in the first place - is there not a better product here that runs on a computer?

I'm already generally just using modellers with no cab at all and running the cab sims on the computer. No importing or exporting IR's, no storage constraints, consistent workflow every time.
 
Is there a discussion to be had about whether all of this control (and associated compromises) need to be 100% stored within a HW unit?

This strikes me as something that would make far more sense to run externally on a computer - you can download and audition appropriate cabs and mic positoons, and then from there decide which ones are loaded on to HW (and in what resolution for interpolation). Do all users require all cabs/mics/positions taking up space on their device?

It already feels as though several compromises have had to take place for it to exist in the first place - is there not a better product here that runs on a computer?

I'm already generally just using modellers with no cab at all and running the cab sims on the computer. No importing or exporting IR's, no storage constraints, consistent workflow every time.
We're already working on that. Should be out soon.
 
Back
Top