Will the Axe have profiling?

I love seeing honest pros and cons, and that the top guy is willing to chime in to clear up places where we users don't understand the inventors work, priorities, or intentions. Good show.

They are different. But back to the post, is it that you specifically want the technology or means of getting the result to be exactly the same? Or is it just the result you're after? Not criticizing, sincerely just asking.
 
FractalAudio said:
It now has Tone Matching which is arguably the most important part of profiling.

Profiling consists of four parts:
1. Finding the input EQ.
2. Finding the "shape" and bias point of the nonlinear transfer function.
3. Finding the output EQ.
4. Finding the compression, or sag, characteristics.

The Axe-Fx II with V6 uses a hybrid modeling/profiling approach. The Axe-Fx modeling is much more complex in that it does not use a single waveshaper with adjustable shape and bias point. It uses multiple dynamic nonlinearities including preamp and power amp modeling. Profilers distill everything down to a single static waveshaper and then adjust the shape (probably a x+k/x-k waveshaper) and bias point of that waveshaper to try and match the measured transfer function. They then find the input and output EQ and the compression.

With V6 the Axe-Fx II uses a combination of modeling and profiling. The amps are modeled using our exclusive multiple dynamic nonlinearities that very accurately replicate actual tube triodes and our new power amp modeling which fully recreates the behavior of a tube power amp and output transformer. The models are then refined by applying test tones to the actual amp to find deviations between the real amp and the model. For example, traditional circuit based modeling cannot account for things like parasitics since these are not represented by the schematic. So we now augment our traditional circuit modeling with measurements from the actual amp and store that data in the model.

With Tone Matching you can morph a model to match other amps. This works best if the model and amp are similar.

There are numerous advantages to this approach. These advantages include full control matching. The drive, tone, etc. controls behave just like the real amp. You're not just getting a snapshot of the amp at some setting. Another advantage is full separability of the amp from the cabinet. Due to inseparability of linear responses, profiling lumps the amplifier output with the cab response. The multiple nonlinearities also capture the complexities of amps that rely on both preamp and power amp distortion. Single waveshaper approaches cannot capture the complex interactions of multiple distorting stages and the concomitant duty-cycle modulation, EQ modulation (which produces note bloom and swirl) and the vagaries of feel.

The one disadvantage is that you only get the amps that we have modeled. If you have an amp that is very unique that we haven't modeled then Tone Matching may not fully capture the essence of that amp. Profiling allows you to capture that amp at your favorite settings with your favorite cab. Another disadvantage is that modeling is very labor intensive. We have to enter all the circuit data, measure the control tapers, verify the model accuracy and then apply all the fancy test tones and capture the refinement data.
Click to expand...

Wow. Thanks, Cliff. This is an excellent comparison...really clears up some of the guess work as to what the actual, nitty-gritty differences are. Much appreciated.
 
  1. Record your actual amp in mono, high resolution.
  2. Shoot Custom IR of your own cabinet.
  3. Choose amp model closest in spirit to your actual amp.
  4. Shoot Tone Match Block.
Enjoy.
 
I like the totally control and the superior effects you have afterwards for sure. AxeFx the clear winner in my book for sure.
 
I really believe what Fractal is doing is the best way to go about it, combining modeling with tone matching, I think in the end better results can be achieved this way then just profiling, Axe II is staying, but I probably won't keep my other device much longer.. :)
 
JoshuaLogan said:
It's obviously not the same thing going on with the Kemper, nor have I heard it produce results as impressive sounding. I'd like to though...
Click to expand...

Produce results with accurate testtone like a sinesweep you can get from this little tool: Impulse response deconvolution tool software - Voxengo Deconvolver - Voxengo

Instead of shredding similiar on your guitar, you`ll go perfect and with same level through all frequencies. With sine sweep i`ve made a very accurate match of my real tube amp! Hear and believing: http://forum.fractalaudio.com/ultra...axe-standard-ultra-comparison.html#post627993

Technical vice - i`ve done the same process, like now the TMA block support.... and because of this i wish having a sine sweep as testtone (like it`s there in the cab block for creating own IRs...) for for the matching block.... tried to forumlate this wish in the "wish forum" ...
 
javajunkie said:
Agreed on the differences, the Axe-fx II only does the linear aspects at this point.

However, I don't believe you have enough information to definitively state (at this point) when or if it will have this one day (which was the main question).
Click to expand...

Well, using that point, then no one has said anything of any use on this thread yet because the only one who can answer that question is Cliff. :) Of course we don't know what "one day" will bring. Hell, we don't even know if we're going to wake up tomorrow. All we can do in the meantime is sit around tone matching Justin Bieber's best guitar tones and enjoy the journey. :)
 
First of all thanks to Cliff for "explaining" ;) some bits were a bit difficult not to say not understandable but if I knew how to program axe myself I would probably release some unofficial versions lol ;) but what I was about to say is that today I've turned on my mark IV after two month absence just to check how it compares to v6 and to my ears Mesa sounds digital!haha honestly I prefer Axe II and I never tried new "Fat" feature yet but I'm curious and willing to check it tomorrow!
 
FractalAudio said:
It now has Tone Matching which is arguably the most important part of profiling.

Profiling consists of four parts:
1. Finding the input EQ.
2. Finding the "shape" and bias point of the nonlinear transfer function.
3. Finding the output EQ.
4. Finding the compression, or sag, characteristics.

The Axe-Fx II with V6 uses a hybrid modeling/profiling approach. The Axe-Fx modeling is much more complex in that it does not use a single waveshaper with adjustable shape and bias point. It uses multiple dynamic nonlinearities including preamp and power amp modeling. Profilers distill everything down to a single static waveshaper and then adjust the shape (probably a x+k/x-k waveshaper) and bias point of that waveshaper to try and match the measured transfer function. They then find the input and output EQ and the compression.

With V6 the Axe-Fx II uses a combination of modeling and profiling. The amps are modeled using our exclusive multiple dynamic nonlinearities that very accurately replicate actual tube triodes and our new power amp modeling which fully recreates the behavior of a tube power amp and output transformer. The models are then refined by applying test tones to the actual amp to find deviations between the real amp and the model. For example, traditional circuit based modeling cannot account for things like parasitics since these are not represented by the schematic. So we now augment our traditional circuit modeling with measurements from the actual amp and store that data in the model.

With Tone Matching you can morph a model to match other amps. This works best if the model and amp are similar.

There are numerous advantages to this approach. These advantages include full control matching. The drive, tone, etc. controls behave just like the real amp. You're not just getting a snapshot of the amp at some setting. Another advantage is full separability of the amp from the cabinet. Due to inseparability of linear responses, profiling lumps the amplifier output with the cab response. The multiple nonlinearities also capture the complexities of amps that rely on both preamp and power amp distortion. Single waveshaper approaches cannot capture the complex interactions of multiple distorting stages and the concomitant duty-cycle modulation, EQ modulation (which produces note bloom and swirl) and the vagaries of feel.

The one disadvantage is that you only get the amps that we have modeled. If you have an amp that is very unique that we haven't modeled then Tone Matching may not fully capture the essence of that amp. Profiling allows you to capture that amp at your favorite settings with your favorite cab. Another disadvantage is that modeling is very labor intensive. We have to enter all the circuit data, measure the control tapers, verify the model accuracy and then apply all the fancy test tones and capture the refinement data.
Click to expand...


Great post. Informative and unbiased. I have downloaded 6.0 but am on vacation so have not had a chance to try it. From all accounts it looks like a killer upgrade. Thanks!

I may still have to get a KPA just out of curiosity and to have another toy. Plus it may lead me to ebay some weird amps to try profiing. I am thinking you won't be modeling a Peavey Mace any time soon ;-) Of course the odds are decent that I sell my ultra and get another II.

It does look like you took a giant leap in making dialing in the tones much easier. To me that was the one downside to the axe. Now if I can combine the crazy cool routing and power of the FX, Controllers, filters, envelopes etc with easy to dial amps. . . . .I'll be good to go.

Thanks to all at FAS for an incredible job!
 
peterc52 said:
Hi everyone!

I'm new to this forum, so hello. :D

I've just bought the Kemper to try it out, and it sounds good, but I would really like to try out the Axe-FX II as well, so I ordered one from G66 a little over a week ago and it should be here in about a month, but here is my question:

I love the thought that I can profile my own amps and record with them, and I believe that's the KPA's strenght! But will the Axe-FX have that one day? Have Cliff stated anything?

Best regards Peter from Denmark :p
Click to expand...

In a short answer...it won't. The Kemper can start from zero and model your amp
or any amp. The AXEFX you have to try and find an amp model that is close to the one you are modeling. I've tried to model some of my preamps and it's just not happening.
 
FractalAudio said:
It now has Tone Matching which is arguably the most important part of profiling.

Profiling consists of four parts:
1. Finding the input EQ.
2. Finding the "shape" and bias point of the nonlinear transfer function.
3. Finding the output EQ.
4. Finding the compression, or sag, characteristics.

The Axe-Fx II with V6 uses a hybrid modeling/profiling approach. The Axe-Fx modeling is much more complex in that it does not use a single waveshaper with adjustable shape and bias point. It uses multiple dynamic nonlinearities including preamp and power amp modeling. Profilers distill everything down to a single static waveshaper and then adjust the shape (probably a x+k/x-k waveshaper) and bias point of that waveshaper to try and match the measured transfer function. They then find the input and output EQ and the compression.

With V6 the Axe-Fx II uses a combination of modeling and profiling. The amps are modeled using our exclusive multiple dynamic nonlinearities that very accurately replicate actual tube triodes and our new power amp modeling which fully recreates the behavior of a tube power amp and output transformer. The models are then refined by applying test tones to the actual amp to find deviations between the real amp and the model. For example, traditional circuit based modeling cannot account for things like parasitics since these are not represented by the schematic. So we now augment our traditional circuit modeling with measurements from the actual amp and store that data in the model.

With Tone Matching you can morph a model to match other amps. This works best if the model and amp are similar.

There are numerous advantages to this approach. These advantages include full control matching. The drive, tone, etc. controls behave just like the real amp. You're not just getting a snapshot of the amp at some setting. Another advantage is full separability of the amp from the cabinet. Due to inseparability of linear responses, profiling lumps the amplifier output with the cab response. The multiple nonlinearities also capture the complexities of amps that rely on both preamp and power amp distortion. Single waveshaper approaches cannot capture the complex interactions of multiple distorting stages and the concomitant duty-cycle modulation, EQ modulation (which produces note bloom and swirl) and the vagaries of feel.

The one disadvantage is that you only get the amps that we have modeled. If you have an amp that is very unique that we haven't modeled then Tone Matching may not fully capture the essence of that amp. Profiling allows you to capture that amp at your favorite settings with your favorite cab. Another disadvantage is that modeling is very labor intensive. We have to enter all the circuit data, measure the control tapers, verify the model accuracy and then apply all the fancy test tones and capture the refinement data.
Click to expand...

This is why I've never understood the frenzy some people have for "profiling" - judging by the reactions of people before the KPA was even released, profiling is somehow able to magically recreate your amp in perfect detail, far more accurately than the mere "modeling" in the Axe, and it can do all this in record time. I've even seen people considering getting both a KPA and an Axe, but using the Axe "just for effects", as if that were the one point the KPA couldn't match.

Profiling just does what Cliff and his team do for each of the amp models in the Axe, except that (a) you can do it yourself in a matter of minutes and (b) the results of an algorithm crunching data for a minute or two, based on a simple series of test tones run through amp, cab, and microphone as a black box system, and then applied to a generic waveshaper-based model, are pretty much guaranteed to be less sophisticated, less flexible, and less exacting than what Cliff does manually. It can't possibly account for all the dynamic interactions of a real system, can't possibly be more than a rough approximation of the behavior of the multi-dimensional, nonlinear space represented by all of the different settings an amp is capable of. That's not conjecture, that's the laws of physics. If you think it's just a calculation problem, you don't even begin to understand what's actually involved.

Profiling - or the EQ matching part of it at least - is convenient, especially if you're having difficulty dialing in an EQ curve using your ears, and it's an effective way to get reasonably close results without doing all the gruntwork that Cliff does for his models, but it baffles me that anyone really believes it's a superior approach than what the Axe already does. Christoph Kemper is a smart guy, but he's not a visitor from the eleventh dimension. Profiling is a shortcut to the ballpark, not the yellow brick road.
 
chase said:
This is why I've never understood the frenzy some people have for "profiling" - judging by the reactions of people before the KPA was even released, profiling is somehow able to magically recreate your amp in perfect detail, far more accurately than the mere "modeling" in the Axe, and it can do all this in record time. I've even seen people considering getting both a KPA and an Axe, but using the Axe "just for effects", as if that were the one point the KPA couldn't match.

Profiling just does what Cliff and his team do for each of the amp models in the Axe, except that (a) you can do it yourself in a matter of minutes and (b) the results of an algorithm crunching data for a minute or two, based on a simple series of test tones run through amp, cab, and microphone as a black box system, and then applied to a generic waveshaper-based model, are pretty much guaranteed to be less sophisticated, less flexible, and less exacting than what Cliff does manually. It can't possibly account for all the dynamic interactions of a real system, can't possibly be more than a rough approximation of the behavior of the multi-dimensional, nonlinear space represented by all of the different settings an amp is capable of. That's not conjecture, that's the laws of physics. If you think it's just a calculation problem, you don't even begin to understand what's actually involved.

Profiling - or the EQ matching part of it at least - is convenient, especially if you're having difficulty dialing in an EQ curve using your ears, and it's an effective way to get reasonably close results without doing all the gruntwork that Cliff does for his models, but it baffles me that anyone really believes it's a superior approach than what the Axe already does. Christoph Kemper is a smart guy, but he's not a visitor from the eleventh dimension. Profiling is a shortcut to the ballpark, not the yellow brick road.
Click to expand...

The Kemper is 'digital alchemy'. (Their words, not mine).

The Axe-FX is the 'magic black box' (My words).
 
You must log in or register to reply here.
Back
Top Bottom