Axe-Fx II "Quantum" 2.00 Public Beta Deux

scook said:
I have a question that is sincerely not meant as a challenge. Let's assume for the sake of argument that 99/100 people could not tell the difference between a real world recorded amp and that amp's equivalent Q1.06 modeled counterpart. That would mean that only 1 out of 100 people can hear the difference, and the "gap" between real and modeled is incredibly small.

Now, let's also assume for the sake of argument that Q2.0 does indeed narrow that "gap." Let's say it even closes that gap! Wouldn't it stand to reason that only the 1 person out of 100 who could hear the difference at F/W Q1.06 could hear the improvements in the jump from 1.06 to 2.0 (the narrowed "gap")?

How is that everyone can tell a difference when 1.06 was already insanely close to the real thing-- close enough to fool 99/100?

These are theoretical numbers, so feel free to invalidate my assumptions. I'm just wondering what all of your thoughts are.
Click to expand...

The problem is your question is based on an unproven assumption, in my opinion.
 
Jason Scott said:
The problem is your question is based on an unproven assumption, in my opinion.
Click to expand...
While you are 100% correct, I have a hard time believing that most of the people who are hearing a difference (and make no mistake, I believe them) could pick out a "real" tube amp in a blind A/B test with Axe-FX on Q1.06. Perhaps I'm just wrong about that. It just seems like the "gap" was already incredibly small, and it makes me wonder why it's so easy for people to hear the improvements to the point of calling them "major."
 
You don't have to be an audiologist to detect qualitative differences in an amp model you're familiar with, in my opinion.
 
You are right, we tend to overuse the word major because we all get excited and it blows are minds that it still can get better. ( I have not said major this go round). With these more recent firmware's its been more about feel than anything else. G3 and Quantum in my opinion were major updates though. Before that I'd say 13 (ultrares) 10 or 11 (Mimic?)...etc. The gap is very much small but I suppose back when we used to say the gap was 95% in the ultra days we were in fact much more off. Are we at 95% now? Maybe. Now its just not about AB tests, its about feel mostly and subtle things that we don't realize we're missing until cliff again updates this awesome modeler.
 
It's simple: Cliff should release an entirely new version, maybe Q3, which is actually exactly like Q2.**. I bet that many will hear an improvement! There's nothing wrong with this, it's just how our brain works. I know that this would work for me aswell, simply because I did my own "tests". I loaded a new firmware (a while ago, maybe from 18 to 19 or so) and immediately found it better than the previous release. But as soon as I took a closer look I realized that I loaded the wrong file, it was actually the firmware that was currently loaded, so nothing changed :D Then of course, when I loaded the real firmware 19, I found this one even betterer :D

I'm not saying that this applies for everybody, but I KNOW that this placebo effect exists.


And there's something else: Feeling! When the first beta of Q1 was released, I first tried it over headphones. I could hear an improvement in the low frequency area, but didn't think it was so big. But when I played through my guitar cab, the difference was huge! So what I'm saying is, it's possible that if you can't really HEAR a difference, you could still FEEL it.
 
scook said:
While you are 100% correct, I have a hard time believing that most of the people who are hearing a difference (and make no mistake, I believe them) could pick out a "real" tube amp in a blind A/B test with Axe-FX on Q1.06. Perhaps I'm just wrong about that. It just seems like the "gap" was already incredibly small, and it makes me wonder why it's so easy for people to hear the improvements to the point of calling them "major."
Click to expand...

It's just people getting super hyped about the unit, man. Not much more. Like I said, many people are probably just hearing the amps louder as there is a volume increase with this fw, and they are assuming "huge" differences.

People have been saying this unit sounds 100% since fw one. People also say they can't hear the difference between a mic on a cab and an IR in the axe. People also say they can't hear the difference between long and short IRs. People can't hear the difference between a direct recorded axe and a pro studio recorded live amp, but somehow... They hear massive differences when there's an update. But it was 100% before...

It's just people getting hyped up.
 
FractalAudio said:
The previous firmwares were based on theoretical tube models. One thing I began to notice is that the theoretical models simply weren't agreeing with measurements. So after much research we realized that the theoretical models are just that, theoretical and that real tubes do not behave exactly as theory predicts. I can't reveal the details of what aspects don't agree and why those aspects are important.

Q2 is based on the actual tube curves which gives a different character to the distortion and a more bouncy and responsive feel.
Click to expand...


I must begin saying thank you very much for all your work and effort to offer us the best experience with Axe FX.

Cliff, have you considered real tubes statistical sampling variability when changing modeling approach? did you measured just one or two tubes and used the results, or did you get 10 or more (including different brands) and built a mean (standard deviation and so)?

What about the other electronic devices? I remember when I was at college labs, RC filters and transistor circuits never behave as the books predicted :) frequency response curves never were smooth curves and the parameters never were the same as the nominal values - in some case they were out of the tolerance range (example: a 470ohm +/- 2% resistor was in fact 450ohm). Maybe you should compare theoretical models of resistors, capacitors, inductors, etc, to the real ones too, as you did with tubes.

Another aspect related to this is the changes that occur in these devices parameters when time passes by.

I beg your pardon for my not so good english.
 
Sonic differences between firmware versions is quantifiable. Degree of qualitative differences, however, is quite subjective. Major to one may be minor to another.
 
I mean, I believe there are improvements, and I'm not saying people are being disingenuous. It must come down to feel. If you simply recorded and normalized the same loop through each firmware, fewer people could probably hear the differences. But feel is where you and you alone, as the player, know exactly what's being "input," so you know how well the amp is responding. Perhaps 1.06 was 99% "there" when listening to a recording of someone else playing, but less than 99% "there" in terms of amp response to your direct, sensory input.
 
FractalAudio said:
If you're a medium gain person and like Class-A amps, i.e. AC-30, then you should hear decent improvements as there is an entire separate algorithm for those types of amps now.
Click to expand...
One of my two main amps is the AC20 DLX so you really have me interested in hearing this!!!
 
scook said:
I mean, I believe there are improvements, and I'm not saying people are being disingenuous. It must come down to feel. If you simply recorded and normalized the same loop through each firmware, fewer people could probably hear the differences. But feel is where you and you alone, as the player, know exactly what's being "input," so you know how well the amp is responding. Perhaps 1.06 was 99% "there" when listening to a recording of someone else playing, but less than 99% "there" in terms of amp response to your direct, sensory input.
Click to expand...

In my opinion, there's a difference between claiming to hear major leaps in authenticity vs. qualitative sonic differences in an amp model you're familiar with.
 
scook said:
How is that everyone can tell a difference when 1.06 was already insanely close to the real thing-- close enough to fool 99/100?

These are theoretical numbers, so feel free to invalidate my assumptions. I'm just wondering what all of your thoughts are.
Click to expand...
Actually they're made up numbers, a theoretical number would be based on some kind of line of specific reasoning or empirical but not fully quantified observation. These are simply pulled out of the air. So, you answered your own question... this is self-selecting data. For the most part, people are only going to comment on the FW if it is a big deal to them. And also I would wager (based on the number of posters in this thread versus how many FAS units must have been sold over their time in business and apparent sales activity) the number of AFX users who regularly come here and obsess over FW updates and their delta effects, and also post, is minuscule. So saying 'everyone' claims to hear differences in FW is simply baseless. Only people who post in this thread claiming to hear differences are the only ones you know for sure to say this and you have no idea how many users that constitutes as a percentage of the population.

FWIW, I have been playing Q2 for two days and have not noticed anything significant in my core set of presets. That is a good thing because they rocked. You could have loaded it without telling me and I would have only noticed a few changes on specific amps. i.e. the Nitrous.

Across 200+ amps a modeling change is very likely only going to majorly affect some small number of those (at this point in the algos evolution). My tones are not relying on power amp distortion for the most part. They have some and are enhanced by it, but it does not dominate. Others are all about it. Others are a mix. I have clean and mild crunch patches but I do not listen and tune them obsessively like I do my my bread and butter face melters.

Some amps are class A, some are AB, etc. Some are more subject to some secondary affect based on the topology, etc. If your gotos all get changed by a FW it seems huge. If your mainstays do not rely on the thing that changed or did not get one of the bug fixes a lot of what you're hearing is likely placebo or you really don't notice much diff (like me with this FW). Also, for the most part, I would assume most do not have 2 AFXs side by side. This is going to make the chances of meaningful observation from ANYONE much lower. I'm sure it is natural for many to load a FW spin through everything and say to themselves, "what's different? what's different?" well your head is going to come up with an answer whether it is meaningful or not.

IMHO.

That said, new amps and bug fixes are awesome. And much respect to FAS for continuing to pursue perfection even if the improvements sometimes fail to blow away my uncultured ears/tastes. I have huge faith in the methods and abilities in use here.
 
Last edited:
scook said:
I mean, I believe there are improvements, and I'm not saying people are being disingenuous. It must come down to feel. If you simply recorded and normalized the same loop through each firmware, fewer people could probably hear the differences. But feel is where you and you alone, as the player, know exactly what's being "input," so you know how well the amp is responding. Perhaps 1.06 was 99% "there" when listening to a recording of someone else playing, but less than 99% "there" in terms of amp response to your direct, sensory input.
Click to expand...

I don't think people are disingenuous. I think people are mostly just totally full of it. It doesn't help when fw changes aren't normalized and things get louder. People will automatically hear "improvements" when they load it up. Every time.

The axe fx 2 is good gear, and while it's a class gear, it's not an amp. It never will be. Twenty five or so fw haven't made direct tones sound like a truly well-recorded amp, and twenty five more aren't going to either. You can't have a system that apparently sounds exactly like the real thing and still have drastic improvements. It's contrary to basic logic.
 
Jason Scott said:
In my opinion, there's a difference between claiming to hear major leaps in authenticity vs. qualitative sonic differences in an amp model you're familiar with.
Click to expand...
Well... sure, obviously. But the claim isn't just that there are qualitative sonic differences, but that there is an increase in authenticity. So I'm not sure what your point is.
 
scook said:
Well... sure, obviously. But the claim isn't just that there are qualitative sonic differences, but that there is an increase in authenticity. So I'm not sure what your point is.
Click to expand...

After reading through this thread, I don't see most participants claiming major improvements in modeling authenticity, however. Some, but not most.
 
Further, it's been assumed that the people who claimed that the modeling was 99% authentic a year ago are the same people who are now claiming that the new update has introduced major improvements in authenticity.
 
zenaxe said:
Actually they're made up numbers, a theoretical number would be based on some kind of line of specific reasoning or empirical but not fully quantified observation. These are simply pulled out of the air. So, you answered your own question... this is self-selecting data. For the most part, people are only going to comment on the FW if it is a big deal to them. And also I would wager (based on the number of posters in this thread versus how many FAS units must have been sold over their time in business and apparent sales activity) the number of AFX users who regularly come here and obsess over FW updates and their delta effects, and also post, is minuscule. So saying 'everyone' claims to hear differences in FW is simply baseless. Only people who post in this thread claiming to hear differences are the only ones you know for sure to say this and you have no idea how many users that constitutes as a percentage of the population.

FWIW, I have been playing Q2 for two days and have not noticed anything significant in my core set of presets. That is a good thing because they rocked. You could have loaded it without telling me and I would have only noticed a few changes on specific amps. i.e. the Nitrous.

Across 200+ amps a modeling change is very likely only going to majorly affect some small number of those (at this point in the algos evolution). My tones are not relying on power amp distortion for the most part. They have some and are enhanced by it, but it does not dominate. Others are all about it. Others are a mix. I have clean and mild crunch patches but I do not listen and tune them obsessively like I do my my bread and butter face melters.

Some amps are class A, some are AB, etc. Some are more subject to some secondary affect based on the topology, etc. If your gotos all get changed by a FW it seems huge. If your mainstays do not rely on the thing that changed or did not get one of the bug fixes a lot of what you're hearing is likely placebo or you really don't notice much diff (like me with this FW). Also, for the most part, I would assume most do not have 2 AFXs side by side. This is going to make the chances of meaningful observation from ANYONE much lower. I'm sure it is natural for many to load a FW spin through everything and say to themselves, "what's different? what's different?" well your head is going to come up with an answer whether it is meaningful or not.

IMHO.

That said, new amps and bug fixes are awesome. And much respect to FAS for continuing to pursue perfection even if the improvements sometimes fail to blow away my uncultured ears/tastes. I have huge faith in the methods and abilities in use here.
Click to expand...
There are A/B tests between real amps and modelers all over the forums, and most people can't consistently tell the difference, especially with the top of the line modelers like Axe or Kemper. Yeah, it's not a perfect sample, but it's not completely meaningless. There are also numerous, albeit anecdotal, examples of producers tricking guitarists into using a modeler when they think they're playing an amp. I don't think I'm making a baseless claim when I say the modeling was already really darn close to being real. And when a gap is very small relative to the whole, changes in that gap are harder to perceive. I certainly take your point, but I don't think I'm that far off.
 
You must log in or register to reply here.
Back
Top Bottom