Further proof that Fractal blows everything else out of the water

I watched the results new video and the only I disliked was the helix.
By the way, in the new video he redid the kemper's sound because he did a mistake with an OD and it sounds far better.
The AX8 sound is great, I'm devoted to fractal and it'll stay my home for amp and effect modelling :)
 
I am curious if he used an actual tone match block to do this test, or if he just dialed it in to be exactly like the real amp?
 
Light years away from my kind of music but thought AX8 and BIAS were really close. Didnt hear the updated version of Kemper, but in this clip the Kemper blew pretty much...
 
And he turned power amp modeling off in the AX-8. If he had turned it on and used a proper IR it probably would've sounded even better.

Ola is a great player, does great videos and has a great ear for metal tones. However I have a problem with tests like these. They are really only testing a unit's ability to match the EQ which isn't really that great a measure of how well a modeler works. When there's that much gain and you boost with a Tube Screamer the input EQ is basically a bandpass filter and all you're hearing is the output EQ. Pretty much all distortion algorithms sound the same with this much gain. You could use a simple hard clipper, i.e if x>k then x = k.

The real tests of a modeler are things like edge-of-breakup, what's the distortion texture like, what does the distortion sound like as the notes decay, does it clean up the same when you roll off the volume, do the tone controls behave accurately, etc., etc. These are the things that we've literally spent a decade perfecting. Getting the EQ right is easy, the hard part is getting the distortion characteristics. Tube amps have complex, multi-band distortion and (in my obviously biased opinion) nothing matches our algorithms in that regard.

He turned the power amp modeling off on the Helix as well and I don't think this was fair to the Helix.
 
FractalAudio said:
Pretty much all distortion algorithms sound the same with this much gain. You could use a simple hard clipper, i.e if x>k then x = k.

The real tests of a modeler are things like edge-of-breakup, what's the distortion texture like, what does the distortion sound like as the notes decay, does it clean up the same when you roll off the volume, do the tone controls behave accurately, etc., etc. These are the things that we've literally spent a decade perfecting. Getting the EQ right is easy, the hard part is getting the distortion characteristics. Tube amps have complex, multi-band distortion and (in my obviously biased opinion) nothing matches our algorithms in that regard.
Click to expand...

I thought I was the only one thinking "hey! those all sound the same!". FAS gear stands apart the most in the boutique/special amp arena with the captured uniqueness of each amplifier's character. With heavily distorted tones, yeah, we can get those too but it doesn't differentiate from the competition that much. My $.02 about the comparison. Agreed that Ola's videos are helpful but can be questionable on the desired outcome from the comparison.

Kind of you to be that honest about the Helix Cliff-anator!
 
For the AX8, I think it is kind of cool that he used the intrinsic amp model and just dialed it the same as the real amp and got so close.
 
To establish that FAS gear is the best all I have to do is look in my studio. Cue the Jeopardy theme... Woop, there it is. I see an FXII and an AX8. No Helix. Settled.
 
FractalAudio said:
And he turned power amp modeling off in the AX-8. If he had turned it on and used a proper IR it probably would've sounded even better.

Ola is a great player, does great videos and has a great ear for metal tones. However I have a problem with tests like these. They are really only testing a unit's ability to match the EQ which isn't really that great a measure of how well a modeler works. When there's that much gain and you boost with a Tube Screamer the input EQ is basically a bandpass filter and all you're hearing is the output EQ. Pretty much all distortion algorithms sound the same with this much gain. You could use a simple hard clipper, i.e if x>k then x = k.

The real tests of a modeler are things like edge-of-breakup, what's the distortion texture like, what does the distortion sound like as the notes decay, does it clean up the same when you roll off the volume, do the tone controls behave accurately, etc., etc. These are the things that we've literally spent a decade perfecting. Getting the EQ right is easy, the hard part is getting the distortion characteristics. Tube amps have complex, multi-band distortion and (in my obviously biased opinion) nothing matches our algorithms in that regard.

He turned the power amp modeling off on the Helix as well and I don't think this was fair to the Helix.
Click to expand...

The mild distortion, edge of breakup and decays have always separated the men from the boys in the modeling world for me. That's why the earlier ones/PODs/etc always went for the heavy distortion tones and preamp gain models like the Mesas, and they tried to make up for the others in EQ. And they rarely sounded accurate, or even listenable, when rolling the volume knob down. I've waited a long time to hear a decent Vox/Matchless grind and am immensely happy to have this realism in my Axe.
 
Sigh... cue the high gain hate brigade. Seriously, this was a fun exercise, the guy explained what he did clearly, stated it was not meant to be the definitive word, and was complimentary of all the products. He is a well known and accomplished player in the 'metal' field; I don't want hear him playing 75 year old blues licks.

The internet, collectively, is incapable of enjoying anything, apparently.
 
The problem with capturing an IR using the amp's power section is that an IR is linear. So while the power amp's EQ gets "baked in" the nonlinear aspects are not captured. Even at relatively low master volume settings most power amps are distorting to some degree. And with virtually ever tube head ever made it's about getting the amp into the "sweet spot" which is all about getting the right amount of power amp distortion. So if you're just using a preamp model into a "baked" IR you're doing a disservice to the modeler. When you get to the sweet spot the highs are compressed a bit which warms things up. The AX-8 clip sounds exactly like what I would expect if not using the power amp modeling. The highs are not compressed enough. I suspect this is also what's going on with the Helix model. Now the AX-8 clip sounds pretty darn good but if you use a proper IR and the power amp modeling it should be really close.

The reason I'm even commenting on this is because there is a profound lack of knowledge about these sort of things in the community. The more people that understand these things, the better for everyone.
 
FractalAudio said:
The reason I'm even commenting on this is because there is a profound lack of knowledge about these sort of things in the community. The more people that understand these things, the better for everyone.
Click to expand...

Totally agree, and I know for certain I am one of people who needs more help in this area. I think have a reasonable grasp but what would really help me is if one of the experts (or guys who know more than I do) would do a video on it someday. Then whenever someone moans that they cannot get their tone right we can just whack the link to the video to them.
 
Scottckr said:
Haven't seen this posted yet, but this video is a pretty damn good demonstration by Ola Englund of how close to the real thing Axe-Fx and AX8 gets:
Click to expand...


Well, it wasn't very close. And neither was Helix. And, quite surprisingly for me, Kemper was way off. Now, there were reasons for that (power amp modeling turned off, that whole thing with screamer pedal not working as it should into Kemper's input), but the most expensive devices in the comparison did worse that Bias. If by "good" you mean "close to the original", of course.
 
You must log in or register to reply here.
Back
Top Bottom