Fractal Talk

[laxu]

Well again.... if people don't want authenticity, why aren't they using a Pod 2.0 and just calling it a day?

I think there are levels in this. POD 2.0 was not there. Neither was the PODXT. Helix is certainly good enough, even though charts & graphs might show it doesn't do something right.

Then you have e.g Helix's custom amps that have no real world counterpart. There the only authenticity level is "it sounds and behaves similar to a good tube amp" but at the same time those models can streamline away things real amp circuits do, because there is no need to emulate them since there's no real world counterpart.

Most users are going to be more than happy when the level reached is "sounds and behaves similar to a good tube amp". Any increased realism is just gravy on top.

 

[MirrorProfiles]

The Rectifier adventures thread shows that Fractal's model matches very closely to their reference amp, but users have raised concerns vs their own amps. Both sides are right.

tbf, that thread revealed a mistake with the MV pot taper, and a bug with how SIC curves are loaded. And when speaker modelling parameters are turned to zero, it should match even closer. In that instance, the product improves and the userbase has a better understanding of how it all works. If we accept “that’s just how it sounds” there is no scope to improve. Whether or not products need improving at all is down to personal interpretation but I’d say that’s an even more vague criteria than whether 2 amps of the same circuit should sound similar.

For my needs, any emulation of an amp needs to be able to match a real amp 1:1, otherwise I’ll just use the real thing. Granted that’s not the case for everyone, but if there is room to make stuff more accurate, who loses out?

 

[Orvillain]

POD 2.0 was not there. Neither was the PODXT

That can't be true. So many records were made with them both. The early 2000's was practically the green light for modelling as an entire pursuit. I reckon with a Pod 2.0 even today, you could still make awesome sounds, and they'd be absolutely fine for putting out music.

So what are we actually after here if it isn't authenticity? Even a lot of your UI complaints with the Fractal boil down to it not being authentic enough to the experience of using a real piece of gear, and the ease of use that comes with it.

 

[laxu]

That can't be true. So many records were made with them both. The early 2000's was practically the green light for modelling as an entire pursuit. I reckon with a Pod 2.0 even today, you could still make awesome sounds, and they'd be absolutely fine for putting out music.

Yes, they were good enough to end up on a lot of tracks, but they weren't a particularly accurate tube amp sim. That aspect has changed over the years. Let's not forget that a Rockman is also on many guitar tracks, or pure DI guitar.

Michael Nielsen actually had a video just recently revisiting the POD 2.0. Add some 3rd party IRs and I think you'd get pretty good results out of it.

So what are we actually after here if it isn't authenticity? Even a lot of your UI complaints with the Fractal boil down to it not being authentic enough to the experience of using a real piece of gear, and the ease of use that comes with it.

When you have a real world amp that the digital model is supposed to represent, authenticity tends to become a factor - up to a point. There's a lot of "I think that is how the real deal sounds, therefore the model is accurate" thinking, and trusting when e.g Fractal says a model is accurate even if there's a bug found later that shows it wasn't. I'll gladly leave developers of the device and some invested individuals to chase that part of it because it can only make it all better, but it does not really change my enjoyment of the gear.

My UI complaints have never been about whether some knob taper of an amp works like the real deal though. Or if the amp looks like the real deal on screen. While real amps and pedals are still the yardstick - dedicated knobs for almost everything that you can just grab and turn - my UI complaints mostly deal with how to get navigation and parameter adjustment of modelers closer to that ideal, however they want to achieve it.

I'd say we are after tones that we enjoy playing. When a piece of gear does that, then you are going to be happy with it regardless of its "authenticity". When Helix 3.7 got released, I gave it a spin in Helix Native and ended up playing the Aristocrat + 4x12 GB cab setup for a good time because I liked that. I didn't think about whether it was authentic, I just enjoyed it as its own thing.

I'd gladly ditch the whole "this is a model of this real amp" paradigm and go for purely custom models so digital modelers simply become another amp option evaluated on its own merits rather than endless accuracy comparisons. But I know I'm definitely in the minority here.

 

[jellodog]

I've often wondered if perhaps you need to get your amp accuracy game near-perfect first, before using that code as a foundation for satisfying novel or idealized amp models.

But perhaps it doesn't really work like that at all.

 

[Orvillain]

My UI complaints have never been about whether some knob taper of an amp works like the real deal though. Or if the amp looks like the real deal on screen. While real amps and pedals are still the yardstick - dedicated knobs for almost everything that you can just grab and turn - my UI complaints mostly deal with how to get navigation and parameter adjustment of modelers closer to that ideal, however they want to achieve it.

Yes. I know. Hence authenticity - not in terms of looks, not in terms of colours or pretty knobs. But in terms of the actual workflow of the real gear.

Like @MirrorProfiles said before; an amp... most of the time you can just plug into it, turn a few knobs, and be up and running.

Thinking about it ... I think modellers have kind of given people these wild expectations for features and possibilities... when even a 4 channel valve amp is a really simple thing to work with compared to this world of Kemper, Fractal, Line6.

I simply don't believe it is possible to create a modeller that gives you the ideals of a valve amp and the possibilities of modelling all at the same time. Spidervalve was probably closest, but as I understand it, it didn't sell that well... and it was still quite a complicated bit of kit.

 

[Orvillain]

I'd say we are after tones that we enjoy playing. When a piece of gear does that, then you are going to be happy with it regardless of its "authenticity".

Well what I can tell you is... in the room pound for pound I still prefer my amps to a modeller going through a poweramp. With the proviso that it gets much closer when I use a valve poweramp, but at which point the benefits of a modeller are more or less lost on me. I don't really care about having 1000's of "preamp" models.

 

[Orvillain]

Yes, they were good enough to end up on a lot of tracks, but they weren't a particularly accurate tube amp sim. That aspect has changed over the years. Let's not forget that a Rockman is also on many guitar tracks, or pure DI guitar.

Yes that is precisely my point. They weren't accurate, but people loved and used them.

Yet we've got people who didn't stick with them, and who jumped through all the hoops over the decades and eventually ended up with an Axe FX III... and who are bemoaning the accuracy goal. I'm simply questioning... if they didn't care about accuracy... why didn't they stick with their Pod 2.0's or the HD500's or their Kemper (lol)

 

[laxu]

I simply don't believe it is possible to create a modeller that gives you the ideals of a valve amp and the possibilities of modelling all at the same time. Spidervalve was probably closest, but as I understand it, it didn't sell that well... and it was still quite a complicated bit of kit.

IMO the Quad Cortex actually got a lot of things right for the basic editing experience for an already built preset. It achieved this through a couple of things:

1. The touchscreen gives you instant access to any block in your preset since they are all on the same screen. No scrolling involved. Just tap a block and there it is.
2. The knob/switches allow for a lot more immediately accessible control. It does take a bit to build a mental map between screen and knob tho.
3. The amount of knobs meant that there are few pages to jump between in any given block. So overall less back and forth navigation required.
4. The knobs had a good feel to them that was closer to knobs on amps or pedals than what you find on e.g Helix or Fractal. This can be hard to get right. I sold my Helix before they updated the acceleration curves on it, but it required a ton of knob spinning to adjust parameters. Fractal is somewhere in between where it is either a bit sluggish feeling or tends to overshoot the value. Delay/reverb times are the worst offenders when the scale is a ridiculous "a few milliseconds - several seconds".

It's the closest to that amp/pedals workflow I've gotten in a modeler I've used.

 

[laxu]

Yes that is precisely my point. They weren't accurate, but people loved and used them.

Yet we've got people who didn't stick with them, and who jumped through all the hoops over the decades and eventually ended up with an Axe FX III... and who are bemoaning the accuracy goal. I'm simply questioning... if they didn't care about accuracy... why didn't they stick with their Pod 2.0's or the HD500's or their Kemper (lol)

We are all after multiple things.

• Form factor.
• User experience.
• Convenience.
• Better tones/feel.
• Specific set of features.

When I had the Helix Floor, FM3 and QC at the same time, it was the form factor that ruled out the Helix (wanted a desktop unit, but HX Stomp is too gimped for onboard UI), and the effects quality that ruled out the QC - which surprised me as I expected it might be e.g amp sounds.

I got the used Axe-Fx 3 Mk2 only because it cost me about as much as a new FM9 in EU back then. FM9 was months away from European release at the time. The FM3 required enough block tetris to annoy me often enough to start to want the big boys. Later Fractal did improve its CPU usage to a point where I could have stuck with it tho.

 

[2112]

There's modeling the physical properties of electrical components and their interactions (let's call this algorithmic accuracy) and then parameters in those algorithms with specific tolerances like pot tapers (call this algorithmic detail).

Using these terms; Algorithmic Accuracy - yes please. This is why playing through a POD is like eating 3 day old cold pizza and playing through an Axe-Fx is like having the chef's table at a great restaurant. Algorithmic Detail - nice to have but I can certainly make ridiculous guitar noises without it. See pretty much every firmware since Quantum in Fractal land.

 

[2112]

Having said that nothing has blown my mind as much as plugging into this on my 14th birthday.

 

[Orvillain]

IMO the Quad Cortex actually got a lot of things right for the basic editing experience for an already built preset. It achieved this through a couple of things:

1. The touchscreen gives you instant access to any block in your preset since they are all on the same screen. No scrolling involved. Just tap a block and there it is.
2. The knob/switches allow for a lot more immediately accessible control. It does take a bit to build a mental map between screen and knob tho.
3. The amount of knobs meant that there are few pages to jump between in any given block. So overall less back and forth navigation required.
4. The knobs had a good feel to them that was closer to knobs on amps or pedals than what you find on e.g Helix or Fractal. This can be hard to get right. I sold my Helix before they updated the acceleration curves on it, but it required a ton of knob spinning to adjust parameters. Fractal is somewhere in between where it is either a bit sluggish feeling or tends to overshoot the value. Delay/reverb times are the worst offenders when the scale is a ridiculous "a few milliseconds - several seconds".

It's the closest to that amp/pedals workflow I've gotten in a modeler I've used.

The touchscreen fails to react to every other touch. Which is a huge fail. Any potential benefits are completely lost the moment I try to name a patch when saving it, and I have to repeatedly press the N key because it didn't register the first two times.

The poo encoders being so far away from the screen causes your eye to have to flit around too much, and visually joining up the knobs on the screen with the right poo encoder, at least for me, I found quite difficult and distracting.

Because their encoders have detents, there are values you cannot access from them. They might be small floating point values, but they matter when comparing to a real amplifier that doesn't have them (not looking at you, Orange Rockerverb MKIII !!)

It might be the closest something has gotten, but it was still so so so so far away.

I think about it like this.... my ideal rig is a 4 channel amp, a row of boosts... a row of delays.... a row of reverbs.... everything with its own dedicated pedal and set of knobs.... all plumbed into an analog switching matrix to bring everything in and out whilst retaining trails. GigRig on acid basically. But the floor space and power requirements for that would be quite off the charts, and it would be noisy at the end of the day too. It may even have a bit of latency due to the build up AD/DA conversion stages.

Breaking that kind of setup down into a multi-effect... you're almost certain to lose some of the immediacy. Because what gives you the immediacy is the fact that the "real thing" has dedicated knobs.... but what you gain from a multi-effect is not having to have tons of power, tons of cables, tons of noise, and tons of floor real estate taken up because you had the temerity to want one pedal to have its mix parameter at 10 o'clock, and the other at noon.

All I'm saying is ... there aren't any right and wrong answers when it comes to UI, UX, and implementation. We've just gotta pick what jives the best with us, and what limitations/obstacles sit the best with us as well.

For me, out of everything I've tried, it is actually the Helix that sits best with me from a UI+UX perspective.


Godamnit that is a lot of nonsense words on a Monday morning. Guhhhhh.

 

[Orvillain]

Using these terms; Algorithmic Accuracy - yes please. This is why playing through a POD is like eating 3 day old cold pizza and playing through an Axe-Fx is like having the chef's table at a great restaurant. Algorithmic Detail - nice to have but I can certainly make ridiculous guitar noises without it. See pretty much every firmware since Quantum in Fractal land.

Bad analogy. 3 day old cold pizza is awesome.

 

[Orvillain]

We are all after multiple things.

• Form factor.
• User experience.
• Convenience.
• Better tones/feel.
• Specific set of features.

When I had the Helix Floor, FM3 and QC at the same time, it was the form factor that ruled out the Helix (wanted a desktop unit, but HX Stomp is too gimped for onboard UI), and the effects quality that ruled out the QC - which surprised me as I expected it might be e.g amp sounds.

I got the used Axe-Fx 3 Mk2 only because it cost me about as much as a new FM9 in EU back then. FM9 was months away from European release at the time. The FM3 required enough block tetris to annoy me often enough to start to want the big boys. Later Fractal did improve its CPU usage to a point where I could have stuck with it tho.

I think me and you are just polar opposites, coz I love the Helix form factor!

 

[2112]

Bad analogy. 3 day old cold pizza is awesome.

 

[laxu]

I think me and you are just polar opposites, coz I love the Helix form factor!

The Helix is great as an all-in-one floor unit. But I wanted a desktop unit because crouching down to the floor to adjust stuff is no good for me. Even my pedalboard sits on top of a few boxes right now and is operated remotely for any footswitching.

 

[laxu]

The touchscreen fails to react to every other touch. Which is a huge fail. Any potential benefits are completely lost the moment I try to name a patch when saving it, and I have to repeatedly press the N key because it didn't register the first two times.

The poo encoders being so far away from the screen causes your eye to have to flit around too much, and visually joining up the knobs on the screen with the right poo encoder, at least for me, I found quite difficult and distracting.

I found the QC touchscreen to work well enough. Not as good as the best smartphones, but not enough to be a problem for me.

The encoders took me a while to mentally map them to reach the right one based on what I see on screen. Then it worked much better. Some of the blocks like the cab block were harder because the order wasn't intuitive. I think they improved that with some fw update.

It's of course not perfect by any means, but still better than the endless paging back and forth on other units. I'm interested to see what Line6 specifically does for their next gen units because I woudl expect they aren't just going to copy QC and TMP in this regard.

 

[dean701]

That can't be true. So many records were made with them both. The early 2000's was practically the green light for modelling as an entire pursuit. I reckon with a Pod 2.0 even today, you could still make awesome sounds, and they'd be absolutely fine for putting out music.

So what are we actually after here if it isn't authenticity? Even a lot of your UI complaints with the Fractal boil down to it not being authentic enough to the experience of using a real piece of gear, and the ease of use that comes with it.

To be fair, there have been other advancements besides the amps sounds since the pod 2. Effects, reamping, digital i/o... plus the sounds. I used a pod into a mark Iv for years for small bars. It was passable for sure

 

[Achilles]

I found the QC touchscreen to work well enough. Not as good as the best smartphones, but not enough to be a problem for me.

Same. Certainly not enough to list it as a legitimate gripe. If we're gonna get that anal then each of these platforms will end up with a LONG list of issues.