Line 6 Helix Stadium Pre-Launch Discussion

[AlbertA]

One of the things in the launch video that @Digital Igloo said was words to the effect of, sample accurate communication between the amp and cab block for speaker impedance curves.

That means - in my reading of it anyway - Agoura is powerful enough to communicate the individual sample-stream across blocks, rather than relying on block-based processing. Which means more accurate representation of signals from sample-buffer to sample-buffer.

In practical terms, this means the two blocks can communicate in "real-time" rather than waiting for the next processing buffer to finish before exchanging data.

Other systems process audio in fixed-size blocks, say 32, 64, or 128 samples at a time. That means the amp block would process its chunk of samples, then hand off the results to the cab block in the next cycle.

That tiny delay introduces latency and temporal “coarseness” in the interaction between the amp and cab, so the simulated impedance curve only updates once per buffer, not per individual sample.

So.... you can assume that Helix Stadium will be more precise AND more accurate than other "older" systems.

I suspect this would've been the big jump between Axe FX II and Axe FX III as well. But that's just a gut feeling thing.

I think that's just reading too much into it.

It's likely just "Oh you switched to CAB X, ok yeah that has a SIC curve Y" in the power amp model. Amp simulation should be done with an ODE solver, which will have its own time step (not to mention running at the oversampling rate).

 

[AZG]

One of the things in the launch video that @Digital Igloo said was words to the effect of, sample accurate communication between the amp and cab block for speaker impedance curves.

That means - in my reading of it anyway - Agoura is powerful enough to communicate the individual sample-stream across blocks, rather than relying on block-based processing. Which means more accurate representation of signals from sample-buffer to sample-buffer.

In practical terms, this means the two blocks can communicate in "real-time" rather than waiting for the next processing buffer to finish before exchanging data.

Other systems process audio in fixed-size blocks, say 32, 64, or 128 samples at a time. That means the amp block would process its chunk of samples, then hand off the results to the cab block in the next cycle.

That tiny delay introduces latency and temporal “coarseness” in the interaction between the amp and cab, so the simulated impedance curve only updates once per buffer, not per individual sample.

So.... you can assume that Helix Stadium will be more precise AND more accurate than other "older" systems.

I suspect this would've been the big jump between Axe FX II and Axe FX III as well. But that's just a gut feeling thing.

It will be interesting to see the end results. Listening to what has been released so far it largely sounds like they are upgrading their modeling engine to be very similar to what Fractal has been doing for years. I expect it will sound and feel better than the original Helix, but will it match or overtake the competition? Guess we will know more in a few months.

 

[Sascha Franck]

but will it match or overtake the competition?

As long as it's "close enough" (whatever that might mean for whomever), IMO there's more than enough pretty compelling reasons to go for a Stadium instead of whatever competitor's products. The general featureset (outside of any sound related things) presented so far to us mere mortals is absolutely outstanding already and I'd bet we'll see much more coming.

I said so before, in terms of onboard editing, this will likely mop the floor with anything else out there, then there's a whole lotta things that seem to be incredibly well thought out already (such as the output mixer), so if you're looking for a killer allround unit, these things will have to be taken into consideration by pretty much anyone interested in modelers.

I'm quite sure that there will still be people prefering FAS products for the variety of amps, their tweakability and what not, but for most people, a good assortment of great sounding amps (being able to compete with the best) should be sufficient - and at that point, other aspects will likely become more important for many folks.

 

[Mongillo19]

I've never done a A/B with a two year version of Fractal firmware so I can't say for sure if it's really better. But I can't help but think that touch/feel has gotten notably better over the past few years. Maybe placebo, but I do enjoy the tones/feel I get out of the FM3 more than ever today.

Oh for sure. I'm not trying to knock cliffs work. If anything, it shows the fractal line has been fantastic for years and at this point, I have to think we are hitting diminishing returns. But I'm sure that's what they said for color TV

 

[Orvillain]

I think that's just reading too much into it.

It's likely just "Oh you switched to CAB X, ok yeah that has a SIC curve Y" in the power amp model. Amp simulation should be done with an ODE solver, which will have its own time step (not to mention running at the oversampling rate).

I strongly suspect you are wrong.

It’s not just about the amp “loading” the right static impedance curve when you switch cabs.

Sample-accurate communication strongly implies that the amp and cab blocks are exchanging data on a per-sample basis during processing, not just referencing a fixed curve.

Even if the amp modeling itself is running with its own internal solver at an oversampled rate, in a traditional block-based DSP chain that solver only sees the load update once per buffer - say, every 64 samples. This results in a coarsening of the data, and thus less precision.

If Agoura eliminates that buffer-level latency, so the solver is aware of the instantaneous state of the cab/impedance at every single sample. That allows tighter, more accurate interaction between the virtual power amp and speaker model. This is a crucial improvement when it comes to the final tone and the non-linear processes that produce it.

 

[Orvillain]

but will it match or overtake the competition?

Honestly, it doesn't even need to. If you were to put numbers on it... let's put Kemper at 10% worthwhile-ness... and let's put Fractal at 90% worthwhileness... Helix regular at 70% .... if Helix Stadium gets to 83.4584% worthwhile-ness.... it's going to clean house. Absolutely clean house.

 

[Mongillo19]

83.4584%

 

[mikah912]

While I do have a Stadium XL on preorder, couldn't pass up 20% off, whether I keep it will largely come down to overall sound and feel quality. The UI looks like a nice refinement of the original Helix and the modeling improvements sound promising. But much of the Stadium is carried over from the original Helix good and bad. So while I would love a unit with a much better UI, I'm not sure I will be willing to give up Fractals massive high quality amp and effects library. Honestly I feel a little bit of deja vu from 10 years ago. Both companies have moved forward a lot, but they both still seem to have the same strengths and weaknesses. Not sure where I will end up this time round, but looking forward to giving the XL a go.

The precise reason I made a Stadium XL preorder as soon as the Jun 11th prezzie ended was that Line 6 has been making significant forward progress on sound quality and UX/quality of life/usability since the launch of the Helix in 2015. One Helix SKU or another has remained at the heart of my studio since then. I consider it a non-negotiable.

On the other side of the aisle, I've had two FM3s and currently rock an FM9. While there have been wonderful engine improvements and amp/FX added in that time, the UX/usability/QoL front has been pindrop quiet with the exception of the Dynacabs - which didn't really solve any problems due to the wealth of IRs it came with at launch. Even when Stadium launched, Cliff's kneejerk reaction was to brag about the supercomputer-esque processing power of the next gen....not ANY user experience lessons learned or intent to revamp usability. Suggestion after suggestion to optimize functions just using current gen hardware buttons lies ignored in their Wish List forums, while they cross off things like adding a Stealth or Supro Black Magik amp.

So viewed through that lens, I disagree that both companies "still seem to have the same strengths and weaknesses". Line 6 decided to try to address all of theirs in one fell swoop, while Fractal has largely taken a "We good" stance.

And make no mistake - The current Fractal hardware is good. It's great, even.

But it requires sacrifices, and I'm not seeing anything I'd be sacrificing with Stadium XL, while I'd be gaining tons of additional capabilities on top.

 

[Orvillain]

And make no mistake - The current Fractal hardware is good. It's great, even.

In 4cm with an amp, the Axe3 is definitely the leader when it comes to pure noise-floor, latency, and signal quality measurements. Helix Floor comes 2nd place, and nothing else really comes close. But I hope Stadium will be the new number 1, which is why I pre-ordered.

 

[Sascha Franck]

In 4cm with an amp, the Axe3 is definitely the leader when it comes to pure noise-floor, latency, and signal quality measurements.

Possibly not in terms of latency.
Edit: They seem to have improved on it quite a bit.

 

[Tritium3H]

In 4cm with an amp, the Axe3 is definitely the leader when it comes to pure noise-floor, latency, and signal quality measurements. Helix Floor comes 2nd place, and nothing else really comes close. But I hope Stadium will be the new number 1, which is why I pre-ordered.

I imagine that the Boss GT-1000 would strongly disagree.

 

[JiveTurkey]

 

[Orvillain]

I imagine that the Boss GT-1000 would strongly disagree.

Had one. Wasn't as bad as the GT-100 was, but it didn't sound as good as the Axe3.

 

[AlbertA]

I strongly suspect you are wrong.

Well I strongly suspect you are wrong - I think you are misremembering what was said in the introduction videos. I just went back, he said:

so we can sample accurately capture the interaction between signal and power supply.

<goes on, then> Double precision tone stacks. Interblock communications so cabs block can communicate back to the amp block for dynamic speaker impedance curves.

So yeah I'm sticking with "the power amp model loads a SIC curve, based on what cabinet you have selected"

It’s not just about the amp “loading” the right static impedance curve when you switch cabs.

I mean we are in speculation land - but why wouldn't it be?

Sample-accurate communication strongly implies that the amp and cab blocks are exchanging data on a per-sample basis during processing, not just referencing a fixed curve.

I assume you are hinting at non-linear speaker modeling aspects? I mean there's thermal effects that could affect the impedance curve but that's certainly not changing in the tens of microseconds (sample accuracy) and if you have a power amp model, being solved by an ODE, it would not make sense to split this processing/speaker model into a separate block.

Even if the amp modeling itself is running with its own internal solver at an oversampled rate, in a traditional block-based DSP chain that solver only sees the load update once per buffer - say, every 64 samples. This results in a coarsening of the data, and thus less precision.

Same point as above - it'd be weird for the power amp model to be split up like that. But again, we are in speculation land.

If Agoura eliminates that buffer-level latency, so the solver is aware of the instantaneous state of the cab/impedance at every single sample.

Again it wouldn't make sense to split that up in separate blocks if that was the case. But in a more general note, what exactly is changing in tens of microseconds?

 

[AZG]

The precise reason I made a Stadium XL preorder as soon as the Jun 11th prezzie ended was that Line 6 has been making significant forward progress on sound quality and UX/quality of life/usability since the launch of the Helix in 2015. One Helix SKU or another has remained at the heart of my studio since then. I consider it a non-negotiable.

On the other side of the aisle, I've had two FM3s and currently rock an FM9. While there have been wonderful engine improvements and amp/FX added in that time, the UX/usability/QoL front has been pindrop quiet with the exception of the Dynacabs - which didn't really solve any problems due to the wealth of IRs it came with at launch. Even when Stadium launched, Cliff's kneejerk reaction was to brag about the supercomputer-esque processing power of the next gen....not ANY user experience lessons learned or intent to revamp usability. Suggestion after suggestion to optimize functions just using current gen hardware buttons lies ignored in their Wish List forums, while they cross off things like adding a Stealth or Supro Black Magik amp.

So viewed through that lens, I disagree that both companies "still seem to have the same strengths and weaknesses". Line 6 decided to try to address all of theirs in one fell swoop, while Fractal has largely taken a "We good" stance.

And make no mistake - The current Fractal hardware is good. It's great, even.

But it requires sacrifices, and I'm not seeing anything I'd be sacrificing with Stadium XL, while I'd be gaining tons of additional capabilities on top.

Without question the UI is Fractals biggest weakness and if they continue to ignore it I believe it will hurt them in the long run. I don't see them ever competing with Line 6 for UI and I don't think they need to. But they do need something more competitive again. Nevertheless, as I said before I can't stop feeling deja vu from with the last Helix release. Seems both companies are in similar places to each other, just ten years further down the road.

 

[Orvillain]

so we can sample capture accurately the interaction between signal and power supply.

You've got that sentence wrong. Go back and re-listen.

 

[AlbertA]

You've got that sentence wrong. Go back and re-listen.

Fixed - he still not talking about the communication between amp block and cab block in that particular sentence.

 

[Orvillain]

Fixed - he still not talking about the communication between amp block and cab block in that particular sentence.

Do you understand what sample-accurate communication means?

 

[AlbertA]

Possibly not in terms of latency.
Edit: They seem to have improved on it quite a bit.

Wait, what is being referenced here?

 

[AlbertA]

Do you understand what sample-accurate communication means?

Where did he say that? And yes I do, more than you care to know.