QC vs Helix

[MirrorProfiles]

I like Helix’s modelling a lot, but Neural’s amps mostly sound better to me. That doesn’t mean Helix’s are crap, or that the differences are big. But Neural’s amps mostly stuff feels nicer to play, the distortion sounds less congested and more realistic and there is usually less artifacts.

Helix has a ton of great stuff in there and as a complete package it’s a modern day classic. But the next iteration of Helix absolutely must take a step up to at least the level of the competition, and hopefully beyond it.

The QC rectifier is also not as bad or inaccurate as people say, IMO. It’s just maybe not as intuitive to dial in. People also said the same about Helix’s Recto and 5150 and they are both as good as anything else in Helix

 

[mercifulfuzziness]

You're one puzzling dude. The Helix models amps extremely accurately. Can't emphasis this enough.

If you want your Helix DR to sound exactly like the DR in the store, get a Deluxe reverb CAB and Speaker and try your Helix with that. You'll find next to no difference.

If you're playing through ""FRFR"", it WILL sound and feel different. That's the bottom line. Some other companies apply compensations for speaker modeling (speaker breakup, inertia etc.) and that can make a difference with ""FRFR"".

It is not at all hard to tweak a Helix - I could dial up almost anything in less than 5 min. 10 min at most. I don't know why you keep going in circles about what other people say.

Don't keep going by what others say - use your ears. I dial a modeler pretty much just how I dial in an amp. It ain't rocket science.

What circle? I am just saying I have no idea how an Dumble should sound so what do I know. With the real amp I am sure.

 

[jellodog]

Is someone paying @laxu to say “Hotone” in as many threads as possible?

 

[GTR37]

I like Helix’s modelling a lot, but Neural’s amps mostly sound better to me. That doesn’t mean Helix’s are crap, or that the differences are big. But Neural’s amps mostly stuff feels nicer to play, the distortion sounds less congested and more realistic and there is usually less artifacts.

Helix has a ton of great stuff in there and as a complete package it’s a modern day classic. But the next iteration of Helix absolutely must take a step up to at least the level of the competition, and hopefully beyond it.

The QC rectifier is also not as bad or inaccurate as people say, IMO. It’s just maybe not as intuitive to dial in. People also said the same about Helix’s Recto and 5150 and they are both as good as anything else in Helix

I agree w your first statement
NDSP just handles the low end attack better and the sound is more open less congested as you said

That is why I posted the above with Leon Todd on the latest Helix and John on the latest QC updates
I can audibly hear the differences you may like one better but to me they are different and it’s obvious enough

As you also said that does Not make Helix bad

 

[laxu]

Is someone paying @laxu to say “Hotone” in as many threads as possible?

 

[mbenigni]

 

[littlebadboy]

Quoting, "Who cares?"

I have to say that most listeners (if not all) wouldn't care what you are you using.

 

[Frodebro]

To be fair, any report from FAS/L6 may have some bias to it. Not saying it isn't factual but it's good to remember where the data is coming from.

Same would be said if NDSP made claims about L6 and FAS

That's why we have graphs.

 

[GTR37]

Another interesting thing is John saying in the Andertons video that all NDSP models are essentially a huge amount of captures shot together

Capture is like a photo graph
Model is like a video with panoramic view

I thought that was a really easy to understand explanation of the difference

 

[Orvillain]

Another interesting thing is John saying in the Andertons video that all NDSP models are essentially a huge amount of captures shot together

Capture is like a photo graph
Model is like a video with panoramic view

I thought that was a really easy to understand explanation of the difference

This is not correct.

Principally, if you took a ton of captures and interpolated between them, or otherwise had a parametric approach, you would have a model that is as detailed as any SPICE or otherwise component modelled amplifier model.

None of us truly know how Neural are doing their models, but I think your comparison is overly simplistic.

They are both modelling approaches. They just use different underlying tech to get there.

 

[GTR37]

This is not correct.

Principally, if you took a ton of captures and interpolated between them, or otherwise had a parametric approach, you would have a model that is as detailed as any SPICE or otherwise component modelled amplifier model.

None of us truly know how Neural are doing their models, but I think your comparison is overly simplistic.

They are both modelling approaches. They just use different underlying tech to get there.

John works with / for Neural
And he says it’s all done AI
Maybe originally they did some ?? Component modelling but he is stating in the video it’s all Tina now

 

[Orvillain]

And he says it’s all done AI

I never said otherwise.

 

[laxu]

Another interesting thing is John saying in the Andertons video that all NDSP models are essentially a huge amount of captures shot together

That's an oversimplification.

https://arxiv.org/html/2403.08559v1 details their current approach.

As far as I understand, TINA captures a sensible data set which is then fed to a neural network model which will churn out the final model. They figured a novel way to get enough data for different knob positions that they don't need a massive data set.

It's not individual captures interpolate or anything like that. It's just based on a data set from tons of test signals.

 

[Orvillain]

That's a cool link, I'd not seen that before I don't think. Man.... my farts really stink.

 

[GTR37]

That's an oversimplification.

https://arxiv.org/html/2403.08559v1 details their current approach.

As far as I understand, TINA captures a sensible data set which is then fed to a neural network model which will churn out the final model. They figured a novel way to get enough data for different knob positions that they don't need a massive data set.

It's not individual captures interpolate or anything like that. It's just based on a data set from tons of test signals.

Maybe he was trying to dumb it down but I think he was saying there really is no component modelling
No one at NDSP has 10 probes in an amp measuring caps and impedance ,

The thing is while this data would match the tonestack it would not account for the feel
And the QC does feel good under the fingers so they are introducing compression somehow

Also as Lee says in the video the Models sound idealized
He does not hear that fizz and rasp that the captures exhibit
So perhaps Bishop from Line6 is not all wrong and they do in fact trim the fact and eliminate some undesired freq in the full ampmodels

 

[MirrorProfiles]

Also as Lee says in the video the Models sound idealized
He does not hear that fizz and rasp that the captures exhibit
So perhaps Bishop from Line6 is not all wrong and they do in fact trim the fact and eliminate some undesired freq in the full ampmodels

I don’t think this is the case at all personally. I think it’s just down to them using reactive loads sometimes rather than cab loads. It can affect the overall frequency response so things sound darker/smoother like that.

 

[Sleezy E]

We're getting close boys... I can feel it

 

[Orvillain]

Maybe he was trying to dumb it down but I think he was saying there really is no component modelling

No one here said they did.

The thing is while this data would match the tonestack it would not account for the feel

How? Why?

 

[GTR37]

No one here said they did.


How? Why?

Well you would think measurements like fractal uses impedance , voltages that will have more impact on feel than eq
For instance a Bogner has that chew in the mids because of a low voltage in the preamp
So technically without seeing what the data being collected by Tina is it is hard to know ,
How are they getting the power amp data , things like Fractal is collecting transformers data like # of winds
I don’t know how Tina would do that with having a probe in the amp

 

[Orvillain]

Well you would think measurements like fractal uses impedance , voltages that will have more impact on feel than eq
For instance a Bogner has that chew in the mids because of a low voltage in the preamp
So technically without seeing what the data being collected by Tina is it is hard to know ,
How are they getting the power amp data , things like Fractal is collecting transformers data like # of winds
I don’t know how Tina would do that with having a probe in the amp

I don't understand why you're talking about EQ? The paper that Laxu linked is pretty clear about the methodology. Because I'm roughly half an idiot, I can't speak to specifics. But.... it seems to me:

- They send sounds (test signals) into the actual amplifier and record what comes out at different knob settings (like gain, bass, treble).
- They use a type of AI (a neural network) to learn the patterns between the input sounds, the amp’s knobs, and the output sounds.
- This process captures the amp’s full behavior, including complex interactions between the input signal and the controls.
- Once trained, this AI can take any input sound, simulate how the amp would change it, and let you tweak the virtual knobs just like the real amp.

So in what world would that not account for the feel of the original amp???

This little side-quest discussion has nothing to do with the original EQ point being parsed from what bishopgame "guessed" in his original comments, that has somehow become a kind of gospel, with no evidence whatsoever I might add.