Line 6 Helix Stadium

[Chocol8]

It would be very cool if Proxy could internally self-capture (no AD/DA), that way you could perhaps make an internal snapshot of a dual amp setup you've dialled in, then load that into one block. This may then help reduce *CPU cost compared to using 2x Agoura amp models, *if it worked out that 1x proxy model was less computationally expensive than 2x Agoura amps.

If Proxy can capture effects as well as or better than the current tech, the possibility for cpu reduction could be even larger. That won’t be a big deal for everyone, but it could open up the ability for more complex signal chains, multiple instruments through a single Stadium, and processor heavy effects like longer convolution reverbs.

 

[Lysander]

It would be very cool if Proxy could internally self-capture (no AD/DA), that way you could perhaps make an internal snapshot of a dual amp setup you've dialled in, then load that into one block.

That's a good question for @Digital Igloo but, given what we know about HXS' routing capabilities, i think there's a good chance this will be possible.

 

[ian_dissonance]

I mean, if I could make a signal chain of a pre-eq, a split boost section (full scream-style blended with a 20% HM2/80% clean), and amp and a cab and a post EQ “proxied” into a single block, even if it didn’t free up ALL the DSP you’d still have more slots to work with.

 

[Chocol8]

So what emulates the preamp stages or the poweramp or the transformers? Is it just magic? EQ doesn’t create distortion.

Machine learning captures are basically a complex transfer function where if input = a then output = x and if input = b then output = y with variation of frequency level and potentially other variables on the input. A matching EQ would be the same concept, but only varying frequency on the input and output. Capture tech does much more and can capture distortion and other non-linear aspects of an amp.

Basically it feeds the amp, pedals, or pedals and amp a bunch of test signals and learns what comes out. This is all done as a single stage no matter how many devices or amp stages are in the signal chains. The capture doesn’t know if it is a simple amp, an amp with multiple preamp gain stages, a dirt pedal, or a combination unless it is stored in meta data.

 

[Chocol8]

That's a good question for @Digital Igloo but, given what we know about HXS' routing capabilities, i think there's a good chance this will be possible.

I am not sure. It depends on how much of the Proxy capture process is done on the device vs offloaded to a PC or cloud and how. Hopefully it is possible to do it all without analog conversion. It would be a killer feature that Kemper, QC, Tonex, and obviously Fractal currently don’t offer.

 

[mikah912]

I am not sure. It depends on how much of the Proxy capture process is done on the device vs offloaded to a PC or cloud and how. Hopefully it is possible to do it all without analog conversion. It would be a killer feature that Kemper, QC, Tonex, and obviously Fractal currently don’t offer.

From what I understand, it's pretty much cloud-based.

 

[Chocol8]

From what I understand, it's pretty much cloud-based.

It sounds like the final processing will be, but what matters is how the training files are created before they get sent to the cloud for processing.

 

[paisleywookiee]

Same as with Kemper, ToneX, and NAM, I’ve got zero desire to sift through a ton of shit I don’t like to find the one I do. The only use I’ve got the capture tech is capturing amps I’ve personally dialed in. It’s the only way I’ve begun to enjoy NAM.
So “access to a big capture library” isn’t really a selling point.

 

[EOengineer]

Proxy being able to reliably/consistently capture tones that are 100% transferable to other Hx devices will be a huge win. I can’t wait.

 

[Sascha Franck]

If Proxy can capture effects as well as or better than the current tech

I guess we're quite far away from that. Or rather, so far it seems only static settings cane be captured. And would you really want, say, a chorus without control over its speed or depth? I wouldn't.
Right now, we're at the beginning of amp/dirt things being captured parametrically, and even that seems to be no easy task. FX might take quite some more time for that to happen. And a lot more "user input" would be required as well, because you'd actually have to create captures of all kinda different settings.
And I also doubt that this would save you too much CPU juice as the playback engine doing kinda realtime intermodulation between parameters would have to do quite some work.

 

[timbuck3]

 

[Sascha Franck]

Already seen (not watched) it. As long as you can generate clicks, it doesn't seem to matter whether you make a tool out of yourself anymore.

 

[bigmenace]

I might not care now but I sure will once it releases and I have one in my studio.
At the moment all I can do is imagine.

 

[Lysander]

Already seen (not watched) it. As long as you can generate clicks, it doesn't seem to matter whether you make a tool out of yourself anymore.

"No one cares about this new modelling tech", says guy who sells captures for a living

 

[Gary_W]

Same as with Kemper, ToneX, and NAM, I’ve got zero desire to sift through a ton of shit I don’t like to find the one I do. The only use I’ve got the capture tech is capturing amps I’ve personally dialed in. It’s the only way I’ve begun to enjoy NAM.
So “access to a big capture library” isn’t really a selling point.

TL DR : I think you need capture tech on a device to bring others over from competitor products. Here’s why:-

Because I hadn’t tried modelling since the Pod XT (I went real amps for a while then Kemper for years) I had it burned into my brain that capturing an amp is better than modelling an amp. I know I’m out of touch but I doubt I’m the only person on the planet who has this mindset; if one of the goals of digital gear is reality, believing that something that captures reality is superior to something that ‘just models it’ seemed logical to me. It hasn’t helped that there’s been lots of buzz about NAM, Tonex etc over the last few year and, although I’ve lived under a rock in terms of trying them, it’s further fed into my brain the message that capturing is superior to modelling.

I’ve been playing with Helix Native over the last week and have realised that my thinking was flawed. I’ve fallen in love with the sound of lots of models but alongside than that, I am enjoying the workflow more. Your mentioning the sifting through loads to find the thing you want? That was my normal…. I’d favourite the couple of good ones that my ears liked in a big pack of captures and just use them. Instead I’m now going straight to an amp and having the ability to tweak it *in the mindset of the actual amp* to interact with the little quirks. As one of (many!) examples, channel 2 on the matchless having the 6 different tone positions and then a cut control. I spent a while just playing with that yesterday and enjoying most of what I heard. Although it’s absolutely possible to get an awesome capture of an amp and really enjoy playing through it, I’m not sure you can capture the interesting control interactions in the same way.

I’m looking forward to seeing what Line 6 does with capture technology but I’m far less interested in it now that Native has given me a different perspective on modelling. Proxy was necessary to get someone with my mindset interested in the Stadium in the first place but I think anyone coming from a capture-only workflow who has an open mind will enjoy the modelling side of things far more than they expect.

 

[MirrorProfiles]

Capturing is a feature that people are coming to expect in 2025. I’m excited for Line 6’s take on what that experience should be. Even if DI doesn’t see it as his favoured approach, they have the benefit of standing back and watching how others have done it first before making their move.

Pretty certain the user experience will be beyond what we’ve seen from Kemper, NDSP, Headrush, Valeton etc.

The more existing formats it supports the better - if someone can easily recreate a rig that currently requires using 3 pedals into one that can do way more, including an entire band’s mixing and routing, then what’s the loss?

Line 6 has cloud training available, even if it’s as inefficient as making a proxy model of a kemper/tonex/QC file, I don’t see why it wouldn’t be possible. If it is a fork of NAM itself, perhaps that’ll be even easier to translate to Proxy’s own format (although not all NAM models use the same internal settings so some translation may be required).

 

[Mooselake]

This guy comes across as an idiot.

So, future Stadium users,who haven't had the opportunity to try out the new Agoura models yet, aren't saying things in response to having tried Agoura models before getting to try them out like Fractal and QC users did after trying out past new models. I wonder why that is.

 

[FuzzyAce]

The corniness aside, I actually agree with his overall premise. I do think modeled tones have reached a sort of plateau and the interest lies more in features/processes, routing capabilities, and user interface. Many are very content with the tones they are currently getting from Fractal or HX, or any number of other modelers. But, the Stadium hasn't been released yet either, so maybe the Agoura amps will be so much "better" that you can actually hear and feel it? Do they even need to be better for some users? The efx side of things, especially verbs and delays, seems to be where HX could make significant gains in the tonal department, though. Time will tell.

 

[JiveTurkey]

Watched the whole video and didn't get triggered crew, checking in.

 

[Lysander]

But, the Stadium hasn't been released yet either, so maybe the Agoura amps will be so much "better" that you can actually hear and feel it? Do they even need to be better for some users?

You know, that's the kind of question anyone outside YouTube would ask themselves before claiming no one cares about something.