Line 6's stance on profiling

[JiveTurkey]

Imagine being a serial killer, and you finally manage to make that high-end skin suit you've dreamt about since you were 9. Except you put it on, split the gussett, and realize that the whole thing just itches.

Yeah, that's profiling.

I knew there was a reason I invited you!

 

[Orvillain]

This hit me in the feels because not a week goes by where I don't say "Ugh, I can't do this in Illustrator? I gotta use Photoshop? Pixels, really?"

I've switched to Affinity Photo and Affinity Designer, coz fuck Adobe - mostly.

I use Substance3D quite a lot, and Adobe acquired them through necrotic means last year, and there isn't really anything else that gives you the breadth of stuff. So I have to use that. Le sigh.

Rest of the time it is Balsamiq Mockups (good for putting quick stuff together), Sketch, and Figma.

 

[pipelineaudio]

Am I the only one who finds something really different about the distortion characteristics of different companies? Even the Kemper?

Even with these new open source profilers, there's something quite different between them as well.

I would love to see someone who can really work the helix get something similar to Bias or LePou or TSE. Is it just that those things used such an earlier version of where amp sim DSP is at now?

Is anyone that good at getting these types of tones?

As a counter to how it seems to me, I swear Choptones Metallica presets sound pretty much identical on Helix as on Bias, so I keep coming back to user error. I just don;t know how to fix it if that's the case

 

[Alex Kenivel]

Am I the only one who finds something really different about the distortion characteristics of different companies? Even the Kemper?

I think people have been singing this song for a while

 

[Orvillain]

It's not surprising really. Distortion is the product of components running outside of their expected tolerances. Each component and each stage of a circuit is going to produce different harmonics and different frequency responses according to what was fed into them.

Also if you're just taking a bog standard digital waveshaper and slapping an ENGL skin on top of it.. I mean... that isn't a guitar amp really.

 

[pipelineaudio]

Also if you're just taking a bog standard digital waveshaper and slapping an ENGL skin on top of it.. I mean... that isn't a guitar amp really.

Chances are (especially since my using our compressor plugin as an amp sim seemed to work so well at creating the same tones) that this is close to the mark

However, I don't agree its not a guitar amp. If its a sound you like that's really all that matters.

I mean, until as some expert on here hits us as hard or as long or as influentially as say "In the City" by Joe Walsh, a sound made by plugging right into a console (he insisted it wasn't even a DI), I'll stick with "whatever gets you there"

 

[Orvillain]

However, I don't agree its not a guitar amp. If its a sound you like that's really all that matters.

These two things aren't entirely congruent. Something not being a guitar amp doesn't mean I am saying it couldn't be a sound you like. Sound is subjective, turning analog schematics into digital counterparts is not.

It really comes down to accuracy, and almost something akin to a quantization problem. The more you quantize your data, the less is it going to be accurate to what happens in the real world.

Which isn't the same thing as saying it sounds bad. But for amplifier modelling, the goal is certainly to bring the real world into the digital one.

 

[jay mitchell]

Distortion is the product of components running outside of their expected tolerances.

No. Components in a nonlinear system can have zero tolerance for their values and parameters and still produce nonlinear behavior.

Automating the tweaking process ("profiling," "capturing," or whatever a manufacturer may choose to call it) is still modeling. It can only make use of the available tools within the modeler. If the available tools have limitations - they always do - then an automated process can never get any closer to the target than those limitations allow. Furthermore, there is no way a "black box" process - which is being discussed here - could ever come close to capturing all the behaviors of a tube amp over the range of control settings, input levels and spectral content, and variations in guitar pickup parameters that occur in the real world. Overcoming this limitation need not involve modeling every circuit component in the target amp - I'd wager that nobody goes that far - but it does require modeling various stages in an amplifier and the ways in which those stages interact with each other. Absent the ability to extract signals from intermediate stages in an amplifier - essentially treating each stage as a black box - profiling will always have an element of hit or miss.

 

[Orvillain]

No. Components in a nonlinear system can have zero tolerance for their values and parameters and still produce nonlinear behavior.

Automating the tweaking process ("profiling," "capturing," or whatever a manufacturer may choose to call it) is still modeling. It can only make use of the available tools within the modeler. If the available tools have limitations - they always do - then an automated process can never get any closer to the target than those limitations allow. Furthermore, there is no way a "black box" process - which is being discussed here - could ever come close to capturing all the behaviors of a tube amp over the range of control settings, input levels and spectral content, and variations in guitar pickup parameters that occur in the real world. Overcoming this limitation need not involve modeling every circuit component in the target amp - I'd wager that nobody goes that far - but it does require modeling various stages in an amplifier and the ways in which those stages interact with each other. Absent the ability to extract signals from intermediate stages in an amplifier - essentially treating each stage as a black box - profiling will always have an element of hit or miss.

I don't disagree with that. I was being a bit careless with my words. But fundamentally it is true that the distortion characteristics we all tend to love in guitar amps come from how each stage cascades into the next. That's my layman's understanding anyway. Not an engineer.

There has been a huge improvement in guitar amplifier modeller accuracy over the last 15 years or so, maybe a bit more. And that's because plugins and DSP engineers are no longer doing things like this:

Waveshaper — Musicdsp.org documentation

And then slapping a pretty picture on top of it. I think that was my wider point. And I'm reasonably sure that some of the earlier amp models from some big name companies were doing exactly this, because I was told so by a DSP engineer and industry veteran.

 

[dk_ace]

Walks in…. Confidently heads to the bar and orders a drink… looks around and sees a lot of cool people and knows this is gonna be a good hang…

D

 

[SeeD]

Yeaaaah, unless the goal is to be provocative, "Blatant ripping off of GUI layout, design language, use of color, signal flow, block allocation, DSP allocation, snapshot implementation, footswitch modes, and more is no different than modelers existing!" holds no water.

Freaking out implies concern. No one's freaking out; just not afraid to point out what others (including numerous friends at other MI companies) have noticed. Cliff calls it as he sees it too; that's a big reason why I like him.

Line 6 has been blatantly copied for decades. We've been copied for so long that people now accuse us of copying our copiers. (Flextone > Flextone II > Flextone III > Katana > Catalyst). Sometimes we're first to the multieffects market with certain features, but who cares, because another company would've eventually gotten there anyway (color screen, RGB switches, maaaybe scribble strips). In other cases, it's literally impossible for a company to have landed on a massive portion (not "a few things") of their UX and feature implementation unless Product Management specifically dictated "just copy how Line 6 does it." When there are literally a thousand ways to do one thing, doing dozens of things in a nearly identical manner as someone else is not "design." At best, it's lazy opportunism.

Eh, fine, whatever—The Mooers and B€#®!ng€®s of this world will always exist. And Line 6 will continue to sell hundreds of thousands of Helix/HX SKUs because we do our own thing.

Just thinking out loud.

ahh, i did not mean you were freeking out, or even L6
I agree with a lot of what you say.

But the thing i asked ( to anyone) is it okey to make 1:1 copys of amps and effects ? Is this not a bit of copying `stuff` and using there `name` to sell products ?

 

[KingsXJJ]

Fake snark aside; that's actually a good suggestion if you want that tone. Combine with a Stomp or HX FX and you have a giggable version of the tone you are hunting with the FX you are using. If you can find a good deal on it; it's a win-win I'd think?

I agree. Tone first. Everything else is icing.

 

[Orvillain]

ahh, i did not mean you were freeking out, or even L6
I agree with a lot of what you say.

But the thing i asked ( to anyone) is it okey to make 1:1 copys of amps and effects ? Is this not a bit of copying `stuff` and using there `name` to sell products ?

Depends what you mean by copying.

Taking someone else's code ... no. IP infringement. Illegal.

Looking at something cool someone has done and going "I can do that!" .... fine. Not illegal, provided you're not violating any patents. Could be seen as ethically wrong, depending on the skeeziness/Music Tribe factor.

 

[KingsXJJ]

..or you could buy a car without a sun roof and then constantly complain about the lack of sunroof.

Seems legit

Or buy a car with a sun roof and never use it. But still glad it’s there.

me

 

[KingsXJJ]

It's one thing to get a modern digital device with clean converters and loads of but depth…

but depth is underrated.

 

[KingsXJJ]

I look at profiles as something a person would need to use if they don't understand how amps work.

The more tools and approaches the better. Not everyone is a gear head or has or is a great tech.

I think bringing great tone to the more the merrier. I’m not just a player… I’m a listener too.

 

[jay mitchell]

But the thing i asked ( to anyone) is it okey to make 1:1 copys of amps and effects ? Is this not a bit of copying `stuff` and using there `name` to sell products ?

Is it OK for a car manufacturer to make a 1:1 copy of the performance envelope of a competitor's car? That's exactly what modeling tries to do (with varying degrees of success). You can't copyright, trademark, or patent a sound. Harley-Davidson actually tried to register the sound of their bikes as a trademark, were challenged in court, and eventually gave up. Whether naming the target of a model is ethical is another question, but there are numerous ways to get around that....

 

[Lysander]

Imagine being a serial killer, and you finally manage to make that high-end skin suit you've dreamt about since you were 9. Except you put it on, split the gussett, and realize that the whole thing just itches.

Yeah, that's profiling.

I'm bookmarking this one for future use

 

[Lysander]

The more tools and approaches the better. Not everyone is a gear head or has or is a great tech.

This. I much prefer modelers to profilers myself, but anyone stating one is clearly better than the other is deluded, IMHO. Very different use cases and pros/cons.

 

[KingsXJJ]

Is it OK for a car manufacturer to make a 1:1 copy of the performance envelope of a competitor's car? That's exactly what modeling tries to do (with varying degrees of success). You can't copyright, trademark, or patent a sound. Harley-Davidson actually tried to register the sound of their bikes as a trademark, were challenged in court, and eventually gave up. Whether naming the target of a model is ethical is another question, but there are numerous ways to get around that....

Exactly why 3 knob versions were allowed after the Gibson lawsuit era on Les Pauls. It can quack exactly like a duck. But it can not be an exact copy in physical form or name. It can be super close though…