Brian Wampler on capturing (a la NAM, Proteus)

[Sascha Franck]

Whenever there's a sufficiently complex amp with highly interactive controls, the whole capture thing falls apart

Depends on whether you need all those tonal options. Most people I know, don't, even Mk owners.

 

[paisleywookiee]

Depends on whether you need all those tonal options. Most people I know, don't, even Mk owners.

It's not whether you need all the options. It's which options. Hence why it's difficult to buy third party caps, as opposed to owning the amp and the profiler. For example, my gain/EQ/GEQ/mode switch settings won't be the same as yours, or anyone else's. Also, I dial the amp in differently, depending upon what guitar is used.

 

[Elric]

I don't know if this will ever happen, but if one capturing product comes out victorious and we end up with a de-facto standard then yes, i can totally see captures going the way of IRs.

NAM is likely the best candidate. Becoming an industry standard comes with drawbacks though; namely, the fact that protocols evolve (much) slower than platforms, so improvements on the NAM core would be quite difficult to materialize.

Yes... look at how slow IRs as a technology have evolved, too. If that is indicative, I would not expect it to evolve at super fast rate... the MI industry != tech. There's no standard for metadata, everyone is trying to move to a model where it is like an abstraction of of a speaker cab and interpolating between different positions but all the manufacturers are doing it independently of one another so you have a bunch of proprietary stuff on top of it, so nobodies stuff works with anyone else's etc. Jay is still the only person in the world that seems able to do a proper Far Field IR. etc.

The core technology is really cool and captures are useful in and of themselves, but having that core tech does not mean you are going to move to stuff that requires significant metadata, data management, much more complex/costly measurement paradigms quickly (i.e. shooting a cap for 'every' knob position on an an amp, taking measurements, etc).

I do think it could make capture capability a commodity item on H/W units in the near future. I.e. "NPD: Nux Mini Capture Player Pedal" . And I do think that is a big deal. A couple of caps is enough for a lot folks.

 

[Will Chen]

much more complex/costly measurement paradigms quickly (i.e. shooting a cap for 'every' knob position on an an amp, taking measurements, etc).

That's not what will happen. IMHO what will happen is the merging of component based modeling with front side profiling/capturing. Clearly it won't work for everything immediately, but having the user input metadata into a system would allow the algorithm to select an appropriate backend model (or even build one on the fly) and mirror the input control settings. Theoretically the controls would match real world counterparts.

For example, user inputs JCM 800, bass 4, middle 8, treble 8, Master 10. The underlying engine selects the model with the controls at the given positions and then does a tone match to create an IR on the fly. But what if it's an amp the engine's never seen? You look up the specs and input them: 5X12AX7 gain stages, Mesa tone stack, 2XEL-84 power amp, etc.

 

[Elric]

That's not what will happen. IMHO what will happen is the merging of component based modeling with front side profiling/capturing. Clearly it won't work for everything immediately, but having the user input metadata into a system would allow the algorithm to select an appropriate backend model (or even build one on the fly) and mirror the input control settings. Theoretically the controls would match real world counterparts.

For example, user inputs JCM 800, bass 4, middle 8, treble 8, Master 10. The underlying engine selects the model with the controls at the given positions and then does a tone match to create an IR on the fly. But what if it's an amp the engine's never seen? You look up the specs and input them: 5X12AX7 gain stages, Mesa tone stack, 2XEL-84 power amp, etc.

I dunno... I guess that would be cool... but I don't see any reason to think it'd yield better results than a component model. It might be a good approach to making profiles 'less sucky' to manipulate after the fact. Going much beyond basic frequency response effects, there are going to be a lot of technical problems with manipulating the profile in a realistic way; even with a component model of the tone stack and other pieces.

In solving those problems, you are basically going to end up building a legit component model and using it to interpolate changes to the capture. And if you have good component model, WTF do you need the capture for... The developer might be likely better off shooting a bajillion captures and trying to morph between them or something. Maybe have your 'AI' listen to you sweep the knobs while a test signal plays or something.

It's also pretty naive to think you could just put in a little bit of data like that and characterize some unknown amp. And it would be delusional to think a significant number of users on some exchange like tone.net or similar are going to provide all the metadata about their caps, etc. Some commercial folks would be willing to do so but I think, like IRs, there's never going to be much money in this market and you are not going to see any kind of industry rally around that. Everyone's going to try to monetize their own thing and the whole market is small potatoes.

 

[Sascha Franck]

I think component modeling will stay for quite a while but I'm pretty sure that once capturing is widely spread (ideally using one common format...), we will see a whole bunch of smaller blocks you'll be able to combine. At least that's how I would utilize things.

 

[Will Chen]

I dunno... I guess that would be cool... but I don't see any reason to think it'd yield better results than a component model. It might be a good approach to making profiles 'less sucky' to manipulate after the fact. Going much beyond basic frequency response effects, there are going to be a lot of technical problems with manipulating the profile in a realistic way; even with a component model of the tone stack and other pieces.

In solving those problems, you are basically going to end up building a legit component model and using it to interpolate changes to the capture. And if you have good component model, WTF do you need the capture for... The developer might be likely better off shooting a bajillion captures and trying to morph between them or something. Maybe have your 'AI' listen to you sweep the knobs while a test signal plays or something.

It's also pretty naive to think you could just put in a little bit of data like that and characterize some unknown amp. And it would be delusional to think a significant number of users on some exchange like tone.net or similar are going to provide all the metadata about their caps, etc. Some commercial folks would be willing to do so but I think, like IRs, there's never going to be much money in this market and you are not going to see any kind of industry rally around that. Everyone's going to try to monetize their own thing and the whole market is small potatoes.

There's no interpolation here. What I'm essentially talking about is an extension of Fractal's Tone Match capability. Honestly, if the analysis is good enough it could potentially be smart enough to set the gain by itself, EQ would absolutely need to be manually input due to the mic type/position and room having so much influence on the tone

And based on my experience with the Kemper, I have absolutely zero interest in the crowd sourced concept here and way more interest in known producers/artists selling packs. And I say that as someone who generally feels buying presets is a waste of money.

 

[Whizzinby]

I dunno... I guess that would be cool... but I don't see any reason to think it'd yield better results than a component model. It might be a good approach to making profiles 'less sucky' to manipulate after the fact. Going much beyond basic frequency response effects, there are going to be a lot of technical problems with manipulating the profile in a realistic way; even with a component model of the tone stack and other pieces.

In solving those problems, you are basically going to end up building a legit component model and using it to interpolate changes to the capture. And if you have good component model, WTF do you need the capture for... The developer might be likely better off shooting a bajillion captures and trying to morph between them or something. Maybe have your 'AI' listen to you sweep the knobs while a test signal plays or something.

It's also pretty naive to think you could just put in a little bit of data like that and characterize some unknown amp. And it would be delusional to think a significant number of users on some exchange like tone.net or similar are going to provide all the metadata about their caps, etc. Some commercial folks would be willing to do so but I think, like IRs, there's never going to be much money in this market and you are not going to see any kind of industry rally around that. Everyone's going to try to monetize their own thing and the whole market is small potatoes.

100%. If a model is going to be used to fill in the gaps of a capture, just use the model and save the effort!

Having used two capture devices in the last couple years, I’m already over capture tech in its current form. It’s just a pain in the ass unless you are looking to capture your own amps.

 

[paisleywookiee]

There's no interpolation here. What I'm essentially talking about is an extension of Fractal's Tone Match capability. Honestly, if the analysis is good enough it could potentially be smart enough to set the gain by itself, EQ would absolutely need to be manually input due to the mic type/position and room having so much influence on the tone

And based on my experience with the Kemper, I have absolutely zero interest in the crowd sourced concept here and way more interest in known producers/artists selling packs. And I say that as someone who generally feels buying presets is a waste of money.

I continue to have zero desire to keep buying packs, just to get what I want. It's the "in-app purchase" of the gear world, and it sucks.

 

[Coalface1971]

Going against the grain.

I'm in my 50's and find this stuff frightening. It makes me feel old.

I nothing invested here, other than owning a bunch of real amps and some older-tech modelers, a ton of rack gear, pedals, and all sorts of shit that this "AI" tech will eventually make "obsolete". I've never bought gear as an investment. But with AI emergence I could have a worthless pile junk I've spent 35 years collecting. Quite a smack in chops, tbh. Like watching a motor car go by when the mode of transport was primarily horse and sulky.

I don't have any doubt that this technology will become very prominent, and though at present in relatively formative development stages, is already way beyond a concept.

Yep agree, it is another tool in the war-chest if you want to treat it like that.

But can't help agreeing with BW (and Will Chen) that this one has a lot more hair on it going forward. The market will drive this tech onwards and having very little corporate constraint (at this stage) to stifle it, and will be very good before we know it.

The power in this stuff is the virtually unlimited access for the individual.

If this is a way to "cheat" and get "that" sound, actual musicians usually have deep dreams, and empty pockets, and are very adept at the 5-finger discount. Doesn't matter if it's a form of piracy or not. It's going to happen.

 

[Sascha Franck]

Going against the grain. [...]

You could've said all this about any modern technology, especially once it comes to digital.
Computers render tape machines a worthless pile of junk - true. Plugins render many HW synths obsolete - true. Plugins render many outboard FX obsolete - true. Etc. Be it so.
They're all just some of the major steps in the democratization of music production. And I applaud to that as much as it gets.
And fwiw, I'm saying that as someone who had spent his last dime on f*cking expensive tapes for a mere Fostex G-16. Or another truckload of cash for a mere Atari ST. Heck, the update from 1 to 2 MB (yes, MB!) of RAM, just so I could run Cubase's Score modul, cost me more than a very decent PC these days.
I neither miss these old days (well, I partially do, but for other reasons) nor did I ever expect to get any of my money back.
In fact, I absolutely love being able to more or less fully enjoy all this modern technology.

 

[Mike Oliphant]

I nothing invested here, other than owning a bunch of real amps and some older-tech modelers, a ton of rack gear, pedals, and all sorts of s**t that this "AI" tech will eventually make "obsolete". I've never bought gear as an investment.

I think there will always be a market for classic gear - even if (as with a lot of HW synths) it has been reproduced in software beyond any reasonable doubt.

But technology does march on. You can buy a chip these days for a few dollars that outperforms systems I once spent thousands on. I spent the equivalent of thousands of dollars for my first 20mb hard drive. So it goes.

Has some of the "magic" been lost? Maybe a bit - there is something to be said about working within constraints. Still, I continue to find plenty of magic in the possibilities being opened up by new technological advancements.

 

[Coalface1971]

You could've said all this about any modern technology, especially once it comes to digital.
Computers render tape machines a worthless pile of junk - true. Plugins render many HW synths obsolete - true. Plugins render many outboard FX obsolete - true. Etc. Be it so.
They're all just some of the major steps in the democratization of music production. And I applaud to that as much as it gets.
And fwiw, I'm saying that as someone who had spent his last dime on f*cking expensive tapes for a mere Fostex G-16. Or another truckload of cash for a mere Atari ST. Heck, the update from 1 to 2 MB (yes, MB!) of RAM, just so I could run Cubase's Score modul, cost me more than a very decent PC these days.
I neither miss these old days (well, I partially do, but for other reasons) nor did I ever expect to get any of my money back.
In fact, I absolutely love being able to more or less fully enjoy all this modern technology.

Sorry Pal, not sure if you're trying to argue with me or agree with me

 

[the-trooper]

Sorry Pal, not sure if you're trying to argue with me or agree with me

 

[Sascha Franck]

Sorry Pal, not sure if you're trying to argue with me or agree with me

I'm not trying to argue, but I don't agree, either. I just happen to love modern music related technology for a whole lot of reasons. For me it has been liberating in many, many aspects.

 

[Stone]

One has to wonder what Fractal and Line 6 are thinking based on this free open source profiling ?