-
NAM works on the premises that it’s modelling a linear system. The system is not linear at all. Perhaps Kemper deals better with different volumes because it doesn’t accurately replicate each frequency and amplitude in a single way.
Originally Posted by Spook410
Sent from my iPhone using Tapatalk
-
06-26-2026 08:45 PM
-
it doesn't really matter if the capture is more accurate. It's more important that the amp is captured at close to the user settings that a particular user wants to hear. I bought a twin reverb capture once that folks were raving about and to me, it sounds horrible. I asked the seller what settings he used, "all knobs dimed" he replied. Originally Posted by Spook410
Oh, ok...
Since there are few jazz guitarists capturing these things, most of the captures are with amp settings that are more usable for rock/blues. I think we've also reached a point of diminishing returns for most of this stuff.
-
You are, of course, absolutely correct. And that's where I'm at now: sorting through captures. There is enough to start getting an idea of the tech but not nearly as many good ones as I was expecting. As you say, there are few captures at the settings many jazz players want. Even on the amps we want to use. Still, there's a decent Super Reverb, Twin, and a couple different Deluxe's. Could live with that but was hoping for more. Maybe I'm wrong about NAM open source establishing some sense of standardization and we'll see more white box modeling or even something none of us are expecting. Originally Posted by jzucker
Thing is, players need to try new things.
-
Neural amp modeling is not linear and it does not assume the amp to be linear. Originally Posted by Eck
-
The creator compared it with IR capture "Atkinson:The simplest way is to think of it like making an impulse response of a speaker cabinet, but instead of just capturing how a cab sounds, NAM can capture the whole behavior of an amp or pedal, or the combination of those things. " Originally Posted by Spook410
IR capture is linear. If I made something that was non-linear, clearly a feature, I would be careful to draw the distinction between it and a linear capture.
-
Why don't you download a NAM file and use it in you computer DAW.
See if you like the sound or not.
Fender Deluxe Reverb NAM file link below.
Fender Deluxe Reverb (A2) NAM Profiles by @sdatkinson
It's as easy as that.
-
The DR preview doesn't produce sound for me, Originally Posted by jzucker
my speaker makes the faint "dup" sound same
as when I load any page with a sound playback.
Windows 7 sp2 Home Premium 2009
Opera Browser version 95.0.4635.90
-
maybe time to upgrade? Originally Posted by pauln
-
First you have to add a DAW plugin. Running these models requires significant processing power and a reasonably fast processing chain (chipset, bus speeds, cache size & speed etc). The more capable your computer, the better NAM models will run in your DAW. So the simpler and “lighter” (ie less demanding of computer power) your DAW is, the more resources your computer can devote to modeling. The fidelity and sound quality of modeled tone depend on having enough computing power. But interestingly enough, NAM models use very little RAM. Originally Posted by GuyBoden
The original NAM plugin is excellent but does not have built in effects and requires a separate IR loader to add mics, speakers & cabinets. It’s an excellent NAM renderer that’s easy on the CPU and runs well even on older or lower powered computers. This is the first one I tried.
But there are easier and more complete alternatives now. I like NAM Universal, which is complete with IR annd effects and makes pretty efficient use of the CPU - and it’s free. You can download it at wavemind.com
-
If this is addressed to me, I fail to see how it's relevant to anything I've said in this thread. Originally Posted by GuyBoden
Why don't you download a NAM file and use it in you computer DAW.
See if you like the sound or not.
Fender Deluxe Reverb NAM file link below.
Fender Deluxe Reverb (A2) NAM Profiles by @sdatkinson
It's as easy as that.
-
Yes - it is. But amp modeling is not. It has to capture the real world behavior of the electronics in front of the speaker, and they’re definitely not linear. All the things that give our amps their tonal character are nonlinear, eg compression, harmonic distortion, and power supply voltage sag. Pick response is a classic nonlinear aspect of an amplifier’s behavior that challenges amp modeling. Originally Posted by CliffR
Impulse response captures sonic content like room reverb, mic output (up to the maximum SPL the capsule can handle), phase shift, frequency response, and room resonance. These are fairly linear in that they all get louder in proportion to signal level (e,g, SPL in the case of speakers, and voltage from microphones).
But amplifier behavior like OD and other distortion based effects is nonlinear. They have thresholds, and their characteristics change constantly but not linearly as signal level (ie pickup output) changes. For example, edge of breakup becomes crunch and then fuzz, and the transition points change if you change EQ, add compression, or use a limiter. They also interact to make those changes dependent on each other
-
No, it's only a general comment. All good. Originally Posted by CliffR
-
Sure, I get that. You have a much deeper understanding of how amps work than I do, but I do know they're non-linear. But there's nothing in the NAM's guy's description of what he does that suggests he's capturing that non-linearity. In fact, he seems to imply the opposite. But this is just supposition based on some statements he made. Originally Posted by nevershouldhavesoldit
-
I first used NAM files in my computer DAW 2-3 years ago. Originally Posted by nevershouldhavesoldit
I used them everyday.
-
Originally Posted by CliffR
As you posted.. he said "but instead of just capturing how a cab sounds, NAM can capture the whole behavior of an amp or pedal, or the combination of those things."
The whole point of the neural net is to model an amp's behavior. It's not linear. It's dynamic.
Simple internet search can provide a tutorial on the tech.
-
What NAM plugin do you use? Originally Posted by GuyBoden
-
Yep, it seems I was wrong Originally Posted by Spook410
.
-
One thing I liked about Fractal's technology is that the design is not based on capturing. It's based on circuit recreations. I don't know the details but from what I gather, they have modeled the actual components down to the capacitor and resistor level. They then feed the software a digital version of the circuit diagram. From there they are able to change component values and various other modifiers to "dial in the sound". But what makes this compelling to me is that they are able to create new designs that live in the virtual world only. They are not reliant on captures and on the imperfection of modern or vintage tubes. And you can vary the amount of gain in the circuits in a way that is beyond what a capture can do, after-the-fact. Originally Posted by Spook410
OTOH, when I had to send my fractal in for repair, I bought a kemper because I was in the middle of a recording project. I had been using a few amps, primarily:
- Fender Vibroverb,
- Marshall Plexi,
- Dumble
I thought the kemper had a feel that was much closer to the real thing. And when I got the fractal back and continued with my recording project, the difference was enough that I ended up keeping the kemper.
The thing is, I purchased 300 profiles in order to find the 3 that were really stellar. I have since found a couple dozen that I really like but rarely use anything other than those 3. I do really like the tonejunkie John Mayer Continuum capture with my strat and have used it on a bunch of my recordings.
But out of 300 profiles i use 4 or 5. Between 1 & 2 %.
-
Is the only purpose of digital "modeling" of amplifiers, IRs, etc., to merely re-create old technology in a more convenient and portable package, or is there an opportunity here to develop something completely new in the tonal structure of guitar music? It seems to me that if the only use of NAM and other digital technologies is to reproduce 50-60-70 year old amps, we are missing the potentials of these tools. I don't personally know what that would be or sound like, but some creative person out there might be able to do something.
Our fundamental conservatism as a musical genre and technologically shackles us to the past. If the only good sound is a Fender Twin, as some would have us believe, then there is simply no point to further technological development. The pinnacle has already been reached and it's all downhill from here. It's like the creed that all jazz has to be rooted in the bebop language; future development is shut off.
Had Allan Holdsworth, Kurt Rosenwinkel, Bill Frisell, etc., believed those things, they could not do what they do with their instruments. Will jazz remain mired in its past, incapable of further development and instead only elaborating and repeating what has been done before? This applies to technology as well as to technique, theory, etc. Holdsworth's creativity demanded technological development and technological development in turn fueled his creativity. The great visionary jazz musicians (Charlie Parker, Miles Davis, John Coltrane, Ornette Coleman, etc.) were resolutely forward thinking. Is there a parallel to this in guitar application technology with digital tools like NAM?
-
I thought I just answered that
-
Innovation does matter. But maybe it's like learning to paint still life and landscapes before you get into your Pollack meets Picasso phase. Get the tech working well enough to invite experimentation. Also, and I'm likely more the norm around here in this, I am probably not going to be breaking new ground when it comes to jazz tones.
I will say captures, just like some modeling and IR's, can disappoint. For example, I was really looking forward to trying out a Fender Panoverb (15" plus 10" tube amp). Loaded up the capture and.. meh. Sounded like the mic was just on the 10" speaker and even though it was supposed to be a clean profile, seemed the amp was set both louder and more 'brassy' than I would choose. Not usable and probably not the best the amp can do for a good jazz tone. That being said, I've been playing a free Super Reverb capture from Tone3000 (I believe it's their most downloaded nam file) and it's pretty good even on my $64 Valeton GP-5 running nam conversions. There's also a DeLuxe and Twin that have potential and that's before exploring nam files for purchase. I found with IR's there was quite a difference between what I could download free and the stuff from York Audio. It's possible better captures can be had with a small investment.
It's a bit to unpack but not really hard. And the only way to know is to try new things. Something many around here seem quite averse to even when it's cheap and easy.
-
Innovation only matters up to the point of diminishing returns and we've already reached that. The kemper is more than good enough quality for what we are trying to do. Guitar amps and guitar speakers have an extremely narrow frequency range. We are grasping at straws here. The tones that I get out of the kemper are as good as any real tube amp that I could mic up in my studio. Maybe a $200k studio in LA could get better tones but the tones I get are way better than what I could get with a real tube amp in the room. Originally Posted by Spook410
Reply With Quote
Guitar Amps built into case?
Yesterday, 05:03 PM in Guitar, Amps & Gizmos