IK Multimedia TONEX

*cough*

What's the verdict on this one? Worth the $50 for SE?

Been trying to play around with the CS version but just about everything has white noise. Seems cool, especially with Amplitube 5, for recording.
 
*cough*

What's the verdict on this one? Worth the $50 for SE?

Been trying to play around with the CS version but just about everything has white noise. Seems cool, especially with Amplitube 5, for recording.
I'm 100% yes on this.
 
*cough*

What's the verdict on this one? Worth the $50 for SE?

Been trying to play around with the CS version but just about everything has white noise. Seems cool, especially with Amplitube 5, for recording.
Fully agree with everyone else. 100% yes it's worth the $50. Included profiles didn't really click with me (although I probably should give them another go), but there are some fantastic free user profiles.
 
SE is worth $50 if you get some of the best freebies on the share site. Tim(LRS has been a staple of my Kemper for awhile) has some great freebies on Tone.net. Also check out AmalgamAudio(search as typed), and SinMix(another Kemper profiler of HQ). Jason Sadites of Helix fame has some captures for sale on his site that you import as presets.

More and more will been coming out, when the pedal arrives. IMHO.

I port the Tonex amp only into Blue Cat Patchwork (a great VST host, like Gig Performer) and use other brands of VST FX for delay, verb, mods.
 
Picked up both the Amplitube 5 (standard version) and Tonex SE while they were $49 each. Some of the Tonex captures are really good, others not so much, will take some time to dig through. Amplitube is already bugging me with the paywalls for half the effects though. But I wanted to see how good it was in standalone with Tonex for additional effects.

Also it's funny on the website it's cheaper to buy the Tonex SE ($49) and then the upgrade to Max ($149) than it is to buy Max outright ($249).
 
Open-source neural amp modeler, seems to do better than the ToneX:



Yep. Saw that. Quite amazing .... lets hope one of the "big players" license this tech so we can see and use it one day soon-ish :)

In the meantime, as I "quickly reviewed" in the other place, I am stunned at how accurate Tonex is compared to the KPA and QC.

See this video below at the 19min 11 sec mark for the "null testing".



It is simply astonishing how much more accurate the Tonex is than the KPA and QC.

As usual, good times :)

Ben
 
Last edited:
Steve’s doing awesome stuff with NAM and there’s a great community growing around it. IMO focussing on amp only captures is preferable.

It’s also quite easy to make your own and the process is getting simpler with every update. Still not quite like ToneX where you get walked through it but it’s not too difficult. I think what he’s doing should keep commercial devs on their heels.

I will say that even despite these being far more accurate than Kemper, Kemper was already in a realm of being close enough not to matter - hence why it’s found in so many studios and bands back lines. It’s reliable, offers lots of features, profiles quickly, has all the I/O built in. The tech may not be cutting edge any more but the product itself is solid despite the compromises.
 
In the meantime, as I "quickly reviewed" in the other place, I am stunned at how accurate Tonex is compared to the KPA and QC.
I did write this article a while back now:


ToneX was measureably better than KPA and QC. QC was measurably better than KPA. Which is also what my ears tell me.

Kemper was already in a realm of being close enough not to matter
That's not true for me tbh Ed. After 11 odd years of owning 5 different iterations of the Kemper... I had a moment of clarity with it - I've used the Kemper on some albums, but I've never liked the low end, and it always sounded different to the real amp to me. People blamed my DI box, so I got a posh one. Then people blamed my microphone, so I tried others. Then people blamed my refining process, so I tried loads of different options. Then people blamed my ears, and I stopped short of cutting them off and realised that most of the guys on the Kemper forums are just total fangurls and apologists.

Kemper v2.0 if/when it happens will most likely be amazing, taking advantage of machine learning techniques developed in the last 10 years. But the Kemper as it is... dead duck to me.
 
That's not true for me tbh Ed
You’re not alone, (I know plenty of others who feel the same) but that doesn’t negate all the professional situations that tons of others have been using them without problems. There’s definitely a compromise on tone quality but it’s made it’s way on to countless albums and touring rigs without anyone giving a shit or even noticing.

There’s all kinds of gear that fits that category, that I may personally not like but I can’t deny it being used by lots of people just fine. Is Waves SSL the best SSL channelstrip out there? is it still being used constantly by the biggest mixers and all over songs in the charts?
 
Open-source neural amp modeler, seems to do better than the ToneX:

Maybe, but you need a software engineering PhD to train a model.

Download the 'repo', install Anaconda for package management, then activate the environment, modify a bunch .json files, run a million lines of code, export model with more lines of code, 1000 epochs is typically overkill and keeping an eye out for when the ESR drops below 0.01.

 
Last edited:
Open-source neural amp modeler, seems to do better than the ToneX:
Neural Amp Modeler is cool, no doubt about that, but a lot of the heavy lifting is done by the PyTorch machine learning library. That's where all the "magic" lies. NAM is basically just about setting things up to feed it data, setting how the library should handle that data and storing the model.

This is not meant to diminish the effort of its developer, but just to highlight how much of it some smart people have built as is by making the PyTorch library and others like it. Even I could - with a good bit of effort - make something like NAM despite the only machine learning programming I've done is make Google's Tensorflow be able to determine if an image is a cat or not, which is basically tutorial level stuff.

By comparison I would not know even where to begin for something like what Line6 or Fractal do with component based modeling. I absolutely do not have the math skills for it. I've never studied computer science, my degree is in digital communications but I've worked as a software developer for about 15 years on all kinds of web/mobile based systems whether it's user interfaces, mobile apps, backends to process data, databases or cloud services.

Even if Fractal says that his component based approach gives more accurate results, ML-based amp modeling is good enough for most users, especially if it comes with a significantly lower price tag. The problem with it is that it can't really do anything new, just copy existing things.

I am quite confident for example UA pedals are ML models because the UA Ruby's treble/bass controls only work on the Brilliant channel - just like the real vintage Vox AC30s. A sensible approach on a digital modeler would be to say "hey, turn those to this setting for authentic operation or use them as you prefer" and offer both because "these knobs do nothing unless you have this channel selected" is not really a great user experience.

By comparison a component modeling system like Strymon Iridium can say "hey, if you want to push this Fender Deluxe Reverb based model to more of a tweed sound, just turn up the mids!" so it can be more than the original. Which IMO is the beauty of digital modeling - it can be authentic if you want, but also go beyond the limits of the gear modeled by giving users control over aspects of the sound.
 
Last edited:
I am quite confident for example UA pedals are ML models because the UA Ruby's treble/bass controls only work on the Brilliant channel - just like the real vintage Vox AC30s
UA are typically very much component modelling based, and only do black box type modelling where it makes sense (and for small portions of the circuit). I don’t think even their black box approach is ML based at all, I don’t think it’s how they operate as a company or where their tech is at. There’s quite a few interviews with their devs scattered around online and also their old blog posts that talk a bit about their process.

Maybe, but you need a software engineering PhD to train a model.

Download the 'repo', install Anaconda for package management, then activate the environment, modify a bunch .json files, run a million lines of code, export model with more lines of code, 1000 epochs is typically overkill and keeping an eye out for when the ESR drops below 0.01.

tbf, it’s got a lot simpler. if I can manage to do it then I’d say anyone can. Was essentially a case of dragging the corresponding .wav files into my browser and checking a few boxes. There’s a plugin now and only one file per capture (previously it was two).

 
Back
Top