IK Multimedia TONEX

If you don’t think the EVH packs capture the EVH sound, I think there is a problem somewhere on your end, or Eddies tone just isn’t to your taste.
Yep. I could live off the variety just in these brown packs.
For reference, I play an ebmm axis into an audient id44. Audient at 0db, tonex software input at 4.6db. Amp gain flat. Bad gal 68v lr hot. I can get eob with guitar volume all the way to rippin. Adjust presence as needed but mine is at noon.
 
Must be the weekend for it, also knocked up a "How well is ToneX V2" video. TLDR its a massive update in terms of capture results and capturing workflow. It's gone from uninstalled to "this is pretty damn good", big turnaround. The results are pretty much splitting hairs with NAM now.

I've uploaded 45 captures like in the video to Tone3000 (NAM) and ToneNet... grab the caps and/or watch the video if you want to see them in action. But yeah, well done to IK. I only wish it launched this polished, but we're here now at least




Thanks for sharing these! They sound fantastic with the York Audio Mesa 4x12 Oversized IR. I haven't been a fan of much of the EVH/5150 model captures thus far, but these are my favorite now. Really good idea to capture all of the gain levels - pretty cool to be able to "dial back" the gain by selecting another capture.
 
And here's the tone talk interview with the guy who made them.

 
Objectively the Brown Sound packs nail the EVH album tones, which are actually much thinner and brighter than people expect when heard outside the context of the album mix.

I found some of them to be quite nice with a LP or SG. I don’t care about EVH in the least but I do like some of those caps.
 
Objectively the Brown Sound packs nail the EVH album tones, which are actually much thinner and brighter than people expect when heard outside the context of the album mix.

I found some of them to be quite nice with a LP or SG. I don’t care about EVH in the least but I do like some of those caps.



Let her rip Ed

 
I’m so appreciative of the recent and consistent updates of Tonex by IK, but there are times where my dev/engineering brain kicks in and I wonder whether there’s a long term plan for Tonex. Here’s my issue:

IK (via tonenet) is in possession of a database with probably hundreds of thousands of rows of captures, training data, etc. I have to wonder what kind of dynamic next level neural next gen whatever amp model their engineers might create if all that training data had been normalized with calibrated levels and more detailed input about the amp settings/configuration.
 
IK (via tonenet) is in possession of a database with probably hundreds of thousands of rows of captures, training data, etc. I have to wonder what kind of dynamic next level neural next gen whatever amp model their engineers might create if all that training data had been normalized with calibrated levels and more detailed input about the amp settings/configuration.

I would say none. Even with calibration and better input and meta data, a crowd sourced data set is going to have way too much garbage in it. There will always be a lot of people who don't follow the directions because they are stupid, because they are lazy, or because they think they know better (and might be right or wrong). And then the amps being used, plus all the other gear, is not going to be comparable. There is way too much variation between older amps of the same model to learn anything of use by combining captures made of different amps by different people and hoping AI can find patterns.
 
I would say none. Even with calibration and better input and meta data, a crowd sourced data set is going to have way too much garbage in it. There will always be a lot of people who don't follow the directions because they are stupid, because they are lazy, or because they think they know better (and might be right or wrong). And then the amps being used, plus all the other gear, is not going to be comparable. There is way too much variation between older amps of the same model to learn anything of use by combining captures made of different amps by different people and hoping AI can find patterns.
Respectfully, I doubt you or anyone here (including myself) has enough awareness of where AI is heading to know what may or may not be feasible in the next few years, especially had IK taken better measures to normalize some of these variables up front.

I’d be very curious to see what kind of information could be gathered from a couple hundred or thousand direct captures, for example. Perhaps it’s nothing, perhaps it ends up being helpful.
 
I already said so in the other thread, IKs own collections seem to expect a very high input level. This can be sort of verified by the internal player's file leveling. All the prerecorded DI files peak very close to 0dB, regardless of the guitars that were used for these files. This isn't exactly in line with real life where different guitars result in different peak levels. It also isn't exactly in line with the all the interface input calibration databases, which, if you follow them, at least with anything that isn't a guitar equipped with super hot pickups, will never result in peaks that high.
It's also kinda verified by IK going for +8.5dB as the default on their Tonex pedals.
That kinda level range is in fact working fine with most of IK's own collections, but a lot of other captures (and by now I tried freebies of pretty much all commercial vendors, obviously the highly regarded ones included) will come out somewhat "overgained".

For myself, the "solution" is to keep the hardware at 0dB, which seems to work best for the very captures I think I will be using on my live board, then adjust the plugin input gain accordingly, so there's at least gain parity between the two scenarios. I will then just use the amp block's gain knob for things I plan to use on either platform. Fortunately, when trying out Tone Models, this stays in position, so one should only have to do this once per collection and then possibly save the interesting captures as presets, which will have the gain settings saved as well.

Very obviously, none of all of that is ideal.
 
For myself, the "solution" is to keep the hardware at 0dB, which seems to work best for the very captures I think I will be using on my live board, then adjust the plugin input gain accordingly, so there's at least gain parity between the two scenarios.
This is exactly what I’ve been doing for the last year+ and it’s working ok. It’s flawed. I get it and I don’t blame folks for staying away or having bad experiences.

The hope is that we see a more mature strategy in time. Ideally sooner than later. Revalver 4 had auto leveling calibration on their garbage profiling feature a decade ago.
 
I tried my hand at creating a capture. In this case it's a stomp profile of the Helix (Native) Cali Q - I wanted to try and use my Tonex One as a smaller replacement for the HX Stomp in the FX Loop of my Mark III that has no on board EQ. I think it might be a little quieter than unity gain but it seems to work ok. I uploaded it to Tone.net in case anyone cares :ROFLMAO:

 
Ok, did some tests for myself, to find out whether running my dirt pedals in one loop of the GT-1000, followed by the second loop with the Tonex One, would be a decent option. One might suspect latency to become an issue, but it's not. Not even by far. Latency of the GT-1000 with both loops running serially and a Tonex One in loop 2 kicks in at pretty much exactly 3ms! Yes, you're reading right. That's 4 full ADDA cycles and some processing. At 3-fucking-milliseconds!
I crosschecked everything multiple times because I absolutely couldn't believe it myself. A bit more about it here.

So, it's now time to get one of those MIDI-to-Tonex boxes and have a killer setup.
 
Well, I thought to get out my Quilter Superblock US and use it as a preamp, and… [boost > Quilter > Tonex power amp capture] ended up being one of my favorite guitar sounds in a while. I forgot how well the Quilter takes pedals, and the power amp capture really takes it to the next level and smoothes out the brittleness it can sometimes have.

Now I’m almost definitely in the market for a tube preamp… looking at the Kingsley, Tubesteader, and AS stuff…

So, I grabbed a Tubesteader Lightkeeper v2 (high voltage tube Dumble-esque preamp type thing like the Kingsley and AS stuff) and used it this way, and... damn haha. Really, really good. I recommend people trying this out if you haven't.

It's soooo nice to be able to hit it with loud fuzzes and boosts without worrying about clipping.

Also, off topic, but the built-in boost on this pedal sounds great, and the EQ is really flexible. I'm definitely going to be looking at more of their pedals (the Magnatone one sounds amazing).

At this point I'm thinking of trading my Tonex in for a Tonex One and keeping a couple power amps in one bank and a few full amp captures in another (I've gotta look into exactly how the bank switching works on the One, but yeah). Preamp pedal + Tonex One power amp capture seems like the ideal setup for me.
 
You know what's really annoying with the Tonex plugin? All the UI "locations" aren't saved within a DAW. They're even already getting lost as soon as you close and reopen the UI.
Now, I could perhaps understand that for the ToneNET part of the plugin, so you don't stay there all the time, possibly getting interruptions by updated capture lists or whatever. But in case I'm only browsing local files, I would really, really like to stay where I've been the last time I used it, especially when checking out larger collections - but no, it's always back to "All Tone Models".

@IK Multimedia, you could possibly forward that to your development team.
 
Hi everyone! I know this question has probably been asked a thousand times, but I’m really confused about setting the input level for ToneX App (not the pedal).


I mainly used NAM before, and there was no problem with calibration there. But with ToneX I’m not sure what approach I should follow. I have an Audient iD4 MK2 (12 dBu), and I just want to get the correct input level for listening to different capture packs—or at least understand a general approach to aim for.


At what level should I set the Input in the ToneX VST plugin for Amalgam captures if the gain knob on my Audient is at zero?
Or should I instead adjust the hardware gain on my interface to a specific point?
Or maybe I should follow IK Multimedia’s approach—turn it up until it almost clips, then back it off slightly?


Basically, I’m not sure where I should set the input level: in the VST plugin or on the interface itself.
I just need the simplest but reasonably accurate method to play through different profiles and hear how the captured amps sound.


Thanks for your understanding and help!
 
I mainly used NAM before, and there was no problem with calibration there. But with ToneX I’m not sure what approach I should follow. I have an Audient iD4 MK2 (12 dBu), and I just want to get the correct input level for listening to different capture packs—or at least understand a general approach to aim for.

Have a look here:
Your interface is covered.
 
Back
Top