Tonex -vs- NAM ..... the Tonex "Aliasing" Low Pass Filter is at ~ 14kHz

[Richard_G]

(…) I heard a difference and then tried to reconcile it (…)

FWIW, we don’t know if Karlis at Amalgam made the NAM and Tonex captures @BenIfin used/compared here at the same time with the same excact setup and conditions, do we?

My impression is that his Tonex and NAM captures often are made and/or released at separate times.

This could be a source of differences between the NAM and Tonex captures as well.

(Edit: I didn’t jump in here to «defend» Tonex or anything. I don’t question whether NAM is more accurate. And I’d wish both had better aliasing performance.)

 

[Ed DeGenaro]

Cliff, perhaps?

Ya think a rep has the same interest as an owner?

 

[Richard_G]

I stated no such thing.

 

[BenIfin]

FWIW, we don’t know if Karlis at Amalgam made the NAM and Tonex captures @BenIfin used/compared here at the same time with the same excact setup and conditions, do we?

My impression is that his Tonex and NAM captures often are made and/or released at separate times.

This could be a source of differences between the NAM and Tonex captures as well.

(Edit: I didn’t jump in here to «defend» Tonex or anything. I don’t question whether NAM is more accurate. And I’d wish both had better aliasing performance.)

By way of clarification - this was all confirmed by Karlis [AmalgamAudio] at T.O.P when he started to release his DI only sets:-

=> its always the same Amp
=> each Tonex and NAM Capture have the exact same settings
=> he does not use a Load Box of any kind
=> he taps of the Speaker so the Amp/Speaker Impedance is baked-in to his Tonex and NAM Captures

So in short, they are going to be as close to 1:1 as you can get.

Having said that, the differences are not really tonal - to me - the top end in the NAM Captures is clearer and cleaner and the mids in the Tonex Captures are more congested. If you have a matching set, just set up a tone with the same capture for each and turn it up to loud/gig levels - its really not hard to hear at all.

After much time trying to dial-in the Tonex Captures to sound like the NAM Captures, I just gave up as I could never get there even though that was what I really, really wanted to do ie: use Tonex not NAM.

ie: you cant add high end definition if the frequencies just aren't there and you can cut mids all you want, but the baked-in congestion does not go away - EQ is great, but it cant add or remove what's baked-in, or not baked-in to the core tone.

Not only is it quite obvious [to me] as you A/B them at increasing volumes as I have done -but- run a Tonex or NAM "dirt" pedal Capture into them - I use the Sultans of Stomp Marshall Bluesbreaker Pedal Captures [brilliant by the way] - and the Tonex Amp Captures gets very-very mid-heavy very quickly, whereas the NAM Captures respond pretty much as you would expect.

Maybe I've got shit hearing - maybe I've got Jesus ears - I don't know - but I know what I hear and what I prefer.

As always - each to their own.

 

[dronerstone]

Well not in my world...but mkay

So, the irony store was closed already?

 

[DLC86]

A 14K LPF is 3db down at 14K, while the guitar amp's signal through a relatively bright Celestion V30/IR is some 30db down at 14K, so you guys are saying that that 33db down vs 30db down is a deal breaker?

OK, so now let's add a relative bright high quality large capsule condenser mic to the equation like a Neumann U87, and it's another 6db down at 14K, so now you guys are saying that 39db down vs 36db down at 14K is a deal breaker?

Also, ToneX has EQ as well, so adding 3db of top end is not hard should the dog ear'd among you really feel the need LOL!

That said, NAM is more accurate so I'm looking forward to watching products using it develop going forward, but there's no technical reason that ToneX can't get the job done IME.

It's not just 3 dB down at 14 kHz... I posted this a while ago:

It's already down 3 dB at 2kHz, roughly 12-15 dB down at 14 kHz and 30 dB down at 20 kHz... and I think this also varies depending on the capture, so might be even worse in some cases

 

[Deadpan]

As always, each to their own. I can hear a difference in the real world when I play live/loud between the same Tonex and NAM Captures, so I know which is nicer to my ears. Its all good.

There are no identical captures on Nam and Tonex. You can capture on the same device twice in a row and get different results.

 

[BenIfin]

It's not just 3 dB down at 14 kHz... I posted this a while ago:

View attachment 40598

It's already down 3 dB at 2kHz, roughly 12-15 dB down at 14 kHz and 30 dB down at 20 kHz... and I think this also varies depending on the capture, so might be even worse in some cases

Yep. There's no doubt about the very severe drop off at ~14kHz. There is also some very strong 'filtering" going on below ~14kHz - hence why in the aliasing charts I posted here it took me to apply a ~5kHz Low Pass Filter to get the NAM aliasing at -90db to be the same as the Tonex aliasing at -90db.

There is clearly some "special juju" filtering going on in the Tonex Capturing process.

Doesn't mean Tonex isn't great - it just means it is what it is.

 

[djd100]

What sample rate are you guys using with ToneX? 44.1 is it's native SR, use that for best aliasing performance it's been said.

Gig volume monitoring begats midrange "congestion"? What exact SPL is your gig level monitoring? With what monitors and in what treated and designed rooms are these critical listening tests happening in?

What are the test procedures, test equipment, monitoring scenarios and control for all these guys getting such different test outcomes?

One guy says ToneX is down at 14K, another guy says it's down at 2K, another says 5K etc, and I know that my ToneX captures are certainly not rolling off at 2K or 5k, and that any aliasing is not an audible issue in real-world usage as guitar tube power amp sims through guitar cabs/IR's etc, so sorry, it still sounds like BS to me (excepting 3db down at 14K which is certainly possible, but it's inaudible in real world usage as guitar tube power amp sims through guitar speakers/IR's).

The ToneX software certainly has/had issues, however the recent upgrade is a big step in the right direction, and for an inexpensive low latency capture device with a minimal footprint, the ToneX One can't be beat.

 

[DLC86]

it still sounds like BS to me

ok

 

[ian_dissonance]

Record a guitar preamp with no cab, add a LPF at 14Khz between the preamp and an IR. If you don’t hear a very noticeable difference your hearing is probably shot. Whether it’s a “good” sound or not is kind of irrelevant when the point is accuracy.

 

[EOengineer]

You guys don’t know what tone is until you’ve low passed your amps at 100hz. Good luck hearing the aliasing!

 

[Orvillain]

Whether it’s a “good” sound or not is kind of irrelevant when the point is accuracy.

MuH DyInG HiLL !!!!!!!!!!

 

[ian_dissonance]

You guys don’t know what tone is until you’ve low passed your amps at 100hz. Good luck hearing the aliasing!

Just replace the guitars with synths and then it’s even better.

 

[ian_dissonance]

MuH DyInG HiLL !!!!!!!!!!

I don’t personally care, lol, but that is literally the point of profiling/capturing no?

 

[Orvillain]

I don’t personally care, lol, but that is literally the point of profiling/capturing no?

Yes I agree.

Lots of people here and "abroad" don't feel that way. They'd accept any old dogshit.

 

[djd100]

Record a guitar preamp with no cab, add a LPF at 14Khz between the preamp and an IR. If you don’t hear a very noticeable difference your hearing is probably shot. Whether it’s a “good” sound or not is kind of irrelevant when the point is accuracy.

No, do that and decide whether the ultra high-freq differences actually make a meaningful difference in a music production (seeing they're often LPF'd out of the mix anyway...), as the goal is not accuracy, rather it's whether the sound works for the real world production at hand. 3db down at 14K when it's already 30+ db down in the mud is not an earth shattering sonic problem IMO (YMMV).

NAM is more accurate, never argued that it wasn't, but my ToneX One is not rolling off at 2K or 5K, and it works fine for typical electric guitar tube power amp sim uses in the real world.

Lab development and real world usage are two different things, and each have their place, though all the internet development engineers here posting wildly different test outcomes without publishing all the details of their test procedures, test equipment, and monitoring equipment/environment aren't proving anything.

Good night kids...

 

[EOengineer]

The Tonex rolloff has never stopped me from being able to use Tonex as a productive tool, be it rehearsing, recording, whatever.

With all things modeling/capture though, accuracy is what we’re after. I’d love to see IK improve their accuracy, but I’m fine as is too.

 

[DLC86]

all the internet development engineers here posting wildly different test outcomes without publishing all the details of their test procedures, test equipment, and monitoring equipment/environment aren't proving anything.

I don't see any "wildly different" outcome here, my tests and @BenIfin 's tests are pretty much in agreement (even though he used a spectrogram which is not the best type of graph to show and measure a roll-off).
I've shared details about all the tests I've made during these months/years on this forum, just open my profile and search thru my posts if you're so interested in seeing those.
I surely won't do that for you (or anyone else) and I surely won't write each of my posts containing statements of this kind like it is a research paper... if you expect me to do so then I expect to be payed for it!

 

[ian_dissonance]

No, do that and decide whether the ultra high-freq differences actually make a meaningful difference in a music production (seeing they're often LPF'd out of the mix anyway...), as the goal is not accuracy, rather it's whether the sound works for the real world production at hand. 3db down at 14K when it's already 30+ db down in the mud is not an earth shattering sonic problem IMO (YMMV).

NAM is more accurate, never argued that it wasn't, but my ToneX One is not rolling off at 2K or 5K, and it works fine for typical electric guitar tube power amp sim uses in the real world.

Lab development and real world usage are two different things, and each have their place, though all the internet development engineers here posting wildly different test outcomes without publishing all the details of their test procedures, test equipment, and monitoring equipment/environment aren't proving anything.

Good night kids...

Well no, the whole point of profiling is dead-on accuracy. If “useable for a production” is the only criteria than literally any device out today will get you there. A POD bean from 20 years ago is “useable for a production”. I don’t disagree that it’s a non-issue in most cases, that was never the point of the discussion here. So far I’m actually quite enjoying Boss modeling and as far as accuracy goes that shit sucks, lol. But that’s not a capture of a real-world piece of hardware (in the case of the X-amps anyways).