Very Interesting Comment By James May on Tone Dexter

[gfirob]

I have really enjoyed this thread, especially the information-heavy posts from people who seem to really know their onions (Cuki79, Jonfields45 and of course James May). Reading the patent description was fascinating.

It still makes my head explode.

While the technical explanations are enlightening, they can hardly capture the deep satisfaction of hearing your guitar amplified and sounding like its true acoustic self. Finite Impulse Response Filter and milliseconds not withstanding, in the end we are talking about the human connection between a player and a musical instrument made of wood and bits of metal. The reason we love these things is because of the way they sound and for many of us the fragile and intimate nature of that sound is lost when we amplify the guitar through a pickup, no matter how much we have spent on preamps and other boxes to solve the problem. Thanks to James May’s work with that little red box, this is not longer true, and I really appreciate it and I also appreciate not having to really understand what a Finite Impulse Response Filter is in order to use it…

[James May]

Originally Posted by gfirob

... Reading the patent description was fascinating.

It still makes my head explode.

While the technical explanations are enlightening, they can hardly capture the deep satisfaction of hearing your guitar amplified and sounding like its true acoustic self....

I also appreciate not having to really understand what a Finite Impulse Response Filter is in order to use it…

I really appreciate these comments. We put a lot of effort into designing this technology to be easy to use, and of course we don't want musicians to have to think too hard, or understand digital signal processing, in order to use it. I think we've succeeded, because I have heard from may folks that it is surprisingly easy to use.

I take that as the second best compliment we can receive, after the first best compliment which was just expressed above.

[James May]

Originally Posted by Mr. Jelly

Anyone have an educated explanation of what he is trying to convey by characteristics? Please no guesses. I can guess as well as anyone.

I can elaborate. Think of the instrument as a complex resonant acoustical chamber that amplifies the relatively weak string vibrations, and adds complex frequency response, small echos, and varied radiation patterns.

It is not precisely a linear time invariant system (LTI), but for practical purposes it is ENOUGH of an LTI that ToneDexter can learn the impulse response that represents the difference between the pickup signal and what the mic hears.

Then, that impulse response is used during playback to transform the pickup and thus provides a very convincing picture of what the instrument actually sounds like to our ears. Transform in this case means nothing more than convolution using the IR that was created and stored - what we call a WaveMap.

[jonfields45]

It might be worth pointing out again that the reason an Impulse Response is so interesting is that in the time domain that we live in, they are really easy to generate. An impulse, if you drew one on graph paper, is just a very narrow, very tall (short and loud) spike.

In the frequency domain of equalizer knobs and Fourier's math, they become just a "1" (those center notches on your graphic EQ), which makes the impulse response of a system so useful for quick analysis.

Transform the impulse response with Fourier's math of a system that can be driven by one, an electric guitar amp for example, and you've got the equalizer settings (frequency response) to duplicate it (since the impulse itself disappears into a multiplication by 1).

Until ToneDextor creates it, there is no existing filter that takes you from your pickup tone to your mic's tone. Once that filter exists, it can be excited by an impulse and we can talk about it's impulse response, which is simply its programming. Reminds me of that chicken and egg debate!

[Peter Z]

James,
First of all, thank you for creating the ToneDexter!
As you announced, there will be a way to copy wavemaps from one unit to another with the next firmware. A great feature. Will it then be possible to apply an EQ to the wavemaps (i.e. In Cubase or protools), save it and reimport it to the ToneDexter?
That would open the door to a new world.

All the best
Peter

[Paultergeist]

Originally Posted by caballlero59

.......It's the same reason that you can play with a pick or with your fingers during training and get the same result." - James May

Huh......I play fingerstyle.....AND I use a flat pick....doing both/either at different times because I WANT to get different results......if a modeler is going to take the frequency input and make the output sound identical in either case, then aren't I losing something of the "me" part that I am trying to infuse into the musical expression?

I guess I am just not a fan of having a computer generate my tone.

[James May]

Originally Posted by Paultergeist

Huh......I play fingerstyle.....AND I use a flat pick....doing both/either at different times because I WANT to get different results......if a modeler is going to take the frequency input and make the output sound identical in either case, then aren't I losing something of the "me" part that I am trying to infuse into the musical expression?

I guess I am just not a fan of having a computer generate my tone.

I can see why my statement quoted above may have been confusing. By same result, I meant that the WaveMap learns the resonances and character of your instrument in either case. Then, the techniques you use to stimulate your instrument, such as fingers or pick, new strings or dull strings - all will cause the tone to be different, but all sounding like your instrument.

Put another way, the sonic essence of your instrument is there no matter how you play it. ToneDexter captures that essence.

And, it would be a more accurate statement to say that ToneDexter is modifying your tone, not generating it.

[guitaniac]

I'd like to add that ToneDexter not only trains well with old strings, it can also compensate for a problematic pickup - in some instances, at least.

I have this oddball active UST system (Shadow NanoFlex with DiMarzio preamp) which I bought off ebay a few years back. When I tried the system in one of my guitars, it proved to be very feedback prone and way too strong in the lows and low mids. I quickly replaced it with a more balanced pickup and set it aside, thinking I'd likely never get any good use from it.

Last night I needed to replace an old UST system which had developed a ground hum, and it occurred to me that ToneDexter might well compensate for the excessive lows and low mids with this NanoFlex system. I swapped in the NanoFlex system and trained up a WaveMap for it and presto - I have a noiseless and well-balanced amplified guitar sound once ToneDexter has done its thing. The difference between the pickup's muddy raw sound and the Dexterized sound is like night and day, but ToneDexter manages to pull a silk purse from a sow's ear.