If this can make a DX7 fun to program, what could it do for your synth?

[rylos]

http://us.f2.yahoofs.com/bc/4124db49_cb0e/bc/images/touchram_quad.jpg?bf_FOJBBNmktiCpA

 

Touch controls. Dozens of them.

Pull up a voice. Bargraph displays show the settings. Touch the display where you want the value to be. Jump to a new value, or drag the value with your finger. Touch several control strips at the same time to change several settings at once.

 

What could make it simpler to have dynamic, realtime control. This makes a synth much easier to learn, and a ton of fun to program. Imagine learning the fundamentals of DX7 programming in a half hour. How about being able to get those dynamic changes in your sounds without having to go through a procedure (kill your menus!).

 

These photos are of a simple prototype. A full-featured version would have more control strips (closer together), some touch "buttons", and a display to indicate what each strip is assigned to.

 

Should a product like this be on the market, or should I just keep it to myself?

[rylos]

Darned link won't work now. Will have to find anothe place to put the photos.

[M Peasley]

I've used SoundDiver for editing my K2000 and FS1R for years, so I'm pretty familiar with having a graphical layout that contains all the parameters of a patch.

 

Your approach would add a touch-screen controller to the concept, which would be great. Sounds worthwhile on the face of it.

 

The only thing that comes to mind is the fact that there are quite a few parameters that are simple toggles, so a virtual slider would be wasted on that type. Also, some synths work on a sort of tree diagram, where one choice changes the next series of choices. Like on the Kurzweils, you have to choose a basic wiring diagram to begin with (an "algorithm"), of which there are a set number (32 on my K2000). Once that is selected, the next set of available parameters is determined. I can't see it working to put the choice of algorith on a virtual slider, and then having all the other parameters change to match what is available under that particular algorithm. A bitch to program, if even possible.

 

So I'm thinking your idea would be slicker'n snot on a doorknob for things like filter cutoff, fx depth, EQ, resonance, FM modulation depth, envelope depth, and so on. But the other hard-wired sets of choices may not be amenable to your layout.

 

M Peasley

[marino]

I think it's a very clever idea. I didn't see your images, but I guess you could program virtual sliders *and* pots, and also selectors, switches... You could let the user to decide the screen layout, and let him save his configurations. It would also be a nice alternative to physical control surfaces for controlling softsynths.

[Yoozer]

What would it make more fun to program a DX would be this:

 

- a resynthesis option. Something that analyses a sample, executes a fast Fourier transform, and tries to guess the best algorithm and the closest settings to mimic it.

 

- standard building block options with preset envelopes included. With 6 operators, you can 'think' in terms of 3 x 2 operators, 2 + 4 operators, or 2 x 3 operators. The 3 x 2 and 2 x 3 ones can be used to 'stack' sounds - a Rhodes sound has a bell, a bark, and a midsection component, and 2 operators can make every component work out. So, instead of trying to have a lot of trouble with which algo one should choose plus what ratio one should use, it's better to make partial presets for a whole algorithm. Too bad the DX has constraints in that aspect .

 

- FM7 does it pretty in terms of editability, but making a single knob-laden gridwork with knobs per parameter would be completely useless; there's no decent complete view of the parameters.

 

- visual aides like "brightness curves" - analysing the harmonic content during the wave, so you have the subtractive idea of how "bright" or "dull" a sound is. That already helps a lot, because with FM that's harder to do.

 

Anyway, just some suggestions .

[synthetic]

There was a great interview with Brian Eno in Future Music a year or so ago. He said that even an electric guitar has volume and tone knobs, btu the DX-7 doesn't even have tone, which would have been easy to implement. (Scale down all of the modulators.) I've been trying to use the Sounddiver beta without any luck, I'll try the new version and see if it's better on my system.

[rylos]

http://img30.exs.cx/img30/8244/tquad.jpg

 

Hopefully this image will work...

I can design a circuit with over 100 chips, but posting a photo on the internet is beyond my abilities?

 

Touch-operated buttons can be used for toggling functions.

 

If the controls have a simple display at each slider to indicate what it controls, and if it's set up in a logical fashion, it could be set up with sliders being dynamically allocated to match the particular algorithm's requirements.

 

Any way you look at it, it's still easier than wading through menus...

[Synthoid]

I want one....

[rylos]

These mp3 files will give you a bit of a taste of the possiblities that real-time tweaking can give you. Now the process of pulling up a voice from memory, then tweaking it, is a seamless operation.

These sound files are simply to demostrate the realtime sound tweaking, there is hardly any hint of musical content in them.

 

noodle-1.mp3

 

metabass_mutate-1.mp3

 

epiano1_morph-1.mp3

[RABid]

Well, this could make a lot of VSTi's fun to program.

 

What I want.

 

Robert

[rylos]

I've also built a touchpad with multi-point sensitivity (will recognize all ten fingers at once). I've not yet programmed it for midi operation. The current size is over 5" X 10", but it's possible to go much larger. I expect to build one with a sensing surface large enough to allow it to by played like the continuum .

 

The fun part now is getting this (patent pending) technology into the market, and ultimately into the hands (or fingers...) of musicians.

 

http://img50.exs.cx/img50/1345/jes0011.jpg