What about a signal mapper?

[Tusker]

What it is:
This is an idea based on the tracking generator. The concept is simple, provide an interface which allows a linear control signal to be plotted against a user defined response curve on a cartesian grid. The outcome is a very generalized shaping tool for all kinds of signals.

Related implementations:
1- The Oberheim Matrix 12 had 3 five-point tracking generators, which means the user had 5 points in which to define a curve. (actually a jagged shape comprised of 4 linear segments).
2- The Alesis Ion’s tracking generator can be switehc from a 24 point one to a 32 point one. Again my understanding is that N points, allow for N-1 linear segments.
3- The G1 and G2’s Note Velocity Scaler could be thought of as a 3 point (2 segment) tracking generator.

Why we need this:
Control is the heart of expression, and generalized, intuitive means of shaping control will be the simplest way to improve the expressiveness of a powerful modular synthesizer. While the G1 and G2 have rectifiers and shapers, it would be very cumbersome to develop a full complement of graphical shapes in this manner. A math module (algebra based) would be easier, but it won't work for certain curves (and certain users). However these ideas may have additional merit that a signal mapper may not address.

Implementation Ideas:
I suggest one of two interface choices for this module. One would be the ability to simply draw a graph and have the G2 memorize it. Drawing tools (curves, straight lines, triangle waves, square waves) would make it easier. Yet another interface, is to have the ability to insert shaping components into a pre-existing x=y graph. Components similar to those in a parametric EQ would be placeable anywhere on the graph. (Center value, Q (steepness) , Depth and Shelving (left, right and none) could be parameters for these objects. These objects would morph the graphs from an x=y starting point, to the shape desired by the user. These types of interfaces would be a richer, more flexible language than the historical tracking generator. While the primary purpose is to have a generalized shaping tool for control signals, it may have application to audio waveshaping as well. Comments?

Thx,

Jerry

[Kassen]

Well written.

I have several comments; first your plan sugests a extra level to be build on the G2´s interface. I don´t think we´ll be getting that.

I am in fact sure your idea has purposes in audio waveshaping, this leads to problems with linesegments because those imply corners meaning hard to controll harmonics and probably aliassing.

I also think a successfull canidate will allow for fairly intuitive ways of modulation.

The good news is that one of Clavia´s beta testers cooked up a deliciously versatile plan that actually comes from the graphical field. The problem is that he´s still working on it and so it would be impolite for me to spill the beans on this. I do think it´s reasonable to expect a very good version of this kind of idea to be implemented at some future date in the G2. For now I think the lookuptable sequencers combined with some smoothing will suit you but that means a G2 and not your own Classic.....

[mosc]

I think the control sequencer does this already. If you want more than 16 steps, they can be easilly chained. It is assumed the wish list is for the G2 becaue the NM1 is no longer in development.