FR for Modulator/Transformation-Shape-Graph Enhancement

[dmbaer]

This is a feature request for an improvement to how normal modulators can control some other aspect of processing. Let me explain at some length what I was trying to accomplish in order to make a compelling case for what I’m asking. This is what I believe to be a very generalized need, not an obscure specific situation.

I was attempting to create a simulation of an orbiting sound source in MXXX, somewhat inspired by the Waves Brauer Motion plug-in. The sound source would travel in an orbit that was fully in front of the listener and the orbit would be in the plane of the speakers/listener. In other words, 2D with no attempt at simulating height or surround placement.

I’ll say up front that this experiment never got completed and maybe it would not have worked even the feature already existed, but without that FR, I think no solution is possible.

To do this, several aspects were involved: the main two were sound-source/listener distance controlling volume, and orbit position controlling pan position. The orbit position was such that the furthest distance was the position associated with the start-cycle (and end-cycle) position of the LFO.

Other optional factors were to include band-pass filtering (the amount being controlled by distance), a simulation of Doppler frequency changes, and maybe some reverb where far position was more wet than near position (again controlled by distance).

I generated all the mapping values using some basic trig calculations in in Excel – no rocket science involved here.

In order to do this, I needed to use transformation shape graphs in the modulators. Defining direct modulator waveforms was not an option for several reasons, the main one being that those waveforms are required to start and end at 0%.

The main transformation shape graph needed was for distance. It looks something like a bell-curve from statistics. The panning was controlled by a graph that looks like a somewhat wonky sine wave. The Doppler control is an even wonkier sine wave, one in which the excursions above the line (zero frequency change) and below the line are not the same amounts.

Here’s the problem. Think of these graphs simple functions: y = f(x). The graph shape is the function and x should be the LFO position: 0.0 to .999999 (or maybe 0.000001 to 1.0). To make these graphs work right, they need to have a continuous ramp as the x value. At the end of the cycle, we must go from somewhere very close to 1.0 to somewhere very close to 0.0, and we must do so instantaneously.

Using a saw waveform (inverted to be an upward ramp) simply does not work. It takes way too much time going from 1.0 to 0.0 (or 100% to -100% if you prefer) at the cycle-end/cycle-start transition due to interpolation that is totally desirable under most circumstances. To use a transformation shape graph in a pure fashion, where the graph provides a value based upon the LFO cycle position, does not appear to be possible.

So, two possibilities come to mind. The first is to provide a very basic “LFO” ramp wave that “plays back” with no interpolation. The y value produced by the LFO is the same as the x value (LFO cycle position), maybe normalized to the -100% to 100% range if that makes things clearer. The other possibility is to have the graph definition window in the modulator have an option that says “just feed me the LFO x value, not the y value I would normally be provided, in order to do my value lookup”.

[MeldaProduction]

Ok, it took me a while to understand the only needed paragraphs were at the end . Anyways added "Interpolate between 1 to 0" option, enabled by default, should do the trick . The update should be out today/tomorrow.

[dmbaer]

Ok, it took me a while to understand the only needed paragraphs were at the end .

Yeah, well I know your often-default response "that's an unusual and very specific situation".

Anyways added "Interpolate between 1 to 0" option, enabled by default, should do the trick . The update should be out today/tomorrow.

Way cool!!!! Can't wait to check it out. Thanks.

[MeldaProduction]

Hehe yeah, you know, I have haters complaining about the GUI (in fact more like number of features) and then many people asking for features potentially nobody would use, so I need to be smart about what I say yes to .

[dmbaer]

Ok, it took me a while to understand the only needed paragraphs were at the end . Anyways added "Interpolate between 1 to 0" option, enabled by default, should do the trick . The update should be out today/tomorrow.

I just installed 12.03. I'm not finding this new feature after looking in what I thought were all the obvious places. Could you please point out where the option is set?

[jmg8]

Right click on the wave shape.

[dmbaer]

Right click on the wave shape.

Bingo. Somehow I missed the new switch the first time I looked here.

Took this new feature for a test drive yesterday - works brilliantly for the downward ramp basic waveform. I constructed an upward ramp with the custom shape and the interpolation behavior is still present (i.e., the new option has no effect). This isn't a showstopper, but it does require that the transform map be reversed (right to left rather than left to right as would be more logical and comprehensible) when using the downward ramp. But make no mistake, I'm very pleased we now have a seamless way to use transform maps from an LFO.