ScalaCollider-Voices

A library for ScalaCollider to aid building polyphonic synth graphs. This project is (C)opyright 2016 by the Institute of Electronic Music and Acoustics (IEM), Graz. Written by Hanns Holger Rutz. This software is published under the GNU Lesser General Public License v2.1+.

linking

The following artifact is available from Maven Central:

"at.iem" %% "scalacollider-voices" % v

Not yet released! The current stable version v is "0.1.0".

building

This project builds with sbt 0.13 and Scala 2.11, 2.10. To compile sbt test:compile.

contributing

Please see the file CONTRIBUTING.md

documentation

TODO

Auxiliary graph elements.

There are two auxiliary elements that aid Voices, and which will probably be best put back into ScalaCollider soon.

ChannelRangeProxy is a straight forward extension of ChannelProxy that takes a range of indices instead of a single index. So we can grap sub-groups of channels of a signal.

RepeatChannels aids with the representation of the conditional testing in the voice management, by allowing us to "bypass" the regular multi-channel-expansion and instead eagerly pair the channels of two signals. This way, we have a simple representation (that can be serialised) of the matrix from which we select condition groups in the actual expansion of the voices analysis. For a full explanation, see the scala-doc comments of RepeatChannels, but the following two diagrams should illustrate what is going on. Imagine the regular MCE when two signals are combined:

SynthGraph {
  val in      = In.kr(0, 2)
  val state   = DC.kr(Seq(3, 4, 5))
  val all     = in sig_== state
  Out.kr(0, all)
}

This matches in and state to produce a signal that has the higher of the two signals' channel numbers, so three channels. The expanded UGen graph should be familiar and looks as follows:

We can now enforce a repetition of the channels of signal, for example we can repeat every channel in in for the number of times that corresponds to state's number of channels:

SynthGraph {
  val in      = In.kr(0, 2)
  val state   = DC.kr(Seq(3, 4, 5))
  val num     = 4
  val inComb  = RepeatChannels(in, num)
  val all     = inComb sig_== state
  Out.kr(0, all)
}

The corresponding expanded UGen graph looks as follows:

It is now easy to see that we can "grab" for example, after the fact, the equivalent of (in \ 0) * state which is ChannelRangeProxy(RepeatChannels(in, 3) * state, from = 0, until = 3) and the equivalent of (in \ 1) * state which is ChannelRangeProxy(RepeatChannels(in, 3) * state, from = 3, until = 6).