Korg M3, Parameter Manual

Page: 559 / 647

Cluster Group Overview
549
Cluster Group
Overview
The Cluster Group contains the parameters that control the
“clustering” characteristics of the Generated Effect (the
number of notes generated simultaneously each time a
rhythm event occurs, thereby creating chords and
polyphonic phrases). Also controlled from here is the degree
of strumming on a cluster, and the direction in which a
cluster is strummed.
About Cluster Patterns
A “cluster” is a group of notes that will be generated at the
same time, with a size ranging from 1 to 10. Cluster Patterns
control how many notes at a time will be generated each
time the Rhythm Pattern determines that it is time to
produce some notes (each “rhythm event”). This can be
used to cause chords to be generated at certain times; the
resulting chords may optionally strummed or skewed in
several ways, with differing degrees of looseness. Random
choices of cluster sizes can be made from “Random Pools”
of values as described in detail later on.
If the Cluster Pattern is 1, then single notes are produced
(when “GE Type” (
☞
p.529) = 0: Generated-Riff). A CMaj7
as input source material might produced the following Riff:
If the Cluster Pattern was {3, 1, 1, 3, 1, 2} (and the Index
Pattern “Cluster Mode” 0: Single
☞
p.547), then the
following result would be produced by the same input
notes:
Note that the number of notes in a cluster can be effectively
doubled or quadrupled if “Invert,” “Double” (
☞
p.548) or
both are being used in the Index Group.
A Cluster Pattern will loop as long as note generation
continues. It normally will not reset to the beginning of the
Pattern unless a new Trigger is received, or unless the Phase
Pattern has been configured to restart it at the beginning of
certain Phases. This means that a four step Rhythm Pattern
can be looping while an eight step Velocity Pattern and a
twelve step Cluster Pattern are also independently looping,
for example.
Strum [0…1000 (ms)]
Sets the value in milliseconds that each cluster will be
“strummed” (quickly arpeggiated or flammed). Note that a
single note cannot be strummed, so this only has an effect on
cluster sizes greater than 1. This is the total time for each
cluster, so each note in a cluster of 5 notes will be closer
together than a cluster of 3 notes with the same strum value.
This maintains a proper degree of “looseness” while
changing cluster sizes.
Strums are presently alternated in the up and down
direction. By setting this to high values (so there are not
really strums, but arpeggiations happening), interesting
overlappings of each cluster over subsequent clusters can be
made to occur.
Not available if “GE Type” (
☞
p.529) = 2: Generated-
Drum . Not available if “GE Type” = 1: Generated-
Gated and “Gate Type” = Expression.
Pattern Grid & Associated
Parameters
Cluster Pattern
A Cluster Pattern represents a series of values indicating the
size of clusters of notes to be generated (chords). It has any
number of steps up to 32, with each step being represented
by one column on the grid. The first column always contains
at least one value; unused columns appear disabled. The 10
rows represent the number of notes to be generated at a
time, from 1 to 10.
Pattern grids cannot be viewed or edited on the M3.
Random Weighting Parameters
The Random Weighting Parameters are made available
when at least one step (column) in the Cluster Pattern has
multiple values selected (constituting a “random pool” of
values).
For more information, please see “Rhythm Group” on
page 539, and “Random Weighting Parameters - Pools” on
page 541.
Pools-Random Factor [–99…+99]
For more information, please see “Rhythm Group” on
page 539, and “Pools-Random Factor” on page 541.
Pools-Weighting Curve [0…3]
For more information, please see “Rhythm Group” on
page 539, and “Pools-Weighting Curve” on page 541, and
“Random Weighting Curves” on page 584.
The following table summarizes the effect of the various
Weighting Curves and the Factor field on the choices from
the Cluster Pattern Pools:
Input 1 1 1 1 1 1 1
Cluster Pattern [1]
etc.
Input 3 1 1 3 1 2 3 etc.
Cluster Pattern {3, 1, 1, 3, 1, 2}
0: Exponential 2: Exp-S
1: Logarithmic 3: Log-S