Specifying the waveform segments

Currently, a segment can be one of three types:

wiggle

All the samples have the same amplitude

twiggle

The amplitudes rise to a fluctuating ``peak,'' then return to their starting value. Both rise and fall are linear.

ciggle

The amplitudes rise to and fall from a fluctuating ``peak,'' following the path of two second-order polynomials.

The above-mentioned ``peak'' requires some explanation. The ``peak'' is a point (located between the specified starting and ending samples) that has two variables: an amplitude (-32768 to 32767), and a location (0.0 to 1.0). Both variables of the ``peak'' can fluctuate between their specifiable minima and maxima, each at a specifiable rate. Future plans include the stipulation of a variable number of ``peaks,'' each with its minimum, maximum, and rate of change. Currently, a segment can have one ``peak.''

The disjunction of amplitudes between adjacent segments will be one of the contributing factors to the prominence of particular frequencies in the resulting sound spectrum. The amplitude range used is 16-bits, giving a dynamic range of about 96 dB. Thus the differences possible between adjacent segments range from 0 dB (when adjacent segments have the same amplitude) to 96 dB (when one segment is at -32768 and the other is at +32767).

Since the amplitudes of adjacent segments are continuously varying, the resulting sound spectrum will also be constantly shifting. To better control this phenomena, each segment can be ``slanted,'' that is, its starting amplitude can be the ending amplitude of the previous segment, allowing for a degree of control over unwanted disjunctions