SINEWAVE SYNTHESIS
* Replication
* Tone Combination
* Sentences
* Research
* Information

SineWave Synthesis
SWS Input Parameter Information

The SineWave Synthesizer (SWS) uses ASCII (text) parameter files (known as SineWave Input, or .SWI files), to control the synthesis process.

The first item is an integer called NSW, which is the number of sinewaves, that is, the number of parallel tones that will be created.
(For replicating natural speech using sinewaves, we usually use 3 or 4 tones.)

For each tone, a frequency value (in Hz) and an amplitude value (linear from 0 to 1) are given.

Time is specified by "time slice" values, given in milliseconds.
The first time slice starts at time 0 ms.
For the first time slice, NSW frequency and amplitude values pairs are specified.
Frequency and amplitude values are linearly interpolated (ramped) from time slice to time slice. Because of this, the amplitude values for the first and last time slices should be set to 0.

The file format, thus, is:
LINE 1: NSW
followed by a time slice set, where the first line of the set is:
TIME SLICE VALUE IN MS.

Followed by a number of lines determined by NSW.
Each of these lines has:
FREQUENCY IN HZ, AMPLITUDE.


Sentences


Haskins Laboratories • 300 George Street
New Haven, CT 06511 • 203.865.6163