8

Chapter 14 EVOC 20 PolySynth 171
Achieving the Best Analysis and Synthesis Signals
For good speech intelligibility, please keep these points in mind:
 The spectra of the analysis and synthesis signals should overlap almost completely.
Low male voices with synthesis signals in the treble range do not work well.
 The synthesis signal must be constantly sustained, without breaks. The track should
be played legato, as breaks in the synthesis signal will stop the vocoders output.
Alternatively, the Release parameter of the synthesis signal (not to be confused with
the Release time of the analysis section) can be set to a longer time. Nice effects can
also be achieved by the use of a reverberation signal as a synthesis signal. Note that
the two latter methods can lead to harmonic overlaps.
 Do not overdrive the vocoder. This can happen easily, and distortion will occur.
 Enunciate your speech clearly, if the recording is to be used as an analysis signal.
Spoken words, with a relatively low pitch, work better than sung vocals—even if the
creation of vocoder choirs is your goal! Pronounce consonants well. A nice example is
the rolled R of “We are the Robots,” by Kraftwerk, a classic vocoder track. This
pronunciation was specifically made to cater to the demands of the vocoder.
Feel free to do what you like when setting the Formant parameters. The intelligibility of
speech is surprisingly little affected by shifting, stretching, or compressing the
formants. Even the number of frequency bands used has a minimal impact on the
quality of intelligibility. The reason for this is our ability to intuitively differentiate the
voices of children, women, and men, whose skulls and throats vary vastly by nature.
Such physical differences cause variations in the formants which make up their voices.
Our perception (recognition) of speech is based on an analysis of the relationships
between these formants. In the EVOC 20 plug-ins, these stay intact, even when extreme
formant settings are used.