Creamware’s Modular Synth Vocoder consists of four components: Analysis module, Synthesis module, Vocder matrix and Voiced/Unvoiced Detector module. All ModIII.

dspLoad 44.00/-
generates 11 modulation sources according to the amplitude of the 11 filter bands of the analysed signal.2nd to 10th filterband are 6-pole bandpass filters; the 1st can be switched between Lp and Bp and the 11th can be switched between Bp and Hp (“use BPF”)                

dspLoad 35.90/32.70/-
input signal is analysed into the 11 freq bands. The VCAs are controlled by the respective incoming modulation signals. Again, the 1st and the 11th band can be switched between Bp or Lp resp. Hp.

dspLoad 0.00
Normally, the modOuts of the Analysis module would go straight to the respective inputs of the Analysis module. For special effects, the modulation connections can be routed freely via the Vocoder Matrix module. It can also be used without the analysis/synthesis modules, allowing to route up to 11 modulation or audio sources - plus having the option to randomize their assignment.
disadvantage:No MidiCC assignment (routing). Monophone only.

dspLoad 15.90/12.60/12.60
Basically a xFade module for two inputs V and UV, where the modulation source is the frequency of the Analyse input. In a vocder patch, it categorizes the input signal as “voiced” or “unvoiced”. Vowels will be (hopefully) recognized as voiced, diphtongs as unvoiced signals.

Type:Two algorithms are available: a faster but not-so-accurate standard algo and a more complex alternative with “Hysteresis” control. Aim is to minimize incorrect switching.
Inputs- UV:connect a noiseOSC or the high-freq portion of a speech signal.
V: typically a synth output signal.
Analyse: typically a speech signal.