[sdiy] Anyone built the ELEKTOR VOCODER?

[Magnus Danielson]

From: jhaible at t-online.de (jh.)
Subject: Re: [sdiy] Anyone built the ELEKTOR VOCODER?
Date: Wed, 31 Jan 2001 00:16:06 +0100

> > I actually found the trick in another Elektor article (spectrum analyzer),
> > and it consist in detuning slightly the central freq of each filter, so
> that
> > Fc of filter 1 is the central freq of the band, and the central freq of
> the
> > 2 other filters respectively the 2 cut freqs (high & low) of the band.
> 
> Don't underestimate the problem.
> 
> A spectrum analyzer "only" needs the 3 partial filters of each channel
> to be fitted together in terms of phase and amplitude
> (resulting in different center frequencies and Q factors).
> This is "easy", because you don't sum up adjacent filter bands !
> In a vocoder, *all* partial filters interact, so you don't have to fit
> groups of 3, but groups of 9 at least (if you go for a 6-pole design).
> 
> The slight difference between analyzing (and dropping phase
> information) and doing synthesis.

I agree. If I am going to spend money and effort on filters in a phase
sence, then it is going to be the synthesis section. If it is going to
be filters with just deep slopes, then resonant bandpass filters is
pretty fine in the analysis section of a vocoder.

For each added pole-pair in the sythesis section, the better phase
linearity you could get, but the harder it becomes to trim, so you
need to spend more thought on your topology and all that. It's just an
endless story.

Cheers,
Magnus