I guess you can get similar results with cheaper hardware, but if you have money and you have it around... ¯\_(ツ)_/¯
His German WP page also claims that he sold a VSM 201 to Neil Young in 1982! https://de.wikipedia.org/wiki/Kai_Krause
English WP has less details on that part of his life, especially the VSM 201 :) https://en.wikipedia.org/wiki/Kai_Krause
Just in time for https://en.wikipedia.org/wiki/Trans_(album) !!!
I was surprised about how much better the Sennheiser sounds compared to the others. From the audio comparison in the first YouTube video. I did expect minor variations in the harmonics but the differences are quite significant between the models.
The other vocoder that sound almost as good is quite new and it seems to still be a prototype with a "contact us" price.
I can’t speak on Dromedary Modular’s behalf and I think rising parts costs have been an issue, but buying an Ultimate VoIS should be a fair bit cheaper than the Moog vocoder.
Oh, it also might be of interest that the IVL algorithm isn’t FFT-based. I think their harmonizers sound better than the rest, so maybe FFT isn’t the best way to go.
Has anyone got more details?
If you look at the code of (phone/voice) codecs GSM/Speex/Opus you can see that you can estimate the spectral envelope (or the configuration of a physical tube model for the vocal tract) in time domain with linear prediction coefficients (LPC).
And it is simple, e.g. the often used Levinson-Durbin algorithm is just 22 lines of C code. It is an interesting exercise to build your own vocoder from scratch that fits in a single screen page.
Many of the code snippets I have seen (which likely have already processed your voice) are just translations of the Fortran code of the book "Linear Prediction of Speech" by Markel and Gray (1976).
The shape of the filters, the smoothing between the filters and the synthesis section, and (on some models) the patchability all create a very different result.
The reason the best analog vocoders are so expensive is because the filter for each band is much more complex than a plain old bandpass filter, with a much higher component count. Typically there's a flatter passband and a steeper slope than you'd expect.
You can do digital convolution with thousands of bins and it sounds nothing like analog vocoding. It's much cleaner, doesn't have those lovely harmonically spaced filter resonances, and creates sounds that can feel more acoustic than electronic.
Above all, the biggest help from his father was probably to insist that they keep the ownership of their music when signing with any label, regardless of any money, because as a producer he knew that this is how artists get screwed by record labels.