A DECtalk Express stands in for Monty Python actor Michael Palin, with an Intex Talker (Votrax SC-01A) voicing the role of John Cleese. Retro speech synthesis just sounds better. Not more intelligible, necessarily, but absolutely better.
And it’s hilarious. Eat your heart out, Transmediale / Ars Electronica / etc.:
If all of this sounds familiar, it’s because of how influential the work of a handful of early speech synthesis researcher-inventors has been. Dennis Klatt worked on the original MITalk project on which DECtalk’s speech synthesis was based, including the model Perfect Paul — the Minimoog of speech synthesis, if you will. (See this history.) US Americans of a certain will particularly know that voice, as it was the synthetic voice of NOAA weather radio.
For its part, Detroit-born Votrax — awesomely originally known as the vocal division of Federal Screw Works — built models on the work of one Richard T. Gagnon. That system saw action on arcade games, Gottlieb System 80 pinball machines, talking terminals, the VIC-30, and even the first generation Kurzweil Reading Machine for the Blind. The company is defunct, too, but there’s a digital gravesite to visit with more history.
What’s especially cool about these models and early voice synthesis in general is the way that they could become expressive — played like an instrument. We really lose something if we only ever let AI do that for us. (There are many theories that speech itself arose together with or even out of musical communication — see diverse percussion communication systems across the African continent, for instance.)
For an example of that, watch this early Voder demo — and note the use of the word “expression”:
And if you’re into this sort of thing, see the wonderful Chipspeech from Plogue, which turns vintage voice models into a playable instrument.
I recently came across this track, which I’d forgotten I had made. I made it with a prerelease version of chipspeech for a launch compilation; the libretto of sorts is from Claude Shannon.
I remember specifically keying all of this in by hand with a MIDI keyboard — Voder style — mod wheel for vibrato and all, because I later did a live show of all chipspeech at Berlin’s Panke. I was not asked to do that again, but I enjoyed it!
PS, plenty more where that came from, featuring some legendary chip artists — a who’s who (not me, I’m not worthy!):