Pattern playback
Early speech synthesiser developed at Haskins Laboratories (during and after the second world war). Speech was synthesised by converting patterns (stylised spectrograms) painted on acetate film loops into acoustic signals by a photoelectric system. The synthesiser was used to systematically vary the acoustic parameters of speech to determine the cues used in speech perception.
Links
Utrecht Lexicon of Linguistics More on Pattern Playback and audiodemonstration
