Frederick Sammis invented the “singing Keyboard” in 1936, a precursor of modern samplers, the instrument played electro-optical recordings of audio waves stored on strips of 35mm film.
Let us suppose that we are to use this machine as a special-purpose instrument for making “talkie” cartoons. At once it will be evident that we have a machine with which the composer may try out various combinations of words and music and learn at once just how they will sound in the finished work. The instrument will probably have ten or more sound tracks recorded side by side on a strip of film and featuring such words as “quack” for a duck, “meow” for a cat, “moo” for a cow. . . . It could as well be the bark of a dog or the hum of a human voice at the proper pitch.
(Frederick Sammis, quoted in Rhea [1977]
Sammis had moved to Hollywood in 1929 to lead RCA into the era of film sound. Sammis was already familiar with the Moviola, a sound- and filmediting table that incorporated photoelectric cells. Using methods that were being developed for the new ‘talkies’, he recorded sung and spoken words onto individual strips of film. He then attached the resulting strips to the keyboard in such a way that a specific strip would be drawn across the optical cell when he depressed a corresponding key. More recent instruments such as the Mellotron and Chamberlin use a similar technology of triggered and pitched magnetic tape recordings.
Sources
‘The Computer Music Tutorial’ Curtis Roads
Invention and Technology Magazine. Mathew Nicholl. Volume 8, Issue 4. 1993
The “singing keyboard” seemed to have no real need to be a mass distributed instrument. Intended for talkies, not everyone has a need for it’s use. The main innovation seems to be it’s ability to trigger and then pitch bend. This instrument also seems to have paved the way for a huge market in music technology, leading up to the mellotron and even today’s sampling.