Vox Antiqua

The Smithsonian has restored some early recordings, hitherto unheard, with digital techniques. If there’s an image of a waveform of a sound recorded – by whatever means available at the time (and that’s the novelty here) – then they’ve had a bash at converting it into something we can listen to.

Which is nice.

Years ago there was a theory that whilst painters were a-painting, their voices would be ever-so-slightly deforming the as-yet-undried oil paint as it was being slapped on the canvas. Tiny little ripples in the surface caused by the disturbed air from the voice of the painter would settle into the still malleable paint. And presumably also, though less so, from the voice of any chatty sitter. Take, therefore, a magnified image of any linear cross-section of the microstructure of the paint, and you have yourself a waveform.

Mostly, of course, this amplitude will be nothing to do with sound but will be dominated by the underlying roughness of the canvas, or by the movements of the brush, and – not least – by the lumpiness of the paint itself. But they will be the large-scale variations – relatively speaking. It’s going to be the really tiny variations, like the grass on a mountainside, which would need to be looked at.

So, how about having a go at listening to Mona Lisa?