FRONT PAGE  •  SYLLABUS/SCHEDULE  •  ASSIGNMENTS  •  COURSE RESOURCES   •  YOUR PAGES   •   FEEDBACK

Case Study: Dual Site 

A premiere performance, DUAL SITE, featuring two actors in acts of reclamation and recollection: Paris, 1850, excerpts from the Journals of the Goncourt Brothers - Russia, 1929, two men seek answers in a doll store, just under the heel of Stalinist annihilation, adapted from a short play by A.A. Amal'rik - Warsaw, 1944, drawn from the book DEFIANT GARDENS by Kenneth Helphand. Each night, between sequences, a three-course dinner will be served to the 35 members of the audience in custom produced porcelain bowls, in collaboration with Maine chefs, bakers and farmers.—Leon Johnson, author

After reading the script - -   set in those old European cities - -  I remembered the ancient 10-second fragment that was recently uncovered and released as "The World's Oldest Recording". I decided to engage in my own act of reclamation and recollection, using this sound fragment as a starting point. 

It's from a phonoautograph made in 1860, of apparently a woman singing a few passages from "Claire de Lune" (the folk song, not the Debussy piece. Debussy wouldn't even be born until a few years later). But, it's really a man singing, with the tempo/pitch raised (the man was most likely Edoward-Leon Scott de Martinville, the inventor of the phonoautograph).

Here's what the phonoautogram of the recording originally looked like (go to FirstSounds.org for complete scholarship):

This is the original file:

I added a little reverb to it, to give it just a bit more of a 'halo' here: 

The next 4 files are further digital transformations of the piece, which may or may not work as background for one or more parts of the drama (they don't necessarily need to be part of the doll shop scene).In the first iteration, I brought the file into Sound Hack, and applied Phase Vocoder, to make the sound file 5.8 times as long:

I used the Noise Removal filter to just get rid of some of the distorted noise, not all of it, just the most noisy parts;

Next, I added a few pauses between the phrases, which opens it up more.

Convolve is a bit of a trial and error process (a process we don't hear much about, it seems, in these digital times). The first convolve used the Kaiser window with these settings, and the result, well, failed:

The second try, using a Ramp window (and no Ring Modulation) worked pretty well. This resulted in an ambience in its purest form (using the convolve filter, which is like melting grains of sound, that, like sand, become glass):

Number 4 is a trio of these voices in canon (up and down a tenth from the original). So, these are the voices I found inside that piece of sound-history: