Computer controlled bank of signal generators that speaks by modulating sine waves to match the frequencies of a human voice.

The dials on each box rotate back and forth and at different rates, the boxes emitting the characteristic whooping sound of a signal generator as the dials turn to their designated frequencies. Human voices can clearly be heard saying something in the middle of it all.

The computer takes a recording of spoken text and constructs a table of frequencies for every few milliseconds. Each row in this list has a time-stamp and 4 frequencies split according to range. The first band contains the lowest frequencies up to 127Hz, the second 128 to 169Hz, the third 170 to 211Hz and the fourth from 212Hz up. This table is used like a musical score for one set of signal generators.

Each "voice" consists of a set of 4 generators. Each generator produces waves within its own band of frequencies. The box on the far left handles the lowest band of frequencies (up to 127Hz), the box on the far right the highest (212Hz and above). The boxes in the middle sharing the frequencies in between.

The 16 boxes constitute 4 separate voices.

The dials turn as the signal generators attempt to emit their frequencies at the designated time, something they cannot always achieve. This results in the sound going in and out of phase, of elongated frequencies pulling voices in and out of recognition and sense.

Although the last iteration ended in a regular pulsating sound, a vague burning smell and smoke.

• 04.24: Under Erasure, Galeria Łęctwo, Poznań.
• 10.11: Before the Crash – Art and Science Collide, The Exeter Castle, Exeter.