The object records sounds from the surrounding space, selecting only the loudest ones (those which exceed a certain level). Then, the machine puts together a composition consisting of these sounds organised in the order they were recorded. In fact, the object excludes the silence and composes the music from the surrounding sounds, while compressing the time. The voices, music, city sounds and other random noise are shaped into complex algorithmic compositions, which can be played after it has collected enough of them. It is a kind of reality re-mixer - by simply removing the silence and pauses between loud sounds and words, it creates the sense of very rhythmical and organised aural experience, which sounds very musical to me.
The process has two phases, recording and playback. In the first phase, it records the surrounding sounds until it has gathered 100 samples. In the second phase, the object plays the resulting compositions as a loop through speakers for 1 minute, so that the compositions can be heard by everyone around. In this mode, the recording stops. After that, it erases all sounds and begins a new search.
The object itself is a robot with a rotating stereo microphone, which tends to activate with the help of a motor towards the loudest sounds in the surrounding space. It also has 2 speakers to present the results, and a light torch to show the detected sounds or indicate the sounds during playback mode.
Hardware: Directional stereo microphone, 2 channel sound speaker, LEDs, Arduino uno, Servo motor
Software: Pure data