FEATURES

DESCRIPTION

Filter is a web-based installation that “listens” to real-time expressions and discourse online about noise and silence while simultaneously displaying them within the canvas of the webpage. Using Twitter as its database, the work places tweets at random within the browser window of the viewer to render a visual composition of listening between opposing forms of sound, in addition to visualizing the politics of noise and silence. A message that contains ‘silence’ removes the earliest captured message featuring ‘noise’ (and vice versa). Over the duration that the webpage is viewed, one of the two forms of sound eventually outnumbers the other; echoing the reality that one always seems to be more salient than the other.

The more dominant message of ‘noise’ versus ‘silence’ also tends to change depending on the time of day, based on the stream of messages from all of Twitter's active users in conjunction with the state of the world. The font size of these messages is determined by the tone of its connotations using natural language-processing algorithms; that is, the more positive-sounding messages are measured to be of a larger font size, and conversely the more negative-sounding messages are smaller. This method of display also attempts to imitate the behaviour of listening, hearing, and grey area between the two.