This is a work designed for a conference Chinese Chi 2019. A couple of videos below show various iterations. One made in Processing and one made in OpenFrameworks. Both use SuperCollider for the audio. The piece utilises a Twitter feed attuned to the keyword "Dungeon". Any tweet containing the word "Dungeon" will get into the piece. It also makes use of imagery associated with the tweets.

The finished work will be posted in the next few weeks.

This iteration with OpenFrameworks and SuperCollider.

This version with Processing and SuperCollider

The perlin noise function used in the OpenFrameworks version is written by Stefan Gustavson - the function was slightly modified.  The github address: