An audio backend for the multi-layered soundscape of Beyond Perception: Seeing the Unseen, a permanent exhibition at Scienceworks in Melbourne, Australia.
This implements the first generative movement algorithm to the audio
server. The "Autonomous Agent" movement variant utilises physics-based
steering forces to guide the virtual sound location throughout the
exhibition space in a seemingly organic, self-driven manner.
Behaviour
The autonomous agent has a simple set of goals.
Pick a random target location within one of the assigned
installations.
Travel towards that location.
Once arriving at that location, pick a new random location.
Goto 2.
The autonomous agent has several configurable parameters which affect
the way in which the agent travels between locations:
Maximum speed in metres per second.
Maximum force in metres per second squared.
Initial velocity and orientation.
These parameters can be configured to give agent's different
characteristics, e.g.
high max speed and force can make an agent appear erratic (possibly
well suited to some high-frequency sounds)
high max speed and low max force can make agents swing wildly out of
control as they pass their target location, perhaps briefly swinging
into nearby installations.
low max speed and low max force can give the agent a heavy, slow
feeling (perhaps good for low frequency sources).
TODO
GUI controls for the parameters mentioned above still need to be added.
Sum a subtle distance-based noise-walk onto the agent's target in
order to simulate a more organic approach to reaching its target (rather
than travelling in a straight line once oriented).
A concept of "target location arrival" so that the agent does not
continue travelling full speed through its target but slows down
slightly as it approaches.
Possibly add a "repel" force that causes the agent to avoid other
sounds that were spawned from the same source.
Apart from the agent, some other small bug fixes are included. In
particular, vector orientation is now calculated using a shared function
and in turn is consistent across all uses for channel layout.
This implements the first generative movement algorithm to the audio server. The "Autonomous Agent" movement variant utilises physics-based steering forces to guide the virtual sound location throughout the exhibition space in a seemingly organic, self-driven manner.
Behaviour
The autonomous agent has a simple set of goals.
The autonomous agent has several configurable parameters which affect the way in which the agent travels between locations:
These parameters can be configured to give agent's different characteristics, e.g.
TODO
Apart from the agent, some other small bug fixes are included. In particular, vector orientation is now calculated using a shared function and in turn is consistent across all uses for channel layout.