Abstract
A tremendous amount of research has been done into the visualization of three dimensional vector fields, however very little work has ever been done into how to sonify the same data. This project focused on how to sample various points within a vector field and convert it into sound that the user could use to help understand the properties of a given vector field.
The primary metaphor used here is that of wind. The technique we created tried to simulate a wind-like noise based on the local properties of the vector field. The first step is to be able to generate a sound for a given vector field sample, and the second stage is to be able to interpolate the sounds from successive samples to create the impression of a moving sound source (and thus the impression of a shifting wind).
To create the initial sound sample, we choose a vector sampled from a position within the vector field. The direction of this vector defines the relative direction of wind flow in relation to the user's head. Whichever direction the vector is pointing is the direction that the wind is flowing towards. The direction of the wind is used to chose a 3D position for the resulting sound source using spatial audio algorithms.
The magnitude of the vector defines the speed of the wind (and thus the volume and pitch of the sound). The greater the magnitude, the greater the volume, and the higher the pitch. By placing a sound source nearer or farther to the listener, the volume can be varied, and by using doppler shifting, the pitch of the sound can be altered.
In order to move the sound over time, two consecutive sound samples are taken (using random sampling in a uniform volume surrounding the user's head). The sound positions for these two samples are then computed. Once the sound positions are known, a hermite spline path is defined that the sound source can follow through space to simulate the shifting wind position/sound. By altering the speed of sampling based on vector magnitude, a very convincing wind sound can be generated using this techinque. This second stage of changing the sound position over time by using samples is critical for conveying information about the vorticity of the dataset in a given region.
The wind sound that results from the application of this technique can be used to convey information about flow direction, magnitude, vorticity and general flow patterns within the dataset.
Publications
