We found this youtube tutorial on how to make images into a mesh and affect it with audio input. This takes the brightest points and extrude outward to the viewer. see here
We altered this by taking image from the webcam instead of using images as in the youtube tutorial. For the webcam image, we got it to randomise brightness, contrast and saturation every 5 seconds.
We also had to deal with positioning and orientation of the webcam input because we wanted it to mirror the viewer. So had to make the mesh orientate in the same way as the matrix we get from webcam. Matrix can be re-orientated with DimMap object, but for the mesh we had to play around with the values of its position in the Jit.World objects.
We also got it to display lines and grids based on the peak amplitude and pitch of the sounds being generated.