Virtual WFS

Miguel Negrao wrote:

> Is there any source of information on simulating wavefield synthesis

> inside another wavefield synthesis setup ?

I guess this isn't exactly what you are looking for, but Génther Theile and his colleagues have been researching WFS. They have created so called "virtual panning points", which are basicly stereo bases within the wavefield area, a bit in the direction you are talking about. Check out the references at Hauptmikrofon: Hauptmikrofon.de.

Eero

> Is there any source of information on simulating wavefield
> synthesis inside another wavefield synthesis setup ?

I'm not aware of publications documenting such an approach. However, binaural synthesis of WFS systems has e.g. been used by Helmit Wittek in order to do subjective experiments.

> Let's say for example I have real WFS setup which is a
> square with many speakers covering that square. And let's say
> I have the audio outputs for another WFS setup, for example
> rectangle bigger than the square I'm working on. If I was to
> place each infividual speaker of the rectangle as a virtual
> source in the square WFS system, would this recreate the
> original rectangle WFS system inside the square WFS ?

Theoretically you could simulate a larger WFS system on a smaller one by representing the loudspeakers of the larger system by virtual point sources placed at the simulated loudspeaker positions driven by the approriate signals for the larger setup. However, as a first guess I would say that this will not exactly be the same as using the larger system for reproduction. This is due to the fact that the amplitude errors, the aliasing and truncation artifacts, and the famous sqrt(jk) equalization will depend on the smaller (physically estisting) system. If you want to use this for subjective experiments I would advise to carfully investigate the resulting artifacts and limitations.

> What about the inverse situation where the virtual WFS system
> is smaller than the real WFS system and is simulated with
> sources which are inside the listening area ?

You could try to use focused sources at the virtual loudspeaker positions for this pupose. The properties of focused sources have not been studied in detail so far, especially with respect to spatial aliasing and equalization. However, it is known that focused sources have some undesirable properties e.g. in terms of coloration.

Both situations can be easily realized by a state-of-the art WFS implementation. From my point of view its worth trying it. I can also offer to try it in the next days and give a short report. However, I would be careful about the resulting properties, like coloration, of the virtual WFS system. These may not match a real existing larger/smaller system.

Sacha Spors