[Verse-dev] using material node for audio

[Pontus Nyman]

I have some thoughts about using the material node fragments for acoustic
properties (as discussed in the "Using tags and C++ issues"-thread):

The acoustic behaviour for a material differ a lot for different
frequencies, so you really need to separate acoustic properties over the
frequency spectra. This seems a bit tricky to represent in the current
material node and with the fragment types provided. Using only one "audio"
output fragment as hinted in the specification wont solve the problem.

One idea is to have multiple fragment trees in the material node (is that
possible?) where different trees correspond to different frequency bands.
The output fragments could have labels like "audio120", "audio500",
"audio1000", etc, where the numbers indicate the "frequency band
locations" in Hz. I am not sure that this solution would be less
application-specific than using tags.

If multiple fragment trees aren't allowed you could off course split up
the acoustic properties for a material over several nodes with one
fragment tree in each. This seems a bit ugly though. You really should
have all material properties for a specific material collected in one
node.

Any other suggestions how to deal with this in a neat way?

Pontus