Multiple Nodes
The thing to remember if you add a lot of different sounds to a VRML scene is the effect that it will have on the processor usage. Adding additional sounds slows down the system, but only if they are audible. If you design your scene so that the viewpoint is never in range of more than four or five sounds at a time, it should be able to run well on most machines, with plenty of processing power left over for rendering the graphics.

Synchronized Nodes
Another way to implement sound involves the synchronization of multiple sound nodes together. In order to synchronize several sounds, just use a start time for all of the sounds in the "synchronization group". They can either have a common start time or be tied to a common sensor.

Using a common start time, you must also make sure that it is sufficiently long enough to allow all of the sounds to download before starting. Therefore, you might want to test your world on the modem speed of your projected audience (14.4? T1?) in order to get an idea of the best length to choose. For that reason, a common sensor is an attractive method, eliminating the potential for miscues.

Also, when synchronizing loops, make sure that all of the sounds in the set are exactly the same length in time. Otherwise, they will all be out of sync after the first time through the loop.

In this example, we have three nodes that are set up so you can hear one, two, or all three sounds at once to see how it affects your CPU cycles.

< < < < < < _ > >