Audiokinetic Wwise Knowledge Base

Granular synthesis is a basic sound synthesis method that operates on the micro sound time scale. In granular synthesis, sound samples are usually split in small pieces called grains. Multiple grains are then combined together and transformed in various ways to yield the synthesized output. The Source effect SDK allows for custom third party tools (such as your existing granular synthesis technology) to be integrated into Wwise.

You may simulate granular synthesis with relatively big grains using a continuous random container with trigger rate transitions ("smear"). Note that the trigger rate time must be at least 20.3 milliseconds. You may then randomize the trigger rate duration and sounds pitch.

Furthermore, SoundSeed Impact may be used as a transformation tool to create more variety in sound grains, reduce the number of grains required to get good results, and gain control over sound quality (e.g. explicit control of bandwidth scaling for damping).

Be aware that each grain will require one voice pipeline in the sound engine, so you need to be careful not to overload the CPU. Also, sounds should be stored in banks, and not streamed, because of the inherent streaming latency. Finally, note that when sounds are auditioned in the authoring tool, they are always streamed from disk, regardless of their "streaming" property. In order to test your design accurately, you need to generate a soundbank, load it in your game, connect to Wwise and play your sound structure from the Transport panel or Soundcaster.