What is best practices to use FMOD?

We are trying to achieve:

  1. rich musical content based on 3-10 parameters … my opinion is one single event modified by those parameters and the musician would like more events with fewer parameters. When I instantiate the main event it takes few frames to start so I guess it would be better to instantiate all in the beginning

  2. step sounds - we have many different floor types and we aren’t mixing them together, we also have different creek sounds for those floors and we have different kinds of steps like walk, run, jump etc., there are also body sounds but they don’t depend on the floor, just on the step type … my opinion is (as before) one big event with parameter step type and floor type and sound designer would like to have different event for each floor type and define the step type in it by one parameter. This means instantiating or pooling events for each floor type which might take a lot of memory.

My question are:

  1. how heavy is FMOD on event instancing? does it depend on the content, so less stuff in it means faster init?

  2. how heavy is FMOD on memory? does it depend on the event content, so more in it means more RAM occupied even if it’s playing one sample?

  3. when there’s volume 0 on some subtrack in FMOD event, does it stop the sound or does it use the channel / cpu / memory anyway?

Thanks for answer.