0
1

Hi,

I’m working on a new application for modular audio processing (along the lines of Bidule, Audiomulch) and considering FMOD as a basis for the DSP. The application would need to be able to connect audio processing ‘modules’ (effects) in a directed graph with arbitrary branching between inputs and outputs of each module. An example might be:

Audio Input -> Granular Delay -> Harmonizer -> Reverb -> Audio Output
-> Distortion -> Ping-pong Delay ———-/

Looking at the API, it seems like I could manage this in FMOD if I provided my own DSP callbacks for the effects that aren’t already implemented in the library. Is that correct?

Also, how would FMOD handle larger DSP graphs, with the graph being modified at runtime? Is this possible with FMOD? Any potential problems such as clicks or efficiency issues?

Thanks,

Jamie


[url:286sduzk]http://www.jamiebullock.com[/url:286sduzk]

  • You must to post comments
0
0

Hi postlude, welcome to the FMOD forum.

[quote:15pt1spk]Looking at the API, it seems like I could manage this in FMOD if I provided my own DSP callbacks for the effects that aren’t already implemented in the library. Is that correct?
[/quote:15pt1spk]
Yes.

[quote:15pt1spk]Also, how would FMOD handle larger DSP graphs, with the graph being modified at runtime? [/quote:15pt1spk]
It handles it quite well as that is what it is designed for. You can modify the graph from the main thread using the FMOD API and those modifications will be protected from race conditions in the mixer.

[quote:15pt1spk]Any potential problems such as clicks or efficiency issues?[/quote:15pt1spk]
FMOD is very efficient, it all comes down to the performance of your custom DSPs.

  • You must to post comments
Showing 1 result
Your Answer

Please first to submit.