Are there ways to manipulate audio features like pitch and bass in real-time?

[braunale]

Hey,
I was testing the realtime API with livekit and noticed, that GPT is struggeling to correclty understand the voices of children, I guess this is a typical domain shift problem and I managed to partially solve this by lowering the pitch and manipulating some other voice features like bass and treble. However this was only possible by applying the changes to an audio file and then play this to my application.

I wanted to ask, if livekit provides any build-in functions to make these kind of audio manipulations on real-time to the audio-stream befor it is sent to openai?

In the docs, I did found some functions for removing background noise or voice enhancement but this was not really usefull for me.

Thanks and best regards!