Route Real time Audio input to an object location in 3D world coordinates

Question how to make the “singer effect” by that i mean to talk as Input (mic) in a submix probably or VOIP talker then sent this submix to a an object like a speaker 3D asset around the 3D world and hear my voice generating from there… something similar to walkie talkie effect in a multiplayer (i am struggling a week in blueprints to achieve any of these two i cant get it working). I cant convert the real time sound to a location sound.

This is complicated due to the microphone input component. The old VOIP talker API is not utilized widely and is definitely an older system. You’d need to write some C++ code to attach a source effect to the VOIP talker output.

We use the VOIP system in EOS (Epic Online Services) for Fortnite. I implemented some gameplay code that allows us to attach a DSP source effect chain on the voice audio so players can hear processed audio, so it’s possible.

You could check out the Microphone Capture Component but that is not hooked up to VOIP and was implemented as a means to get captured audio into Sequencer (the Sequencer Take Recorder uses it under the hood) – it’s not really designed for the use-case you’re talking about.

More broadly speaking, the way to arbitrarily route audio around is either through Audio Buses (which is essentially an audio buffer wrapped by an asset that you can use as a handle) or through Source Buses (which are audio buses turned into sound sources so you can hear them). Source Buses have an implicit internal audio bus if one is not supplied to them by the user. Source Buses allow you to control and play audio routed to it from other sound sources as another sound source (i.e. via audio components, etc). In this way, they can be thought of as 3D submixes. We utilize source buses quite a bit in Fortnite for a variety of different applications.

Thanks for the Reply. That’s actually accurate I am making a video tutorial and blueprint asset about this and I manage to do it via VOIP talker and Begin talking event (audio capture component in that class instead) override in the component parent. The problem was indeed I was using audioCapture (as component in character) which is not replicated but VOIP talker as a warper class I guess? it gets replicated and the audio signal its gets transmitted via sourceBuses instead of submixes which I did not know the proper use case of them. For anyone reading this I will post the link when I will make the video!.

As I promised here is the blueprint asset available in the marketplace (Advanced VoIP Voice Chat System in Blueprints - UE Marketplace) and a video explanation (Unreal Engine VoIP Voice Chat System for Multiplayer #3 Blueprints: How it works? - YouTube)