Hi!
I dont really work with fmod but our sound designer does so I dont want to split the systems, The plan is to have a solution that works both on lan and online so Vivox doesnt work for audio collection. Instead I used Opus, an open source voice encoder/decoder to send a byte array over RPC. Ive finaly got this solution to work with unity audiosource and am trying to get it to work with fmod events for better spatial settings and so on but no matter what i do the sound comes out horribly staticky or doesnt come out at all. I was wondering if there is some secret trick I dont know about to correctly bundle sound data in fmod. (current unity code)
> using Concentus;
using Concentus.Structs;
using System;
using System.Collections.Generic;
using Unity.Netcode;
using UnityEngine;
public class VoiceReceiverBehaviour : NetworkBehaviour
{
private OpusDecoder decoder;
private short[] decodeBuffer;
private float[] floatBuffer;
public AudioSource audioSource;
private static Dictionary<ulong, VoiceReceiverBehaviour> receivers = new();
private Queue<float> playbackBuffer = new(); // Buffer PCM samples for playback
private const int sampleRate = 16000;
private const int channels = 1;
private AudioClip playbackClip;
private const int playbackClipLengthSeconds = 2; // 1-second clip buffer
private object bufferLock = new object();
public static void Register(ulong clientId, VoiceReceiverBehaviour receiver)
{
receivers[clientId] = receiver;
}
public static VoiceReceiverBehaviour GetReceiver(ulong clientId)
{
receivers.TryGetValue(clientId, out var r);
return r;
}
public override void OnNetworkSpawn()
{
if (IsClient)
{
Register(OwnerClientId, this);
}
}
void Awake()
{
decoder = (OpusDecoder)OpusCodecFactory.CreateDecoder(sampleRate, channels);
decodeBuffer = new short[320];
floatBuffer = new float[320];
// Create a streaming AudioClip for continuous playback
playbackClip = AudioClip.Create("VoicePlayback", sampleRate * playbackClipLengthSeconds, channels, sampleRate, true, OnAudioRead, OnAudioSetPosition);
if (audioSource != null)
{
audioSource.clip = playbackClip;
audioSource.loop = true;
audioSource.Play();
}
Debug.Log("VoiceReceiverBehaviour initialized with playback buffer");
}
// This is called on the audio thread by Unity to fill the audio buffer
private void OnAudioRead(float[] data)
{
lock (bufferLock)
{
for (int i = 0; i < data.Length; i++)
{
if (playbackBuffer.Count > 0)
data[i] = playbackBuffer.Dequeue();
else
data[i] = 0f; // Silence if no data buffered
}
}
}
private void OnAudioSetPosition(int newPosition)
{
// Reset playback buffer on position change
lock (bufferLock)
{
playbackBuffer.Clear();
}
}
public void ReceiveVoiceData(byte[] encoded, GameObject source)
{
if(source != transform.gameObject) return;
if (decoder == null)
{
Debug.LogWarning("Decoder not initialized!");
return;
}
int decodedSamples = decoder.Decode(encoded, 0, encoded.Length, decodeBuffer, 0, decodeBuffer.Length, false);
for (int i = 0; i < decodedSamples; i++)
floatBuffer[i] = decodeBuffer[i] / 32768f;
lock (bufferLock)
{
// Enqueue decoded samples for playback
foreach (var sample in floatBuffer.AsSpan(0, decodedSamples))
playbackBuffer.Enqueue(sample);
}
Debug.Log($"Received voice data: {encoded.Length} bytes, decoded to {decodedSamples} samples. Buffer length: {playbackBuffer.Count}");
}
private void OnDestroy()
{
receivers.Remove(OwnerClientId);
}
}
`
