UniVoice

An extensible voice chat/VoIP solution for Unity.

Features

👥 Group Voice Chat
- Multiple peers can join a chatroom and exchange audio.
- APIs to mute audio on a per peer basis
📦 Plug-And-Play
- No need to write audio and networking code for most use cases
- Support for multiple network types, including PUN2 and WLAN
🎛️ Customizable
- UniVoice is pretty much collection of interfaces, you can make your own implementations
- For more info, see the Customization section below

Installation

The recommended installation is via the UniVoice UPM package named com.adrenak.univoice available on npmjs.org

Add the NPM registry with UniVoice as a scope. To do this, go to Edit/Project Settings/Package Manager, add the NPM scope registry with the URL https://registry.npmjs.org and add com.adrenak.univoice as a scope.
Install the package from NPM registry To do this, go to Window/Package Manager and refresh packages. Select in the My Registries view (located at the top left, the default selection is Unity Registry), locate UniVoice and click install. After installation, the package will show up in the In Project view as well.

Usage

Creating a chatroom agent

To be able to host and join voice chatrooms, you need a ChatroomAgent instance.

var agent = new ChatroomAgent(IChatroomNetwork network, IAudioInput audioInput, IAudioOutputFactory audioOutputFactory);

Hosting and joining chatrooms

agent.Network is the most commonly accessed object to do things with UniVoice.

// Host a chatroom 
agent.Network.HostChatroom(optional_data);

// Join an existing
agent.Network.JoinChatroom(optional_data);

// Leave the chatroom
agent.Network.LeaveChatroom(optional_data);

// Closes a chatroom
agent.Network.CloseChatroom(optional_data);

The current mode of the ChatroomAgent can be accessed using agent.Network.CurrentMode, this will return a ChatroomAgentMode enum object, the possible values are:

Unconnected: The agent is neither hosting or currently a guest of a chatroom
Host: The agent is currently hosting a chatroom
Guest: The agent is currently a guest in a chatroom

Interacting with peers

Everyone in the chatroom is assigned an ID by the host. And everyone has a list of IDs of their peers. An ID in UniVoice is a C# short that is unique for each member in the chatroom

To get your ID

short myID = agent.Network.OwnID;

To get a list of the other peers in the chatroom, use this:

short others = agent.Network.PeerIDs

To mute everyone in the chatroom, use agent.MuteOthers = true;, set it to false to unmute. This will stop audio from every peer in the chatroom.

To mute yourself to everyone in the chatroom use agent.MuteSelf = true; set it to false to unmute. This will stop sending your audio to the peers in the chatroom.

For muting a specific peer, first get the peers settings object using this:

agent.PeerSettings[id].muteThem = true; // where id belongs to the peer you want to mute

If you want to mute yourself towards a specific peer, use this:

agent.PeerSettings[id].muteSelf = true; // where id belongs to the peer you don't want to hear you

Events

agent.Network provides several network related events. Refer to the API reference for them.

⚙️ Customization

UniVoice is a plug-and-play solution with several pre-existing modules. The most common use case (which is probably 3D/spatial audio group chat) doesn't need you to write any audio or networking code yourself

But it also offers you ways to extend and modify it.

🎤 Audio Input

UniVoice is audio input source agnostic. This means, it doesn't care where it is getting the audio from. It only cares about getting the audio data, giving you freedom to change where it gets that from.

transmit real-time mic input
send audio by reading an mp3 file from disk
send the audio track of a video file.
send in-game audio

The most common input source is real-time mic input. If this is what you're looking for, an official input implementation based on UniMic is available here which can be used for this.

The IAudioInput interface API reference is here

🔊 Audio Output

UniVoice is also audio output agnostic. This means, it doesn't care what you do with the audio you receive from peers. It just gives you the data and forgets about it, leaving it up to you what you want to do with it.

you can play it back in Unity using an AudioSource
you can save it to disk
you can use it as input for a speech-to-text
stream it to a server

The most common use output source is playing it inside your app. An official output implementation that plays peer audio on Unity AudioSource is available here

Creating an audio output requires implementing two interfaces. They are IAudioOutput and IAudioOutputFactory

🌐 Network

You guessed it. UniVoice is network agnostic. This means, it doesn't care how the audio is exchanged between peers. It just needs a networking implementation of its IChatroomNetwork interface, which allows you to adapt it to different kinds of network infra.

you can send audio over WLAN
or webrtc
or popular networking solution providers such as Photon
or your own custom backend solution

Creating your own network requires you to implement the IChatroomNetwork interface

Docs and samples

Manual is not available yet. For the API reference, please visit http://www.vatsalambastha.com/univoice

A sample app is available here

License and Support

This project is under the MIT license.

Updates and maintenance are not guaranteed and the project is maintained by the original developer in his free time. Community contributions are welcome.

Commercial consultation and development can be arranged but is subject to schedule and availability.

Contact

The developer Vatsal Ambastha can be reached at the following links:

Website
LinkedIn
GitHub
Twitter

adrenak/univoice