User Guide

Purpose

NsVoice gives proximity voice chat to a multiplayer game. A component on the player pawn reads the microphone, encodes the voice with Opus, and sends it through the host. The other machines play the voice from that component, with its attenuation. The player can pick the microphone, the gain, and the volumes.

NsVoice does not bind input, draw UI, or save options. The game binds push to talk, draws the meter, and saves the picked microphone, the gain, and the volumes. NsVoice does not use the voice chat of the online subsystem.

Setup

1. Copy the NsVoice folder into the Plugins folder of your project.
2. Add "NsVoice" to the dependencies of the game module that uses the components.
3. Open Config/DefaultEngine.ini.
4. Add bEnabled=true to the [Voice] section.
5. Generate the project files.
6. Enable the plugin in Edit > Plugins. The engine plugin AudioCapture turns on with it.

The [Voice] line is required. Without it, the engine does not create the Opus encoder and decoder.

These lines are a common setup for this kind of voice chat. They are not required by the code:

[OnlineSubsystem]
bHasVoiceEnabled=false

[/Script/Engine.Player]
ConfiguredInternetSpeed=300000
ConfiguredLanSpeed=300000

The first line keeps the engine voice chat off. The other two lines give the connection more bandwidth for the voice packets.

The plugin has no project settings page.

Quick Start

In Blueprint:

1. Open the Blueprint of the player character.
2. In the Components panel, click Add and search for Ns Voice Chat.
3. Attach the new component to the head socket of the mesh.
4. In the Details panel of the component, set an Attenuation Settings asset, or turn on Override Attenuation.
5. Start Play In Editor with two players and Net Mode set to Play As Listen Server.
6. Talk into the microphone. The other player hears the voice from the head of your character.

In C++:

#include "NsVoiceChatComponent.h"

AMyHero::AMyHero()
{
    VoiceChat = CreateDefaultSubobject<UNsVoiceChatComponent>(TEXT("VoiceChat"));
    VoiceChat->SetupAttachment(GetMesh(), TEXT("head"));
}

Transmit When Locally Controlled is on by default. The microphone opens one time, when this machine controls the pawn.

Common Tasks

Use push to talk

1. Select the Ns Voice Chat component.
2. Turn off Transmit When Locally Controlled.
3. On the key press, call Set Transmitting with New Transmitting on.
4. On the key release, call Set Transmitting with New Transmitting off.

Is Transmitting returns true while the microphone is open.

Pick the microphone (C++)

1. Call UNsMicSubsystem::GetDevices() to get the list. Each FNsMicDevice has an Id and a Name.
2. Show the Name values to the player.
3. Call UNsMicSubsystem::Get(this)->SetDeviceId(Device.Id) with the picked device.
4. Save the ID in your options. On the next start, call SetDeviceId again.

An empty ID picks the default microphone. An open stream opens again on the new device at once. GetDeviceName returns the name of a saved ID.

Set the microphone gain

1. Get the subsystem with the Get Game Instance Subsystem node and the class NsMicSubsystem.
2. Call Set Gain with a value from 0 to 4.

The gain applies to the meter, the talk level, and the voice that the other players hear.

Show a microphone meter

1. Get NsMicSubsystem.
2. Read Get Level each frame. The value goes from 0 to 1, after the gain.

Get Level returns 0 while the microphone is closed. In C++, call OpenListener to keep the microphone open for a meter, and CloseListener when the meter closes.

Set the volumes

1. Get the subsystem with the Get Game Instance Subsystem node and the class NsVoiceSubsystem.
2. Call Set Voice Volume with a value from 0 to 2. It applies to every voice.
3. Call Set Player Volume with the player name and a value from 0 to 2. It applies to one player.

The player name is the name in the player state of the speaker. A player volume lasts for the session.

Show who speaks

1. Add an Ns Voice Activity component to the same pawn as the Ns Voice Chat component.
2. Bind the event On Speaking Changed of the activity component.
3. Read Is Speaking at any time.

The event fires on every machine. The voice chat component gives the microphone level to the activity component while the microphone is open.

Move a mouth or a light with the voice

1. Read Get Talk Level on the Ns Voice Chat component each frame.
2. Drive the mouth, the head, or the light with the value from 0 to 1.

On the speaker, the value comes from the microphone. On the other machines, it comes from the voice that the machine hears. Get Voice Level gives the raw level of the heard voice.

Change the volume with a game rule

1. Make a C++ child class of UNsVoiceChatComponent.
2. Override GetPlaybackVolume_Implementation.
3. Return the volume for your rule, for example full volume for a radio.
4. Call Super::GetPlaybackVolume_Implementation() to keep the voice volume and the player volume.

Settings

Ns Voice Chat component (UNsVoiceChatComponent), category Voice:

Ns Voice Chat component, category Voice > Talk Level:

Ns Voice Activity component (UNsVoiceActivityComponent), category Voice:

Subsystems, set at runtime:

Debug

• Log category LogNsVoice. Set it to Verbose with Log LogNsVoice Verbose to see each received packet, with its size and peak.
• At Log level, LogNsVoice shows the sample rate when the microphone opens, and warnings when a microphone does not open.

Troubleshooting

Limits

• The owner of both components must be a pawn.
• One microphone stream for each game instance.
• No echo cancellation and no noise gate.
• The speaker does not hear its own voice. A dedicated server does not play voice.
• The host relays every packet to every machine where the pawn is relevant.
• Player volumes use the player name. Two players with the same name share one volume.
• The plugin does not save the picked microphone, the gain, or the volumes.
• GetDevices, SetDeviceId, OpenListener, and CloseListener are C++ only.

Contact

E-mail LinkedIn GitHub Bluesky