Introduction

Generally, Agora SDKs use default audio modules for capturing and rendering in real-time communications.

However, these default modules might not meet your development requirements, such as in the following scenarios:

• Your app has its own audio module.
• You need to process the captured audio with a pre-processing library.
• You need flexible device resource allocation to avoid conflicts with other services.

Agora provides a solution to enable a custom audio source and/or renderer in the above scenarios. This article describes how to do so using the Agora Native SDK.

Sample project

Agora provides an open-source demo project on GitHub. You can view the source code on Github or download the project to try it out.

Implementation

Before proceeding, ensure that you have implemented the basic real-time communication functions in your project. For details, see Start a Voice Call or Start Interactive Live Audio Streaming.

Custom audio source

Refer to the following steps to implement a custom audio source in your project:

1. Before calling joinChannel, call setExternalAudioSource to specify the custom audio source.
2. Implement audio capture and processing yourself using methods from outside the SDK.
3. Call pushExternalAudioFrame to send the audio frames to the SDK for later use.

API call sequence

Refer to the following diagram to implement the custom audio source:

Audio data transfer

The following diagram shows how the audio data is transferred when you customize the audio source:

• You need to implement the capture module yourself using methods from outside the SDK.
• Call pushExternalAudioFrame to send the captured audio frames to the SDK.

Code samples

Refer to the following code samples to implement the custom audio source in your project.

1. Before the local user joins the channel, specify the custom audio source.

// Specifies the custom audio source
engine.setExternalAudioSource(true, DEFAULT_SAMPLE_RATE, DEFAULT_CHANNEL_COUNT);
// The local user joins the channel
int res = engine.joinChannel(accessToken, channelId, "Extra Optional Data", 0);

2. Implement your own audio capture module. After the local user joins the channel, enable the capture module to start capturing audio frames from the custom audio source.

 public class RecordThread extends Thread
    {
        private AudioRecord audioRecord;
        public static final int DEFAULT_SAMPLE_RATE = 16000;
        public static final int DEFAULT_CHANNEL_COUNT = 1, DEFAULT_CHANNEL_CONFIG = AudioFormat.CHANNEL_IN_MONO;
        private byte[] buffer;

        RecordThread()
        {
            int bufferSize = AudioRecord.getMinBufferSize(DEFAULT_SAMPLE_RATE, DEFAULT_CHANNEL_CONFIG,
                    AudioFormat.ENCODING_PCM_16BIT);
            audioRecord = new AudioRecord(MediaRecorder.AudioSource.MIC, DEFAULT_SAMPLE_RATE, DEFAULT_CHANNEL_COUNT,
                    AudioFormat.ENCODING_PCM_16BIT, bufferSize);
            buffer = new byte[bufferSize];
        }
        // Starts audio capture. Reads and sends the captured frames until audio capture stops.
        @Override
        public void run()
        {
            try
            {
                // Start audio recording
                audioRecord.startRecording();
                while (!stopped)
                {
                    // Reads the captured audio frames
                    int result = audioRecord.read(buffer, 0, buffer.length);
                    if (result >= 0)
                    {
                        // Sends the captured audio frames to the SDK
                        CustomAudioSource.engine.pushExternalAudioFrame(
                                buffer, System.currentTimeMillis());
                    }
                    else
                    {
                        logRecordError(result);
                    }
                    Log.e(TAG, "Data size:" + result);
                }
                release();
            }
            catch (Exception e)
            {e.printStackTrace();}
        }

        ...
    }

API reference

• setExternalAudioSource
• pushExternalAudioFrame

Custom audio renderer

Refer to the following steps to implement a custom audio renderer in your project:

1. Before calling joinChannel, call setExternalAudioSink to enable and configure the external audio renderer.
2. After joining the channel, call pullPlaybackAudioFrame to retrieve the audio data sent by a remote user.
3. Use your own audio renderer to process the audio data, then play the rendered data.

API call sequence

Refer to the following diagram to implement the custom audio renderer in your project:

Audio data transfer

The following diagram shows how the audio data is transferred when you customize the audio renderer:

• You need to implement the rendering module yourself using methods from outside the SDK.
• Call pullPlaybackAudioFrame to retrieve the audio data sent by a remote user.

Code samples

Refer to the following code samples to implement the custom audio renderer in your project:

// Enables the custom audio renderer
rtcEngine.setExternalAudioSink(
    true,      // Enables external audio rendering
    44100,     // Sampling rate (Hz). You can set this value as 8000, 16000, 32000, 441000, or 48000
    1          // The number of channels of the external audio source. This value must not exceed 2
);

// Retrieves remote audio frames for playback
rtcEngine.pullPlaybackAudioFrame(
    data,             // The data type is byte[]
    lengthInByte      // The size of the audio data (byte)
);

API reference

• setExternalAudioSink
• pullPlaybackAudioFrame

Considerations

Performing the following operations requires you to use methods from outside the Agora SDK:

• Manage the capture and processing of audio frames when using a custom audio source.
• Manage the processing and playback of audio frames when using a custom audio renderer.