Audio Recorder plugin for Xamarin and Windows

Features:

• Records audio on a device's microphone input
• Allows access to recorded audio via file or Stream
• Configurable silence detection to automatically end recording
• Simple event and Task-based APIs
• Cross platform AudioPlayer included

Setup

• Available on NuGet: https://www.nuget.org/packages/Plugin.AudioRecorder
• Install into your platform-specific projects (iOS/Android/UWP), and any PCL/.NET Standard 2.0 projects required for your app.

Platform Support

Platform	Supported	Version	Notes
Xamarin.iOS	Yes	iOS 7+
Xamarin.Android	Yes	API 16+	Project should target Android framework 8.1+
Windows UWP	Yes	10.0	Build 15063 and up

Notes:

• Supports both native Xamarin.iOS / Xamarin.Android and Xamarin.Forms projects.
• Contains reference assemblies to use the library from PCL projects (profile 111) and .NET Standard 2.0 projects.
  • Please note the PCL/.NET Standard support is via "bait & switch", and this will ONLY work alongside the proper platform-specific DLLs/NuGet packages in place.

Required Permissions & Capabilities

The following permissions/capabilities are required to be configured on each platform:

Android

<uses-permission android:name="android.permission.MODIFY_AUDIO_SETTINGS" />
<uses-permission android:name="android.permission.READ_EXTERNAL_STORAGE" />
<uses-permission android:name="android.permission.WRITE_EXTERNAL_STORAGE" />
<uses-permission android:name="android.permission.RECORD_AUDIO" />

Additionally, on OS versions Marshmallow and above, you may need to perform a runtime check to ask the user to access their microphone.

Example:

Do this in your main activity or at the point you'll be needing access to record audio:

if (ContextCompat.CheckSelfPermission (this, Manifest.Permission.RecordAudio) != Permission.Granted)
{
	ActivityCompat.RequestPermissions (this, new String [] { Manifest.Permission.RecordAudio }, 1);
}

iOS

For iOS 10 and above, you must set the NSMicrophoneUsageDescription in your Info.plist:

<key>NSMicrophoneUsageDescription</key>
<string>The [app name] wants to use your microphone to record audio.</string>

UWP

You must check the Internet and Microphone capabilities in your app's Package.appxmanifest file.

Usage

In a controller/activity/page, initialize a new AudioRecorderService.

Example:

recorder = new AudioRecorderService
{
	StopRecordingOnSilence = true, //will stop recording after 2 seconds (default)
	StopRecordingAfterTimeout = true,  //stop recording after a max timeout (defined below)
	TotalAudioTimeout = TimeSpan.FromSeconds (15) //audio will stop recording after 15 seconds
};

More settings and properties are defined below

Recording

To begin recording, use the StartRecording () and StopRecording () methods as shown:

async void RecordButton_Click (object sender, EventArgs e)
{
	await RecordAudio ();
}

async Task RecordAudio ()
{
	try
	{
		if (!recorder.IsRecording)
		{
			await recorder.StartRecording ();
		}
		else
		{
			await recorder.StopRecording ();
		}
	}
	catch (Exception ex)
	{
	...
	}
}

In lieu of calling StopRecording (), you can also make use of the StopRecordingAfterTimeout and/or StopRecordingOnSilence settings, which are explained below.

Using the Audio Data

Once recording has begun, there are two different ways to determine when recording has finished:

Task-based API

To use the Task-based API, you can grab the returned Task from the call to StartRecording (). This allows you to await the result of the Task, which will complete when recording is complete and return the path to the recorded audio file.

Example:

var recordTask = await recorder.StartRecording ();

... maybe do some other things like toggle your 'mic' button off while recording

//await the returned Task... this will complete once recording has been stopped
var audioFile = await recordTask;

if (audioFile != null) //non-null audioFile indicates audio was successfully recorded
{
	//do something with the file
}

Event-based API

The AudioInputReceived is raised when recording is complete, and the full filepath of the recorded audio file is passed along.

Example:

recorder.AudioInputReceived += Recorder_AudioInputReceived;

...

await recorder.StartRecording ();

...

private async void Recorder_AudioInputReceived(object sender, string audioFile)
{
	//do something with the file
}

NOTE: This event is raised on a background thread to allow for further file processing as needed. If the audioFile is null or empty, no audio was recorded.

--

There are also multiple ways to use the recorded (or recording) audio data:

Accessing the Recorded File

There are multiple ways to access the recorded audio file path:

• The Task-based API will return the file path when the task completes. The Task can be awaited or use standard Task continuation APIs.
• The Event-based API will return the full path to the recorded audio file in the audioFile parameter of the AudioInputReceived event handler.
• The GetAudioFilePath () method on the AudioRecorderService class will return the recorded audio file path.

These will all return null in the case that no audio has been recorded yet or no audio was recorded/detected in the last recording session.

Once you have the path to the recorded audio file, you can use standard file operations (for native/.NET Standard) and/or a cross platform file system abstraction like PCLStorage to get a stream to the file data.

Concurrent Streaming

It's also possible to get a stream to the recording audio data as it's being recorded, once StartRecording () has been called.

To access this readonly stream of audio data, you may call the GetAudioFileStream () method. This is useful in the case you want to immediately begin streaming the audio data to a server or other consumer.

NOTE: Since the WAV header is written after recording, once the audio length is known, the provided Stream data will contain the PCM audio data only and will not contain a WAV header. If your use case requires a WAV header, you can call AudioFunctions.WriteWaveHeader (Stream stream, int channelCount, int sampleRate, int bitsPerSample), which will write a WAV header to the stream with an unknown length.

Since GetAudioFileStream () will return a Stream that is also being populated concurrently, it can be useful to know when the recording is complete - the Stream will continue to grow! This can be accomplished with either the Event-based API or the Task-based API (which is often more useful).

An example of the Task-based API and concurrent writing and reading of the audio data is shown in the sample accompanying the Xamarin.Cognitive.Speech library. This speech client will stream audio data to the server until the AudioRecordTask completes, signaling that the recording is finished.

Properties and Settings

• IsRecording

  bool IsRecording

  Returns a value indicating if the AudioRecorderService is currently recording audio.

• StopRecordingAfterTimeout / TotalAudioTimeout

  bool StopRecordingAfterTimeout

  Gets/sets a value indicating if the AudioRecorderService should stop recording after a certain amount of time. Default is true.

  TimeSpan TotalAudioTimeout

  If StopRecordingAfterTimeout is set to true, this TimeSpan indicates the total amount of time to record audio for before recording is stopped. Defaults to 30 seconds.

• StopRecordingOnSilence / AudioSilenceTimeout

  bool StopRecordingOnSilence

  Gets/sets a value indicating if the AudioRecorderService should stop recording after silence (low audio signal) is detected. Default is true.

  TimeSpan AudioSilenceTimeout

  If StopRecordingOnSilence is set to true, this TimeSpan indicates the amount of 'silent' time is required before recording is stopped. Defaults to 2 seconds.

• SilenceThreshold

  float SilenceThreshold

  Gets/sets a value indicating the signal threshold that determines silence. If the recorder is being over or under aggressive when detecting silence for your use case, you can alter this value to achieve different results. Defaults to .15. Value should be between 0 and 1.

• FilePath

  string FilePath

  Gets/sets the desired file path. If null it will be set automatically to a temporary file.

• ConfigureAVAudioSession (iOS only)

  static bool ConfigureAVAudioSession

  Gets/sets whether the AudioRecorderService should attempt to control the shared AVAudioSession category. This can be set from the AppDelegate (example) or other iOS-specific class. When set to true, the AudioRecorderService will attempt to set the category to Record before recording, and restore the category to its previous value after recording is complete. This can help this plugin coexist with other media plugins that try to unilaterally set the category. Defaults to false.

Limitations

• Currently this is only recording in WAV audio format (due to original use case this was developed for).

Samples

Complete samples demonstrating audio recording (AudioRecorderService) and playback (AudioPlayer) of the recorded file are available in the /Samples folder:

• Xamarin.Forms (.NET Standard) containing iOS, Android, and UWP apps.
• Native iOS, Android, and UWP apps.

Contributing

Contributions are welcome. Feel free to file issues and pull requests on the repo and they'll be reviewed as time permits.

Building

This solution requires the following environment/config:

• Visual Studio 2017 or above
  • You may be able to load and build the iOS/Android targets on Visual Studio for Mac
• If you want to build the NuGet package (for some reason):
  • Nugetizer 3000 setup:
    • Install the Visual Studio extension
    • Nugetizer 3000 CI NuGet package source for development builds of NuGet.Build.Packaging

Debugging

The easiest way to debug this library is to clone this repo and copy the shared library + platform project (of the platform you're debugging with) into your solution folder, then include the projects in your solution and reference them directly, rather than referencing the NuGet package.

About

• Created by Nate Rickard
• AudioRecorderService concept and some original code from SmartCoffee by Pierce Boggan
• Audio stream and wave recorder classes adapted from SimplyMobile by Sami M. Kallio

License

Licensed under the MIT License (MIT). See LICENSE for details.