UE5 MicrophoneRecorder plugin document

1. Introduction

MicrophoneRecorder is an Unreal Engine plugin that provides microphone recording, streaming audio capture, audio format conversion, and playback. All features are exposed as Blueprint nodes and can be used directly in the Blueprint editor. It has the following features:

1. Streaming and non-streaming mode
2. Both USoundWave and PCM binary data are provided in the output pins.
3. The main user scenarios are streaming playback and cloud ASR audio (audio chunks) input.
4. Support Android/iOS/Win/Linux/Mac devices.

1.1 Sample files:

• Copiable blueprint sample can be found here: https://blueprintue.com/blueprint/n2oy88a2/

• Blueprint sample file can be found here: https://github.com/bingothreed/LocalASRService/blob/main/UEBPs/MicrophoneRecorder/MicrophoneRecorderBasicUsage.uasset

1.2 Component preview

1. Add MicrophoneReocrder component
   

2. Details pannel preview
   

1.3 Typical usage blueprint screenshots:

1. Grant permission
   

2. Batch Recording
   

3. Streaming Recording with chunks
   

4. Streaming Recording with realtime playback
   

2. UMicrophoneRecorderComponent

BlueprintSpawnableComponent. Add it to any Actor via “Add Component → Audio”.

2.1 Permission

Node	Signature	Type	Description
Has Microphone Permission	bool HasMicrophonePermission() const	BlueprintPure	Returns true if microphone permission is already granted. Always true on Windows/Linux/Mac
Request Microphone Permission	void RequestMicrophonePermission()	BlueprintCallable	Triggers the OS permission dialog on Android/iOS. Fires OnMicrophonePermissionResult when the user responds. Fires immediately on platforms that need no permission

2.2 Batch Recording

Start once, stop once — get the complete recording.

Node	Signature	Type	Description
Start Record	bool StartRecord()	BlueprintCallable	Starts batch recording. Returns false if no permission, if streaming is active, or if the audio device fails to initialize
Stop Record	void StopRecord(USoundWave*& RecordAudio, TArray<uint8>& BinaryData)	BlueprintCallable	Stops batch recording and returns the recorded audio as both a transient USoundWave and a WAV byte array

Mutual exclusion: batch recording and streaming capture cannot run simultaneously.

---

2.3 Streaming Capture

Captures audio continuously and delivers it in real-time chunks.

Node	Signature	Type	Description
Start Stream Capture	bool StartStreamCapture()	BlueprintCallable	Starts streaming capture. Fires OnAudioChunkReady each time a chunk duration elapses. Returns false if already streaming, no permission, or batch recording is active
Stop Stream Capture	void StopStreamCapture(USoundWave*& RecordAudio, TArray<uint8>& BinaryData)	BlueprintCallable	Stops streaming and returns the full accumulated recorded audio as both a USoundWave and WAV byte array
Is Stream Capturing	bool IsStreamCapturing() const	BlueprintPure	Returns whether streaming capture is currently active
Get Stream Sample Rate	int32 GetStreamSampleRate() const	BlueprintPure	Returns the sample rate of the streaming capture (thread-safe)
Get Stream Num Channels	int32 GetStreamNumChannels() const	BlueprintPure	Returns the number of channels in the streaming capture (thread-safe)

---

2.4 Delegates

Delegate	Signature	Trigger
On Microphone Permission Result	FOnMicPermissionResult(bool bGranted)	Fired when the user responds to the microphone permission dialog. bGranted indicates whether permission was granted
On Audio Chunk Ready	FOnAudioChunkReady(const TArray<uint8>& PCMInt16Data, int32 SampleRate)	Fired on the game thread each time a PCM chunk is ready during streaming capture. Data is mono int16 little-endian at the native device sample rate

---

2.5 Configurable Properties

Property	Type	Default	Description
Stream Chunk Duration Ms	int32 (BlueprintReadWrite)	100	Duration of each chunk delivered via OnAudioChunkReady, in milliseconds

---

3. UAudioUtils (Audio Utility Function Library)

UBlueprintFunctionLibrary. All methods are static and can be called directly in Blueprints.

3.1 Import / Export / Conversion

Node	Signature	Type	Description
Import Audio From Binary	USoundWave* ImportAudioFromBinary(const TArray<uint8>& AudioData)	BlueprintPure	Imports a WAV byte array into a transient USoundWave. Returns nullptr on failure
Sound Wave To Bytes	TArray<uint8> SoundWaveToBytes(const USoundWave* Audio)	BlueprintPure	Serialises a USoundWave’s RawPCMData into a WAV byte array
Convert Record	TArray<uint8> ConvertRecord(const TArray<uint8>& PCMBytes, int32 SampleRate = 48000, int32 NumChannels = 1, int32 BitDepth = 16)	BlueprintPure	Wraps raw PCM bytes in a WAV header. No sample-rate conversion is performed
Resample PCM Int16	TArray<uint8> ResamplePCMInt16(const TArray<uint8>& PCMInt16Data, int32 SourceRate, int32 TargetRate)	BlueprintPure	Resamples mono int16 PCM from source rate to target rate using linear interpolation. Suitable for downsampling speech audio before ASR (e.g. 48000 Hz to 16000 Hz). Returns input unchanged if rates are equal
Save TArray As WAV File	bool SaveTArrayAsWAVFile(const TArray<uint8>& AudioData, const FString& FilePath)	BlueprintCallable	Writes a WAV byte array directly to a file on disk. Returns whether the write succeeded
Save Sound Wave As Asset	USoundWave* SaveSoundWaveAsAsset(const TArray<uint8>& BinaryData, const FString& AssetPath)	BlueprintCallable	Saves WAV binary data into the Content Browser as a persistent .uasset. AssetPath is the full content path, e.g. "/Game/Recordings/MySound". Editor-only (DevelopmentOnly); returns nullptr at runtime

---

3.2 Real-time Playback

Node	Signature	Type	Description
Create Streaming Playback Wave	USoundWaveProcedural* CreateStreamingPlaybackWave(int32 SampleRate, int32 NumChannels = 1)	BlueprintCallable	Creates a USoundWaveProcedural ready to receive streaming PCM chunks. Use with SpawnSound2D to start playback
Queue PCM Chunk For Playback	void QueuePCMChunkForPlayback(USoundWaveProcedural* Wave, const TArray<uint8>& PCMInt16Data)	BlueprintCallable	Queues one PCM chunk (mono int16 raw bytes) into a procedural wave for playback. Typically called inside OnAudioChunkReady for seamless streaming audio output

---

4. Typical Blueprint Usage Flows

Batch Recording

HasMicrophonePermission → RequestMicrophonePermission → [wait for OnMicrophonePermissionResult]
→ StartRecord → ... → StopRecord → (optional) ImportAudioFromBinary / SaveTArrayAsWAVFile

Streaming Capture + Real-time Playback

HasMicrophonePermission → RequestMicrophonePermission → [wait for OnMicrophonePermissionResult]
→ CreateStreamingPlaybackWave → SpawnSound2D
→ StartStreamCapture → [call QueuePCMChunkForPlayback in OnAudioChunkReady callback]
→ StopStreamCapture → SaveTArrayAsWAVFile

Speech Recognition Preprocessing

... → StopRecord → ResamplePCMInt16 (48000 → 16000) → ConvertRecord → feed to ASR interface

---

5. Node Index by Category

Category	Nodes
MicrophoneRecorder\|Permission	HasMicrophonePermission, RequestMicrophonePermission, OnMicrophonePermissionResult
MicrophoneRecorder	StartRecord, StopRecord, ImportAudioFromBinary, ConvertRecord, SoundWaveToBytes, ResamplePCMInt16, SaveTArrayAsWAVFile, SaveSoundWaveAsAsset
MicrophoneRecorder\|Streaming	StartStreamCapture, StopStreamCapture, IsStreamCapturing, GetStreamSampleRate, GetStreamNumChannels, OnAudioChunkReady, StreamChunkDurationMs
MicrophoneRecorder\|Playback	CreateStreamingPlaybackWave, QueuePCMChunkForPlayback

---

6. Issues

6.1 Permission

By defaule, the permission should be automatically integrated in the plugin. But if you still met the permission:

1. If you want to pack it into iPhone, you can add this line to Project Settings-> IOS-> Extra Plist Data-> Additional Plist Data: <key>NSMicrophoneUsageDescription</key> <string>Record Audio</string>
2. If you want to pack it into Android, you can add this line to Project Settings-> Platform->Android-> Advanced APK Packing-> Extra Permissions: android.permission.RECORD_AUDIO

Plugin Version: 1.0

Engine Version: 5.7.0/5.6.0/5.5.0

Platforms: Win64 / Linux / LinuxArm64 / Mac / iOS / Android