Using OpenSL ES for Audio Recording

OpenSL ES, short for Open Sound Library for Embedded Systems, is an embedded, cross-platform audio processing library that is free of charge. It provides high-performance and low-latency APIs for you to develop applications running on embedded mobile multimedia devices. OpenHarmony have implemented certain native APIs based on OpenSL ES 1.0.1 API specifications developed by the Khronos Group. You can use these APIs through <OpenSLES.h> and <OpenSLES_OpenHarmony.h>.

OpenSL ES on OpenHarmony

Currently, OpenHarmony implements parts of OpenSL ES APIs to implement basic audio recording functionalities.

If an API that has not been implemented on OpenHarmony is called, SL_RESULT_FEATURE_UNSUPPORTED is returned.

The following lists the OpenSL ES APIs that have been implemented on OpenHarmony. For details, see the OpenSL ES specifications.

  • Engine APIs implemented on OpenHarmony

    • SLresult (*CreateAudioPlayer) (SLEngineItf self, SLObjectItf * pPlayer, SLDataSource *pAudioSrc, SLDataSink *pAudioSnk, SLuint32 numInterfaces, const SLInterfaceID * pInterfaceIds, const SLboolean * pInterfaceRequired)
    • SLresult (*CreateAudioRecorder) (SLEngineItf self, SLObjectItf * pRecorder, SLDataSource *pAudioSrc, SLDataSink *pAudioSnk, SLuint32 numInterfaces, const SLInterfaceID * pInterfaceIds, const SLboolean * pInterfaceRequired)
    • SLresult (*CreateOutputMix) (SLEngineItf self, SLObjectItf * pMix, SLuint32 numInterfaces, const SLInterfaceID * pInterfaceIds, const SLboolean * pInterfaceRequired)
  • Object APIs implemented on OpenHarmony

    • SLresult (*Realize) (SLObjectItf self, SLboolean async)
    • SLresult (*GetState) (SLObjectItf self, SLuint32 * pState)
    • SLresult (*GetInterface) (SLObjectItf self, const SLInterfaceID iid, void * pInterface)
    • void (*Destroy) (SLObjectItf self)
  • Recorder APIs implemented on OpenHarmony

    • SLresult (*SetRecordState) (SLRecordItf self, SLuint32 state)
    • SLresult (*GetRecordState) (SLRecordItf self,SLuint32 *pState)
  • BufferQueue APIs implemented on OpenHarmony

    The APIs listed below can be used only after <OpenSLES_OpenHarmony.h> is introduced.

API Description
SLresult (*Enqueue) (SLOHBufferQueueItf self, const void *buffer, SLuint32 size) Adds a buffer to the corresponding queue.
For an audio playback operation, this API adds the buffer with audio data to the filledBufferQ_ queue. For an audio recording operation, this API adds the idle buffer after recording data storage to the freeBufferQ_ queue.
The self parameter indicates the BufferQueue object that calls this API.
The buffer parameter indicates the pointer to the buffer with audio data or the pointer to the idle buffer after the recording data is stored.
The size parameter indicates the size of the buffer.
SLresult (*Clear) (SLOHBufferQueueItf self) Releases a BufferQueue object.
The self parameter indicates the BufferQueue object that calls this API.
SLresult (*GetState) (SLOHBufferQueueItf self, SLOHBufferQueueState *state) Obtains the state of a BufferQueue object.
The self parameter indicates the BufferQueue object that calls this API.
The state parameter indicates the pointer to the state of the BufferQueue object.
SLresult (*RegisterCallback) (SLOHBufferQueueItf self, SlOHBufferQueueCallback callback, void* pContext) Registers a callback.
The self parameter indicates the BufferQueue object that calls this API.
The callback parameter indicates the callback to be registered for the audio playback or recording operation.
The pContext parameter indicates the pointer to the audio file to be played for an audio playback operation or the pointer to the audio file to be recorded for an audio recording operation.
SLresult (*GetBuffer) (SLOHBufferQueueItf self, SLuint8** buffer, SLuint32* size) Obtains a buffer.
For an audio playback operation, this API obtains an idle buffer from the freeBufferQ_ queue. For an audio recording operation, this API obtains the buffer that carries recording data from the filledBufferQ_ queue.
The self parameter indicates the BufferQueue object that calls this API.
The buffer parameter indicates the double pointer to the idle buffer or the buffer carrying recording data.
The size parameter indicates the size of the buffer.

Sample Code

Linking the Dynamic Library in the CMake Script

target_link_libraries(sample PUBLIC libOpenSLES.so)

Refer to the sample code below to record an audio file.

  1. Add the header files.

    #include "SLES/OpenSLES.h"
    #include "SLES/OpenSLES_OpenHarmony.h"
    #include "SLES/OpenSLES_Platform.h"
    
  2. Use the slCreateEngine API to create and instantiate an engine object.

    SLObjectItf engineObject = nullptr;
    slCreateEngine(&engineObject, 0, nullptr, 0, nullptr, nullptr);
    (*engineObject)->Realize(engineObject, SL_BOOLEAN_FALSE);
    
  3. Obtain the engineEngine instance of the SL_IID_ENGINE API.

    SLEngineItf engineItf = nullptr;
    (*engineObject)->GetInterface(engineObject, SL_IID_ENGINE, &engineItf);
    
  4. Configure the recorder information (including the input source audiosource and output source audiosink), and create a pcmCapturerObject instance.

    SLDataLocator_IODevice io_device = {
        SL_DATALOCATOR_IODEVICE,
        SL_IODEVICE_AUDIOINPUT,
        SL_DEFAULTDEVICEID_AUDIOINPUT,
        NULL
    };
    SLDataSource audioSource = {
        &io_device,
        NULL
    };
    SLDataLocator_BufferQueue buffer_queue = {
        SL_DATALOCATOR_BUFFERQUEUE,
        3
    };
    // Configure the parameters based on the audio file format.
    SLDataFormat_PCM format_pcm = {
        SL_DATAFORMAT_PCM,           // Input audio format.
        1,                                              // Mono channel.
        SL_SAMPLINGRATE_44_1,        // Sampling rate, 44100 Hz.
        SL_PCMSAMPLEFORMAT_FIXED_16, // Audio sampling format, a signed 16-bit integer in little-endian format.
        0,
        0,
        0
    };
    SLDataSink audioSink = {
        &buffer_queue,
        &format_pcm
    };
    
    SLObjectItf pcmCapturerObject = nullptr;
    (*engineItf)->CreateAudioRecorder(engineItf, &pcmCapturerObject,
        &audioSource, &audioSink, 0, nullptr, nullptr);
    (*pcmCapturerObject)->Realize(pcmCapturerObject, SL_BOOLEAN_FALSE);
    
    
  5. Obtain the recordItf instance of the SL_IID_RECORD API.

    SLRecordItf  recordItf;
    (*pcmCapturerObject)->GetInterface(pcmCapturerObject, SL_IID_RECORD, &recordItf);
    
  6. Obtain the bufferQueueItf instance of the SL_IID_OH_BUFFERQUEUE API.

    SLOHBufferQueueItf bufferQueueItf;
    (*pcmCapturerObject)->GetInterface(pcmCapturerObject, SL_IID_OH_BUFFERQUEUE, &bufferQueueItf);
    
  7. Register the BufferQueueCallback function.

    static void BufferQueueCallback(SLOHBufferQueueItf bufferQueueItf, void *pContext, SLuint32 size)
    {
        // Obtain the user information passed in during the registration from pContext.
        SLuint8 *buffer = nullptr;
        SLuint32 pSize = 0;
        (*bufferQueueItf)->GetBuffer(bufferQueueItf, &buffer, &pSize);
        if (buffer != nullptr) {
            // The recording data can be read from the buffer for subsequent processing.
            (*bufferQueueItf)->Enqueue(bufferQueueItf, buffer, size);
        }
    }
    void *pContext; // This callback can be used to obtain the custom context information passed in.
    (*bufferQueueItf)->RegisterCallback(bufferQueueItf, BufferQueueCallback, pContext);
    
  8. Start audio recording.

    (*recordItf)->SetRecordState(recordItf, SL_RECORDSTATE_RECORDING);
    
  9. Stop audio recording.

    (*recordItf)->SetRecordState(recordItf, SL_RECORDSTATE_STOPPED);
    (*pcmCapturerObject)->Destroy(pcmCapturerObject);