Download source code of this article.
My previous article was about recording audio using MediaRecorder class. This class is very easy to use and can be used for quick and simple audio recording, also this class saves the audio data either MPEG 4 or 3GPP compressed format. But if you need raw data that can be used to process audio data, then you cannot use MediaRecorder class but you have to use AudioRecord class. AudioRecord class provides you the raw data in uncompressed format. You can use this data to write to a file, display waveform, etc..
To record audio using AudioRecord class:
- Create an instance of AudioRecord
- Start recording using startRecording() method
- Read uncompressed data using AudioRecord.read() method
- Stop recording using stop() method.
Creating AudioRecord instance
The constructor of AudioRecord class is:
int audioSource, int sampleRateInHz,
int channelConfig, int audioFormat, int bufferSizeInBytes)
The first parameter is the audio source, this can be one of the AudioSource values. Second parameter is the sample rate in Hz. This can be 44100, 22050, 11025, 8000, etc.. Third parameter is channel value, this can be one of the AudioFormat values (normally CHANNEL_IN_MONO or CHANNEL_IN_STEREO). Third is the audio encoder format, this can be one of the following values:
- ENCODING_PCM_16BIT – 16 bit per sample
- ENCODING_PCM_8BIT – 8 bit per sample
- ENCODING_DEFAUL – default encoding
Forth parameter is the buffer size. This should be calculated using AudioRecord.getMinBufferSize() static method. The syntax of getMinBufferSize() method is:
public static int getMinBufferSize (
int sampleRateInHz, int channelConfig, int audioFormat)
The parameters are same as that of the constructor. The method calculate the minimum buffer size needed to store one chunk of audio. Anything less than this number will result in failure while creating the AudioRecord object.
Once we create an instance of the AudioRecord object, startRecording() method is used to start recording audio.
Reading uncompressed data
Once the recording is started, it is the responsibility of the application to read the audio data and store it for further processing. We can read the audio data using one of the read method of AudioRecord class. The audio data will be raw PCM_16BIT or PCM_8BIT format depending on how you initialized the object. The application can use this data to do further processing of any sort. In our sample application we just save this data to a temporary file, further when the recording is stopped we read this data and write to WAV file. The application has to continuously read the data otherwise the previous chunk of data will be overwritten by the new one.
Recording can be stopped using stop() method. Remember to call release() method to release the AudioRecord object.
The following code snippet shows the typical usage:
AudioRecord recorder = new AudioRecord(MediaRecorder.AudioSource.MIC,
The sample application of this article records audio and save it to a WAV file. The WAV file will be placed in “/SDCard/AudioRecorder” folder with current millisecond as the file name.
Writing WAV file – RingDroid sample application, RehearsalAssist application