A C++ node.js module that allows manipulating audio streams from/to soundcards and work with the fft of the signal etc.
npm install soundengineA C++ addon for node.js that allows interacting with soundcards to e.g. play, record or process live microphone samples and send them back to the speakers.
The module uses PortAudio to interact with the different host apis.
Soundengine requires the shared portaudio library.
``sh`
$ brew install portaudio fftw
`sh`
$ sudo apt-get install libasound-dev libportaudio2 portaudio19-dev libfftw3-3 libfftw3-dev libfftw3-double3
...not supported yet.
`sh`
$ npm i soundengine
`javascript
const soundengine = require('soundengine')
// Start live transmission from the default input device to the default output device at 22kHz
var engine = new soundengine.engine({sampleRate: 22050})
// Start recording
engine.startRecording()
// Apply a beep to the output when recording has stopped
engine.on('recording_stopped', () => {
engine.beep({frequency: 300})
})
// Stop recording after 5 seconds
setTimeout(() => {
// Stop the recording
engine.stopRecording()
// Playback of the recording
engine.startPlayback()
}, 5000)
// Let the programm run forever
process.stdin.resume()
`
To list the available devices
`javascript`
const soundengine = require('soundengine')
var devices = soundengine.getDevices()
Property | Type | Description
-------------------------|-----------|-------------------------------------------------------
id | number | The device id to use in the engine options.
name | string | The device name.
defaultSampleRate | number | The default sample rate of the device.
maxInputChannels | number | The maximum supported input channels for this device.
maxOutputChannels | number | The maximum supported output channels for this device.
defaultLowInputLatency | number | See PortAudio docs
defaultLowOutputLatency | number | See PortAudio docs
defaultHighInputLatency | number | See PortAudio docs
defaultHighOutputLatency | number | See PortAudio docs
The engine class actually has almost all the methods of a nodejs EventEmitter to interact with the upcomming events. Furthermore these methods exist:
* loadRecording(file: string) - Loads a wave file that is then playable with startPlayback(). (1)startPlayback()
* - Starts playback of the last recording or loaded file.stopPlayback()
* - Stops playback.pausePlayback()
* - Pauses playback.isPlaying(): boolean
* - Returns if playback is active.startRecording()
* - Starts recording.stopRecording()
* - Stops recording.deleteRecording()
* - Deletes the recording that is currently held in memory.saveRecording(file: string)
* - Saves the recording to a wave file. (2)isRecording(): boolean
* - Returns if recording is active.getRecordingSamples(): number
* - Return the number of total samples.getPlaybackPosition(): number
* - Returns the current sample index of the playback.getRecordingSampleAt(index: number): number
* - Returns a specific sample (between -1..1) at index.getPlaybackProgress(): number
* - Returns the relative playback progress (between 0..1).setPlaybackProgress(progress: number)
* - Sets the relative playback progress (between 0..1).beep(options?: beepOptions)
* - Applies a beep to the output.getVolume(): number
* - Returns the volume (between 0..1).setVolume(volume: number)
* - Sets the volume (between 0..1).getMute(): boolean
* - Returns if the output is muted or not.setMute(mute?: boolean)
* - Mutes or unmutes the output.getOptions(): engineOptions
* - Returns the current engine options.setOptions(options?: engineOptions)
* - Sets the engine options.synchronize()
* - Clears the internal buffer queues. If there for example is a large delay between the input and the output after initializing a new engine, calling synchronize could potentially minimize this delay.
Notes:
(1) Currently only 32bit floating point waves with the same samplerate of the current engine can be loaded (and the header will not be checked).
(2) The wave files are 32bit floating point.
Option | Type | Default | Description
----------------|-----------|-----------------------------|------------
sampleRate | number | 44100 | Samples per second for each channel.
bufferSize | number | 1024 | The count of samples for each processing iteration.
inputChannels | number | 1 | The number of input channels.
outputChannels | number | 1 | The number of output channels (should equal inputChannels).
inputDevice | number | default input device | The id of the input device to use or -1 for a output only stream.
outputDevice | number | default output device | The id of the output device to use or -1 for a input only stream.
inputLatency | number | default high input latency | See PortAudio docs
outputLatency | number | default high output latency | See PortAudio docs
Option | Type | Default | Description
----------------|-----------|-----------------------|------------
duration | number | 200 | The duration of the applied beep in ms.
frequency | number | 700 | The frequency of the beep (sine wave).
level | number | 1.0 | The volume of the beep (between 0..1).
EventName | Signature | Description
---------------------|--------------------------------------|------------
data | (inputBuffer: number[]): number[] | Will be called when a new inputBuffer is available to be processed and returned. Note: If the processing function takes to long to process the buffer you might experience dropouts.loadRecording
info | ({min: number[], max: number[]}) | This event gets fired with messurements of the inputBuffer such as peaks (min) for every channel.
playback_started | | Gets fired when playback started.
playback_stopped | | Gets fired when playback stopped.
playback_paused | | Gets fired when playback paused.
playback_progress | (progress: number) | Gets fired when playback progressed with the relative progress.
playback_finished | | Gets fired when playback reached the end of the recording or loaded wave.
recording_loaded | | Gets fired when the loaded a file successfully.saveRecording
recording_started | | Gets fired when recording started.
recording_stopped | | Gets fired when recording stopped.
recording_progress | | Gets fired when recording progressed.
recording_saved | | Gets fired when the recording was saved with .beep
recording_deleted | | Gets fired when the recording in memory was deleted.
beep_started | | Gets fired when the method started apply a beep to the output.beep` method stopped apply a beep to the output.
beep_stopped | | Gets fired when the
* Implement fft stuff