BaseAudioContext.createPeriodicWave()

The createPeriodicWave() method of the BaseAudioContext Interface is used to create a PeriodicWave, which is used to define a periodic waveform that can be used to shape the output of an OscillatorNode.

A PeriodicWave.

real

An array of cosine terms (traditionally the A terms).

imag

An array of sine terms (traditionally the B terms).

The real and imag arrays have to have the same length, otherwise an error is thrown.

constraints Optional

An dictionary object that specifies whether normalization should be disabled (if not specified, normalization is enabled by default.) It takes one property:

• disableNormalization: If set to true, normalization is disabled for the periodic wave. The default is false.

The following example illustrates simple usage of createPeriodicWave(), to create a PeriodicWave object containing a simple sine wave.

var real = new Float32Array(2);
var imag = new Float32Array(2);
var ac = new AudioContext();
var osc = ac.createOscillator();

real[0] = 0;
imag[0] = 0;
real[1] = 1;
imag[1] = 0;

var wave = ac.createPeriodicWave(real, imag, {disableNormalization: true});

osc.setPeriodicWave(wave);

osc.connect(ac.destination);

osc.start();
osc.stop(2);

This works because a sound that contains only a fundamental tone is by definition a sine wave

Here, we create a PeriodicWave with two values. The first value is the DC offset, which is the value at which the oscillator starts. 0 is good here, because we want to start the curve at the middle of the [-1.0; 1.0] range.

The second and subsequent values are sine and cosine components. You can think of it as the result of a Fourier transform, where you get frequency domain values from time domain value. Here, with createPeriodicWave(), you specify the frequencies, and the browser performs an inverse Fourier transform to get a time domain buffer for the frequency of the oscillator. Here, we only set one component at full volume (1.0) on the fundamental tone, so we get a sine wave.

The coefficients of the Fourier transform should be given in ascending order (i.e. $\left(a+bi\right)e^{i} , \left(c+di\right)e^{2i} , \left(f+gi\right)e^{3i}$etc.) and can be positive or negative.  A simple way of manually obtaining such coefficients (though not the best) is to use a graphing calculator.

• Using the Web Audio API