This specification describes a high-level JavaScript API for processing and synthesizing audio in web applications. The primary paradigm is of an audio routing graph, where a number of AudioNode objects are connected together to define the overall audio rendering. The actual processing will primarily take place in the underlying implementation (typically optimized Assembly / C / C++ code), but direct JavaScript processing and synthesis is also supported.

The introductory section covers the motivation behind this specification.

This API is designed to be used in conjunction with other APIs and elements on the web platform, notably: XMLHttpRequest [[XHR]] (using the responseType and response attributes). For games and interactive applications, it is anticipated to be used with the canvas 2D [[2dcontext]] and WebGL [[WEBGL]] 3D graphics APIs.

Introduction

Audio on the web has been fairly primitive up to this point and until very recently has had to be delivered through plugins such as Flash and QuickTime. The introduction of the audio element in HTML5 is very important, allowing for basic streaming audio playback. But, it is not powerful enough to handle more complex audio applications. For sophisticated web-based games or interactive applications, another solution is required. It is a goal of this specification to include the capabilities found in modern game audio engines as well as some of the mixing, processing, and filtering tasks that are found in modern desktop audio production applications.

The APIs have been designed with a wide variety of use cases in mind. Ideally, it should be able to support any use case which could reasonably be implemented with an optimized C++ engine controlled via JavaScript and run in a browser. That said, modern desktop audio software can have very advanced capabilities, some of which would be difficult or impossible to build with this system. Apple's Logic Audio is one such application which has support for external MIDI controllers, arbitrary plugin audio effects and synthesizers, highly optimized direct-to-disk audio file reading/writing, tightly integrated time-stretching, and so on. Nevertheless, the proposed system will be quite capable of supporting a large range of reasonably complex games and interactive applications, including musical ones. And it can be a very good complement to the more advanced graphics features offered by WebGL. The API has been designed so that more advanced capabilities can be added at a later time.

Features

The API supports these primary features:

Modular Routing

Modular routing allows arbitrary connections between different AudioNode objects. Each node can have inputs and/or outputs. A source node has no inputs and a single output. A destination node has one input and no outputs, the most common example being AudioDestinationNode the final destination to the audio hardware. Other nodes such as filters can be placed between the source and destination nodes. The developer doesn't have to worry about low-level stream format details when two objects are connected together; the right thing just happens. For example, if a mono audio stream is connected to a stereo input it should just mix to left and right channels appropriately.

In the simplest case, a single source can be routed directly to the output. All routing occurs within an AudioContext containing a single AudioDestinationNode:

modular routing
A simple example of modular routing.

Illustrating this simple routing, here's a simple example playing a single sound:


var context = new AudioContext();

function playSound() {
    var source = context.createBufferSource();
    source.buffer = dogBarkingBuffer;
    source.connect(context.destination);
    source.start(0);
}

Here's a more complex example with three sources and a convolution reverb send with a dynamics compressor at the final output stage:

modular routing2
A more complex example of modular routing.

var context = 0;
var compressor = 0;
var reverb = 0;

var source1 = 0;
var source2 = 0;
var source3 = 0;

var lowpassFilter = 0;
var waveShaper = 0;
var panner = 0;

var dry1 = 0;
var dry2 = 0;
var dry3 = 0;

var wet1 = 0;
var wet2 = 0;
var wet3 = 0;

var masterDry = 0;
var masterWet = 0;

function setupRoutingGraph () {
    context = new AudioContext();

    // Create the effects nodes.
    lowpassFilter = context.createBiquadFilter();
    waveShaper = context.createWaveShaper();
    panner = context.createPanner();
    compressor = context.createDynamicsCompressor();
    reverb = context.createConvolver();

    // Create master wet and dry.
    masterDry = context.createGain();
    masterWet = context.createGain();

    // Connect final compressor to final destination.
    compressor.connect(context.destination);

    // Connect master dry and wet to compressor.
    masterDry.connect(compressor);
    masterWet.connect(compressor);

    // Connect reverb to master wet.
    reverb.connect(masterWet);

    // Create a few sources.
    source1 = context.createBufferSource();
    source2 = context.createBufferSource();
    source3 = context.createOscillator();

    source1.buffer = manTalkingBuffer;
    source2.buffer = footstepsBuffer;
    source3.frequency.value = 440;

    // Connect source1
    dry1 = context.createGain();
    wet1 = context.createGain();
    source1.connect(lowpassFilter);
    lowpassFilter.connect(dry1);
    lowpassFilter.connect(wet1);
    dry1.connect(masterDry);
    wet1.connect(reverb);

    // Connect source2
    dry2 = context.createGain();
    wet2 = context.createGain();
    source2.connect(waveShaper);
    waveShaper.connect(dry2);
    waveShaper.connect(wet2);
    dry2.connect(masterDry);
    wet2.connect(reverb);

    // Connect source3
    dry3 = context.createGain();
    wet3 = context.createGain();
    source3.connect(panner);
    panner.connect(dry3);
    panner.connect(wet3);
    dry3.connect(masterDry);
    wet3.connect(reverb);

    // Start the sources now.
    source1.start(0);
    source2.start(0);
    source3.start(0);
}

API Overview

The interfaces defined are:

The following conformance classes are defined by this specification:

conforming implementation

A user agent is considered to be a conforming implementation if it satisfies all of the MUST-, REQUIRED- and SHALL-level criteria in this specification that apply to implementations.

User agents that use ECMAScript to implement the APIs defined in this specification must implement them in a manner consistent with the ECMAScript Bindings defined in the Web IDL specification [[!WEBIDL]] as this specification uses that specification and terminology.

The Audio API

The AudioContext Interface

This interface represents a set of AudioNode objects and their connections. It allows for arbitrary routing of signals to the AudioDestinationNode (what the user ultimately hears). Nodes are created from the context and are then connected together. In most use cases, only a single AudioContext is used per document.

suspended
The AudioContext is currently suspended (context time is not proceeding, audio hardware may be powered down/released).
running
Audio is being processed.
closed
The AudioContext has been released, and can no longer be used to process audio. All system audio resources have been released. Attempts to create new Nodes on the AudioContext will throw InvalidStateError. (AudioBuffers may still be created, through createBuffer or decodeAudioData.)
readonly attribute AudioDestinationNode destination

An AudioDestinationNode with a single input representing the final destination for all audio. Usually this will represent the actual audio hardware. All AudioNodes actively rendering audio will directly or indirectly connect to destination.

readonly attribute float sampleRate

The sample rate (in sample-frames per second) at which the AudioContext handles audio. It is assumed that all AudioNodes in the context run at this rate. In making this assumption, sample-rate converters or "varispeed" processors are not supported in real-time processing.

readonly attribute double currentTime

This is a time in seconds which starts at zero when the context is created and increases in real-time. All scheduled times are relative to it. This is not a "transport" time which can be started, paused, and re-positioned. It is always moving forward. A GarageBand-like timeline transport system can be very easily built on top of this (in JavaScript). This time corresponds to an ever-increasing hardware timestamp.

readonly attribute AudioListener listener

An AudioListener which is used for 3D spatialization.

readonly attribute AudioContextState state
Describes the current state of the AudioContext. The context state MUST begin in "suspended", and transitions to "running" when system resources are acquired and audio has begun processing. For OfflineAudioContexts, the state will remain in "suspended" until startRendering() is called, at which point it will transition to "running", and then to "closed" once audio processing has completed and oncomplete has been fired.

When the state is "suspended", a call to resume() will cause a transition to "running", or a call to close() will cause a transition to "closed".

When the state is "running", a call to suspend() will cause a transition to "suspended", or a call to close() will cause a transition to "closed".

When the state is "closed", no further state transitions are possible.

Promise suspend()
Suspends the progression of time in the audio context, allows any current context processing blocks that are already processed to be played to the destination, and then allows the system to release its claim on audio hardware. This is generally useful when the application knows it will not need the AudioContext for some time, and wishes to let the audio hardware power down. The promise resolves when the frame buffer is empty (has been handed off to the hardware), or immediately (with no other effect) if the context is already suspended. The promise is rejected if the context has been closed. This method will cause an INVALID_STATE_ERR exception to be thrown if called on an OfflineAudioContext.

While the system is suspended, MediaStreams will have their output ignored; that is, data will be lost by the real time nature of media streams. HTMLMediaElements will similarly have their output ignored until the system is resumed. Audio Workers and ScriptProcessorNodes will simply not fire their onaudioprocess events while suspended, but will resume when resumed. For the purpose of AnalyserNode window functions, the data is considered as a continuous stream - i.e. the resume()/suspend() does not cause silence to appear in the AnalyserNode's stream of data.

Promise resume()
Resumes the progression of time in an audio context that has been suspended, which may involve re-priming the frame buffer contents. The promise resolves when the system has re-acquired (if necessary) access to audio hardware and has begun streaming to the destination, or immediately (with no other effect) if the context is already running. The promise is rejected if the context has been closed. If the context is not currently suspended, the promise will resolve. This method will cause an INVALID_STATE_ERR exception to be thrown if called on an OfflineAudioContext.
Promise close()
Closes the audio context, releasing any system audio resources used by the AudioContext. This will not automatically release all AudioContext-created objects, unless other references have been released as well; however, it will forcibly release any system audio resources that might prevent additional AudioContexts from being created and used, suspend the progression of audio time in the audio context, and stop processing audio data. The promise resolves when all AudioContext-creation-blocking resources have been released. This method will cause an INVALID_STATE_ERR exception to be thrown if called on an OfflineAudioContext.
attribute EventHandler onstatechange
A property used to set the EventHandler for an event that is dispatched to AudioContext when the state of the AudioContext has changed (i.e. when the corresponding promise would have resolved). An event of type Event will be dispatched to the event handler, which can query the AudioContext's state directly. A newly-created AudioContext will always begin in the "suspended" state, and a state change event will be fired whenever the state changes to a different state.
AudioBuffer createBuffer()
Creates an AudioBuffer of the given size. The audio data in the buffer will be zero-initialized (silent). An NotSupportedError exception MUST be thrown if any of the arguments is negative, zero, or outside its nominal range.
unsigned long numberOfChannels
Determines how many channels the buffer will have. An implementation must support at least 32 channels.
unsigned long length
Determines the size of the buffer in sample-frames.
float sampleRate
Describes the sample-rate of the linear PCM audio data in the buffer in sample-frames per second. An implementation must support sample-rates in at least the range 22050 to 96000.
Promise<AudioBuffer> decodeAudioData()
Asynchronously decodes the audio file data contained in the ArrayBuffer. The ArrayBuffer can, for example, be loaded from an XMLHttpRequest's response attribute after setting the responseType to "arraybuffer". Audio file data can be in any of the formats supported by the audio element.
ArrayBuffer audioData
An ArrayBuffer containing compressed audio data
optional DecodeSuccessCallback successCallback
A callback function which will be invoked when the decoding is finished. The single argument to this callback is an AudioBuffer representing the decoded PCM audio data.
optional DecodeErrorCallback errorCallback
A callback function which will be invoked if there is an error decoding the audio file.

Although the primary method of interfacing with this function is via its promise return value, the callback parameters are provided for legacy reasons.

The following steps must be performed:

  1. Let promise be a new promise.
  2. If audioData is null or not a valid ArrayBuffer:
    1. Let error be a DOMException whose name is NotSupportedError.
    2. Reject promise with error.
    3. If errorCallback is not missing, invoke errorCallback with error.
    4. Terminate this algorithm.
  3. Neuter the audioData ArrayBuffer in such a way that JavaScript code may not access or modify the data anymore.
  4. Queue a decoding operation to be performed on another thread.
  5. Return promise.
  6. In the decoding thread:
    1. Attempt to decode the encoded audioData into linear PCM.
    2. If a decoding error is encountered due to the audio format not being recognized or supported, or because of corrupted/unexpected/inconsistent data, then, on the main thread's event loop:
      1. Let error be a DOMException whose name is "EncodingError".
      2. Reject promise with error.
      3. If errorCallback is not missing, invoke errorCallback with error.
    3. Otherwise:
      1. Take the result, representing the decoded linear PCM audio data, and resample it to the sample-rate of the AudioContext if it is different from the sample-rate of audioData.
      2. On the main thread's event loop:
        1. Let buffer be an AudioBuffer containing the final result (after possibly sample-rate converting).
        2. Resolve promise with buffer.
        3. If successCallback is not missing, invoke successCallback with buffer.
AudioBufferSourceNode createBufferSource()
Creates an AudioBufferSourceNode.
MediaElementAudioSourceNode createMediaElementSource()
Creates a MediaElementAudioSourceNode given an HTMLMediaElement. As a consequence of calling this method, audio playback from the HTMLMediaElement will be re-routed into the processing graph of the AudioContext.
HTMLMediaElement mediaElement
The media element that will be re-routed.
MediaStreamAudioSourceNode createMediaStreamSource()
MediaStream mediaStream
The media stream that will act as source.
MediaStreamAudioDestinationNode createMediaStreamDestination()
Creates a MediaStreamAudioDestinationNode
AudioWorker createAudioWorker()
The createAudioWorker method creates an AudioWorker and its associated AudioWorkerGlobalScope for direct audio processing using JavaScript. An AudioWorker acts as a factory for AudioWorkerNode instances, allowing the nodes to be created synchronously on demand. Because script loading must be performed asynchronously, newly created nodes may not be able to process audio immediately on creation. However once the AudioWorker's script has been loaded and run, which can be detected via the event callback, nodes can be created with the assurance of no script-induced delays.
DOMString scriptURL
This parameter represents the URL of the script to be loaded as an AudioWorker.
ScriptProcessorNode createScriptProcessor()
This method is DEPRECATED, as it is intended to be replaced by createAudioWorker. Creates a ScriptProcessorNode for direct audio processing using JavaScript. An IndexSizeError exception MUST be thrown if bufferSize or numberOfInputChannels or numberOfOutputChannels are outside the valid range.
optional unsigned long bufferSize = 0
The bufferSize parameter determines the buffer size in units of sample-frames. If it's not passed in, or if the value is 0, then the implementation will choose the best buffer size for the given environment, which will be constant power of 2 throughout the lifetime of the node. Otherwise if the author explicitly specifies the bufferSize, it must be one of the following values: 256, 512, 1024, 2048, 4096, 8192, 16384. This value controls how frequently the audioprocess event is dispatched and how many sample-frames need to be processed each call. Lower values for bufferSize will result in a lower (better) latency. Higher values will be necessary to avoid audio breakup and glitches. It is recommended for authors to not specify this buffer size and allow the implementation to pick a good buffer size to balance between latency and audio quality.
optional unsigned long numberOfInputChannels = 2
This parameter determines the number of channels for this node's input. Values of up to 32 must be supported.
optional unsigned long numberOfOutputChannels = 2
This parameter determines the number of channels for this node's output. Values of up to 32 must be supported.
It is invalid for both numberOfInputChannels and numberOfOutputChannels to be zero.
AnalyserNode createAnalyser()
Create an AnalyserNode.
GainNode createGain()
Create an GainNode.
DelayNode createDelay()
Creates a DelayNode representing a variable delay line. The initial default delay time will be 0 seconds.
optional double maxDelayTime = 1.0
The maxDelayTime parameter is optional and specifies the maximum delay time in seconds allowed for the delay line. If specified, this value MUST be greater than zero and less than three minutes or a NotSupportedError exception MUST be thrown.
BiquadFilterNode createBiquadFilter()
Creates a BiquadFilterNode representing a second order filter which can be configured as one of several common filter types.
WaveShaperNode createWaveShaper()
Creates a WaveShaperNode representing a non-linear distortion.
PannerNode createPanner()
Creates a PannerNode.
StereoPannerNode createStereoPanner()
Creates a StereoPannerNode.
ConvolverNode createConvolver()
Creates a ConvolverNode.
ChannelSplitterNode createChannelSplitter()
Creates an ChannelSplitterNode representing a channel splitter. An IndexSizeError exception MUST be thrown for invalid parameter values.
optional unsigned long numberOfOutputs = 6
The number of outputs. Values of up to 32 must be supported. If not specified, then 6 will be used.
ChannelMergerNode createChannelMerger()
Creates a ChannelMergerNode representing a channel merger. An IndexSizeError exception MUST be thrown for invalid parameter values.
optional unsigned long numberOfInputs = 6
The numberOfInputs parameter determines the number of inputs. Values of up to 32 must be supported. If not specified, then 6 will be used.
DynamicsCompressorNode createDynamicsCompressor()
Creates a DynamicsCompressorNode
OscillatorNode createOscillator()
Creates an OscillatorNode
PeriodicWave createPeriodicWave()
Creates a PeriodicWave representing a waveform containing arbitrary harmonic content. The real and imag parameters must be of type Float32Array (described in [[!TYPED-ARRAYS]]) of equal lengths greater than zero and less than or equal to 4096 or an IndexSizeError exception MUST be thrown. These parameters specify the Fourier coefficients of a Fourier series representing the partials of a periodic waveform. The created PeriodicWave will be used with an OscillatorNode and will represent a normalized time-domain waveform having maximum absolute peak value of 1. Another way of saying this is that the generated waveform of an OscillatorNode will have maximum peak value at 0dBFS. Conveniently, this corresponds to the full-range of the signal values used by the Web Audio API. Because the PeriodicWave will be normalized on creation, the real and imag parameters represent relative values.

As PeriodicWave objects maintain their own representation, any modification of the arrays uses as the real and imag parameters after the call to createPeriodicWave() will have no effect on the PeriodicWave object.

Float32Array real
The real parameter represents an array of cosine terms (traditionally the A terms). In audio terminology, the first element (index 0) is the DC-offset of the periodic waveform and is usually set to zero. The second element (index 1) represents the fundamental frequency. The third element represents the first overtone, and so on.
Float32Array imag
The imag parameter represents an array of sine terms (traditionally the B terms). The first element (index 0) should be set to zero (and will be ignored) since this term does not exist in the Fourier series. The second element (index 1) represents the fundamental frequency. The third element represents the first overtone, and so on.
AudioBuffer decodedData
The AudioBuffer containing the decoded audio data.
DOMException error
The error that occurred while decoding.

Lifetime

Once created, an AudioContext will continue to play sound until it has no more sound to play, or the page goes away.

Lack of introspection or serialization primitives

The Web Audio API takes a fire-and-forget approach to audio source scheduling. That is, source nodes are created for each note during the lifetime of the AudioContext, and never explicitely removed from the graph. This is incompatible with a serialization API, since there is no stable set of nodes that could be serialized.

Moreover, having an introspection API would allow content script to be able to observe garbage collections.

The OfflineAudioContext Interface

OfflineAudioContext is a particular type of AudioContext for rendering/mixing-down (potentially) faster than real-time. It does not render to the audio hardware, but instead renders as quickly as possible, fulfilling the returned promise with the rendered result as an AudioBuffer.

Each OfflineAudioContext instance has an associated rendering started flag that is initially false.

Promise<AudioBuffer> startRendering()

Given the current connections and scheduled changes, starts rendering audio.

Although the primary method of getting the rendered audio data is via its promise return value, the instance will also fire an event named complete for legacy reasons.

The following steps must be performed:

  1. If this instance's rendering started flag is true, return a promise rejected with a DOMException whose name is "InvalidStateError".
  2. Set this instance's rendering started flag to true.
  3. Let promise be a new promise.
  4. Asynchronously perform the following steps:
    1. Let buffer be a new AudioBuffer, with a number of channels, length and sample rate equal respectively to the numberOfChannels, length and sampleRate parameters used when this instance's constructor was called.
    2. Given the current connections and scheduled changes, start rendering length sample-frames of audio into buffer.
    3. Once the rendering is complete,
      1. Resolve promise with buffer.
      2. Fire a simple event named complete at this instance, using an instance of OfflineAudioCompletionEvent whose renderedBuffer property is set to buffer.
  5. Return promise.
attribute EventHandler oncomplete

An EventHandler of type OfflineAudioCompletionEvent.

The OfflineAudioCompletionEvent Interface

This is an Event object which is dispatched to OfflineAudioContext for legacy reasons.

readonly attribute AudioBuffer renderedBuffer

An AudioBuffer containing the rendered audio data.

The AudioNode Interface

AudioNodes are the building blocks of an AudioContext. This interface represents audio sources, the audio destination, and intermediate processing modules. These modules can be connected together to form processing graphs for rendering audio to the audio hardware. Each node can have inputs and/or outputs. A source node has no inputs and a single output. An AudioDestinationNode has one input and no outputs and represents the final destination to the audio hardware. Most processing nodes such as filters will have one input and one output. Each type of AudioNode differs in the details of how it processes or synthesizes audio. But, in general, an AudioNode will process its inputs (if it has any), and generate audio for its outputs (if it has any).

Each output has one or more channels. The exact number of channels depends on the details of the specific AudioNode.

An output may connect to one or more AudioNode inputs, thus fan-out is supported. An input initially has no connections, but may be connected from one or more AudioNode outputs, thus fan-in is supported. When the connect() method is called to connect an output of an AudioNode to an input of an AudioNode, we call that a connection to the input.

Each AudioNode input has a specific number of channels at any given time. This number can change depending on the connection(s) made to the input. If the input has no connections then it has one channel which is silent.

For each input, an AudioNode performs a mixing (usually an up-mixing) of all connections to that input. Please see for more informative details, and the section for normative requirements.

For performance reasons, practical implementations will need to use block processing, with each AudioNode processing a fixed number of sample-frames of size block-size. In order to get uniform behavior across implementations, we will define this value explicitly. block-size is defined to be 128 sample-frames which corresponds to roughly 3ms at a sample-rate of 44.1KHz.

AudioNodes are EventTargets, as described in DOM [[!DOM]]. This means that it is possible to dispatch events to AudioNodes the same way that other EventTargets accept events.

max
computedNumberOfChannels is computed as the maximum of the number of channels of all connections. In this mode channelCount is ignored
clamped-max
Same as “max” up to a limit of the channelCount
explicit
computedNumberOfChannels is the exact value as specified in channelCount
speakers
use up-down-mix equations for mono/stereo/quad/5.1. In cases where the number of channels do not match any of these basic speaker layouts, revert to "discrete".
discrete
Up-mix by filling channels until they run out then zero out remaining channels. down-mix by filling as many channels as possible, then dropping remaining channels.
void connect()
AudioNode destination
The destination parameter is the AudioNode to connect to.
optional unsigned long output = 0
The output parameter is an index describing which output of the AudioNode from which to connect. If this paremeter is out-of-bound, an IndexSizeError exception MUST be thrown. It is possible to connect an AudioNode output to more than one input with multiple calls to connect(). Thus, "fan-out" is supported.
optional unsigned long input = 0
The input parameter is an index describing which input of the destination AudioNode to connect to. If this parameter is out-of-bounds, an IndexSizeError exception MUST be thrown. It is possible to connect an AudioNode to another AudioNode which creates a cycle: an AudioNode may connect to another AudioNode, which in turn connects back to the first AudioNode. This is allowed only if there is at least one DelayNode in the cycle or a NotSupportedError exception MUST be thrown.

There can only be one connection between a given output of one specific node and a given input of another specific node. Multiple connections with the same termini are ignored. For example:

    nodeA.connect(nodeB);
    nodeA.connect(nodeB);
    

will have the same effect as

      nodeA.connect(nodeB);
    
void connect()
Connects the AudioNode to an AudioParam, controlling the parameter value with an audio-rate signal.
AudioParam destination
The destination parameter is the AudioParam to connect to.
optional unsigned long output = 0
The output parameter is an index describing which output of the AudioNode from which to connect. If the parameter is out-of-bound, an IndexSizeError exception MUST be thrown.

It is possible to connect an AudioNode output to more than one AudioParam with multiple calls to connect(). Thus, "fan-out" is supported.

It is possible to connect more than one AudioNode output to a single AudioParam with multiple calls to connect(). Thus, "fan-in" is supported.

An AudioParam will take the rendered audio data from any AudioNode output connected to it and convert it to mono by down-mixing if it is not already mono, then mix it together with other such outputs and finally will mix with the intrinsic parameter value (the value the AudioParam would normally have without any audio connections), including any timeline changes scheduled for the parameter.

There can only be one connection between a given output of one specific node and a specific AudioParam. Multiple connections with the same termini are ignored. For example:

      nodeA.connect(param);
      nodeA.connect(param);
    
will have the same effect as
      nodeA.connect(param);
    
void disconnect()
optional unsigned long output = 0
This parameter is an index describing which output of the AudioNode to disconnect. If this parameter is out-of-bounds, an IndexSizeError exception MUST be thrown.
readonly attribute AudioContext context
The AudioContext which owns this AudioNode.
readonly attribute unsigned long numberOfInputs
The number of inputs feeding into the AudioNode. For source nodes, this will be 0.
readonly attribute unsigned long numberOfOutputs
The number of outputs coming out of the AudioNode. This will be 0 for an AudioDestinationNode.
attribute unsigned long channelCount

The number of channels used when up-mixing and down-mixing connections to any inputs to the node. The default value is 2 except for specific nodes where its value is specially determined. This attribute has no effect for nodes with no inputs. If this value is set to zero, the implementation MUST throw a NotSupportedError exception.

See the section for more information on this attribute.

attribute ChannelCountMode channelCountMode

Determines how channels will be counted when up-mixing and down-mixing connections to any inputs to the node. This attribute has no effect for nodes with no inputs.

See the section for more information on this attribute.

attribute ChannelInterpretation channelInterpretation

Determines how individual channels will be treated when up-mixing and down-mixing connections to any inputs to the node. This attribute has no effect for nodes with no inputs.

See the section for more information on this attribute.

Lifetime

An implementation may choose any method to avoid unnecessary resource usage and unbounded memory growth of unused/finished nodes. The following is a description to help guide the general expectation of how node lifetime would be managed.

An AudioNode will live as long as there are any references to it. There are several types of references:

  1. A normal JavaScript reference obeying normal garbage collection rules.
  2. A playing reference for both AudioBufferSourceNodes and OscillatorNodes. These nodes maintain a playing reference to themselves while they are currently playing.
  3. A connection reference which occurs if another AudioNode is connected to it.
  4. A tail-time reference which an AudioNode maintains on itself as long as it has any internal processing state which has not yet been emitted. For example, a ConvolverNode has a tail which continues to play even after receiving silent input (think about clapping your hands in a large concert hall and continuing to hear the sound reverberate throughout the hall). Some AudioNodes have this property. Please see details for specific nodes.

Any AudioNodes which are connected in a cycle and are directly or indirectly connected to the AudioDestinationNode of the AudioContext will stay alive as long as the AudioContext is alive.

When an AudioNode has no references it will be deleted. Before it is deleted, it will disconnect itself from any other AudioNodes which it is connected to. In this way it releases all connection references (3) it has to other nodes.

Regardless of any of the above references, it can be assumed that the AudioNode will be deleted when its AudioContext is deleted.

The AudioDestinationNode Interface

This is an AudioNode representing the final audio destination and is what the user will ultimately hear. It can often be considered as an audio output device which is connected to speakers. All rendered audio to be heard will be routed to this node, a "terminal" node in the AudioContext's routing graph. There is only a single AudioDestinationNode per AudioContext, provided through the destination attribute of AudioContext.

      numberOfInputs  : 1
      numberOfOutputs : 0

      channelCount = 2;
      channelCountMode = "explicit";
      channelInterpretation = "speakers";
readonly attribute unsigned long maxChannelCount

The maximum number of channels that the channelCount attribute can be set to. An AudioDestinationNode representing the audio hardware end-point (the normal case) can potentially output more than 2 channels of audio if the audio hardware is multi-channel. maxChannelCount is the maximum number of channels that this hardware is capable of supporting. If this value is 0, then this indicates that channelCount may not be changed. This will be the case for an AudioDestinationNode in an OfflineAudioContext and also for basic implementations with hardware support for stereo output only.

channelCount defaults to 2 for a destination in a normal AudioContext, and may be set to any non-zero value less than or equal to maxChannelCount. An IndexSizeError exception MUST be thrown if this value is not within the valid range. Giving a concrete example, if the audio hardware supports 8-channel output, then we may set channelCount to 8, and render 8-channels of output.

For anAudioDestinationNode in an OfflineAudioContext, the channelCount is determined when the offline context is created and this value may not be changed.

The AudioParam Interface

AudioParam controls an individual aspect of an AudioNode's functioning, such as volume. The parameter can be set immediately to a particular value using the value attribute. Or, value changes can be scheduled to happen at very precise times (in the coordinate system of AudioContext's currentTime attribute), for envelopes, volume fades, LFOs, filter sweeps, grain windows, etc. In this way, arbitrary timeline-based automation curves can be set on any AudioParam. Additionally, audio signals from the outputs of AudioNodes can be connected to an AudioParam, summing with the intrinsic parameter value.

Some synthesis and processing AudioNodes have AudioParams as attributes whose values must be taken into account on a per-audio-sample basis. For other AudioParams, sample-accuracy is not important and the value changes can be sampled more coarsely. Each individual AudioParam will specify that it is either an a-rate parameter which means that its values must be taken into account on a per-audio-sample basis, or it is a k-rate parameter.

Implementations must use block processing, with each AudioNode processing 128 sample-frames in each block.

For each 128 sample-frame block, the value of a k-rate parameter must be sampled at the time of the very first sample-frame, and that value must be used for the entire block. a-rate parameters must be sampled for each sample-frame of the block.

An AudioParam maintains a time-ordered event list which is initially empty. The times are in the time coordinate system of the AudioContext's currentTime attribute. The events define a mapping from time to value. The following methods can change the event list by adding a new event into the list of a type specific to the method. Each event has a time associated with it, and the events will always be kept in time-order in the list. These methods will be called automation methods:

The following rules will apply when calling these methods:

attribute float value

The parameter's floating-point value. This attribute is initialized to the defaultValue. If value is set during a time when there are any automation events scheduled then it will be ignored and no exception will be thrown.

readonly attribute float defaultValue
Initial value for the value attribute.
void setValueAtTime(float value, double startTime)

Schedules a parameter value change at the given time.

The value parameter is the value the parameter will change to at the given time.

The startTime parameter is the time in the same time coordinate system as the AudioContext's currentTime attribute. An InvalidAccessError exception MUST be thrown if startTime is negative or is not a finite number.

If there are no more events after this SetValue event, then for t >= startTime, v(t) = value. In other words, the value will remain constant.

If the next event (having time T1) after this SetValue event is not of type LinearRampToValue or ExponentialRampToValue, then, for t:

      startTime <= t < T1,  v(t) = value
    

In other words, the value will remain constant during this time interval, allowing the creation of "step" functions.

If the next event after this SetValue event is of type LinearRampToValue or ExponentialRampToValue then please see details below.

void linearRampToValueAtTime(float value, double endTime)

Schedules a linear continuous change in parameter value from the previous scheduled parameter value to the given value.

The value parameter is the value the parameter will linearly ramp to at the given time.

The endTime parameter is the time in the same time coordinate system as the AudioContext's currentTime attribute. An InvalidAccessError exception MUST be thrown if endTime is negative or is not a finite number.

The value during the time interval T0 <= t < T1 (where T0 is the time of the previous event and T1 is the endTime parameter passed into this method) will be calculated as:

      v(t) = V0 + (V1 - V0) * ((t - T0) / (T1 - T0))
    

Where V0 is the value at the time T0 and V1 is the value parameter passed into this method.

If there are no more events after this LinearRampToValue event then for t >= T1, v(t) = V1.

void exponentialRampToValueAtTime(float value, double endTime)

Schedules an exponential continuous change in parameter value from the previous scheduled parameter value to the given value. Parameters representing filter frequencies and playback rate are best changed exponentially because of the way humans perceive sound.

The value parameter is the value the parameter will exponentially ramp to at the given time. A NotSupportedError exception MUST be thrown if this value is less than or equal to 0, or if the value at the time of the previous event is less than or equal to 0.

The endTime parameter is the time in the same time coordinate system as the AudioContext's currentTime attribute. An InvalidAccessError exception MUST be thrown if endTime is negative or is not a finite number.

The value during the time interval T0 <= t < T1 (where T0 is the time of the previous event and T1 is the endTime parameter passed into this method) will be calculated as:

        v(t) = V0 * (V1 / V0) ^ ((t - T0) / (T1 - T0))
    

Where V0 is the value at the time T0 and V1 is the value parameter passed into this method.

If there are no more events after this ExponentialRampToValue event then for t >= T1, v(t) = V1

void setTargetAtTime(float target, double startTime, float timeConstant)

Start exponentially approaching the target value at the given time with a rate having the given time constant. Among other uses, this is useful for implementing the "decay" and "release" portions of an ADSR envelope. Please note that the parameter value does not immediately change to the target value at the given time, but instead gradually changes to the target value.

The target parameter is the value the parameter will start changing to at the given time.

The startTime parameter is the time in the same time coordinate system as the AudioContext's currentTime attribute. An InvalidAccessError exception MUST be thrown if start is negative or is not a finite number.

The timeConstant parameter is the time-constant value of first-order filter (exponential) approach to the target value. The larger this value is, the slower the transition will be.

More precisely, timeConstant is the time it takes a first-order linear continuous time-invariant system to reach the value 1 - 1/e (around 63.2%) given a step input response (transition from 0 to 1 value).

During the time interval: T0 <= t < T1, where T0 is the startTime parameter and T1 represents the time of the event following this event (or infinity if there are no following events):

    v(t) = V1 + (V0 - V1) * exp(-(t - T0) / timeConstant)
    

Where V0 is the initial value (the .value attribute) at T0 (the startTime parameter) and V1 is equal to the target parameter.

void setValueCurveAtTime(Float32Array values, double startTime, double duration)

Sets an array of arbitrary parameter values starting at the given time for the given duration. The number of values will be scaled to fit into the desired duration.

The values parameter is a Float32Array representing a parameter value curve. These values will apply starting at the given time and lasting for the given duration. Any modification to the the array used as values argument after the call won't have any effect on the AudioParam.

The startTime parameter is the time in the same time coordinate system as the AudioContext's currentTime attribute. An InvalidAccessError exception MUST be thrown if startTime is negative or is not a finite number.

The duration parameter is the amount of time in seconds (after the time parameter) where values will be calculated according to the values parameter.

During the time interval: startTime <= t < startTime + duration, values will be calculated:

        v(t) = values[N * (t - startTime) / duration]
      

where N is the length of the values array.

After the end of the curve time interval (t >= startTime + duration), the value will remain constant at the final curve value, until there is another automation event (if any).

void cancelScheduledValues(double startTime)

Cancels all scheduled parameter changes with times greater than or equal to startTime.

The startTime parameter is the starting time at and after which any previously scheduled parameter changes will be cancelled. It is a time in the same time coordinate system as the AudioContext's currentTime attribute. An InvalidAccessError exception MUST be thrown if startTime is negative or is not a finite number.

Computation of Value

computedValue is the final value controlling the audio DSP and is computed by the audio rendering thread during each rendering time quantum. It must be internally computed as follows:

  1. An intrinsic parameter value will be calculated at each time, which is either the value set directly to the value attribute, or, if there are any scheduled parameter changes (automation events) with times before or at this time, the value as calculated from these events. If the value attribute is set after any automation events have been scheduled, then these events will be removed. When read, the value attribute always returns the intrinsic value for the current time. If automation events are removed from a given time range, then the intrinsic value will remain unchanged and stay at its previous value until either the value attribute is directly set, or automation events are added for the time range.
  2. An AudioParam will take the rendered audio data from any AudioNode output connected to it and convert it to mono by down-mixing if it is not already mono, then mix it together with other such outputs. If there are no AudioNodes connected to it, then this value is 0, having no effect on the computedValue.
  3. The computedValue is the sum of the intrinsic value and the value calculated from (2).

AudioParam Automation Example

AudioParam automation
An example of parameter automation.

var t0 = 0;
var t1 = 0.1;
var t2 = 0.2;
var t3 = 0.3;
var t4 = 0.4;
var t5 = 0.6;
var t6 = 0.7;
var t7 = 1.0;

var curveLength = 44100;
var curve = new Float32Array(curveLength);
for (var i = 0; i < curveLength; ++i)
    curve[i] = Math.sin(Math.PI * i / curveLength);

param.setValueAtTime(0.2, t0);
param.setValueAtTime(0.3, t1);
param.setValueAtTime(0.4, t2);
param.linearRampToValueAtTime(1, t3);
param.linearRampToValueAtTime(0.15, t4);
param.exponentialRampToValueAtTime(0.75, t5);
param.exponentialRampToValueAtTime(0.05, t6);
param.setValueCurveAtTime(curve, t6, t7 - t6);

The GainNode Interface

Changing the gain of an audio signal is a fundamental operation in audio applications. The GainNode is one of the building blocks for creating mixers. This interface is an AudioNode with a single input and single output:

  numberOfInputs  : 1
  numberOfOutputs : 1

  channelCountMode = "max";
  channelInterpretation = "speakers";

Each sample of each channel of the input data of the GainNode MUST be multiplied by the computedValue of the gain AudioParam.

The implementation must make gain changes to the audio stream smoothly, without introducing noticeable clicks or glitches. This process is called "de-zippering".

readonly attribute AudioParam gain
Represents the amount of gain to apply. Its default value is 1 (no gain change). The nominal minValue is 0, but may be set negative for phase inversion. The nominal maxValue is 1, but higher values are allowed (no exception thrown).This parameter is a-rate

The DelayNode Interface

A delay-line is a fundamental building block in audio applications. This interface is an AudioNode with a single input and single output:

    numberOfInputs  : 1
    numberOfOutputs : 1

    channelCountMode = "max";
    channelInterpretation = "speakers";

The number of channels of the output always equals the number of channels of the input.

It delays the incoming audio signal by a certain amount. Specifically, at each time t, input signal input(t), delay time delayTime(t) and output signal output(t), the output will be output(t) = input(t - delayTime(t)). The default delayTime is 0 seconds (no delay). When the delay time is changed, the implementation must make the transition smoothly, without introducing noticeable clicks or glitches to the audio stream.

readonly attribute AudioParam delayTime

An AudioParam object representing the amount of delay (in seconds) to apply. Its default value is 0 (no delay). The minimum value is 0 and the maximum value is determined by the maxDelayTime argument to the AudioContext method createDelay.

If DelayNode is part of a cycle, then the value of the delayTime attribute is clamped to a minimum of 128 frames (one block).

This parameter is a-rate.

The AudioBuffer Interface

This interface represents a memory-resident audio asset (for one-shot sounds and other short audio clips). Its format is non-interleaved IEEE 32-bit linear PCM with a nominal range of -1 -> +1. It can contain one or more channels. Typically, it would be expected that the length of the PCM data would be fairly short (usually somewhat less than a minute). For longer sounds, such as music soundtracks, streaming should be used with the audio element and MediaElementAudioSourceNode.

An AudioBuffer may be used by one or more AudioContexts.

readonly attribute float sampleRate
The sample-rate for the PCM audio data in samples per second.
readonly attribute long length
Length of the PCM audio data in sample-frames.
readonly attribute double duration
Duration of the PCM audio data in seconds.
readonly attribute long numberOfChannels
The number of discrete audio channels.
Float32Array getChannelData()
Returns the Float32Array representing the PCM audio data for the specific channel.
unsigned long channel
This parameter is an index representing the particular channel to get data for. An index value of 0 represents the first channel. This index value MUST be less than numberOfChannels or an IndexSizeError exception MUST be thrown.
void copyFromChannel()
The copyFromChannel method copies the samples from the specified channel of the AudioBuffer to the destination array.
Float32Array destination
The array the channel data will be copied to.
long channelNumber
The index of the channel to copy the data from. If channelNumber is greater or equal than the number of channel of the AudioBuffer, an IndexSizeError MUST be thrown.
optional unsigned long startInChannel = 0
An optional offset to copy the data from. If startInChannel is greater than the length of the AudioBuffer, an IndexSizeError MUST be thrown.
void copyToChannel()
The copyToChannel method copies the samples to the specified channel of the AudioBuffer, from the source array.
Float32Array source
The array the channel data will be copied from.
long channelNumber
The index of the channel to copy the data to. If channelNumber is greater or equal than the number of channel of the AudioBuffer, an IndexSizeError MUST be thrown.
optional unsigned long startInChannel = 0
An optional offset to copy the data to. If startInChannel is greater than the length of the AudioBuffer, an IndexSizeError MUST be thrown.

The methods copyToChannel and copyFromChannel can be used to fill part of an array by passing in a Float32Array that's a view onto the larger array. When reading data from an AudioBuffer's channels, and the data can be processed in chunks, copyFromChannel should be preferred to calling getChannelData and accessing the resulting array, because it may avoid unnecessary memory allocation and copying.

An internal operation acquire the contents of an AudioBuffer is invoked when the contents of an AudioBuffer are needed by some API implementation. This operation returns immutable channel data to the invoker.

When an acquire the content operation occurs on an AudioBuffer, run the following steps:

  1. If any of the AudioBuffer's ArrayBuffer have been neutered, abort these steps, and return a zero-length channel data buffers to the invoker.
  2. Neuter all ArrayBuffers for arrays previously returned by getChannelData on this AudioBuffer.
  3. Retain the underlying data buffers from those ArrayBuffers and return references to them to the invoker.
  4. Attach ArrayBuffers containing copies of the data to the AudioBuffer, to be returned by the next call to getChannelData.
The acquire the contents of an AudioBuffer operation is invoked in the following cases:

The AudioBufferSourceNode Interface

This interface represents an audio source from an in-memory audio asset in an AudioBuffer. It is useful for playing short audio assets which require a high degree of scheduling flexibility (can playback in rhythmically perfect ways). The start() method is used to schedule when sound playback will happen. The playback will stop automatically when the buffer's audio data has been completely played (if the loop attribute is false), or when the stop() method has been called and the specified time has been reached. Please see more details in the start() and stop() description. start() and stop() may not be issued multiple times for a given AudioBufferSourceNode.

  numberOfInputs  : 0
  numberOfOutputs : 1

The number of channels of the output always equals the number of channels of the AudioBuffer assigned to the .buffer attribute, or is one channel of silence if .buffer is NULL.

attribute AudioBuffer? buffer
Represents the audio asset to be played.
readonly attribute AudioParam playbackRate
The speed at which to render the audio stream. Its default value is 1. This parameter is k-rate.
readonly attribute AudioParam detune
An aditional parameter to modulate the speed at which is rendered the audio stream. Its default value is 0. Its nominal range is [-1200; 1200]. This parameter is k-rate.
attribute boolean loop
Indicates if the audio data should play in a loop. The default value is false.
attribute double loopStart
An optional value in seconds where looping should begin if the loop attribute is true. Its default value is 0, and it may usefully be set to any value between 0 and the duration of the buffer.
attribute double loopEnd
An optional value in seconds where looping should end if the loop attribute is true. Its default value is 0, and it may usefully be set to any value between 0 and the duration of the buffer.
void start()
Schedules a sound to playback at an exact time.
optional double when = 0
The when parameter describes at what time (in seconds) the sound should start playing. It is in the same time coordinate system as the AudioContext's currentTime attribute. If 0 is passed in for this value or if the value is less than currentTime, then the sound will start playing immediately. start may only be called one time and must be called before stop is called or an InvalidStateError exception MUST be thrown. An InvalidAccessError exception MUST be thrown if when is negative or is not a finite number.
optional double offset = 0
The offset parameter describes the offset time in the buffer (in seconds) where playback will begin. If 0 is passed in for this value, then playback will start from the beginning of the buffer. An InvalidAccessError exception MUST be thrown if offset is negative or is not a finite number.
optional double duration
The duration parameter describes the duration of the portion (in seconds) to be played. If this parameter is not passed, the duration will be equal to the total duration of the AudioBuffer minus the offset parameter. Thus if neither offset nor duration are specified then the implied duration is the total duration of the AudioBuffer. An InvalidAccessError exception MUST be thrown if duration is negative or is not a finite number.
void stop()
Schedules a sound to stop playback at an exact time.
optional double when = 0
The when parameter describes at what time (in seconds) the sound should stop playing. It is in the same time coordinate system as the AudioContext's currentTime attribute. If 0 is passed in for this value or if the value is less than currentTime, then the sound will stop playing immediately. An InvalidAccessError exception MUST be thrown if when is negative or is not a finite number. If stop is called again after already have been called, the last invocation will be the only one applied; stop times set by previous calls will not be applied, unless the buffer has already stopped prior to any subsequent calls. If the buffer has already stopped, further calls to stop will have no effect. If a stop time is reached prior to the scheduled start time, the sound will not play.
attribute EventHandler onended
A property used to set the EventHandler (described in HTML[[!HTML]]) for the ended event that is dispatched to AudioBufferSourceNode node types. When the playback of the buffer for an AudioBufferSourceNode is finished, an event of type Event (described in HTML [[!HTML]]) will be dispatched to the event handler.

Both playbackRate and detune are k-rate parameters and are used together to determine a computedPlaybackRate value:

  computedPlaybackRate(t) = playbackRate(t) * pow(2, detune(t) / 1200)
The computedPlaybackRate is the effective speed at which the AudioBuffer of this AudioBufferSourceNode MUST be played.

Looping

If the loop attribute is true when start() is called, then playback will continue indefinitely until stop() is called and the stop time is reached. We'll call this "loop" mode. Playback always starts at the point in the buffer indicated by the offset argument of start(), and in loop mode will continue playing until it reaches the actualLoopEnd position in the buffer (or the end of the buffer), at which point it will wrap back around to the actualLoopStart position in the buffer, and continue playing according to this pattern.

In loop mode then the actual loop points are calculated as follows from the loopStart and loopEnd attributes:

if ((loopStart || loopEnd) && loopStart >= 0 && loopEnd > 0 && loopStart < loopEnd) {
    actualLoopStart = loopStart;
    actualLoopEnd = min(loopEnd, buffer.duration);
} else {
    actualLoopStart = 0;
    actualLoopEnd = buffer.duration;
}

Note that the default values for loopStart and loopEnd are both 0, which indicates that looping should occur from the very start to the very end of the buffer.

Please note that as a low-level implementation detail, the AudioBuffer is at a specific sample-rate (usually the same as the AudioContext sample-rate), and that the loop times (in seconds) must be converted to the appropriate sample-frame positions in the buffer according to this sample-rate.

When scheduling the beginning and the end of playback using the start() and stop() methods, the resulting start or stop time MUST be rounded to the nearest sample-frame in the sample rate of the AudioContext. That is, no sub-sample scheduling is possible.

The MediaElementAudioSourceNode Interface

This interface represents an audio source from an audio or video element.

  numberOfInputs  : 0
  numberOfOutputs : 1

The number of channels of the output corresponds to the number of channels of the media referenced by the HTMLMediaElement. Thus, changes to the media element's .src attribute can change the number of channels output by this node. If the .src attribute is not set, then the number of channels output will be one silent channel.

A MediaElementAudioSourceNode is created given an HTMLMediaElement using the AudioContext createMediaElementSource() method.

The number of channels of the single output equals the number of channels of the audio referenced by the HTMLMediaElement passed in as the argument to createMediaElementSource(), or is 1 if the HTMLMediaElement has no audio.

The HTMLMediaElement must behave in an identical fashion after the MediaElementAudioSourceNode has been created, except that the rendered audio will no longer be heard directly, but instead will be heard as a consequence of the MediaElementAudioSourceNode being connected through the routing graph. Thus pausing, seeking, volume, src attribute changes, and other aspects of the HTMLMediaElement must behave as they normally would if not used with a MediaElementAudioSourceNode.

  var mediaElement = document.getElementById('mediaElementID');
  var sourceNode = context.createMediaElementSource(mediaElement);
  sourceNode.connect(filterNode);

Security with MediaElementAudioSourceNode and cross-origin resources

HTMLMediaElement allows the playback of cross-origin resources. Because Web Audio can allows one to inspect the content of the resource (e.g. using a MediaElementAudioSourceNode, and a ScriptProcessorNode to read the samples), information leakage can occur if scripts from one origin inspect the content of a resource from another origin.

To prevent this, a MediaElementAudioSourceNode MUST output silence instead of the normal output of the HTMLMediaElement if it has been created using an HTMLMediaElement for which the execution of the fetch algorithm labeled the resource as CORS-cross-origin.

The AudioWorker

AudioWorker is an interface obtainable from AudioContext's createAudioWorker method. It represents the main thread's view of an associated audio processing worker script, which is hosted inside a corresponding AudioWorkerGlobalScope and runs in the audio processing thread. AudioWorkers act as factories for creating AudioWorkerNodes.

attribute EventHandler onload

An EventHandler of type Event that is notified after the script of thsw AudioWorker has been loaded and fully executed.

readonly attribute AudioContext context
The AudioContext from which this AudioWorker was created.
AudioParam addParameter(DOMString name, optional float defaultValue)

Causes a correspondingly-named read-only AudioParam to be present on all AudioWorkerNodes created by this AudioWorker, and a correspondingly-named read-only Float32Array to be present on the parameters object exposed on the AudioProcessEvent on subsequent audio processing events. The AudioParam may immediately have its scheduling methods called, its .value set, or AudioNodes connected to it.

The name parameter is the name used for the read-only AudioParam added to the AudioWorkerNode, and the name used for the read-only Float32Array that will be present on the parameters object exposed on subsequent AudioProcessEvents.

The defaultValue parameter is the default value for the AudioParam's value attribute, as well as therefore the default value that will appear in the Float32Array in the worker script (if no other parameter changes or connections affect the value).

void removeParameter(DOMString name)
(Propose removing from spec pending understanding of use case, since this is no longer a per-node action.)
void setNodeConfiguration(inputs, outputs)
Determines the input and output configuration and channel count of all nodes created by this AudioWorker. If not called, the configuration defaults to a single input and single output, each with 2 channels (as if setNodeConfiguration() had been called with no arguments).
optional Array inputs = null

This parameter determines the number of inputs for the worker, as well as the default number of channels for that input. The number of inputs cannot be changed once the node is created, although the number of channels in each input can be (see onprocess for more detail). The parameter is an Array whose length determines the number of inputs; each member of this Array will be coerced to an integer representing the default number of channels for the given input. Channel counts of up to 32 must be supported; values outside the supported range should throw NotSupportedError.

A null Array is treated the same as "[2]" - that is, it will default to a single input with two channels.

An empty Array is explicitly supported, and will cause the node to have no inputs - i.e., any attempt to connect() to this node will fail. (This is useful for nodes that function only as sources, such as an oscillator.)

optional Array outputs = null

This parameter determines the number of outputs for the worker, as well as the default number of channels for each output. The number of outputs cannot be changed once the node is created, although the number of channels in each output can be (see onprocess for more detail). The parameter is an Array whose length determines the number of outputs; each member of this Array will be coerced to an integer representing the default number of channels for the given output. Channel counts of up to 32 must be supported; values outside the supported range should throw NotSupportedError.

A null Array is treated the same as "[2]" - that is, it will default to a single output with two channels.

An empty Array is explicitly supported, and will cause the node to have no inputs - i.e., any attempt to connect() to this node will fail. (This is useful for nodes that function only as sources, such as an oscillator.)

It is invalid for both the number of inputs and number of outputs to be zero (i.e. both Arrays are empty). In this case, implementations should throw NotSupportedError.

Examples of usage:

  • setNodeConfiguration() determines a single stereo input with a single stereo output.
  • setNodeConfiguration( [1], [] ) creates a single mono input with no output.
  • setNodeConfiguration( [1], [1] ) also creates a single mono input with a single mono output.
  • setNodeConfiguration( [1], [2] ) creates a single mono input with a stereo output.
  • setNodeConfiguration( [2,1], [2] ) creates two inputs - one stereo, one mono - and a stereo output.
  • setNodeConfiguration( [1,1,1,1,1,1], [6] ) creates six mono inputs with a single 6-channel output. (This is similar to a 6-channel ChannelMergerNode.)
  • setNodeConfiguration( [6], [1,1,1,1,1,1] ) creates a single 6-channel input with six mono outputs. (This is similar to a 6-channel ChannelSplitterNode.)
AudioWorkerNode createAudioWorkerNode()
Creates a new AudioWorkerNode that exhibits the previously parameters and input/output configuration for this AudioWorker. The node is available immediately for use. Following this call, an AudioNodeCreatedEvent is dispatched that creates the audio thread side of the node; this notification is delayed until the worker's script has loaded and run an any onload events have been dispatched. The node does not actually begin to process audio until all creation event processing has completed and an onaudioprocess handler has been attached to the node.
void terminate()
The terminate() method, when invoked, must cause the "terminate a worker" algorithm to be run on the worker with which the object is associated. This will cease any AudioProcessEvents being dispatched to the AudioWorker, and will cause the destruction of the associated AudioWorkerGlobalScope. In practical terms, this means all nodes created by this AudioWorker will be disconnected from any downstream connections.

The AudioWorkerNode Interface

This interface represents an AudioNode which interacts with an AudioWorker thread to generate, process, or analyse audio directly. An AudioWorkerNode is represented in the audio thread by an AudioWorkerNodeHandle.

Nota bene that if the Web Audio implementation normally runs audio process at higher than normal thread priority, utilizing AudioWorkerNodes may cause demotion of the priority of the audio thread (since user scripts cannot be run with higher than normal priority).

    numberOfInputs  : variable
    numberOfOutputs : variable

    channelCount = numberOfInputChannels;
    channelCountMode = "explicit";
    channelInterpretation = "speakers";

Example usage:

    var node = context.createAudioWorker( "inverter.js" ).createAudioWorkerNode();
  });
void postMessage(any message, optional sequence<Transferable> transfer)
postMessage may be called to send a message to the Audio Worker, via the algorithm defined by the Worker specification.
attribute EventHandler onmessage
The onmessage handler is called whenever the Audio Worker posts a message back to the main thread.

Note that AudioWorkerNode objects will also have read-only AudioParam objects for each named parameter added via the addParameter method. As this is dynamic, it cannot be captured in IDL.

AudioWorkerNodes must implement the Worker interface for communication with the audio worker script.

The AudioWorkerGlobalScope Interface

This interface is a DedicatedWorkerGlobalScope-derived object representing the context in which audio processing scripts are run for nodes created from a given AudioWorker; it is designed to enable the generation, processing, and analysis of audio data directly using JavaScript in a Worker thread.

The AudioWorkerGlobalScope has an audionodecreated event that is dispatched when new AudioNodes are created on the main thread.

attribute EventHandler onaudionodecreated
A property used to set the EventHandler (described in HTML[[!HTML]]) for the audionodecreated event that is dispatched to AudioWorkerGlobalScope when a new AudioWorkerNode is created. An event of type AudioNodeCreatedEvent will be dispatched to the event handler.
readonly attribute float sampleRate
The sample rate of the host AudioContext (since inside the Worker scope, the user will not have direct access to the AudioContext.

The AudioWorkerNodeHandle Interface

This interface represents the audio thread side of an AudioWorkerNode within an AudioWorkerGlobalScope.

The AudioWorkerNodeHandle has an audioprocess event that is dispatched synchronously to process audio frames. audioprocess events are only dispatched if the AudioWorkerNode has at least one input or one output connected.

It also supports communication with its AudioWorkerNode in the main thread using onmessage and postMessage, allowing the node's behavior to be controlled by the application and letting the node send data to the application.

void postMessage(any message, optional sequence<Transferable> transfer)
postMessage may be called to send a message to the main thread, via the algorithm defined by the Worker specification.
attribute EventHandler onmessage
The onmessage handler is called whenever the main thread posts a message to the AudioWorkerNode to which this AudioWorkerNodeHandle corresponds.
attribute EventHandler onaudioprocess
A property used to set the EventHandler (described in HTML[[!HTML]]) for the audioprocess event that is dispatched to AudioWorkerGlobalScope when the associated AudioWorkerNode is connected. An event of type AudioProcessEvent will be dispatched to the event handler. Setting this attribute to null will prevent further audio processing by the node and make the node eligible for garbage collection.

Audio Worker Examples

A Bitcrusher Node

Bitcrushing is a mechanism by which the audio quality of an audio stream is reduced - both by quantizing the value (simulating lower bit-depth in integer-based audio), and by quantizing in time (simulating a lower digital sample rate). This example shows how to use AudioParams (in this case, treated as a-rate) inside an AudioWorker.

Main file javascript (with no script loading wait)
var worker = audioContext.createAudioWorker("bitcrusher_worker.js");
worker.addParameter( "bits", 8 );
worker.addParameter( "frequencyReduction", 0.5 );
var bitcrusherNode = worker.createAudioWorkerNode();   // node will not actually operate until worker script is initialized
bitcrusherNode.bits.value = 4;
bitcrusherNode.frequencyReduction.value = 0.8;
Main file javascript (with script loading wait)
var worker = audioContext.createAudioWorker("bitcrusher_worker.js");
worker.addParameter( "bits", 8 );
worker.addParameter( "frequencyReduction", 0.5 );
worker.onload = function(e) {
  // At this point, the worker is loaded and available for AudioWorkerNode creation with no script loading delay.
  var bitcrusherNode = worker.createAudioWorkerNode();
  bitcrusherNode.bits.value = 4;
  bitcrusherNode.frequencyReduction.value = 0.8;
  /* The node can now be used in the audio processing graph immediately. */
})
bitcrusher_worker.js
// Configure custom AudioParam - frequency reduction, 0-1, default 0.5

onaudionodecreated = function(e) {
  // internal state variables of node -- in this example, we use closure variables rather than dynamically adding attributes to the node handle.
  var phaser = 0;
  var lastDataValue = 0;

  e.nodeHandle.onaudioprocess = function (e) {
    for (var channel=0; channel<e.inputs[0].length; channel++) {
      var inputBuffer = e.inputs[0][channel];
      var outputBuffer = e.outputs[0][channel];
      var bufferLength = inputBuffer.length;
      var bitsArray = e.parameters.bits;
      var frequencyReductionArray = e.parameters.frequencyReduction;

      for (var i=0; i<bufferLength; i++) {
        var bits = bitsArray ? bitsArray[i] : 8;
        var frequencyReduction = frequencyReductionArray ? frequencyReductionArray[i] : 0.5;

        var step = Math.pow(1/2, bits);
        phaser += frequencyReduction;
        if (phaser >= 1.0) {
            phaser -= 1.0;
            lastDataValue = step * Math.floor(inputBuffer[i] / step + 0.5);
        }
        outputBuffer[i] = lastDataValue;
      }
    }
  };
}

A Volume Meter and Clip Detector

Another common need is a clip-detecting volume meter. This example shows how to communicate basic parameters (that do not need AudioParam scheduling) across to a Worker, as well as communicating data back to the main thread. This node does not use any output.

Main file javascript
var worker = audioContext.createAudioWorker("vu_meter_worker.js");

// This kind of node has one mixed-to-mono input, no outputs
worker.setNodeConfiguration([1], []);

worker.onload = function(e) {
  var vuNode = worker.createAudioWorkerNode();

  // This handles communication back from the volume meter
  vuNode.onmessage = function (event) {
    if (event.data instanceof Object ) {
      if (event.data.hasOwnProperty("clip")
        this.clip = event.data.clip;
      if (event.data.hasOwnProperty("volume")
        this.volume = event.data.volume;
    }
  }

  // Set up some configuration parameters
  vuNode.postMessage( 
    { "smoothing": 0.9,   // Smoothing parameter
      "clipLevel": 0.9,   // Level to consider "clipping"
      "clipLag": 750,     // How long to keep "clipping" lit up after clip (ms)
      "updating": 90      // How frequently to update volume and clip param (ms)
    });

  // Set up volume and clip attributes.  These will be updated by our onmessage.
  vuNode.volume = 0;
  vuNode.clip = false;
};
vu_meter_worker.js
onaudionodecreated = function(e) {
 var vu = e.nodeHandle;

 // how many samples between updates - default 100ms
 vu.timeToNextUpdate = 0.1 * this.sampleRate;
 vu.smoothing = 0;
 vu.clipLevel = 0;
 vu.clipLag = 0;

  // This handles setting attribute values on the AudioWorkerNodeHandle
  // when messaged to do so from the main thread. The 'this' variable
  // is equal to the handle, since it's fielding the onaudioprocess event.
  vu.onmessage = function ( event ) {
    if (event.data instanceof Object) {
      if (event.data.hasOwnProperty("smoothing")
        this.smoothing = event.data.smoothing;
      if (event.data.hasOwnProperty("clipLevel")
        this.clipLevel = event.data.clipLevel;
      if (event.data.hasOwnProperty("clipLag")
        this.clipLag = event.data.clipLag / 1000;  // convert to seconds
      if (event.data.hasOwnProperty("updating")    // convert to samples
        this.updatingInterval = event.data.updating * this.sampleRate / 1000 ;
    }
  };

  vu.onaudioprocess = function ( event ) {
    var buf = event.inputs[0][0];  // Node forces mono
    var bufLength = buf.length;
    var sum = 0;
    var x;

    // Do a root-mean-square on the samples: sum up the squares...
    for (var i=0; i<bufLength; i++) {
      x = buf[i];
      if (Math.abs(x)>=this.clipLevel) {
        this.clipping = true;
        this.unsentClip = true;  // Make sure, for every clip, we send a message.
        this.lastClip = event.playbackTime + (i/this.sampleRate);
      }
      sum += x * x;
    }

    // ... then take the square root of the sum.
    var rms =  Math.sqrt(sum / bufLength);

    // Now smooth this out with the smoothing factor applied
    // to the previous sample - take the max here because we
    // want "fast attack, slow release."
    this.volume = Math.max(rms, this.volume*this.smoothing);
    if (this.clipping && (!this.unsentClip) && (event.playbackTime > (this.lastClip + clipLag)))
      this.clipping = false;

    // How long has it been since our last update?
    this.timeToNextUpdate -= this.last
    if (this.timeToNextUpdate<0) {
      this.timeToNextUpdate = this.updatingInterval;
      this.postMessage(
        { "volume": this.volume,
          "clip": this.clipping });
      this.unsentClip = false;
    }
  };
};

Reimplementing ChannelMerger

This worker shows how to merge inputs into a single output channel.

Main file javascript
var mergerWorker = audioContext.createAudioWorker("merger_worker.js");
  mergerWorker.setNodeConfiguration([1,1,1,1,1,1], [6]);
  var merger = mergerWorker.createAudioWorkerNode();
    // place into graph...
  });
merger_worker.js
onaudionodecreated = function(createEvent) {
  createEvent.nodeHandle.onaudioprocess = function ( event ) {
    for (var input=0; input<event.inputs.length; input++)
    event.outputs[0][input].set(event.inputs[input][0]);
  };
};

The ScriptProcessorNode Interface - DEPRECATED

This interface is an AudioNode which can generate, process, or analyse audio directly using JavaScript. This node type is deprecated, to be replaced by the AudioWorkerNode; this text is only here for informative purposes until implementations remove this node type.

    numberOfInputs  : 1
    numberOfOutputs : 1

    channelCount = numberOfInputChannels;
    channelCountMode = "explicit";
    channelInterpretation = "speakers";

The ScriptProcessorNode is constructed with a bufferSize which must be one of the following values: 256, 512, 1024, 2048, 4096, 8192, 16384. This value controls how frequently the audioprocess event is dispatched and how many sample-frames need to be processed each call. audioprocess events are only dispatched if the ScriptProcessorNode has at least one input or one output connected. Lower numbers for bufferSize will result in a lower (better) latency. Higher numbers will be necessary to avoid audio breakup and glitches. This value will be picked by the implementation if the bufferSize argument to createScriptProcessor is not passed in, or is set to 0.

numberOfInputChannels and numberOfOutputChannels determine the number of input and output channels. It is invalid for both numberOfInputChannels and numberOfOutputChannels to be zero.

    var node = context.createScriptProcessor(bufferSize, numberOfInputChannels, numberOfOutputChannels);
attribute EventHandler onaudioprocess
A property used to set the EventHandler (described in HTML[[!HTML]]) for the audioprocess event that is dispatched to ScriptProcessorNode node types. An event of type AudioProcessingEvent will be dispatched to the event handler.
readonly attribute long bufferSize
The size of the buffer (in sample-frames) which needs to be processed each time onaudioprocess is called. Legal values are (256, 512, 1024, 2048, 4096, 8192, 16384).

The AudioNodeCreatedEvent Interface

This is an Event object which is dispatched to AudioWorkerGlobalScope objects to handle the creation of an AudioWorkerNode, when the main thread has called createAudioWorkerNode() on the corresponding AudioWorker.

The event handler is provided with an AudioWorkerNodeHandle that represents the node which is being created. At a minimum, the handle's onaudioprocess EventHandler should be initialized to a function that will perform audio processing on sample blocks.

readonly attribute AudioWorkerNodeHandle nodeHandle
This attribute provides the AudioWorkerNodeHandle for the node being created.

The AudioProcessEvent Interface

This is an Event object which is dispatched to AudioWorkerGlobalScope objects to perform processing.

The event handler processes audio from the input (if any) by accessing the audio data from the inputBuffers attribute. The audio data which is the result of the processing (or the synthesized data if there are no inputs) is then placed into the outputBuffers.

Note that the target attribute of an AudioProcessEvent is the AudioWorkerNodeHandle for the node that is doing the processing.

readonly attribute double playbackTime
The time when the audio will be played in the same time coordinate system as the AudioContext's currentTime. playbackTime allows for very tight synchronization between processing directly in JavaScript with the other events in the context's rendering graph.
readonly attribute sequence inputs

A readonly array of Arrays of Float32Arrays. The top-level sequence is organized by input; each input may contain multiple channels; each channel contains a Float32Array of sample data. The initial size of the channel array will be determined by the number of channels specified for that input in the setNodeConfiguration() method. However, an onprocess handler may alter this number of channels in the input dynamically, either by adding a Float32Array of blocksize length (128) or by reducing the Array (by reducing the Array.length or by using Array.pop() or Array.slice(). The event object, the Array and the Float32Arrays will be reused by the processing system, in order to minimize memory churn.

Any reordering performed on the Array for an input will not reorganize the connections to the channels for subsequent events.

readonly attribute sequence outputs

A readonly array of Arrays of Float32Arrays. The top-level sequence is organized by output; each output may contain multiple channels; each channel contains a Float32Array of sample data. The initial size of the channel array will be determined by the number of channels specified for that output in the setNodeConfiguration() method. However, an onprocess handler may alter this number of channels in the output dynamically, either by adding a Float32Array of blocksize length (128) or by reducing the Array (by reducing the Array.length or by using Array.pop() or Array.slice(). The event object, the Array and the Float32Arrays will be reused by the processing system, in order to minimize memory churn.

Any reordering performed on the Array for an output will not reorganize the connections to the channels for subsequent events.

readonly attribute object parameters
This object attribute exposes a correspondingly-named read-only Float32Array for each parameter that has been added via addParameter. As this is dynamic, this cannot be captured in IDL. The length of this Float32Array will correspond to the length of the inputBuffer. The contents of this Float32Array will be the values to be used for the AudioParam at the corresponding points in time. It is expected that this Float32Array will be reused by the audio engine.

The AudioProcessingEvent Interface - DEPRECATED

This is an Event object which is dispatched to ScriptProcessorNode nodes. It will be removed when the ScriptProcessorNode is removed, as the replacement AudioWorker uses the AudioProcessEvent.

The event handler processes audio from the input (if any) by accessing the audio data from the inputBuffer attribute. The audio data which is the result of the processing (or the synthesized data if there are no inputs) is then placed into the outputBuffer.

readonly attribute double playbackTime
The time when the audio will be played in the same time coordinate system as the AudioContext's currentTime. playbackTime allows for very tight synchronization between processing directly in JavaScript with the other events in the context's rendering graph.
readonly attribute AudioBuffer inputBuffer
An AudioBuffer containing the input audio data. It will have a number of channels equal to the numberOfInputChannels parameter of the createScriptProcessor() method. This AudioBuffer is only valid while in the scope of the onaudioprocess function. Its values will be meaningless outside of this scope.
readonly attribute AudioBuffer outputBuffer
An AudioBuffer where the output audio data should be written. It will have a number of channels equal to the numberOfOutputChannels parameter of the createScriptProcessor() method. Script code within the scope of the onaudioprocess function is expected to modify the Float32Array arrays representing channel data in this AudioBuffer. Any script modifications to this AudioBuffer outside of this scope will not produce any audible effects.

The PannerNode Interface

This interface represents a processing node which positions / spatializes an incoming audio stream in three-dimensional space. The spatialization is in relation to the AudioContext's AudioListener (listener attribute).

    numberOfInputs  : 1
    numberOfOutputs : 1

    channelCount = 2;
    channelCountMode = "clamped-max";
    channelInterpretation = "speakers";

The input of this node is either mono (1 channel) or stereo (2 channels) and cannot be increased. Connections from nodes with fewer or more channels will be up-mixed or down-mixed appropriately, but a NotSupportedError MUST be thrown if an attempt is made to set channelCount to a value greater than 2 or if channelCountMode is set to "max".

The output of this node is hard-coded to stereo (2 channels) and currently cannot be configured.

The PanningModelType enum determines which spatialization algorithm will be used to position the audio in 3D space. The default is "equal-power".

equalpower
A simple and efficient spatialization algorithm using equal-power panning.
HRTF
A higher quality spatialization algorithm using a convolution with measured impulse responses from human subjects. This panning method renders stereo output.

The DistanceModelType enum determines which algorithm will be used to reduce the volume of an audio source as it moves away from the listener. The default is "inverse".

linear
A linear distance model which calculates distanceGain according to:
    1 - rolloffFactor * (distance - refDistance) / (maxDistance - refDistance)
  
inverse

An inverse distance model which calculates distanceGain according to:

    refDistance / (refDistance + rolloffFactor * (distance - refDistance))
  
exponential

An exponential distance model which calculates distanceGain according to:

    pow(distance / refDistance, -rolloffFactor)
  
attribute PanningModelType panningModel
Specifies the panning model used by this PannerNode. Defaults to "HRTF".
void setPosition(float x, float y, float z)

Sets the position of the audio source relative to the listener attribute. A 3D cartesian coordinate system is used.

The x, y, z parameters represent the coordinates in 3D space.

The default value is (0,0,0)

void setOrientation(float x, float y, float z)

Describes which direction the audio source is pointing in the 3D cartesian coordinate space. Depending on how directional the sound is (controlled by the cone attributes), a sound pointing away from the listener can be very quiet or completely silent.

The x, y, z parameters represent a direction vector in 3D space.

The default value is (1,0,0)

void setVelocity(float x, float y, float z)

Sets the velocity vector of the audio source. This vector controls both the direction of travel and the speed in 3D space. This velocity relative to the listener's velocity is used to determine how much doppler shift (pitch change) to apply. The units used for this vector is meters / second and is independent of the units used for position and orientation vectors.

The x, y, z parameters describe a direction vector indicating direction of travel and intensity.

The default value is (0,0,0)

attribute DistanceModelType distanceModel
Specifies the distance model used by this PannerNode. Defaults to "inverse".
attribute float refDistance
A reference distance for reducing volume as source move further from the listener. The default value is 1.
attribute float maxDistance
The maximum distance between source and listener, after which the volume will not be reduced any further. The default value is 10000.
attribute float rolloffFactor
Describes how quickly the volume is reduced as source moves away from listener. The default value is 1.
attribute float coneInnerAngle
A parameter for directional audio sources, this is an angle, in degrees, inside of which there will be no volume reduction. The default value is 360.
attribute float coneOuterAngle
A parameter for directional audio sources, this is an angle, in degrees, outside of which the volume will be reduced to a constant value of coneOuterGain. The default value is 360.
attribute float coneOuterGain
A parameter for directional audio sources, this is the amount of volume reduction outside of the coneOuterAngle. The default value is 0.

The AudioListener Interface

This interface represents the position and orientation of the person listening to the audio scene. All PannerNode objects spatialize in relation to the AudioContext's listener. See the Spatialization/Panning section for more details about spatialization.

attribute float dopplerFactor
A constant used to determine the amount of pitch shift to use when rendering a doppler effect. The default value is 1.
attribute float speedOfSound
The speed of sound used for calculating doppler shift. The default value is 343.3.
void setPosition(float x, float y, float z)

Sets the position of the listener in a 3D cartesian coordinate space. PannerNode objects use this position relative to individual audio sources for spatialization.

The x, y, z parameters represent the coordinates in 3D space.

The default value is (0,0,0)

void setOrientation(float x, float y, float z, float xUp, float yUp, float zUp)

Describes which direction the listener is pointing in the 3D cartesian coordinate space. Both a front vector and an up vector are provided. In simple human terms, the front vector represents which direction the person's nose is pointing. The up vector represents the direction the top of a person's head is pointing. These values are expected to be linearly independent (at right angles to each other). For normative requirements of how these values are to be interpreted, see the spatialization section.

The x, y, z parameters represent a front direction vector in 3D space, with the default value being (0,0,-1).

The xUp, yUp, zUp parameters represent an up direction vector in 3D space, with the default value being (0,1,0).

void setVelocity(float x, float y, float z)

Sets the velocity vector of the listener. This vector controls both the direction of travel and the speed in 3D space. This velocity relative to an audio source's velocity is used to determine how much doppler shift (pitch change) to apply. The units used for this vector is meters / second and is independent of the units used for position and orientation vectors.

The x, y, z parameters describe a direction vector indicating direction of travel and intensity.

The default value is (0,0,0)

The StereoPannerNode Interface

This interface represents a processing node which positions an incoming audio stream in a stereo image using a low-cost equal-power panning algorithm. This panning effect is common in positioning audio components in a stereo stream.

    numberOfInputs  : 1
    numberOfOutputs : 1

    channelCount = 2;
    channelCountMode = "clamped-max";
    channelInterpretation = "speakers";

The input of this node is stereo (2 channels) and cannot be increased. Connections from nodes with fewer or more channels will be up-mixed or down-mixed appropriately, but a NotSupportedError will be thrown if an attempt is made to set channelCount to a value great than 2 or if channelCountMode is set to "max".

The output of this node is hard-coded to stereo (2 channels) and cannot be configured.

readonly attribute AudioParam pan
The position of the input in the output's stereo image. -1 represents full left, +1 represents full right. Its default value is 0, and its nominal range is from -1 to 1. This parameter is a a-rate.

The ConvolverNode Interface

This interface represents a processing node which applies a linear convolution effect given an impulse response. Normative requirements for multi-channel convolution matrixing are described here.

    numberOfInputs  : 1
    numberOfOutputs : 1

    channelCount = 2;
    channelCountMode = "clamped-max";
    channelInterpretation = "speakers";

The input of this node is either mono (1 channel) or stereo (2 channels) and cannot be increased. Connections from nodes with fewer or more channels will be up-mixed or down-mixed appropriately, but a NotSupportedError MUST be thrown if an attempt is made to set channelCount to a value great than 2 or if channelCountMode is set to "max".

attribute AudioBuffer? buffer
A mono, stereo, or 4-channel AudioBuffer containing the (possibly multi-channel) impulse response used by the ConvolverNode. The AudioBuffer must have 1, 2, or 4 channels or a NotSupportedError exception MUST be thrown. This AudioBuffer must be of the same sample-rate as the AudioContext or an NotSupportedError exception MUST be thrown. At the time when this attribute is set, the buffer and the state of the normalize attribute will be used to configure the ConvolverNode with this impulse response having the given normalization. The initial value of this attribute is null.
attribute boolean normalize

Controls whether the impulse response from the buffer will be scaled by an equal-power normalization when the buffer atttribute is set. Its default value is true in order to achieve a more uniform output level from the convolver when loaded with diverse impulse responses. If normalize is set to false, then the convolution will be rendered with no pre-processing/scaling of the impulse response. Changes to this value do not take effect until the next time the buffer attribute is set.

If the normalize attribute is false when the buffer attribute is set then the ConvolverNode will perform a linear convolution given the exact impulse response contained within the buffer.

Otherwise, if the normalize attribute is true when the buffer attribute is set then the ConvolverNode will first perform a scaled RMS-power analysis of the audio data contained within buffer to calculate a normalizationScale given this algorithm:


float calculateNormalizationScale(buffer)
{
    const float GainCalibration = 0.00125;
    const float GainCalibrationSampleRate = 44100;
    const float MinPower = 0.000125;

    // Normalize by RMS power.
    size_t numberOfChannels = buffer->numberOfChannels();
    size_t length = buffer->length();

    float power = 0;

    for (size_t i = 0; i < numberOfChannels; ++i) {
        float* sourceP = buffer->channel(i)->data();
        float channelPower = 0;

        int n = length;
        while (n--) {
            float sample = *sourceP++;
            channelPower += sample * sample;
        }

        power += channelPower;
    }

    power = sqrt(power / (numberOfChannels * length));

    // Protect against accidental overload.
    if (isinf(power) || isnan(power) || power < MinPower)
        power = MinPower;

    float scale = 1 / power;

    // Calibrate to make perceived volume same as unprocessed.
    scale *= GainCalibration;

    // Scale depends on sample-rate.
    if (buffer->sampleRate())
        scale *= GainCalibrationSampleRate / buffer->sampleRate();

    // True-stereo compensation.
    if (buffer->numberOfChannels() == 4)
        scale *= 0.5;

    return scale;
}
      

During processing, the ConvolverNode will then take this calculated normalizationScale value and multiply it by the result of the linear convolution resulting from processing the input with the impulse response (represented by the buffer) to produce the final output. Or any mathematically equivalent operation may be used, such as pre-multiplying the input by normalizationScale, or pre-multiplying a version of the impulse-response by normalizationScale.

The AnalyserNode Interface

This interface represents a node which is able to provide real-time frequency and time-domain analysis information. The audio stream will be passed un-processed from input to output.

    numberOfInputs  : 1
    numberOfOutputs : 1    Note that this output may be left unconnected.

    channelCount = 1;
    channelCountMode = "max";
    channelInterpretation = "speakers";
void getFloatFrequencyData()

Copies the current frequency data into the passed floating-point array. If the array has fewer elements than the frequencyBinCount, the excess elements will be dropped. If the array has more elements than the frequencyBinCount, the excess elements will be ignored.

Float32Array array
This parameter is where the frequency-domain analysis data will be copied.
void getByteFrequencyData()

Copies the current frequency data into the passed unsigned byte array. If the array has fewer elements than the frequencyBinCount, the excess elements will be dropped. If the array has more elements than the frequencyBinCount, the excess elements will be ignored.

Uint8Array array
This parameter is where the frequency-domain analysis data will be copied.
void getFloatTimeDomainData()

Copies the current time-domain (waveform) data into the passed floating-point array. If the array has fewer elements than the value of fftSize, the excess elements will be dropped. If the array has more elements than fftSize, the excess elements will be ignored.

Float32Array array
This parameter is where the time-domain sample data will be copied.
void getByteTimeDomainData()

Copies the current time-domain (waveform) data into the passed unsigned byte array. If the array has fewer elements than the value of fftSize, the excess elements will be dropped. If the array has more elements than fftSize, the excess elements will be ignored.

Uint8Array array
This parameter is where the time-domain sample data will be copied.
attribute unsigned long fftSize
The size of the FFT used for frequency-domain analysis. This must be a non-zero power of two in the range 32 to 2048, otherwise an IndexSizeError exception MUST be thrown. The default value is 2048.
readonly attribute unsigned long frequencyBinCount
Half the FFT size.
attribute float minDecibels
The minimum power value in the scaling range for the FFT analysis data for conversion to unsigned byte values. The default value is -100. If the value of this attribute is set to a value more than or equal to maxDecibels, an IndexSizeError exception MUST be thrown.
attribute float maxDecibels
The maximum power value in the scaling range for the FFT analysis data for conversion to unsigned byte values. The default value is -30. If the value of this attribute is set to a value less than or equal to minDecibels, an IndexSizeError exception MUST be thrown.
attribute float smoothingTimeConstant
A value from 0 -> 1 where 0 represents no time averaging with the last analysis frame. The default value is 0.8. If the value of this attribute is set to a value less than 0 or more than 1, an IndexSizeError exception MUST be thrown.

FFT Windowing and smoothing over time

When the current frequency data are computed, the following operations are to be performed:
  1. Apply a Blackman window to the time domain input data
  2. Apply a Fourier tranform to the windowed time domain input data to get imaginary and real frequency data
  3. Smooth over time the frequency domain data
Applying a Blackman window consists in the following operation on the input time domain data buffer:
function blackmanWindow(buffer) {
  var alpha = 0.16;
  var a0 = 0.5 * (1.0 - alpha);
  var a1 = 0.5;
  var a2 = 0.5 * alpha;

  for (var i = 0; i < buffer.length; i++) {
    var x = i / buffer.length;
    buffer[i] *= a0 - a1 * Math.cos(2 * Math.PI * x) + a2 * Math.cos(4 * Math.PI * x);
  }
}
    
Smoothing over time frequency data consists in the following operation:
  function smoothingOverTime(outputBuffer, lastOutputBuffer,
                             smoothingConstant,
                             real, imag, fftSize) {
    for (var i = 0; i < outputBuffer.length; i++) {
      var magnitude = Math.sqrt(real[i] * real[i] + imag[i] * imag[i]) / fftSize;
      outputBuffer[i] = smoothingTimeConstant * lastOutputBuffer[i] + (1.0 - smoothingTimeConstant) * magnitude;
    }
  }

The ChannelSplitterNode Interface

The ChannelSplitterNode is for use in more advanced applications and would often be used in conjunction with ChannelMergerNode.

    numberOfInputs  : 1
    numberOfOutputs : Variable N (defaults to 6) // number of "active" (non-silent) outputs is determined by number of channels in the input

    channelCountMode = "max";
    channelInterpretation = "speakers";

This interface represents an AudioNode for accessing the individual channels of an audio stream in the routing graph. It has a single input, and a number of "active" outputs which equals the number of channels in the input audio stream. For example, if a stereo input is connected to an ChannelSplitterNode then the number of active outputs will be two (one from the left channel and one from the right). There are always a total number of N outputs (determined by the numberOfOutputs parameter to the AudioContext method createChannelSplitter()), The default number is 6 if this value is not provided. Any outputs which are not "active" will output silence and would typically not be connected to anything.

Example:

channel splitter
A diagram of a ChannelSplitter

Please note that in this example, the splitter does not interpret the channel identities (such as left, right, etc.), but simply splits out channels in the order that they are input.

One application for ChannelSplitterNode is for doing "matrix mixing" where individual gain control of each channel is desired.

The ChannelMergerNode Interface

The ChannelMergerNode is for use in more advanced applications and would often be used in conjunction with ChannelSplitterNode.

  numberOfInputs  : Variable N (default to 6)  // number of connected inputs may be less than this
  numberOfOutputs : 1

  channelCountMode = "max";
  channelInterpretation = "speakers";

This interface represents an AudioNode for combining channels from multiple audio streams into a single audio stream. It has a variable number of inputs (defaulting to 6), but not all of them need be connected. There is a single output whose audio stream has a number of channels equal to the sum of the numbers of channels of all the connected inputs. For example, if an ChannelMergerNode has two connected inputs (both stereo), then the output will be four channels, the first two from the first input and the second two from the second input. In another example with two connected inputs (both mono), the output will be two channels (stereo), with the left channel coming from the first input and the right channel coming from the second input.

Example:

channel merger
A diagram of ChannelMerger

Please note that in this example, the merger does not interpret the channel identities (such as left, right, etc.), but simply combines channels in the order that they are input.

Be aware that it is possible to connect an ChannelMergerNode in such a way that it outputs an audio stream with a large number of channels greater than the maximum supported by the audio hardware. In this case where such an output is connected to the AudioContext destination (the audio hardware), then the extra channels will be ignored. Thus, the ChannelMergerNode should be used in situations where the number of channels is well understood.

The DynamicsCompressorNode Interface

DynamicsCompressorNode is an AudioNode processor implementing a dynamics compression effect.

Dynamics compression is very commonly used in musical production and game audio. It lowers the volume of the loudest parts of the signal and raises the volume of the softest parts. Overall, a louder, richer, and fuller sound can be achieved. It is especially important in games and musical applications where large numbers of individual sounds are played simultaneous to control the overall signal level and help avoid clipping (distorting) the audio output to the speakers.

    numberOfInputs  : 1
    numberOfOutputs : 1

    channelCount = 2;
    channelCountMode = "explicit";
    channelInterpretation = "speakers";
readonly attribute AudioParam threshold
The decibel value above which the compression will start taking effect. Its default value is -24, with a nominal range of -100 to 0.
readonly attribute AudioParam knee
A decibel value representing the range above the threshold where the curve smoothly transitions to the "ratio" portion. Its default value is 30, with a nominal range of 0 to 40.
readonly attribute AudioParam ratio
The amount of dB change in input for a 1 dB change in output. Its default value is 12, with a nominal range of 1 to 20.
readonly attribute float reduction
A read-only decibel value for metering purposes, representing the current amount of gain reduction that the compressor is applying to the signal. If fed no signal the value will be 0 (no gain reduction).
readonly attribute AudioParam attack
The amount of time (in seconds) to reduce the gain by 10dB. Its default value is 0.003, with a nominal range of 0 to 1.
readonly attribute AudioParam release
The amount of time (in seconds) to increase the gain by 10dB. Its default value is 0.250, with a nominal range of 0 to 1.

The BiquadFilterNode Interface

BiquadFilterNode is an AudioNode processor implementing very common low-order filters.

Low-order filters are the building blocks of basic tone controls (bass, mid, treble), graphic equalizers, and more advanced filters. Multiple BiquadFilterNode filters can be combined to form more complex filters. The filter parameters such as frequency can be changed over time for filter sweeps, etc. Each BiquadFilterNode can be configured as one of a number of common filter types as shown in the IDL below. The default filter type is "lowpass".

Both frequency and detune are a-rate parameters and are used together to determine a computedFrequency value:

  computedFrequency(t) = frequency(t) * pow(2, detune(t) / 1200)
    numberOfInputs  : 1
    numberOfOutputs : 1

    channelCountMode = "max";
    channelInterpretation = "speakers";

The number of channels of the output always equals the number of channels of the input.

lowpass

A lowpass filter allows frequencies below the cutoff frequency to pass through and attenuates frequencies above the cutoff. It implements a standard second-order resonant lowpass filter with 12dB/octave rolloff.

frequency
The cutoff frequency
Q
Controls how peaked the response will be at the cutoff frequency. A large value makes the response more peaked. Please note that for this filter type, this value is not a traditional Q, but is a resonance value in decibels.
gain
Not used in this filter type
highpass

A highpass filter is the opposite of a lowpass filter. Frequencies above the cutoff frequency are passed through, but frequencies below the cutoff are attenuated. It implements a standard second-order resonant highpass filter with 12dB/octave rolloff.

frequency
The cutoff frequency below which the frequencies are attenuated
Q
Controls how peaked the response will be at the cutoff frequency. A large value makes the response more peaked. Please note that for this filter type, this value is not a traditional Q, but is a resonance value in decibels.
gain
Not used in this filter type
bandpass

A bandpass filter allows a range of frequencies to pass through and attenuates the frequencies below and above this frequency range. It implements a second-order bandpass filter.

frequency
The center of the frequency band
Q
Controls the width of the band. The width becomes narrower as the Q value increases.
gain
Not used in this filter type
lowshelf

The lowshelf filter allows all frequencies through, but adds a boost (or attenuation) to the lower frequencies. It implements a second-order lowshelf filter.

frequency
The upper limit of the frequences where the boost (or attenuation) is applied.
Q
Not used in this filter type.
gain
The boost, in dB, to be applied. If the value is negative, the frequencies are attenuated.
highshelf

The highshelf filter is the opposite of the lowshelf filter and allows all frequencies through, but adds a boost to the higher frequencies. It implements a second-order highshelf filter

frequency
The lower limit of the frequences where the boost (or attenuation) is applied.
Q
Not used in this filter type.
gain
The boost, in dB, to be applied. If the value is negative, the frequencies are attenuated.
peaking

The peaking filter allows all frequencies through, but adds a boost (or attenuation) to a range of frequencies.

frequency
The center frequency of where the boost is applied.
Q
Controls the width of the band of frequencies that are boosted. A large value implies a narrow width.
gain
The boost, in dB, to be applied. If the value is negative, the frequencies are attenuated.
notch

The notch filter (also known as a band-stop or band-rejection filter) is the opposite of a bandpass filter. It allows all frequencies through, except for a set of frequencies.

frequency
The center frequency of where the notch is applied.
Q
Controls the width of the band of frequencies that are attenuated. A large value implies a narrow width.
gain
Not used in this filter type.
allpass

An allpass filter allows all frequencies through, but changes the phase relationship between the various frequencies. It implements a second-order allpass filter

frequency
The frequency where the center of the phase transition occurs. Viewed another way, this is the frequency with maximal group delay.
Q
Controls how sharp the phase transition is at the center frequency. A larger value implies a sharper transition and a larger group delay.
gain
Not used in this filter type.

All attributes of the BiquadFilterNode are k-rate AudioParam.

attribute BiquadFilterType type
The type of this BiquadFilterNode. The exact meaning of the other parameters depend on the value of the type attribute.
readonly attribute AudioParam frequency
The frequency at which the BiquadFilterNode will operate, in Hz. Its default value is 350Hz, and its nominal range is from 10Hz to half the Nyquist frequency.
readonly attribute AudioParam detune
A detune value, in cents, for the frequency. Its default value is 0.
readonly attribute AudioParam Q
The Q factor has a default value of 1, with a nominal range of 0.0001 to 1000.
readonly attribute AudioParam gain
The gain has a default value of 0, with a nominal range of -40 to 40.
void getFrequencyResponse()

Given the current filter parameter settings, calculates the frequency response for the specified frequencies.

Float32Array frequencyHz
This parameter specifies an array of frequencies at which the response values will be calculated.
Float32Array magResponse
This parameter specifies an output array receiving the linear magnitude response values.
Float32Array phaseResponse
This parameter specifies an output array receiving the phase response values in radians.

Filters characteristics

There are multiple ways of implementing the type of filters available through the BiquadFilterNode each having very different characteristics. The formulas in this section describe the filters that a conforming implementation MUST implement, as they determine the characteristics of the different filter type. They are derived from formulas found in the Audio EQ Cookbook.

The transfer function for the filters implemented by the BiquadFilterNode is: H ( z ) = ( b 0 a 0 ) + ( b 1 a 0 ) z - 1 + ( b 2 a 0 ) z - 2 1 + ( a 1 a 0 ) z - 1 + ( a 2 a 0 ) z - 2     The initial filter state is 0.

The coefficients in the transfer function above are different for each node type. The following intermediate variable are necessary for their computation, based on the computedValue of the AudioParams of the BiquadFilterNode.

The six coefficients ( b0, b1, b2, a0, a1, a2) for each filter type, are:

lowpass
b 0 = 1 - cos ( ω 0 ) 2 b 1 = 1 - cos ( ω 0 ) b 2 = 1 - cos ( ω 0 ) 2 a 0 = 1 + α Q a 1 = - 2 cos ( ω 0 ) a 2 = 1 - α Q
highpass
b 0 = 1 + cos ( ω 0 ) 2 b 1 = - ( 1 + cos ( ω 0 ) ) b 2 = 1 + cos ( ω 0 ) 2 a 0 = 1 + α Q a 1 = - 2 cos ( ω 0 ) a 2 = 1 - α Q
bandpass
b 0 = α BW b 1 = 0 b 2 = - α BW a 0 = 1 + α BW a 1 = - 2 cos ( ω 0 ) a 2 = 1 - α BW
notch
b 0 = 1 b 1 = - 2 cos ( ω 0 ) b 2 = 1 a 0 = 1 + α BW a 1 = - 2 cos ( ω 0 ) a 2 = 1 - α BW
allpass
b 0 = 1 - α Q b 1 = - 2 cos ( ω 0 ) b 2 = 1 + α Q a 0 = 1 + α Q a 1 = - 2 cos ( ω 0 ) a 2 = 1 - α Q
peaking
b 0 = 1 + α BW A b 1 = - 2 cos ( ω 0 ) b 2 = 1 - α BW A a 0 = 1 + α BW A a 1 = - 2 cos ( ω 0 ) a 2 = 1 - α BW A
lowshelf
b 0 = A ( ( A + 1 ) - ( A - 1 ) cos ( ω 0 ) + 2 A α S ) b 1 = 2 A ( ( A - 1 ) - ( A + 1 ) cos ( ω 0 ) ) b 2 = A ( ( A + 1 ) - ( A - 1 ) cos ( ω 0 ) - 2 A α S ) a 0 = ( A + 1 ) + ( A - 1 ) cos ( ω 0 ) + 2 A α S a 1 = - 2 ( ( A - 1 ) + ( A + 1 ) cos ( ω 0 ) ) a 2 = ( A + 1 ) + ( A - 1 ) cos ( ω 0 ) - 2 A α S
highshelf
b 0 = A ( ( A + 1 ) + ( A - 1 ) cos ( ω 0 ) + 2 A α S ) b 1 = - 2 A ( ( A - 1 ) + ( A + 1 ) cos ( ω 0 ) ) b 2 = A ( ( A + 1 ) + ( A - 1 ) cos ( ω 0 ) - 2 A α S ) a 0 = ( A + 1 ) - ( A - 1 ) cos ( ω 0 ) + 2 A α S a 1 = 2 ( ( A - 1 ) - ( A + 1 ) cos ( ω 0 ) ) a 2 = ( A + 1 ) - ( A - 1 ) cos ( ω 0 ) - 2 A α S

The WaveShaperNode Interface

WaveShaperNode is an AudioNode processor implementing non-linear distortion effects.

Non-linear waveshaping distortion is commonly used for both subtle non-linear warming, or more obvious distortion effects. Arbitrary non-linear shaping curves may be specified.

    numberOfInputs  : 1
    numberOfOutputs : 1

    channelCountMode = "max";
    channelInterpretation = "speakers";

The number of channels of the output always equals the number of channels of the input.

none
Don't oversample
2x
Oversample two times
4x
Oversample four times
attribute Float32Array? curve

The shaping curve used for the waveshaping effect. The input signal is nominally within the range [-1; 1]. Each input sample within this range will index into the shaping curve, with a signal level of zero corresponding to the center value of the curve array if there are an odd number of entries, or interpolated between the two centermost values if there are an even number of entries in the array. Any sample value less than -1 will correspond to the first value in the curve array. Any sample value greater than +1 will correspond to the last value in the curve array.

The implementation must perform linear interpolation between adjacent points in the curve. Initially the curve attribute is null, which means that the WaveShaperNode will pass its input to its output without modification.

Values of the curve are spread with equal spacing in the [-1; 1] range. This means that a curve with a even number of value will not have a value for a signal at zero, and a curve with an odd number of value will have a value for a signal at zero.

A InvalidAccessError MUST be thrown if this attribute is set with a Float32Array that has a length less than 2.

attribute OverSampleType oversample

Specifies what type of oversampling (if any) should be used when applying the shaping curve. The default value is "none", meaning the curve will be applied directly to the input samples. A value of "2x" or "4x" can improve the quality of the processing by avoiding some aliasing, with the "4x" value yielding the highest quality. For some applications, it's better to use no oversampling in order to get a very precise shaping curve.

A value of "2x" or "4x" means that the following steps must be performed:

  1. Up-sample the input samples to 2x or 4x the sample-rate of the AudioContext. Thus for each processing block of 128 samples, generate 256 (for 2x) or 512 (for 4x) samples.
  2. Apply the shaping curve.
  3. Down-sample the result back to the sample-rate of the AudioContext. Thus taking the 256 (or 512) processed samples, generating 128 as the final result.

The exact up-sampling and down-sampling filters are not specified, and can be tuned for sound quality (low aliasing, etc.), low latency, and performance.

The OscillatorNode Interface

OscillatorNode represents an audio source generating a periodic waveform. It can be set to a few commonly used waveforms. Additionally, it can be set to an arbitrary periodic waveform through the use of a PeriodicWave object.

Oscillators are common foundational building blocks in audio synthesis. An OscillatorNode will start emitting sound at the time specified by the start() method.

Mathematically speaking, a continuous-time periodic waveform can have very high (or infinitely high) frequency information when considered in the frequency domain. When this waveform is sampled as a discrete-time digital audio signal at a particular sample-rate, then care must be taken to discard (filter out) the high-frequency information higher than the Nyquist frequency (half the sample-rate) before converting the waveform to a digital form. If this is not done, then aliasing of higher frequencies (than the Nyquist frequency) will fold back as mirror images into frequencies lower than the Nyquist frequency. In many cases this will cause audibly objectionable artifacts. This is a basic and well understood principle of audio DSP.

There are several practical approaches that an implementation may take to avoid this aliasing. Regardless of approach, the idealized discrete-time digital audio signal is well defined mathematically. The trade-off for the implementation is a matter of implementation cost (in terms of CPU usage) versus fidelity to achieving this ideal.

It is expected that an implementation will take some care in achieving this ideal, but it is reasonable to consider lower-quality, less-costly approaches on lower-end hardware.

Both .frequency and .detune are a-rate parameters and are used together to determine a computedFrequency value:

  computedFrequency(t) = frequency(t) * pow(2, detune(t) / 1200)

The OscillatorNode's instantaneous phase at each time is the time integral of computedFrequency.

  numberOfInputs  : 0
  numberOfOutputs : 1 (mono output)
sine
A sine wave
square
A square wave of duty period 0.5
sawtooth
A sawtooth wave
triangle
A triangle wave
custom
A custom periodic wave
attribute OscillatorType type
The shape of the periodic waveform. It may directly be set to any of the type constant values except for "custom". The setPeriodicWave() method can be used to set a custom waveform, which results in this attribute being set to "custom". The default value is "sine". When this attribute is set, the phase of the oscillator MUST be conserved.
readonly attribute AudioParam frequency
The frequency (in Hertz) of the periodic waveform. Its default value is 440. This parameter is a-rate.
readonly attribute AudioParam detune
A detuning value (in Cents) which will offset the frequency by the given amount. Its default value is 0. This parameter is a-rate.
void start(optional double when = 0)
Defined the same as the when parameter of the AudioBufferSourceNode
void stop(optional double when = 0)
Defined as in AudioBufferSourceNode.
void setPeriodicWave(PeriodicWave periodicWave)
Sets an arbitrary custom periodic waveform given a PeriodicWave.
attribute EventHandler onended
A property used to set the EventHandler (described in HTML[[HTML]]) for the ended event that is dispatched to OscillatorNode node types. When the OscillatorNode has finished playing (i.e. its stop time has been reached), an event of type Event (described in HTML[[HTML]]) will be dispatched to the event handler.

Basic waveform phase

The idealized mathematical waveforms for the various oscillator types are defined here. In summary, all waveforms are defined mathematically to be an odd function with a positive slope at time 0. The actual waveforms produced by the oscillator may differ to prevent aliasing affects.

"sine"
The waveform for sine oscillator is sin(t).
"square"
The waveform for the square wave oscillator is 1 for 0≤ t < π and -1 for -π < t < 0.
"sawtooth"
The waveform for the sawtooth oscillator is the ramp t/π for -π < t ≤ π
"triangle"
The waveform for the triangle oscillator is 2/π*t for 0 ≤ t ≤ π/2 and 1-2/π*(t-π/2) for π/2 < t ≤ π. This is extended to all t by using the fact that the waveform is an odd function with period 2π.

The PeriodicWave Interface

PeriodicWave represents an arbitrary periodic waveform to be used with an OscillatorNode. Please see createPeriodicWave() and setPeriodicWave() and for more details.

The MediaStreamAudioSourceNode Interface

This interface represents an audio source from a MediaStream. The first AudioMediaStreamTrack from the MediaStream will be used as a source of audio. Those interfaces are described in [[!mediacapture-streams]].

    numberOfInputs  : 0
    numberOfOutputs : 1

The number of channels of the output corresponds to the number of channels of the AudioMediaStreamTrack. If there is no valid audio track, then the number of channels output will be one silent channel.

The MediaStreamAudioDestinationNode Interface

This interface is an audio destination representing a MediaStream with a single AudioMediaStreamTrack. This MediaStream is created when the node is created and is accessible via the stream attribute. This stream can be used in a similar way as a MediaStream obtained via getUserMedia(), and can, for example, be sent to a remote peer using the RTCPeerConnection (described in [[!webrtc]]) addStream() method.

    numberOfInputs  : 1
    numberOfOutputs : 0

    channelCount = 2;
    channelCountMode = "explicit";
    channelInterpretation = "speakers";

The number of channels of the input is by default 2 (stereo). Any connections to the input are up-mixed/down-mixed to the number of channels of the input.

readonly attribute MediaStream stream
A MediaStream containing a single AudioMediaStreamTrack with the same number of channels as the node itself.

Mixer Gain Structure

Background

One of the most important considerations when dealing with audio processing graphs is how to adjust the gain (volume) at various points. For example, in a standard mixing board model, each input bus has pre-gain, post-gain, and send-gains. Submix and master out busses also have gain control. The gain control described here can be used to implement standard mixing boards as well as other architectures.

Summing Inputs

The inputs to AudioNodes have the ability to accept connections from multiple outputs. The input then acts as a unity gain summing junction with each output signal being added with the others:

unity gain summing junction
A graph showing Source 1 and Source 2 output summed at the input of Destination

In cases where the channel layouts of the outputs do not match, a mix (usually up-mix) will occur according to the mixing rules.

No clipping is applied at the inputs or outputs of the AudioNode to allow a maximum of dynamic range within the audio graph.

Gain Control

In many scenarios, it's important to be able to control the gain for each of the output signals. The GainNode gives this control:

mixer architecture new
A graph featuring volume control for each voice

Using these two concepts of unity gain summing junctions and GainNodes, it's possible to construct simple or complex mixing scenarios.

Example: Mixer with Send Busses

In a routing scenario involving multiple sends and submixes, explicit control is needed over the volume or "gain" of each connection to a mixer. Such routing topologies are very common and exist in even the simplest of electronic gear sitting around in a basic recording studio.

Here's an example with two send mixers and a main mixer. Although possible, for simplicity's sake, pre-gain control and insert effects are not illustrated:

mixer gain structure
A graph showing a full mixer with send busses.

This diagram is using a shorthand notation where "send 1", "send 2", and "main bus" are actually inputs to AudioNodes, but here are represented as summing busses, where the intersections g2_1, g3_1, etc. represent the "gain" or volume for the given source on the given mixer. In order to expose this gain, an GainNode is used:

Here's how the above diagram could be constructed in JavaScript:


  var context = 0;
  var compressor = 0;
  var reverb = 0;
  var delay = 0;
  var s1 = 0;
  var s2 = 0;

  var source1 = 0;
  var source2 = 0;
  var g1_1 = 0;
  var g2_1 = 0;
  var g3_1 = 0;
  var g1_2 = 0;
  var g2_2 = 0;
  var g3_2 = 0;

  // Setup routing graph
  function setupRoutingGraph() {
      context = new AudioContext();

      compressor = context.createDynamicsCompressor();

      // Send1 effect
      reverb = context.createConvolver();
      // Convolver impulse response may be set here or later

      // Send2 effect
      delay = context.createDelay();

      // Connect final compressor to final destination
      compressor.connect(context.destination);

      // Connect sends 1 & 2 through effects to main mixer
      s1 = context.createGain();
      reverb.connect(s1);
      s1.connect(compressor);

      s2 = context.createGain();
      delay.connect(s2);
      s2.connect(compressor);

      // Create a couple of sources
      source1 = context.createBufferSource();
      source2 = context.createBufferSource();
      source1.buffer = manTalkingBuffer;
      source2.buffer = footstepsBuffer;

      // Connect source1
      g1_1 = context.createGain();
      g2_1 = context.createGain();
      g3_1 = context.createGain();
      source1.connect(g1_1);
      source1.connect(g2_1);
      source1.connect(g3_1);
      g1_1.connect(compressor);
      g2_1.connect(reverb);
      g3_1.connect(delay);

      // Connect source2
      g1_2 = context.createGain();
      g2_2 = context.createGain();
      g3_2 = context.createGain();
      source2.connect(g1_2);
      source2.connect(g2_2);
      source2.connect(g3_2);
      g1_2.connect(compressor);
      g2_2.connect(reverb);
      g3_2.connect(delay);

      // We now have explicit control over all the volumes g1_1, g2_1, ..., s1, s2
      g2_1.gain.value = 0.2;  // For example, set source1 reverb gain

      // Because g2_1.gain is an "AudioParam",
      // an automation curve could also be attached to it.
      // A "mixing board" UI could be created in canvas or WebGL controlling these gains.
  }

   

Dynamic Lifetime

Background

This section is non-normative. Please see AudioContext lifetime and AudioNode lifetime for normative requirements.

In addition to allowing the creation of static routing configurations, it should also be possible to do custom effect routing on dynamically allocated voices which have a limited lifetime. For the purposes of this discussion, let's call these short-lived voices "notes". Many audio applications incorporate the ideas of notes, examples being drum machines, sequencers, and 3D games with many one-shot sounds being triggered according to game play.

In a traditional software synthesizer, notes are dynamically allocated and released from a pool of available resources. The note is allocated when a MIDI note-on message is received. It is released when the note has finished playing either due to it having reached the end of its sample-data (if non-looping), it having reached a sustain phase of its envelope which is zero, or due to a MIDI note-off message putting it into the release phase of its envelope. In the MIDI note-off case, the note is not released immediately, but only when the release envelope phase has finished. At any given time, there can be a large number of notes playing but the set of notes is constantly changing as new notes are added into the routing graph, and old ones are released.

The audio system automatically deals with tearing-down the part of the routing graph for individual "note" events. A "note" is represented by an AudioBufferSourceNode, which can be directly connected to other processing nodes. When the note has finished playing, the context will automatically release the reference to the AudioBufferSourceNode, which in turn will release references to any nodes it is connected to, and so on. The nodes will automatically get disconnected from the graph and will be deleted when they have no more references. Nodes in the graph which are long-lived and shared between dynamic voices can be managed explicitly. Although it sounds complicated, this all happens automatically with no extra JavaScript handling required.

Example

dynamic allocation
A graph featuring a subgraph that will be releases early.

The low-pass filter, panner, and second gain nodes are directly connected from the one-shot sound. So when it has finished playing the context will automatically release them (everything within the dotted line). If there are no longer any JavaScript references to the one-shot sound and connected nodes, then they will be immediately removed from the graph and deleted. The streaming source, has a global reference and will remain connected until it is explicitly disconnected. Here's how it might look in JavaScript:


var context = 0;
var compressor = 0;
var gainNode1 = 0;
var streamingAudioSource = 0;

// Initial setup of the "long-lived" part of the routing graph
function setupAudioContext() {
    context = new AudioContext();

    compressor = context.createDynamicsCompressor();
    gainNode1 = context.createGain();

    // Create a streaming audio source.
    var audioElement = document.getElementById('audioTagID');
    streamingAudioSource = context.createMediaElementSource(audioElement);
    streamingAudioSource.connect(gainNode1);

    gainNode1.connect(compressor);
    compressor.connect(context.destination);
}

// Later in response to some user action (typically mouse or key event)
// a one-shot sound can be played.
function playSound() {
    var oneShotSound = context.createBufferSource();
    oneShotSound.buffer = dogBarkingBuffer;

    // Create a filter, panner, and gain node.
    var lowpass = context.createBiquadFilter();
    var panner = context.createPanner();
    var gainNode2 = context.createGain();

    // Make connections
    oneShotSound.connect(lowpass);
    lowpass.connect(panner);
    panner.connect(gainNode2);
    gainNode2.connect(compressor);

    // Play 0.75 seconds from now (to play immediately pass in 0)
    oneShotSound.start(context.currentTime + 0.75);
}

Channel up-mixing and down-mixing

This section is normative.

describes how an input to an AudioNode can be connected from one or more outputs of an AudioNode. Each of these connections from an output represents a stream with a specific non-zero number of channels. An input has mixing rules for combining the channels from all of the connections to it. As a simple example, if an input is connected from a mono output and a stereo output, then the mono connection will usually be up-mixed to stereo and summed with the stereo connection. But, of course, it's important to define the exact mixing rules for every input to every AudioNode. The default mixing rules for all of the inputs have been chosen so that things "just work" without worrying too much about the details, especially in the very common case of mono and stereo streams. Of course, the rules can be changed for advanced use cases, especially multi-channel.

To define some terms, up-mixing refers to the process of taking a stream with a smaller number of channels and converting it to a stream with a larger number of channels. down-mixing refers to the process of taking a stream with a larger number of channels and converting it to a stream with a smaller number of channels.

An AudioNode input use three basic pieces of information to determine how to mix all the outputs connected to it. As part of this process it computes an internal value computedNumberOfChannels representing the actual number of channels of the input at any given time:

The AudioNode attributes involved in channel up-mixing and down-mixing rules are defined above. The following is a more precise specification on what each of them mean.

For each input of an AudioNode, an implementation must:

  1. Compute computedNumberOfChannels.
  2. For each connection to the input:
    • up-mix or down-mix the connection to computedNumberOfChannels according to channelInterpretation.
    • Mix it together with all of the other mixed streams (from other connections). This is a straight-forward mixing together of each of the corresponding channels from each connection.

Speaker Channel Layouts

When channelInterpretation is "speakers" then the up-mixing and down-mixing is defined for specific channel layouts.

Mono (one channel), stereo (two channels), quad (four channels), and 5.1 (six channels) MUST be supported. Other channel layout may be supported in future version of this specification.

Channel ordering

    Mono
      0: M: mono

    Stereo
      0: L: left
      1: R: right
    
  Quad
      0: L:  left
      1: R:  right
      2: SL: surround left
      3: SR: surround right

    5.1
      0: L:   left
      1: R:   right
      2: C:   center
      3: LFE: subwoofer
      4: SL:  surround left
      5: SR:  surround right
  

Up Mixing speaker layouts

Mono up-mix:

    1 -> 2 : up-mix from mono to stereo
        output.L = input;
        output.R = input;

    1 -> 4 : up-mix from mono to quad
        output.L = input;
        output.R = input;
        output.SL = 0;
        output.SR = 0;

    1 -> 5.1 : up-mix from mono to 5.1
        output.L = 0;
        output.R = 0;
        output.C = input; // put in center channel
        output.LFE = 0;
        output.SL = 0;
        output.SR = 0;

Stereo up-mix:

    2 -> 4 : up-mix from stereo to quad
        output.L = input.L;
        output.R = input.R;
        output.SL = 0;
        output.SR = 0;

    2 -> 5.1 : up-mix from stereo to 5.1
        output.L = input.L;
        output.R = input.R;
        output.C = 0;
        output.LFE = 0;
        output.SL = 0;
        output.SR = 0;

Quad up-mix:

    4 -> 5.1 : up-mix from quad to 5.1
        output.L = input.L;
        output.R = input.R;
        output.C = 0;
        output.LFE = 0;
        output.SL = input.SL;
        output.SR = input.SR;

Down Mixing speaker layouts

A down-mix will be necessary, for example, if processing 5.1 source material, but playing back stereo.

  Mono down-mix:

      2 -> 1 : stereo to mono
          output = 0.5 * (input.L + input.R);

      4 -> 1 : quad to mono
          output = 0.25 * (input.L + input.R + input.SL + input.SR);

      5.1 -> 1 : 5.1 to mono
          output = 0.7071 * (input.L + input.R) + input.C + 0.5 * (input.SL + input.SR)


  Stereo down-mix:

      4 -> 2 : quad to stereo
          output.L = 0.5 * (input.L + input.SL);
          output.R = 0.5 * (input.R + input.SR);

      5.1 -> 2 : 5.1 to stereo
          output.L = L + 0.7071 * (input.C + input.SL)
          output.R = R + 0.7071 * (input.C + input.SR)

  Quad down-mix:

      5.1 -> 4 : 5.1 to quad
          output.L = L + 0.7071 * input.C
          output.R = R + 0.7071 * input.C
          output.SL = input.SL
          output.SR = input.SR

  

Channel Rules Examples

  // Set gain node to explicit 2-channels (stereo).
  gain.channelCount = 2;
  gain.channelCountMode = "explicit";
  gain.channelInterpretation = "speakers";

  // Set "hardware output" to 4-channels for DJ-app with two stereo output busses.
  context.destination.channelCount = 4;
  context.destination.channelCountMode = "explicit";
  context.destination.channelInterpretation = "discrete";

  // Set "hardware output" to 8-channels for custom multi-channel speaker array
  // with custom matrix mixing.
  context.destination.channelCount = 8;
  context.destination.channelCountMode = "explicit";
  context.destination.channelInterpretation = "discrete";

  // Set "hardware output" to 5.1 to play an HTMLAudioElement.
  context.destination.channelCount = 6;
  context.destination.channelCountMode = "explicit";
  context.destination.channelInterpretation = "speakers";

  // Explicitly down-mix to mono.
  gain.channelCount = 1;
  gain.channelCountMode = "explicit";
  gain.channelInterpretation = "speakers";
  

Spatialization / Panning

Background

A common feature requirement for modern 3D games is the ability to dynamically spatialize and move multiple audio sources in 3D space. Game audio engines such as OpenAL, FMOD, Creative's EAX, Microsoft's XACT Audio, etc. have this ability.

Using an PannerNode, an audio stream can be spatialized or positioned in space relative to an AudioListener. An AudioContext will contain a single AudioListener. Both panners and listeners have a position in 3D space using a right-handed cartesian coordinate system. The units used in the coordinate system are not defined, and do not need to be because the effects calculated with these coordinates are independent/invariant of any particular units such as meters or feet. PannerNode objects (representing the source stream) have an orientation vector representing in which direction the sound is projecting. Additionally, they have a sound cone representing how directional the sound is. For example, the sound could be omnidirectional, in which case it would be heard anywhere regardless of its orientation, or it can be more directional and heard only if it is facing the listener. AudioListener objects (representing a person's ears) have an orientation and up vector representing in which direction the person is facing. Because both the source stream and the listener can be moving, they both have a velocity vector representing both the speed and direction of movement. Taken together, these two velocities can be used to generate a doppler shift effect which changes the pitch.

During rendering, the PannerNode calculates an azimuth and elevation. These values are used internally by the implementation in order to render the spatialization effect. See the Panning Algorithm section for details of how these values are used.

Azimuth and Elevation

The following algorithm must be used to calculate the azimuth and elevation: for the PannerNode

  // Calculate the source-listener vector.
  vec3 sourceListener = source.position - listener.position;

  if (sourceListener.isZero()) {
      // Handle degenerate case if source and listener are at the same point.
      azimuth = 0;
      elevation = 0;
      return;
  }

  sourceListener.normalize();

  // Align axes.
  vec3 listenerFront = listener.orientation;
  vec3 listenerUp = listener.up;
  vec3 listenerRight = listenerFront.cross(listenerUp);
  listenerRight.normalize();

  vec3 listenerFrontNorm = listenerFront;
  listenerFrontNorm.normalize();

  vec3 up = listenerRight.cross(listenerFrontNorm);

  float upProjection = sourceListener.dot(up);

  vec3 projectedSource = sourceListener - upProjection * up;
  projectedSource.normalize();

  azimuth = 180 * acos(projectedSource.dot(listenerRight)) / PI;

  // Source in front or behind the listener.
  float frontBack = projectedSource.dot(listenerFrontNorm);
  if (frontBack < 0)
      azimuth = 360 - azimuth;

  // Make azimuth relative to "front" and not "right" listener vector.
  if ((azimuth >= 0) && (azimuth <= 270))
      azimuth = 90 - azimuth;
  else
      azimuth = 450 - azimuth;

  elevation = 90 - 180 * acos(sourceListener.dot(up)) / PI;

  if (elevation > 90)
      elevation = 180 - elevation;
  else if (elevation < -90)
      elevation = -180 - elevation;
  

Panning Algorithm

mono->stereo and stereo->stereo panning must be supported. mono->stereo processing is used when all connections to the input are mono. Otherwise stereo->stereo processing is used.

The following algorithms must be implemented:

Distance Effects

Sounds which are closer are louder, while sounds further away are quieter. Exactly how a sound's volume changes according to distance from the listener depends on the distanceModel attribute.

During audio rendering, a distance value will be calculated based on the panner and listener positions according to:

  v = panner.position - listener.position
  
  distance = sqrt(dot(v, v))
  

distance will then be used to calculate distanceGain which depends on the distanceModel attribute. See the distanceModel section for details of how this is calculated for each distance model.

As part of its processing, the PannerNode scales/multiplies the input audio signal by distanceGain to make distant sounds quieter and nearer ones louder.

Sound Cones

The listener and each sound source have an orientation vector describing which way they are facing. Each sound source's sound projection characteristics are described by an inner and outer "cone" describing the sound intensity as a function of the source/listener angle from the source's orientation vector. Thus, a sound source pointing directly at the listener will be louder than if it is pointed off-axis. Sound sources can also be omni-directional.

The following algorithm must be used to calculate the gain contribution due to the cone effect, given the source (the PannerNode) and the listener:

  if (source.orientation.isZero() || ((source.coneInnerAngle == 360) && (source.coneOuterAngle == 360)))
      return 1; // no cone specified - unity gain

  // Normalized source-listener vector
  vec3 sourceToListener = listener.position - source.position;
  sourceToListener.normalize();

  vec3 normalizedSourceOrientation = source.orientation;
  normalizedSourceOrientation.normalize();

  // Angle between the source orientation vector and the source-listener vector
  float dotProduct = sourceToListener.dot(normalizedSourceOrientation);
  float angle = 180 * acos(dotProduct) / PI;
  float absAngle = fabs(angle);

  // Divide by 2 here since API is entire angle (not half-angle)
  float absInnerAngle = fabs(source.coneInnerAngle) / 2;
  float absOuterAngle = fabs(source.coneOuterAngle) / 2;
  float gain = 1;

  if (absAngle <= absInnerAngle)
      // No attenuation
      gain = 1;
  else if (absAngle >= absOuterAngle)
      // Max attenuation
      gain = source.coneOuterGain;
  else {
      // Between inner and outer cones
      // inner -> outer, x goes from 0 -> 1
      float x = (absAngle - absInnerAngle) / (absOuterAngle - absInnerAngle);
      gain = (1 - x) + source.coneOuterGain * x;
  }

  return gain;
  

Doppler Shift

The following algorithm must be used to calculate the doppler shift value which is used as an additional playback rate scalar for all AudioBufferSourceNodes connecting directly or indirectly to the PannerNode:

  float dopplerShift = 1; // Initialize to default value
  float dopplerFactor = listener.dopplerFactor;

  if (dopplerFactor > 0) {
      float speedOfSound = listener.speedOfSound;

      // Don't bother if both source and listener have no velocity.
      if (!source.velocity.isZero() || !listener.velocity.isZero()) {
          // Calculate the source to listener vector.
          vec3 sourceToListener = source.position - listener.position;

          float sourceListenerMagnitude = sourceToListener.length();

          float listenerProjection = sourceToListener.dot(listener.velocity) / sourceListenerMagnitude;
          float sourceProjection = sourceToListener.dot(source.velocity) / sourceListenerMagnitude;

          listenerProjection = -listenerProjection;
          sourceProjection = -sourceProjection;

          float scaledSpeedOfSound = speedOfSound / dopplerFactor;
          listenerProjection = min(listenerProjection, scaledSpeedOfSound);
          sourceProjection = min(sourceProjection, scaledSpeedOfSound);

          dopplerShift = ((speedOfSound - dopplerFactor * listenerProjection) / (speedOfSound - dopplerFactor * sourceProjection));
          fixNANs(dopplerShift); // Avoid illegal values

          // Limit the pitch shifting to 4 octaves up and 3 octaves down.
          dopplerShift = min(dopplerShift, 16);
          dopplerShift = max(dopplerShift, 0.125);
      }
  }
  

Linear Effects using Convolution

Background

Convolution is a mathematical process which can be applied to an audio signal to achieve many interesting high-quality linear effects. Very often, the effect is used to simulate an acoustic space such as a concert hall, cathedral, or outdoor amphitheater. It can also be used for complex filter effects, like a muffled sound coming from inside a closet, sound underwater, sound coming through a telephone, or playing through a vintage speaker cabinet. This technique is very commonly used in major motion picture and music production and is considered to be extremely versatile and of high quality.

Each unique effect is defined by an impulse response. An impulse response can be represented as an audio file and can be recorded from a real acoustic space such as a cave, or can be synthetically generated through a great variety of techniques.

Motivation for use as a Standard

A key feature of many game audio engines (OpenAL, FMOD, Creative's EAX, Microsoft's XACT Audio, etc.) is a reverberation effect for simulating the sound of being in an acoustic space. The code used to generate this effect has generally been custom and algorithmic (generally using a hand-tweaked set of delay lines and allpass filters which feedback into each other). In nearly all cases, not only is the implementation custom, but the code is proprietary and closed-source, each company adding its own "black magic" to achieve its unique quality. Each implementation being custom with a different set of parameters makes it impossible to achieve a uniform desired effect. And the code being proprietary makes it impossible to adopt a single one of the implementations as a standard. Additionally, algorithmic reverberation effects are limited to a relatively narrow range of different effects, regardless of how the parameters are tweaked.

A convolution effect solves these problems by using a very precisely defined mathematical algorithm as the basis of its processing. An impulse response represents an exact sound effect to be applied to an audio stream and is easily represented by an audio file which can be referenced by URL. The range of possible effects is enormous.

Implementation Guide

Linear convolution can be implemented efficiently. Here are some notes describing how it can be practically implemented.

Reverb Effect (with matrixing)

This section is normative.

In the general case the source has N input channels, the impulse response has K channels, and the playback system has M output channels. Thus it's a matter of how to matrix these channels to achieve the final result.

The subset of N, M, K below must be implemented (note that the first image in the diagram is just illustrating the general case and is not normative, while the following images are normative). Without loss of generality, developers desiring more complex and arbitrary matrixing can use multiple ConvolverNode objects in conjunction with an ChannelMergerNode.

Single channel convolution operates on a mono audio input, using a mono impulse response, and generating a mono output. To achieve a more spacious sound, 2 channel audio inputs and 1, 2, or 4 channel impulse responses will be considered. The following diagram, illustrates the common cases for stereo playback where N and M are 1 or 2 and K is 1, 2, or 4.

reverb matrixing
A graphical representation of the different input and output channel count possibilities when using a ConvolverNode.

Performance Considerations

Latency

latency
Use cases in which the latency can be important

For web applications, the time delay between mouse and keyboard events (keydown, mousedown, etc.) and a sound being heard is important.

This time delay is called latency and is caused by several factors (input device latency, internal buffering latency, DSP processing latency, output device latency, distance of user's ears from speakers, etc.), and is cummulative. The larger this latency is, the less satisfying the user's experience is going to be. In the extreme, it can make musical production or game-play impossible. At moderate levels it can affect timing and give the impression of sounds lagging behind or the game being non-responsive. For musical applications the timing problems affect rhythm. For gaming, the timing problems affect precision of gameplay. For interactive applications, it generally cheapens the users experience much in the same way that very low animation frame-rates do. Depending on the application, a reasonable latency can be from as low as 3-6 milliseconds to 25-50 milliseconds.

Audio buffer copying

When an acquire the content operation is performed on an AudioBuffer, the entire operation can usually be implemented without copying channel data. In particular, the last step should be performed lazily at the next getChannelData call. That means a sequence of consecutive acquire the contents operations with no intervening getChannelData (e.g. multiple AudioBufferSourceNodes playing the same AudioBuffer) can be implemented with no allocations or copying.

Implementations can perform an additional optimization: if getChannelData is called on an AudioBuffer, fresh ArrayBuffers have not yet been allocated, but all invokers of previous acquire the content operations on an AudioBuffer have stopped using the AudioBuffer's data, the raw data buffers can be recycled for use with new AudioBuffers, avoiding any reallocation or copying of the channel data.

Audio Glitching

Audio glitches are caused by an interruption of the normal continuous audio stream, resulting in loud clicks and pops. It is considered to be a catastrophic failure of a multi-media system and must be avoided. It can be caused by problems with the threads responsible for delivering the audio stream to the hardware, such as scheduling latencies caused by threads not having the proper priority and time-constraints. It can also be caused by the audio DSP trying to do more work than is possible in real-time given the CPU's speed.

Hardware Scalability

The system should gracefully degrade to allow audio processing under resource constrained conditions without dropping audio frames.

First of all, it should be clear that regardless of the platform, the audio processing load should never be enough to completely lock up the machine. Second, the audio rendering needs to produce a clean, un-interrupted audio stream without audible glitches.

The system should be able to run on a range of hardware, from mobile phones and tablet devices to laptop and desktop computers. The more limited compute resources on a phone device make it necessary to consider techniques to scale back and reduce the complexity of the audio rendering. For example, voice-dropping algorithms can be implemented to reduce the total number of notes playing at any given time.

Here's a list of some techniques which can be used to limit CPU usage:

CPU monitoring

In order to avoid audio breakup, CPU usage must remain below 100%.

The relative CPU usage can be dynamically measured for each AudioNode (and chains of connected nodes) as a percentage of the rendering time quantum. In a single-threaded implementation, overall CPU usage must remain below 100%. The measured usage may be used internally in the implementation for dynamic adjustments to the rendering. It may also be exposed through a cpuUsage attribute of AudioNode for use by JavaScript.

In cases where the measured CPU usage is near 100% (or whatever threshold is considered too high), then an attempt to add additional AudioNodes into the rendering graph can trigger voice-dropping.

Voice Dropping

Voice-dropping is a technique which limits the number of voices (notes) playing at the same time to keep CPU usage within a reasonable range. There can either be an upper threshold on the total number of voices allowed at any given time, or CPU usage can be dynamically monitored and voices dropped when CPU usage exceeds a threshold. Or a combination of these two techniques can be applied. When CPU usage is monitored for each voice, it can be measured all the way from a source node through any effect processing nodes which apply uniquely to that voice.

When a voice is "dropped", it needs to happen in such a way that it doesn't introduce audible clicks or pops into the rendered audio stream. One way to achieve this is to quickly fade-out the rendered audio for that voice before completely removing it from the rendering graph.

When it is determined that one or more voices must be dropped, there are various strategies for picking which voice(s) to drop out of the total ensemble of voices currently playing. Here are some of the factors which can be used in combination to help with this decision:

  • Older voices, which have been playing the longest can be dropped instead of more recent voices.
  • Quieter voices, which are contributing less to the overall mix may be dropped instead of louder ones.
  • Voices which are consuming relatively more CPU resources may be dropped instead of less "expensive" voices.
  • An AudioNode can have a priority attribute to help determine the relative importance of the voices.

Simplification of Effects Processing

Most of the effects described in this document are relatively inexpensive and will likely be able to run even on the slower mobile devices. However, the convolution effect can be configured with a variety of impulse responses, some of which will likely be too heavy for mobile devices. Generally speaking, CPU usage scales with the length of the impulse response and the number of channels it has. Thus, it is reasonable to consider that impulse responses which exceed a certain length will not be allowed to run. The exact limit can be determined based on the speed of the device. Instead of outright rejecting convolution with these long responses, it may be interesting to consider truncating the impulse responses to the maximum allowed length and/or reducing the number of channels of the impulse response.

In addition to the convolution effect. The PannerNode may also be expensive if using the HRTF panning model. For slower devices, a cheaper algorithm such as EQUALPOWER can be used to conserve compute resources.

Sample Rate

For very slow devices, it may be worth considering running the rendering at a lower sample-rate than normal. For example, the sample-rate can be reduced from 44.1KHz to 22.05KHz. This decision must be made when the AudioContext is created, because changing the sample-rate on-the-fly can be difficult to implement and will result in audible glitching when the transition is made.

Pre-flighting

It should be possible to invoke some kind of "pre-flighting" code (through JavaScript) to roughly determine the power of the machine. The JavaScript code can then use this information to scale back any more intensive processing it may normally run on a more powerful machine. Also, the underlying implementation may be able to factor in this information in the voice-dropping algorithm.

TODO: add specification and more detail here

Authoring for different user agents

JavaScript code can use information about user-agent to scale back any more intensive processing it may normally run on a more powerful machine.

Scalability of Direct JavaScript Synthesis / Processing

Any audio DSP / processing code done directly in JavaScript should also be concerned about scalability. To the extent possible, the JavaScript code itself needs to monitor CPU usage and scale back any more ambitious processing when run on less powerful devices. If it's an "all or nothing" type of processing, then user-agent check or pre-flighting should be done to avoid generating an audio stream with audio breakup.

JavaScript Issues with real-time Processing and Synthesis:

While processing audio in JavaScript, it is extremely challenging to get reliable, glitch-free audio while achieving a reasonably low-latency, especially under heavy processor load.
  • JavaScript is very much slower than heavily optimized C++ code and is not able to take advantage of SSE optimizations and multi-threading which is critical for getting good performance on today's processors. Optimized native code can be on the order of twenty times faster for processing FFTs as compared with JavaScript. It is not efficient enough for heavy-duty processing of audio such as convolution and 3D spatialization of large numbers of audio sources.
  • setInterval() and XHR handling will steal time from the audio processing. In a reasonably complex game, some JavaScript resources will be needed for game physics and graphics. This creates challenges because audio rendering is deadline driven (to avoid glitches and get low enough latency).
  • JavaScript does not run in a real-time processing thread and thus can be pre-empted by many other threads running on the system.
  • Garbage Collection (and autorelease pools on Mac OS X) can cause unpredictable delay on a JavaScript thread.
  • Multiple JavaScript contexts can be running on the main thread, stealing time from the context doing the processing.
  • Other code (other than JavaScript) such as page rendering runs on the main thread.
  • Locks can be taken and memory is allocated on the JavaScript thread. This can cause additional thread preemption.
The problems are even more difficult with today's generation of mobile devices which have processors with relatively poor performance and power consumption / battery-life issues.

Security Considerations

Privacy Considerations

When giving various information on available AudioNodes, the Web Audio API potentially exposes information on characteristic features of the client (such as audio hardware sample-rate) to any page that makes use of the AudioNode interface. Additionally, timing information can be collected through the AnalyserNode or ScriptProcessorNode interface. The information could subsequently be used to create a fingerprint of the client.

Currently audio input is not specified in this document, but it will involve gaining access to the client machine's audio input or microphone. This will require asking the user for permission in an appropriate way, probably via the getUserMedia() API.

Requirements and Use Cases

Please see Example Applications

Acknowledgements

This specification is the collective work of the W3C Audio Working Group.

Members of the Working Group are (at the time of writing, and by alphabetical order):
Adenot, Paul (Mozilla Foundation); Akhgari, Ehsan (Mozilla Foundation); Berkovitz, Joe (Invited Expert); Bossart, Pierre (Intel Corporation); Carlson, Eric (Apple, Inc.); Geelnard, Marcus (Opera Software); Goode, Adam (Google, Inc.); Gregan, Matthew (Mozilla Foundation); Jägenstedt, Philip (Opera Software); Kalliokoski, Jussi (Invited Expert); Lilley, Chris (W3C Staff); Lowis, Chris (Invited Expert. WG co-chair from December 2012 to September 2013, affiliated with British Broadcasting Corporation); Mandyam, Giridhar (Qualcomm Innovation Center, Inc); Noble, Jer (Apple, Inc.); O'Callahan, Robert(Mozilla Foundation); Onumonu, Anthony (British Broadcasting Corporation); Paradis, Matthew (British Broadcasting Corporation); Raman, T.V. (Google, Inc.); Schepers, Doug (W3C/MIT); Shires, Glen (Google, Inc.); Smith, Michael (W3C/Keio); Thereaux, Olivier (British Broadcasting Corporation) – WG Chair; Verdie, Jean-Charles (MStar Semiconductor, Inc.); Wilson, Chris (Google,Inc.); ZERGAOUI, Mohamed (INNOVIMAX)

Former members of the Working Group and contributors to the specification include:
Caceres, Marcos (Invited Expert); Cardoso, Gabriel (INRIA); Chen, Bin (Baidu, Inc.); MacDonald, Alistair (W3C Invited Experts) — WG co-chair from March 2011 to July 2012; Michel, Thierry (W3C/ERCIM); Rogers, Chris (Google, Inc.) – Specification Editor until August 2013; Wei, James (Intel Corporation);

Web Audio API Change Log

See changelog.html.