From ddfc2fdb6c04ef7c28ef8ab9cf253c206704c32f Mon Sep 17 00:00:00 2001 From: Hongchan Choi Date: Thu, 26 Sep 2024 14:43:20 -0700 Subject: [PATCH 01/10] initial commit --- index.bs | 274 +------------------------------------------------------ 1 file changed, 3 insertions(+), 271 deletions(-) diff --git a/index.bs b/index.bs index 1e38c5726..c849daa32 100644 --- a/index.bs +++ b/index.bs @@ -1,11 +1,11 @@ 
 Title: Web Audio API
 Shortname: webaudio
-Level: 1
+Level: 1.1
 Group: audiowg
-Status: ED
+Status: FPWD
 ED: https://webaudio.github.io/web-audio-api/
-TR: https://www.w3.org/TR/webaudio/
+TR: https://www.w3.org/TR/webaudio-11/
 Favicon: favicon.png
 Previous Version: https://www.w3.org/TR/2021/CR-webaudio-20210114/
 Previous Version: https://www.w3.org/TR/2020/CR-webaudio-20200611/
@@ -14175,274 +14175,6 @@ class Vec3 {
 
 

 Change Log

-

-  Changes since Recommendation of 17 Jun 2021
-

-  * Issue 2450: Allow specifying a different render-quantum-size for an AudioContext or OfflineAudioContext
-    

-    

-  * Issue 2512: Specify AnalyserNode treatment of non-fininte values.
-    

-    

-  * Issue 2456: Add a MessagePort to the AudioWorkletGlobalScope
-    

-    

-  * Issue 2444: Add AudioRenderCapacity interface and related classes
-    

-    

-  * Issue 2361: Use new Web IDL buffer primitives
-    

-    

-  * Issue 127: RangeError is thrown only for negative cancelTime
-    

-    

-  * Issue 2359: Fix a typo in the VUMeterNode example
-    

-    

-  * Issue 2373: Fix a typo: initially instead of initialy
-    

-    

-  * Issue 2321: Warn about corrupted fils in decodeAudioData.
-    

-    

-  * Issue 2375: decodeAudioData only decodes the first track of a multi-track audio file.
-    

-    

-  * Issue
-        2475: Add definition for various concepts related to system-level audio callbacks
-    

-    

-    * Issue 2400
-        : Access to a different output device: AudioContext.setSinkId()
-        

-
-

-  Since Proposed Recommendation of 6 May 2021
-

-  * Styling, status and boilerplate updates for Recommendation
-  * Updated links to RFC2119, High-Resolution Time, and Audio EQ Cookbook
-  * A spelling error was corrected
-
-

-  Since Candidate Recommendation of 14 January 2021
-

-
- * PR 2333: Update links to point to W3C versions
- * PR 2334: Use bikeshed to link to ErrorEvent
- * PR 2331: Add MIMESniff to normative references
- * PR 2328: MediaStream must be resampled to match the context sample rate
- * PR 2318: Restore empty of pending processor construction data after successful initialization of AudioWorkProcessor#port.
- * PR 2317: Standardize h3/h4 interface and dictionary markup
- * PR 2312: Rework description of control thread state and rendering thread state
- * PR 2311: Adjust the steps to process a context's regular task queue
- * PR 2310: OscillatorNode output is mono
- * PR 2308: Refine phrasing for "allowed to start"
- * PR 2307: Replace "queue a task" with "queue a media element task"
- * PR 2306: Move some steps from AudioWorkletProcessor constructor to the instantiation algorithm
- * PR 2304: Add required components for ES operations in the rendering loop
- * PR 2301: Define when and how regular tasks are processed wrt the processing model
- * PR 2286: Clean up ABSN start algorithm
- * PR 2277: Fix compression curve diagram
- * PR 2273: Clarify units used in threshold & knee value calculations
- * PR 2256: ABSN extrapolates the last output
- * PR 2250: Use FrozenArray for AudioWorkletProcessor process()
- * PR 2298: Bikeshed HTML validation issues
- 
-

-  Since Candidate Recommendation of 11 June 2020
-

-
- * PR 2202: Fixed wrong optionality of IIRFilterNode options
- * Issue 2191: Restrict sounds beyond normal hearing
- * PR 2210: Return rejected promise when the document is not fully active, for operations returning promises
- * Issue 2191: Destination of request created by `addModule`
- * Issue 2213: The message queue is for message running on the rendering thread.
- * Issue 2216: Use inclusive language in the spec
- * PR 2219: Update more terminology in images and markdown documents
- * Issue 2206: PannerNode.rollOffFactor with "linear" distance model is not clamped to [0, 1] in main browser engines
- * Issue 2169: AudioParamDescriptor has member constraints that are redundant
- * Issue 1457: [privacy] Exposing data to an origin: fingerprinting
- * Issue 2061: Privacy re-review of latest changes
- * Issue 2225: Describe "Planar versus interleaved buffers"
- * Issue 2231: WaveShaper [[curve set]] not defined
- * Issue 2240: Align with Web IDL specification
- * Issue 2242: LInk to undefined instead of using ``
- * Issue 2227: Clarify buffer.copyToChannel() must be called before source.buffer = buffer else nothing is played
- * PR 2253: Fix duplicated IDs for decode callbacks
- * Issue 2252: When are promises in "[[pending resume promises]]" resolved?
- * PR 2266: Prohibit arbitrary termination of AudioWorkletGlobalScopes
-
-

- Since Candidate Recommendation of 18 September 2018
-

-
-* Issue 2193: Incorrect azimuth comparison in spatialization algorithm
-* Issue 2192: Waveshaper curve interpolation algorithm incorrect
-* Issue 2171: Allow not having get parameterDescriptors in an AudioWorkletProcessor
-* Issue 2184: PannerNode refDistance description unclear
-* Issue 2165: AudioScheduledSourceNode start algorithm incomplete
-* Issue 2155: Restore changes accidentally reverted in bikeshed conversion
-* Issue 2154: Exception for changing channelCountMode on ScriptProcessorNode does not match browsers
-* Issue 2153: Exception for changing channelCount on ScriptProcessorNode does not match browsers
-* Issue 2152: close() steps don't make sense
-* Issue 2150: AudioBufferOptions requires throwing NotFoundError in cases that can't happen
-* Issue 2149: MediaStreamAudioSourceNode constructor has weird check for AudioContext
-* Issue 2148: IIRFilterOptions description makes impossible demands
-* Issue 2147: PeriodicWave constructor examines lengths of things that might not be there
-* Issue 2113: BiquadFilter gain lower bound can be lower.
-* Issue 2096: Lifetime of pending processor construction data and exceptions in instantiation of AudioWorkletProcessor
-* Issue 2087: Minor issues with BiquadFilter AudioParams
-* Issue 2083: Missing text in WaveShaperNode?
-* Issue 2082: WaveShaperNode curve interpolation incomplete
-* Issue 2074: Should the AudioWorkletNode constructor invoke the algorithm for initializing an object that inherits from AudioNode?
-* Issue 2073: Inconsistencies in constructor descriptions and factory method initialization
-* Issue 2072: Clarification on `AudioBufferSourceNode` looping, and loop points
-* Issue 2071: cancelScheduledValues with setValueCurveAtTime
-* Issue 2060: Would it be helpful to restrict use of `AudioWorkletProcessor.port().postMessage()` in order to facilitate garbage collection?
-* Issue 2051: Update to constructor operations
-* Issue 2050: Restore ConvolverNode channel mixing configurability (up to 2 channels)
-* Issue 2045: Should the check on `process()` be removed from `AudioWorkletGlobalScope.registerProcessor()`?
-* Issue 2044: Remove `options` parameter from `AudioWorkletProcessor` constructor WebIDL
-* Issue 2036: Remove `options` parameter of `AudioWorkletProcessor` constructor
-* Issue 2035: De-duplicate initial value setting on AudioWorkletNode AudioParams
-* Issue 2027: Revise "processor construction data" algorithm
-* Issue 2021: AudioWorkletProcessor constructor leads to infinite recursion
-* Issue 2018: There are still issues with the setup of an AudioWorkletNode's parameters
-* Issue 2016: Clarify `parameters` in AudioWorkletProcessor.process()
-* Issue 2011: AudioWorkletNodeOptions.processorOptions should not default to null.
-* Issue 1989: Please update to Web IDL changes to optional dictionary defaulting
-* Issue 1984: Handling of exceptions in audio worklet is not very clear
-* Issue 1976: AudioWorkletProcessor's [[node reference]] seems to be write-only
-* Issue 1972: parameterDescriptors handling during AudioWorkletNode initialization is probably wrong
-* Issue 1971: AudioWorkletNode options serialization is underdefined
-* Issue 1970: "active source" flag handling is a weird monkeypatch
-* Issue 1969: It would be clearer if the various validation of AudioWorkletNodeOptions were an explicit step or set of steps
-* Issue 1966: parameterDescriptors is not looked up by the AudioWorkletProcessor constructor
-* Issue 1963: NewTarget check for AudioWorkletProcessor isn't actually possible with a Web IDL constructor
-* Issue 1947: Spec is inconsistent about whether parameterDescriptors is an array or an iterable
-* Issue 1946: Population of "node name to parameter descriptor map" needs to be defined
-* Issue 1945: registerProcessor is doing odd things with threads and JS values
-* Issue 1943: Describe how WaveShaperNode shapes the input with the curve
-* Issue 1935: length of AudioWorkletProcessor.process() parameter sequences with inactive inputs
-* Issue 1932: Make AudioWorkletNode output buffer available for reading
-* Issue 1925: front vs forward
-* Issue 1902: Mixer Gain Structure section not needed
-* Issue 1906: Steps in rendering algorithm
-* Issue 1905: Rendering callbacks are observable
-* Issue 1904: Strange Note in algorithm for swapping a control message queue
-* Issue 1903: Funny sentence about priority and latency
-* Issue 1901: AudioWorkletNode state property?
-* Issue 1900: AudioWorkletProcessor NewTarget undefined
-* Issue 1899: Missing synchronous markers
-* Issue 1897: WaveShaper curve value setter allows multiple sets
-* Issue 1896: WaveShaperNode constructor says curve set is initialized to false
-* Issue #1471: AudioNode Lifetime section seems to attempt to make garbage collection observable
-* Issue #1893: Active processing for Panner/Convolver/ChannelMerger
-* Issue #1894: Funny text in PannerNode.orientationX
-* Issue #1866: References to garbage collection
-* Issue #1851: Parameter values used for BiquadFilterNode::getFrequencyResponse
-* Issue #1905: Rendering callbacks are observable
-* Issue #1879: ABSN playback algorithm offset
-* Issue #1882: Biquad lowpass/highpass Q
-* Issue #1303: MediaElementAudioSourceNode information in a funny place
-* Issue #1896: WaveShaperNode constructor says curve set is initialized to false
-* Issue #1897: WaveShaper curve value setter allows multiple sets.
-* Issue #1880: setOrientation description has confusing paragraph
-* Issue #1855: createScriptProcessor parameter requirements
-* Issue #1857: Fix typos and bad phrasing
-* Issue #1788: Unclear what value is returned by AudioParam.value
-* Issue #1852: Fix error condition of AudioNode.disconnect(destinationNode, output, input)
-* Issue #1841: Recovering from unstable biquad filters?
-* Issue #1777: Picture of the coordinate system for panner node
-* Issue #1802: Clarify interaction between user-invoked suspend and autoplay policy
-* Issue #1822: OfflineAudioContext.suspend can suspend before the given time
-* Issue #1772: Sorting tracks alphabetically is underspecified
-* Issue #1797: Specification is incomplete for AudioNode.connect()
-* Issue #1805: Exception ordering on error
-* Issue #1790: Automation example chart has an error (reversed function arguments
-* Fix rendering algorithm iteration and cycle breaking
-* Issue #1719: channel count changes in filter nodes with tail time
-* Issue #1563: Make decodeAudioData more precise
-* Issue #1481: Tighten spec on ABSN output channels?
-* Issue #1762: Setting convolver buffer more than once?
-* Issue #1758: Explicitly include time-domain processing code for BiquadFilterNode
-* Issue #1770: Link to correct algorithm for StereoPannerNode, mention algorithm is equal-power
-* Issue #1753: Have a single `AudioWorkletGlobalScope` per `BaseAudioContext`
-* Issue #1746: AnalyserNode: Clarify how much time domain data we're supposed to keep around
-* Issue #1741: Sample rate of AudioBuffer
-* Issue #1745: Clarify unit of fftSize
-* Issue #1743: Missing normative reference to Fetch
-* Use "get a reference to the bytes" algorithm as needed.
-* Specify rules for determining output chanel count.
-* Clarified rendering algorithm for AudioListener.
-    

-    Since Working Draft of 19 June 2018
-    

-* Minor editorial clarifications.
-* Update implementation-report.html.
-* Widen the valid range of detune values so that any value that doesn't cause 2^(d/1200) to overflow is valid.
-* PannerNode constructor throws errors.
-* Rephrase algorithm for setting buffer and curve.
-* Refine startRendering algorithm.
-* Make "queue a task" link to the HTML spec.
-* Specify more precisely, events overlapping with SetValueCurveAtTime.
-* Add implementation report to gh-pages.
-* Honor the given value in `outputChannelCount`.
-* Initialize bufferDuration outside of process() in ABSN algorithm.
-* Rework definition of ABSN output behavior to account for playbackRate’s interaction with the start(…duration) argument.
-* Add mention of video element in ultrasonic attack surface.
-
-

-Since Working Draft of 08 December 2015

-* Add AudioWorklet and related interfaces to support custom nodes. This replaces ScriptProcessorNode, which is now deprecated.
-* Explicitly say what the channel count, mode, and interpretation values are for all source nodes.
-* Specify the behavior of Web Audio when a document is unloaded.
-* Merge the proposed SpatialListener interface into AudioListener.
-* Rework and clean up algorithms for panning and spatialization and define "magic functions".
-* Clarify that AudioBufferSourceNode looping is limited by duration argument to start().
-* Add constructors with options dictionaries for all node types.
-* Clarify parameter automation method behavior and equations. Handle cases where automation methods may interact with each other.
-* Support latency hints and arbitrary sample rates in AudioContext constructor.
-* Clear up ambiguities in definitions of start() and stop() for scheduled sources.
-* Remove automatic dezippering from AudioParam value setters which now equate to setValueAtTime().
-* Specify normative behavior of DynamicsCompressorNode.
-* Specify that AudioParam.value returns the most recent computed value.
-* Permit AudioBufferSourceNode to specify sub-sample start, duration, loopStart and loopEnd. Respecify algorithms to say exactly how looping works in all scenarios, including dynamic and negative playback rates.
-* Harmonized behavior of IIRFilterNode with BiquadFilterNode.
-* Add diagram describing mono-input-to-matrixed-stereo case.
-* Prevent connecting an AudioNode to an AudioParam of a different AudioContext.
-* Added Audioparam cancelAndHoldAtTime
-* Clarify behaviour of AudioParam.cancelScheduledValues().
-* Add playing reference to MediaElementAudioSourceNodes and MediaStreamAudioSourceNodes.
-* Refactor BaseAudioContext interface out of AudioContext, OfflineAudioContext.
-* OfflineAudioContext inherits from BaseAudioContext, not AudioContext.
-* "StereoPanner" replaced with the correct "StereoPannerNode".
-* Support chaining on AudioNode.connect() and AudioParam automation methods.
-* Specify behavior of events following SetTarget events.
-* Reinstate channelCount declaration for AnalyserNode.
-* Specify exponential ramp behavior when previous value is 0.
-* Specify behavior of setValueCurveAtTime parameters.
-* Add spatialListener attribute to AudioContext.
-* Remove section titled "Doppler Shift".
-* Added a list of nodes and reason why they can add latency, in an informative section.
-* Speced nominal ranges, nyquist, and behavior when outside the range.
-* Spec the processing model for the Web Audio API.
-* Merge the SpatialPannerNode into the PannerNode, undeprecating the PannerNode.
-* Merge the SpatialListener into the AudioListener, undeprecating the AudioListener.
-* Added latencyHint(s).
-* Move the constructor from BaseAudioContext to AudioContext where it belongs; BaseAudioContext is not constructible.
-* Specified the Behavior of automations and nominal ranges.
-* The playbackRate is widened to +/- infinity.
-* setValueCurveAtTime is modified so that an implicit call to setValueAtTime is made at the end of the curve duration.
-* Make setting the `value` attribute of an `AudioParam` strictly equivalent of calling setValueAtTime with AudioContext.currentTime.
-* Add new sections for AudioContextOptions and AudioTimestamp.
-* Add constructor for all nodes.
-* Define ConstantSourceNode.
-* Make the WaveShaperNode have a tail time, depending on the oversampling level.
-* Allow collecting MediaStreamAudioSourceNode or MediaElementAudioSourceNode when they won't play ever again.
-* Add a concept of 'allowed to start' and use it when creating an AudioContext and resuming it from resume() (closes #836).
-* Add AudioScheduledSourceNode base class for source nodes.
-* Mark all AudioParams as being k-rate.
 
 

 Acknowledgements


From bdddb8b5d209db12751ed53f729773a356a8f5b3 Mon Sep 17 00:00:00 2001
From: Hongchan Choi 
Date: Thu, 26 Sep 2024 15:59:21 -0700
Subject: [PATCH 02/10] first pass

---
 index.bs | 1762 ++++++++++++++++--------------------------------------
 1 file changed, 511 insertions(+), 1251 deletions(-)

diff --git a/index.bs b/index.bs
index c849daa32..ed8479a2a 100644
--- a/index.bs
+++ b/index.bs
@@ -17,8 +17,8 @@ Previous Version: https://www.w3.org/TR/2012/WD-webaudio-20121213/
 Previous Version: https://www.w3.org/TR/2012/WD-webaudio-20120802/
 Previous Version: https://www.w3.org/TR/2012/WD-webaudio-20120315/
 Previous Version: https://www.w3.org/TR/2011/WD-webaudio-20111215/
-Editor: Paul Adenot, Mozilla (https://www.mozilla.org/), padenot@mozilla.com, w3cid 62410
-Editor: Hongchan Choi, Google (https://www.google.com/), hongchan@google.com, w3cid 74103
+Editor: Paul Adenot, Mozilla, padenot@mozilla.com, w3cid 62410
+Editor: Hongchan Choi, Google, hongchan@google.com, w3cid 74103
 Former Editor: Raymond Toy (until Oct 2018)
 Former Editor: Chris Wilson (Until Jan 2016)
 Former Editor: Chris Rogers (Until Aug 2013)
@@ -98,76 +98,6 @@ window.addEventListener('DOMContentLoaded', () => {
 });
 
 
-