Media Source Extensions™

There are distinct mechanisms for attaching aMediaSourceto a media element depending on where theMediaSourceobject was constructed, in aWindowversus in aDedicatedWorkerGlobalScope:

• Attaching aMediaSourcethat was constructed in aWindowcan be done by assigning aMediaSource object URLfor thatMediaSourceto the media elementsrcattribute or the src attribute of a <source> inside a media element. AMediaSource object URLis created by passing a MediaSource object tocreateObjectURL().

  Though implementationsMAYallowMediaSource object URLcreation in a DedicatedWorkerGlobalScopefor aMediaSourceconstructed in that worker, attempting to use thatMediaSource object URLto attach to a media element using either thesrcattribute or the src attribute of a <source> inside a media elementMUSTfail in the media element'sresource fetch algorithm,as extended below.

  Note

  Extending the object URL attachment mechanism to worker MediaSource object URLs would further propagate this idiom that is less preferred versus using srcObject, and would unnecessarily increase user agent interoperability risk and implementation complexity.

• Attaching aMediaSourcethat was constructed in a DedicatedWorkerGlobalScopecan only be done by obtaining a handle from it using handle,transferring thatMediaSourceHandleto theWindow context and assigning it to the media elementsrcObjectattribute. For the purposes of aligning this specification withHTMLMediaElementresource loading and fetching algorithms, the underlyingDedicatedWorkerGlobalScope MediaSourceis the MediaSource object mentioned there, and the MediaSourceHandleobject is the media provider object.

If theresource fetch algorithmwas invoked with a media provider object that is a MediaSourceobject, aMediaSourceHandleobject or a URL record whose object is aMediaSourceobject, then let mode be local, skip the first step in theresource fetch algorithm(which may otherwise set mode to remote) and continue the execution of theresource fetch algorithm.

At the beginning of the "Otherwise (mode is local)" section of theresource fetch algorithm,execute the additional steps, below.

1. If theresource fetch algorithmwas invoked with a media provider object that is aMediaSourceobject, aMediaSourceHandleobject or a URL record whose object is aMediaSourceobject, then:

   If the media provider object is a URL record whose object is aMediaSource that was constructed in aDedicatedWorkerGlobalScope,such as would occur if attempting to use aMediaSource object URLfrom a DedicatedWorkerGlobalScopeMediaSource

   Run the"If the media data cannot be fetched at all, due to network errors, causing the user agent to give up trying to fetch the resource"steps of theresource fetch algorithm'smedia data processing steps list.

   Note

   This prevents usingMediaSource object URLsfor DedicatedWorker MediaSource attachments. TransferringMediaSource'shandlefrom the DedicatedWorker to the Window context and assigning it to the media element's srcObjectattribute is the only way to attach such a MediaSource.

   If the media provider object is aMediaSourceHandlewhose [[Detached]]internal slot is true

   Run the"If the media data cannot be fetched at all, due to network errors, causing the user agent to give up trying to fetch the resource"steps of theresource fetch algorithm'smedia data processing steps list.

   If the media provider object is aMediaSourceHandlewhose underlying MediaSource's[[has ever been attached]]internal slot is true

   Run the"If the media data cannot be fetched at all, due to network errors, causing the user agent to give up trying to fetch the resource"steps of theresource fetch algorithm'smedia data processing steps list.

   Note

   This prevents loading an underlyingMediaSourcemore than once using a MediaSourceHandle,even if theMediaSourcewas constructed on Windowand had been loaded previously using aMediaSource object URL. This doesn't preclude subsequent use of aMediaSource object URLfor a WindowMediaSourcefrom succeeding though.

   IfreadyStateis NOT set to "closed"

   Run the"If the media data cannot be fetched at all, due to network errors, causing the user agent to give up trying to fetch the resource"steps of theresource fetch algorithm'smedia data processing steps list.

   Otherwise

   1. Set theMediaSource's[[has ever been attached]] internal slot to true.
   2. Set the media element'sdelaying-the-load-event-flagto false.

   3. If theMediaSourcewas constructed in a DedicatedWorkerGlobalScope,then setup worker attachment communication and open theMediaSource:

      1. Set[[channel with worker]]to be a new MessageChannel.
      2. Set[[port to worker]]to the port1value of[[channel with worker]].
      3. ExecuteStructuredSerializeWithTransferwith the port2of[[channel with worker]]as both the value and the sole member of thetransferList, and let the result beserialized port2.
      4. Queue a taskon theMediaSource's DedicatedWorkerGlobalScopethat will
         1. ExecuteStructuredDeserializeWithTransferwithserialized port2andDedicatedWorkerGlobalScope'srealm,and set[[port to main]]to be the resulting deserialized clone of the transferred port2value of[[channel with worker]].
         2. Set thereadyStateattribute to "open".
         3. Queue a tasktofire an eventnamedsourceopenat theMediaSource.

      Otherwise, theMediaSourcewas constructed in aWindow:

      1. Set[[channel with worker]]null.
      2. Set[[port to worker]]null.
      3. Set[[port to main]]null.
      4. Set thereadyStateattribute to "open".
      5. Queue a tasktofire an eventnamedsourceopenat the MediaSource.

   4. Continue theresource fetch algorithmby running the remaining "Otherwise (mode is local)"steps, with these requirements:
      1. Text in theresource fetch algorithmor themedia data processing steps listthat refers to "the download", "bytes received", or "whenever new data for the current media resource becomes available" refers to data passed in via appendBuffer().
      2. References to HTTP in theresource fetch algorithmand the media data processing steps listshall not apply because the HTMLMediaElement does not fetch media data via HTTP when a MediaSourceis attached.

The following steps are run in any case where the media element is going to transition toNETWORK_EMPTYandqueue a tasktofire an eventnamed emptiedat the media element. These stepsSHOULDbe run right before the transition.

1. If theMediaSourcewas constructed in aDedicatedWorkerGlobalScope:

   1. Notify theMediaSourceusing an internaldetachmessage posted to [[port to worker]].
   2. Set[[port to worker]]null.
   3. Set[[channel with worker]]null.
   4. The implicit message handler for thisdetachnotification runs the remainder of these steps in theDedicatedWorkerGlobalScopeMediaSource.

   Otherwise, theMediaSourcewas constructed in aWindow:

   Continue the remainder of these steps on theWindowMediaSource.

2. Set[[port to main]]null.
3. Set thereadyStateattribute to "closed".
4. Ifthisis aManagedMediaSource,then setstreaming attribute tofalse.
5. Updatedurationto NaN.
6. Remove all theSourceBufferobjects fromactiveSourceBuffers.
7. Queue a tasktofire an eventnamedremovesourcebufferat activeSourceBuffers.
8. Remove all theSourceBufferobjects fromsourceBuffers.
9. Queue a tasktofire an eventnamedremovesourcebufferat sourceBuffers.
10. Queue a tasktofire an eventnamedsourcecloseat theMediaSource.

Run the following steps as part of the "Wait until the user agent has established whether or not the media data for the new playback position is available, and, if it is, until it has decoded enough data to play back that position "step of the seek algorithm:

1. Note

   The media element looks formedia segmentscontaining thenew playback positionin eachSourceBufferobject in activeSourceBuffers.Any position within aTimeRangesin the current value of theHTMLMediaElement'sbufferedattribute has all necessary media segments buffered for that position.

   Ifnew playback positionis not in anyTimeRangesofHTMLMediaElement's buffered

   1. If theHTMLMediaElement'sreadyStateattribute is greater thanHAVE_METADATA,then set the HTMLMediaElement'sreadyStateattribute to HAVE_METADATA.
      Note

      PerHTMLMediaElement ready states[HTML] logic,HTMLMediaElement's readyStatechanges may trigger events on the HTMLMediaElement.

   2. The media element waits until anappendBuffer()call causes thecoded frame processingalgorithm to set the HTMLMediaElement'sreadyStateattribute to a value greater thanHAVE_METADATA.
      Note

      The web application can usebufferedand HTMLMediaElement'sbufferedto determine what the media element needs to resume playback.

   Otherwise

   Continue

   Note

   If thereadyStateattribute is "ended"and the new playback positionis within aTimeRangescurrently in HTMLMediaElement'sbuffered,then the seek operation must continue to completion here even if one or more currently selected or enabled track buffers' largest range end timestamp is less thannew playback position.This condition should only occur due to logic in bufferedwhenreadyStateis "ended".

2. The media element resets all decoders and initializes each one with data from the appropriateinitialization segment.
3. The media element feedscoded framesfrom theactive track buffersinto the decoders starting with the closestrandom access pointbefore thenew playback position.
4. Resume theseek algorithmat the "Await a stable state" step.

The following steps are periodically run during playback to make sure that all of the SourceBufferobjects inactiveSourceBuffershaveenough data to ensure uninterrupted playback.Changes toactiveSourceBuffersalso cause these steps to run because they affect the conditions that trigger state transitions.

Havingenough data to ensure uninterrupted playbackis an implementation specific condition where the user agent determines that it currently has enough data to play the presentation without stalling for a meaningful period of time. This condition is constantly evaluated to determine when to transition the media element into and out of theHAVE_ENOUGH_DATAready state. These transitions indicate when the user agent believes it has enough data buffered or it needs more data respectively.

If theHTMLMediaElement'sreadyStateattribute equals HAVE_NOTHING:

1. Abort these steps.

IfHTMLMediaElement'sbuffereddoes not contain a TimeRangesfor the current playback position:

1. Set theHTMLMediaElement'sreadyStateattribute to HAVE_METADATA.
   Note

   PerHTMLMediaElement ready states[HTML] logic,HTMLMediaElement's readyStatechanges may trigger events on the HTMLMediaElement.

2. Abort these steps.

IfHTMLMediaElement'sbufferedcontains aTimeRanges that includes the current playback position andenough data to ensure uninterrupted playback:

1. Set theHTMLMediaElement'sreadyStateattribute to HAVE_ENOUGH_DATA.
   Note

   PerHTMLMediaElement ready states[HTML] logic,HTMLMediaElement's readyStatechanges may trigger events on the HTMLMediaElement.

2. Playback may resume at this point if it was previously suspended by a transition toHAVE_CURRENT_DATA.
3. Abort these steps.

IfHTMLMediaElement'sbufferedcontains aTimeRanges that includes the current playback position and some time beyond the current playback position, then run the following steps:

1. Set theHTMLMediaElement'sreadyStateattribute to HAVE_FUTURE_DATA.
   Note

   PerHTMLMediaElement ready states[HTML] logic,HTMLMediaElement's readyStatechanges may trigger events on the HTMLMediaElement.

2. Playback may resume at this point if it was previously suspended by a transition toHAVE_CURRENT_DATA.
3. Abort these steps.

IfHTMLMediaElement'sbufferedcontains aTimeRanges that ends at the current playback position and does not have a range covering the time immediately after the current position:

1. Set theHTMLMediaElement'sreadyStateattribute to HAVE_CURRENT_DATA.
   Note

   PerHTMLMediaElement ready states[HTML] logic,HTMLMediaElement's readyStatechanges may trigger events on the HTMLMediaElement.

2. Playback is suspended at this point since the media element doesn't have enough data to advance themedia timeline.
3. Abort these steps.

During playbackactiveSourceBuffersneeds to be updated if the selectedvideo track, theenabledaudio track(s), or a text trackmodechanges. When one or more of these changes occur the following steps need to be followed. Also, whenMediaSourcewas constructed in a DedicatedWorkerGlobalScope,then each change that occurs to aWindowmirror of a track created previously by the implicit handler for the internalcreate track mirrormessageMUSTalso be made to the correspondingDedicatedWorkerGlobalScope track using an internalupdate track statemessage posted to [[port to worker]]whose implicit handler makes the change and runs the following steps. Likewise, each change that occurs to a DedicatedWorkerGlobalScopetrackMUSTalso be made to the correspondingWindow mirror of the track using an internalupdate track statemessage posted to [[port to main]]whose implicit handler makes the change to the mirror.

If the selected video track changes, then run the following steps:

1. If theSourceBufferassociated with the previously selected video track is not associated with any other enabled tracks, run the following steps:
   1. Remove theSourceBufferfromactiveSourceBuffers.
   2. Queue a tasktofire an eventnamedremovesourcebufferat activeSourceBuffers
2. If theSourceBufferassociated with the newly selected video track is not already inactiveSourceBuffers,run the following steps:
   1. Add theSourceBuffertoactiveSourceBuffers.
   2. Queue a tasktofire an eventnamedaddsourcebufferat activeSourceBuffers

If an audio track becomes disabled and theSourceBufferassociated with this track is not associated with any other enabled or selected track, then run the following steps:

1. Remove theSourceBufferassociated with the audio track from activeSourceBuffers
2. Queue a tasktofire an eventnamedremovesourcebufferat activeSourceBuffers

If an audio track becomes enabled and theSourceBufferassociated with this track is not already inactiveSourceBuffers,then run the following steps:

1. Add theSourceBufferassociated with the audio track to activeSourceBuffers
2. Queue a tasktofire an eventnamedaddsourcebufferat activeSourceBuffers

If a text trackmodebecomes"disabled" and theSourceBufferassociated with this track is not associated with any other enabled or selected track, then run the following steps:

1. Remove theSourceBufferassociated with the text track from activeSourceBuffers
2. Queue a tasktofire an eventnamedremovesourcebufferat activeSourceBuffers

If a text trackmodebecomes"showing"or "hidden"and theSourceBufferassociated with this track is not already inactiveSourceBuffers,then run the following steps:

1. Add theSourceBufferassociated with the text track to activeSourceBuffers
2. Queue a tasktofire an eventnamedaddsourcebufferat activeSourceBuffers

Follow these steps whendurationneeds to change to anew duration.

1. If the current value ofdurationis equal tonew duration,then return.
2. Ifnew durationis less than the highestpresentation timestampof any bufferedcoded framesfor allSourceBufferobjects in sourceBuffers,then throw anInvalidStateErrorexception and abort these steps.
   Note

   Duration reductions that would truncate currently buffered media are disallowed. When truncation is necessary, useremove()to reduce the buffered range before updatingduration.

3. Lethighest end timebe the largesttrack buffer ranges end time across all thetrack buffersacross allSourceBufferobjects in sourceBuffers.
4. Ifnew durationis less thanhighest end time,then
   Note

   This condition can occur because thecoded frame removalalgorithm preserves coded frames that start before the start of the removal range.

   1. Updatenew durationto equalhighest end time.
5. Updatedurationtonew duration.
6. Use themirror if necessaryalgorithm to run the following steps inWindow to update the media element's duration:
   1. Update the media element'sdurationtonew duration.
   2. Run theHTMLMediaElement duration change algorithm.

This algorithm gets called when the application signals the end of stream via an endOfStream()call or an algorithm needs to signal a decode error. This algorithm takes anerrorparameter that indicates whether an error will be signalled.

1. Change thereadyStateattribute value to "ended".
2. Queue a tasktofire an eventnamedsourceendedat theMediaSource.

3. Iferroris not set

   1. Run theduration changealgorithm withnew durationset to the largesttrack buffer rangesend time across all the track buffersacross allSourceBufferobjects in sourceBuffers.
      Note

      This allows the duration to properly reflect the end of the appended media segments. For example, if the duration was explicitly set to 10 seconds and only media segments for 0 to 5 seconds were appended before endOfStream() was called, then the duration will get updated to 5 seconds.

   2. Notify the media element that it now has all of the media data.

   Iferroris set to "network"

   Use themirror if necessaryalgorithm to run the following steps in Window:

   If theHTMLMediaElement'sreadyStateattribute equalsHAVE_NOTHING

   Run the"If the media data cannot be fetched at all, due to network errors, causing the user agent to give up trying to fetch the resource"steps of theresource fetch algorithm'smedia data processing steps list.

   If theHTMLMediaElement'sreadyStateattribute is greater thanHAVE_NOTHING

   Run the "If the connection is interrupted after some media data has been received, causing the user agent to give up trying to fetch the resource" steps of theresource fetch algorithm'smedia data processing steps list.

   Iferroris set to "decode"

   Use themirror if necessaryalgorithm to run the following steps in Window:

   If theHTMLMediaElement'sreadyStateattribute equalsHAVE_NOTHING

   Run the "If the media data can be fetched but is found by inspection to be in an unsupported format, or can otherwise not be rendered at all"steps of theresource fetch algorithm'smedia data processing steps list.

   If theHTMLMediaElement'sreadyStateattribute is greater thanHAVE_NOTHING

   Run themedia data is corruptedsteps of theresource fetch algorithm'smedia data processing steps list.

This algorithm is used to run steps onWindowfrom aMediaSourceattached from either the sameWindowor from aDedicatedWorkerGlobalScope,usually to update the state of the attachedHTMLMediaElement.This algorithm takes asteps parameter that lists the steps to run onWindow.

If theMediaSourcewas constructed in aDedicatedWorkerGlobalScope:

Post an internalmirror on windowmessage to[[port to main]]whose implicit handler inWindowwill runsteps.Return control to the caller without awaiting that handler's receipt of the message.

Note

The purpose of the mirror message mechanism is to ensure that:

1. stepsrun asynchronously as their own task onWindowrather than these stepssomehow happening in the middle of some otherWindowtask's execution, and
2. stepsare run without blocking the synchronous execution and return of this algorithm onDedicatedWorkerGlobalScope.

Otherwise:

Runsteps.

EachSourceBufferobject has an[[append state]]internal slot that keeps track of the high-level segment parsing state. It is initially set toWAITING_FOR_SEGMENTand can transition to the following states as data is appended.

EachSourceBufferobject has an[[input buffer]]internal slot that is a byte buffer that holds unparsed bytes across appendBuffer()calls. The buffer is empty when theSourceBuffer object is created.

EachSourceBufferobject has a[[buffer full flag]]internal slot that keeps track of whetherappendBuffer() is allowed to accept more bytes. It is set to false when theSourceBufferobject is created and gets updated as data is appended and removed.

EachSourceBufferobject has a[[group start timestamp]]internal slot that keeps track of the starting timestamp for a new coded frame groupin the "sequence"mode. It is unset when the SourceBuffer object is created and gets updated when themode attribute equals "sequence"and thetimestampOffset attribute is set, or thecoded frame processingalgorithm runs.

EachSourceBufferobject has a[[group end timestamp]]internal slot that stores the highestcoded frame end timestamp across allcoded framesin the currentcoded frame group.It is set to 0 when the SourceBuffer object is created and gets updated by thecoded frame processing algorithm.

EachSourceBufferobject has a[[generate timestamps flag]]internal slot that is a boolean that keeps track of whether timestamps need to be generated for thecoded framespassed to the coded frame processingalgorithm. This flag is set by addSourceBuffer()when theSourceBufferobject is created and is updated bychangeType().

When the segment parser loop algorithm is invoked, run the following steps:

1. Loop Top:If the[[input buffer]]is empty, then jump to the need more datastep below.
2. If the[[input buffer]]contains bytes that violate the SourceBuffer byte stream format specification,then run theappend error algorithm and abort this algorithm.
3. Remove any bytes that thebyte stream format specificationssayMUSTbe ignored from the start of the[[input buffer]].

4. If the[[append state]]equalsWAITING_FOR_SEGMENT,then run the following steps:

   1. If the beginning of the[[input buffer]]indicates the start of aninitialization segment,set the[[append state]]to PARSING_INIT_SEGMENT.
   2. If the beginning of the[[input buffer]]indicates the start of amedia segment,set[[append state]]to PARSING_MEDIA_SEGMENT.
   3. Jump to theloop topstep above.

5. If the[[append state]]equalsPARSING_INIT_SEGMENT,then run the following steps:

   1. If the[[input buffer]]does not contain a complete initialization segmentyet, then jump to theneed more datastep below.
   2. Run theinitialization segment receivedalgorithm.
   3. Remove theinitialization segmentbytes from the beginning of the [[input buffer]].
   4. Set[[append state]]toWAITING_FOR_SEGMENT.
   5. Jump to theloop topstep above.

6. If the[[append state]]equalsPARSING_MEDIA_SEGMENT,then run the following steps:

   1. If the[[first initialization segment received flag]]is false or the[[pending initialization segment for changeType flag]]is true, then run theappend erroralgorithm and abort this algorithm.
   2. If the[[input buffer]]contains one or more completecoded frames,then run thecoded frame processingalgorithm.
      Note

      The frequency at which the coded frame processing algorithm is run is implementation-specific. The coded frame processing algorithmMAYbe called when the input buffer contains the complete media segment or itMAYbe called multiple times as complete coded frames are added to the input buffer.

   3. If thisSourceBufferis full and cannot accept more media data, then set the[[buffer full flag]]to true.
   4. If the[[input buffer]]does not contain a completemedia segment,then jump to theneed more datastep below.
   5. Remove themedia segmentbytes from the beginning of the [[input buffer]].
   6. Set[[append state]]toWAITING_FOR_SEGMENT.
   7. Jump to theloop topstep above.
7. Need more data:Return control to the calling algorithm.

When the parser state needs to be reset, run the following steps:

1. If the[[append state]]equalsPARSING_MEDIA_SEGMENTand the [[input buffer]]contains some completecoded frames,then run the coded frame processingalgorithm until all of these completecoded frameshave been processed.
2. Unset thelast decode timestampon alltrack buffers.
3. Unset thelast frame durationon alltrack buffers.
4. Unset thehighest end timestampon alltrack buffers.
5. Set theneed random access point flagon alltrack buffersto true.
6. If themodeattribute equals "sequence",then set the[[group start timestamp]]to the[[group end timestamp]]
7. Remove all bytes from the[[input buffer]].
8. Set[[append state]]toWAITING_FOR_SEGMENT.

This algorithm is called when an error occurs during an append.

1. Run thereset parser statealgorithm.
2. Set theupdatingattribute to false.
3. Queue a tasktofire an eventnamederrorat thisSourceBuffer object.
4. Queue a tasktofire an eventnamedupdateendat thisSourceBuffer object.
5. Run theend of streamalgorithm with theerrorparameter set to "decode".

When an append operation begins, the following steps are run to validate and prepare theSourceBuffer.

1. If theSourceBufferhas been removed from thesourceBuffers attribute of theparent media sourcethen throw anInvalidStateErrorexception and abort these steps.
2. If theupdatingattribute equals true, then throw an InvalidStateErrorexception and abort these steps.
3. Letrecent element errorbe determined as follows:

   If theMediaSourcewas constructed in aWindow

   Letrecent element errorbe true if theHTMLMediaElement's errorattribute is not null. If that attribute is null, then letrecent element errorbe false.

   Otherwise

   Letrecent element errorbe the value resulting from the steps for the Windowcase, but run on theWindowHTMLMediaElementon any change to itserrorattribute and communicated by using [[port to worker]]implicit messages. If such a message has not yet been received, then letrecent element errorbe false.

4. Ifrecent element erroris true, then throw anInvalidStateErrorexception and abort these steps.

5. If thereadyStateattribute of theparent media sourceis in the "ended"state then run the following steps:

   1. Set thereadyStateattribute of theparent media sourceto "open"
   2. Queue a tasktofire an eventnamedsourceopenat theparent media source.
6. Run thecoded frame evictionalgorithm.

7. If the[[buffer full flag]]equals true, then throw a QuotaExceededErrorexception and abort these steps.

   Note

   This is the signal that the implementation was unable to evict enough data to accommodate the append or the append is too big. The web applicationSHOULDuse remove()to explicitly free up space and/or reduce the size of the append.

WhenappendBuffer()is called, the following steps are run to process the appended data.

1. Run thesegment parser loopalgorithm.
2. If thesegment parser loopalgorithm in the previous step was aborted, then abort this algorithm.
3. Set theupdatingattribute to false.
4. Queue a tasktofire an eventnamedupdateat thisSourceBuffer object.
5. Queue a tasktofire an eventnamedupdateendat thisSourceBuffer object.

Follow these steps when a caller needs to initiate a JavaScript visible range removal operation that blocks other SourceBuffer updates:

1. Letstartequal the startingpresentation timestampfor the removal range, in seconds measured frompresentation start time.
2. Letendequal the endpresentation timestampfor the removal range, in seconds measured frompresentation start time.
3. Set theupdatingattribute to true.
4. Queue a tasktofire an eventnamedupdatestartat this SourceBufferobject.
5. Return control to the caller and run the rest of the steps asynchronously.
6. Run thecoded frame removalalgorithm withstartandendas the start and end of the removal range.
7. Set theupdatingattribute to false.
8. Queue a tasktofire an eventnamedupdateat thisSourceBuffer object.
9. Queue a tasktofire an eventnamedupdateendat thisSourceBuffer object.

The following steps are run when thesegment parser loopsuccessfully parses a completeinitialization segment:

Each SourceBuffer object has a[[first initialization segment received flag]]internal slot that tracks whether the first initialization segmenthas been appended and received by this algorithm. This flag is set to false when the SourceBuffer is created and updated by the algorithm below.

Each SourceBuffer object has a[[pending initialization segment for changeType flag]]internal slot that tracks whether an initialization segmentis needed since the most recent changeType().This flag is set to false when the SourceBuffer is created, set to true bychangeType()and reset to false by the algorithm below.

1. Update thedurationattribute if it currently equals NaN:

   If the initialization segment contains a duration:

   Run theduration changealgorithm withnew durationset to the duration in the initialization segment.

   Otherwise:

   Run theduration changealgorithm withnew durationset to positive Infinity.

2. If theinitialization segmenthas no audio, video, or text tracks, then run the append erroralgorithm and abort these steps.
3. If the[[first initialization segment received flag]]is true, then run the following steps:
   1. Verify the following properties. If any of the checks fail then run the append erroralgorithm and abort these steps.
      • The number of audio, video, and text tracks match what was in the first initialization segment.
      • If more than one track for a single type are present (e.g., 2 audio tracks), then theTrack IDsmatch the ones in the firstinitialization segment.
      • The codecs for each track are supported by the user agent.
        Note

        User agentsMAYconsider codecs, that would otherwise be supported, as "not supported" here if the codecs were not specified intype parameter passed to (a) the most recently successful changeType()on thisSourceBufferobject, or (b) if no successfulchangeType()has yet occurred on this object, theaddSourceBuffer()that created thisSourceBuffer object. For example, if the most recently successful changeType()was called with'video/webm'or 'video/webm; codecs= "vp8" ',and a video track containing vp9 appears in the initialization segment, then the user agentMAYuse this step to trigger a decode error even if the other two properties' checks, above, pass. Implementations are encouraged to trigger error in such cases only when the codec is indeed not supported or the other two properties' checks fail. Web authors are encouraged to usechangeType(), addSourceBuffer()andisTypeSupported() with precise codec parameters to more proactively detect user agent support.changeType()is required if theSourceBuffer object's bytestream format is changing.

   2. Add the appropriatetrack descriptionsfrom thisinitialization segment to each of thetrack buffers.
   3. Set theneed random access point flagon all track buffers to true.
4. Letactive track flagequal false.

5. If the[[first initialization segment received flag]]is false, then run the following steps:

   1. If theinitialization segmentcontains tracks with codecs the user agent does not support, then run theappend erroralgorithm and abort these steps.
      Note

      User agentsMAYconsider codecs, that would otherwise be supported, as "not supported" here if the codecs were not specified intypeparameter passed to (a) the most recently successfulchangeType()on thisSourceBufferobject, or (b) if no successful changeType()has yet occurred on this object, the addSourceBuffer()that created thisSourceBufferobject. For example,MediaSource.isTypeSupported('video/webm;codecs= "vp8,vorbis" ') may return true, but ifaddSourceBuffer()was called with 'video/webm;codecs= "vp8" 'and a Vorbis track appears in theinitialization segment,then the user agentMAYuse this step to trigger a decode error. Implementations are encouraged to trigger error in such cases only when the codec is indeed not supported. Web authors are encouraged to use changeType(),addSourceBuffer()and isTypeSupported()with precise codec parameters to more proactively detect user agent support.changeType()is required if theSourceBufferobject's bytestream format is changing.

   2. For each audio track in theinitialization segment,run following steps:

      1. Letaudio byte stream track IDbe theTrack IDfor the current track being processed.
      2. Letaudio languagebe a BCP 47 language tag for the language specified in theinitialization segmentfor this track or an empty string if no language info is present.
      3. Ifaudio languageequals the 'und' BCP 47 value, then assign an empty string toaudio language.
      4. Letaudio labelbe a label specified in theinitialization segmentfor this track or an empty string if no label info is present.
      5. Letaudio kindsbe a sequence of kind strings specified in theinitialization segmentfor this track or a sequence with a single empty string element in it if no kind information is provided.
      6. For each value inaudio kinds,run the following steps:
         1. Letcurrent audio kindequal the value fromaudio kinds for this iteration of the loop.
         2. Letnew audio trackbe a newAudioTrackobject.
         3. Generate a unique ID and assign it to theidproperty on new audio track.
         4. Assignaudio languageto thelanguageproperty onnew audio track.
         5. Assignaudio labelto thelabelproperty onnew audio track.
         6. Assigncurrent audio kindto thekindproperty onnew audio track.

         7. If thisSourceBufferobject'saudioTracks's lengthequals 0, then run the following steps:

            1. Set theenabledproperty onnew audio trackto true.
            2. Setactive track flagto true.
         8. Addnew audio trackto theaudioTracksattribute on thisSourceBufferobject.
            Note

            This should triggerAudioTrackList[HTML] logic toqueue a tasktofire an eventnamedaddtrackusing TrackEventwith thetrackattribute initialized to new audio track,at theAudioTrackListobject referenced by the audioTracksattribute on thisSourceBufferobject.

         9. If theparent media sourcewas constructed in a DedicatedWorkerGlobalScope:

            Post an internalcreate track mirrormessage to [[port to main]]whose implicit handler inWindow runs the following steps:

            1. Letmirrored audio trackbe a newAudioTrack object.
            2. Assign the same property values tomirrored audio trackas were determined fornew audio track.
            3. Addmirrored audio trackto the audioTracksattribute on the HTMLMediaElement.

            Otherwise:

            Addnew audio trackto theaudioTracks attribute on the HTMLMediaElement.

            Note

            This should triggerAudioTrackList[HTML] logic toqueue a tasktofire an eventnamedaddtrackusing TrackEventwith thetrackattribute initialized to mirrored audio trackornew audio track,at theAudioTrackList object referenced by theaudioTracksattribute on the HTMLMediaElement.

      7. Create a newtrack bufferto storecoded framesfor this track.
      8. Add thetrack descriptionfor this track to thetrack buffer.

   3. For each video track in theinitialization segment,run following steps:

      1. Letvideo byte stream track IDbe theTrack IDfor the current track being processed.
      2. Letvideo languagebe a BCP 47 language tag for the language specified in theinitialization segmentfor this track or an empty string if no language info is present.
      3. Ifvideo languageequals the 'und' BCP 47 value, then assign an empty string tovideo language.
      4. Letvideo labelbe a label specified in theinitialization segmentfor this track or an empty string if no label info is present.
      5. Letvideo kindsbe a sequence of kind strings specified in theinitialization segmentfor this track or a sequence with a single empty string element in it if no kind information is provided.
      6. For each value invideo kinds,run the following steps:
         1. Letcurrent video kindequal the value fromvideo kinds for this iteration of the loop.
         2. Letnew video trackbe a newVideoTrackobject.
         3. Generate a unique ID and assign it to theidproperty on new video track.
         4. Assignvideo languageto thelanguageproperty onnew video track.
         5. Assignvideo labelto thelabelproperty onnew video track.
         6. Assigncurrent video kindto thekindproperty onnew video track.

         7. If thisSourceBufferobject'svideoTracks's lengthequals 0, then run the following steps:

            1. Set theselectedproperty onnew video trackto true.
            2. Setactive track flagto true.
         8. Addnew video trackto thevideoTracksattribute on thisSourceBufferobject.
            Note

            This should triggerVideoTrackList[HTML] logic toqueue a tasktofire an eventnamedaddtrackusing TrackEventwith thetrackattribute initialized to new video track,at theVideoTrackListobject referenced by the videoTracksattribute on thisSourceBufferobject.

         9. If theparent media sourcewas constructed in a DedicatedWorkerGlobalScope:

            Post an internalcreate track mirrormessage to [[port to main]]whose implicit handler inWindow runs the following steps:

            1. Letmirrored video trackbe a newVideoTrack object.
            2. Assign the same property values tomirrored video trackas were determined fornew video track.
            3. Addmirrored video trackto the videoTracksattribute on the HTMLMediaElement.

            Otherwise:

            Addnew video trackto thevideoTracks attribute on the HTMLMediaElement.

            Note

            This should triggerVideoTrackList[HTML] logic toqueue a tasktofire an eventnamedaddtrackusing TrackEventwith thetrackattribute initialized to mirrored video trackornew video track,at theVideoTrackList object referenced by thevideoTracksattribute on the HTMLMediaElement.

      7. Create a newtrack bufferto storecoded framesfor this track.
      8. Add thetrack descriptionfor this track to thetrack buffer.

   4. For each text track in theinitialization segment,run following steps:

      1. Lettext byte stream track IDbe theTrack IDfor the current track being processed.
      2. Lettext languagebe a BCP 47 language tag for the language specified in theinitialization segmentfor this track or an empty string if no language info is present.
      3. Iftext languageequals the 'und' BCP 47 value, then assign an empty string totext language.
      4. Lettext labelbe a label specified in theinitialization segmentfor this track or an empty string if no label info is present.
      5. Lettext kindsbe a sequence of kind strings specified in theinitialization segmentfor this track or a sequence with a single empty string element in it if no kind information is provided.
      6. For each value intext kinds,run the following steps:
         1. Letcurrent text kindequal the value fromtext kindsfor this iteration of the loop.
         2. Letnew text trackbe a newTextTrackobject.
         3. Generate a unique ID and assign it to theidproperty on new text track.
         4. Assigntext languageto thelanguageproperty onnew text track.
         5. Assigntext labelto thelabelproperty onnew text track.
         6. Assigncurrent text kindto thekindproperty onnew text track.
         7. Populate the remaining properties onnew text trackwith the appropriate information from theinitialization segment.
         8. If themodeproperty onnew text trackequals "showing"or"hidden",then setactive track flagto true.
         9. Addnew text trackto thetextTracksattribute on thisSourceBufferobject.
            Note

            This should triggerTextTrackList[HTML] logic toqueue a tasktofire an eventnamedaddtrackusing TrackEventwith thetrackattribute initialized to new text track,at theTextTrackListobject referenced by the textTracksattribute on thisSourceBufferobject.

         10. If theparent media sourcewas constructed in a DedicatedWorkerGlobalScope:

             Post an internalcreate track mirrormessage to [[port to main]]whose implicit handler inWindow runs the following steps:

             1. Letmirrored text trackbe a newTextTrack object.
             2. Assign the same property values tomirrored text trackas were determined fornew text track.
             3. Addmirrored text trackto the textTracksattribute on the HTMLMediaElement.

             Otherwise:

             Addnew text trackto thetextTracksattribute on the HTMLMediaElement.

             Note

             This should triggerTextTrackList[HTML] logic toqueue a tasktofire an eventnamedaddtrackusing TrackEventwith thetrackattribute initialized to mirrored text trackornew text track,at theTextTrackList object referenced by thetextTracksattribute on the HTMLMediaElement.

      7. Create a newtrack bufferto storecoded framesfor this track.
      8. Add thetrack descriptionfor this track to thetrack buffer.
   5. Ifactive track flagequals true, then run the following steps:
      1. Add thisSourceBuffertoactiveSourceBuffers.
      2. Queue a tasktofire an eventnamedaddsourcebufferat activeSourceBuffers
   6. Set[[first initialization segment received flag]]to true.
6. Set[[pending initialization segment for changeType flag]]to false.
7. If theactive track flagequals true, then run the following steps:
8. Use theparent media source'smirror if necessaryalgorithm to run the following step inWindow:
   1. If theHTMLMediaElement'sreadyStateattribute is greater thanHAVE_CURRENT_DATA,then set the HTMLMediaElement'sreadyStateattribute to HAVE_METADATA.
      Note

      PerHTMLMediaElement ready states[HTML] logic,HTMLMediaElement's readyStatechanges may trigger events on the HTMLMediaElement.

9. If each object insourceBuffersof theparent media sourcehas [[first initialization segment received flag]]equal to true, then use theparent media source'smirror if necessaryalgorithm to run the following step inWindow:
   1. If theHTMLMediaElement'sreadyStateattribute is HAVE_NOTHING,then set theHTMLMediaElement's readyStateattribute toHAVE_METADATA.
      Note

      PerHTMLMediaElement ready states[HTML] logic,HTMLMediaElement's readyStatechanges may trigger events on the HTMLMediaElement. If transition fromHAVE_NOTHINGto HAVE_METADATAoccurs, it should trigger HTMLMediaElement logic toqueue a tasktofire an eventnamed loadedmetadataat the media element.

When completecoded frameshave been parsed by thesegment parser loopthen the following steps are run:

1. For eachcoded framein themedia segmentrun the following steps:

   1. Loop Top:

      If[[generate timestamps flag]]equals true:

      1. Letpresentation timestampequal 0.
      2. Letdecode timestampequal 0.

      Otherwise:

      1. Letpresentation timestampbe a double precision floating point representation of the coded frame'spresentation timestampin seconds.
         Note

         Special processing may be needed to determine the presentation and decode timestamps for timed text frames since this information may not be explicitly present in the underlying format or may be dependent on the order of the frames. Some metadata text tracks, like MPEG2-TS PSI data, may only have implied timestamps. Format specific rules for these situationsSHOULDbe in thebyte stream format specificationsor in separate extension specifications.

      2. Letdecode timestampbe a double precision floating point representation of the coded frame's decode timestamp in seconds.
         Note

         Implementations don't have to internally store timestamps in a double precision floating point representation. This representation is used here because it is the representation for timestamps in the HTML spec. The intention here is to make the behavior clear without adding unnecessary complexity to the algorithm to deal with the fact that adding a timestampOffset may cause a timestamp rollover in the underlying timestamp representation used by the byte stream format. Implementations can use any internal timestamp representation they wish, but the addition of timestampOffsetSHOULDbehave in a similar manner to what would happen if a double precision floating point representation was used.

   2. Letframe durationbe a double precision floating point representation of thecoded frame's durationin seconds.
   3. Ifmodeequals "sequence"and [[group start timestamp]]is set, then run the following steps:
      1. SettimestampOffsetequal to[[group start timestamp]]minuspresentation timestamp.
      2. Set[[group end timestamp]]equal to [[group start timestamp]].
      3. Set theneed random access point flagon alltrack buffersto true.
      4. Unset[[group start timestamp]].

   4. IftimestampOffsetis not 0, then run the following steps:

      1. AddtimestampOffsetto thepresentation timestamp.
      2. AddtimestampOffsetto thedecode timestamp.
   5. Lettrack bufferequal thetrack bufferthat the coded frame will be added to.

   6. Iflast decode timestampfortrack bufferis set anddecode timestamp is less thanlast decode timestamp:

      OR

      Iflast decode timestampfortrack bufferis set and the difference betweendecode timestampandlast decode timestampis greater than 2 timeslast frame duration:

      1. Ifmodeequals "segments":

         Set[[group end timestamp]]topresentation timestamp.

         Ifmodeequals "sequence":

         Set[[group start timestamp]]equal to the [[group end timestamp]].

      2. Unset thelast decode timestampon alltrack buffers.
      3. Unset thelast frame durationon alltrack buffers.
      4. Unset thehighest end timestampon alltrack buffers.
      5. Set theneed random access point flagon alltrack buffersto true.
      6. Jump to theLoop Topstep above to restart processing of the currentcoded frame.

      Otherwise:

      Continue.

   7. Letframe end timestampequal the sum ofpresentation timestampand frame duration.
   8. Ifpresentation timestampis less thanappendWindowStart, then set theneed random access point flagto true, drop the coded frame, and jump to the top of the loop to start processing the next coded frame.
      Note

      Some implementationsMAYchoose to collect some of these coded frames with presentation timestampless thanappendWindowStartand use them to generate a splice at the first coded frame that has apresentation timestampgreater than or equal toappendWindowStarteven if that frame is not arandom access point.Supporting this requires multiple decoders or faster than real-time decoding so for now this behavior will not be a normative requirement.

   9. Ifframe end timestampis greater thanappendWindowEnd,then set theneed random access point flagto true, drop the coded frame, and jump to the top of the loop to start processing the next coded frame.
      Note

      Some implementationsMAYchoose to collect coded frames withpresentation timestampless thanappendWindowEndandframe end timestamp greater thanappendWindowEndand use them to generate a splice across the portion of the collected coded frames within the append window at time of collection, and the beginning portion of later processed frames which only partially overlap the end of the collected coded frames. Supporting this requires multiple decoders or faster than real-time decoding so for now this behavior will not be a normative requirement. In conjunction with collecting coded frames that spanappendWindowStart,implementationsMAY thus support gapless audio splicing.

   10. If theneed random access point flagontrack bufferequals true, then run the following steps:
       1. If the coded frame is not arandom access point,then drop the coded frame and jump to the top of the loop to start processing the next coded frame.
       2. Set theneed random access point flagontrack bufferto false.
   11. Letspliced audio framebe an unset variable for holding audio splice information
   12. Letspliced timed text framebe an unset variable for holding timed text splice information
   13. Iflast decode timestampfortrack bufferis unset andpresentation timestampfalls within thepresentation intervalof acoded framein track buffer,then run the following steps:
       1. Letoverlapped framebe thecoded frameintrack bufferthat matches the condition above.

       2. Iftrack buffercontains audiocoded frames:

          Run theaudio splice framealgorithm and if a splice frame is returned, assign it tospliced audio frame.

          Iftrack buffercontains videocoded frames:

          1. Letremove window timestampequal theoverlapped frame presentation timestampplus 1 microsecond.
          2. If thepresentation timestampis less than theremove window timestamp,then removeoverlapped framefromtrack buffer.
             Note

             This is to compensate for minor errors in frame timestamp computations that can appear when converting back and forth between double precision floating point numbers and rationals. This tolerance allows a frame to replace an existing one as long as it is within 1 microsecond of the existing frame's start time. Frames that come slightly before an existing frame are handled by the removal step below.

          Iftrack buffercontains timed textcoded frames:

          Run thetext splice framealgorithm and if a splice frame is returned, assign it tospliced timed text frame.

   14. Remove existing coded frames intrack buffer:

       Ifhighest end timestampfortrack bufferis not set:

       Remove allcoded framesfromtrack bufferthat have apresentation timestampgreater than or equal topresentation timestampand less than frame end timestamp.

       Ifhighest end timestampfortrack bufferis set and less than or equal topresentation timestamp:

       Remove allcoded framesfromtrack bufferthat have apresentation timestampgreater than or equal tohighest end timestampand less than frame end timestamp.

   15. Remove all possible decoding dependencies on thecoded framesremoved in the previous two steps by removing allcoded framesfromtrack bufferbetween those frames removed in the previous two steps and the nextrandom access point after those removed frames.
       Note

       Removing allcoded framesuntil the nextrandom access pointis a conservative estimate of the decoding dependencies since it assumes all frames between the removed frames and the next random access point depended on the frames that were removed.

   16. Ifspliced audio frameis set:

       Addspliced audio frameto thetrack buffer.

       Ifspliced timed text frameis set:

       Addspliced timed text frameto thetrack buffer.

       Otherwise:

       Add thecoded framewith thepresentation timestamp,decode timestamp,andframe durationto thetrack buffer.

   17. Setlast decode timestampfortrack buffertodecode timestamp.
   18. Setlast frame durationfortrack buffertoframe duration.
   19. Ifhighest end timestampfortrack bufferis unset orframe end timestampis greater thanhighest end timestamp,then sethighest end timestampfortrack buffertoframe end timestamp.
       Note

       The greater than check is needed because bidirectional prediction between coded frames can causepresentation timestampto not be monotonically increasing even though the decode timestamps are monotonically increasing.

   20. Ifframe end timestampis greater than[[group end timestamp]],then set[[group end timestamp]]equal toframe end timestamp.
   21. If[[generate timestamps flag]]equals true, then set timestampOffsetequal toframe end timestamp.

2. If theHTMLMediaElement'sreadyStateattribute is HAVE_METADATAand the newcoded framescause HTMLMediaElement'sbufferedto have aTimeRangesfor the current playback position, then set theHTMLMediaElement's readyStateattribute to HAVE_CURRENT_DATA.

   Note

   PerHTMLMediaElement ready states[HTML] logic,HTMLMediaElement's readyStatechanges may trigger events on the HTMLMediaElement.

3. If theHTMLMediaElement'sreadyStateattribute is HAVE_CURRENT_DATAand the newcoded framescause HTMLMediaElement'sbufferedto have aTimeRangesthat includes the current playback position and some time beyond the current playback position, then set theHTMLMediaElement'sreadyState attribute toHAVE_FUTURE_DATA.

   Note

   PerHTMLMediaElement ready states[HTML] logic,HTMLMediaElement's readyStatechanges may trigger events on the HTMLMediaElement.

4. If theHTMLMediaElement'sreadyStateattribute is HAVE_FUTURE_DATAand the newcoded framescause HTMLMediaElement'sbufferedto have aTimeRangesthat includes the current playback position andenough data to ensure uninterrupted playback,then set theHTMLMediaElement'sreadyState attribute toHAVE_ENOUGH_DATA.

   Note

   PerHTMLMediaElement ready states[HTML] logic,HTMLMediaElement's readyStatechanges may trigger events on the HTMLMediaElement.

5. If themedia segmentcontains data beyond the currentduration, then run theduration changealgorithm withnew durationset to the maximum of the current duration and the[[group end timestamp]].

Follow these steps whencoded framesfor a specific time range need to be removed from the SourceBuffer:

1. Letstartbe the startingpresentation timestampfor the removal range.
2. Letendbe the endpresentation timestampfor the removal range.

3. For eachtrack bufferin thisSourceBuffer,run the following steps:

   1. Letremove end timestampbe the current value of duration

   2. If thistrack bufferhas arandom access pointtimestamp that is greater than or equal toend,then updateremove end timestampto that random access point timestamp.

      Note

      Random access point timestamps can be different across tracks because the dependencies betweencoded frameswithin a track are usually different than the dependencies in another track.

   3. Remove all media data, from thistrack buffer,that contain starting timestamps greater than or equal tostartand less than theremove end timestamp.

      1. For each removed frame, if the frame has adecode timestampequal to thelast decode timestampfor the frame's track, run the following steps:

         Ifmodeequals "segments":

         Set[[group end timestamp]]topresentation timestamp.

         Ifmodeequals "sequence":

         Set[[group start timestamp]]equal to the [[group end timestamp]].

      2. Unset thelast decode timestampon alltrack buffers.
      3. Unset thelast frame durationon alltrack buffers.
      4. Unset thehighest end timestampon alltrack buffers.
      5. Set theneed random access point flagon alltrack buffersto true.
   4. Remove all possible decoding dependencies on thecoded framesremoved in the previous step by removing allcoded framesfrom thistrack buffer between those frames removed in the previous step and the nextrandom access pointafter those removed frames.
      Note

      Removing allcoded framesuntil the nextrandom access pointis a conservative estimate of the decoding dependencies since it assumes all frames between the removed frames and the next random access point depended on the frames that were removed.

   5. If this object is inactiveSourceBuffers,thecurrent playback positionis greater than or equal tostartand less than theremove end timestamp,andHTMLMediaElement's readyStateis greater than HAVE_METADATA,then set theHTMLMediaElement's readyStateattribute toHAVE_METADATA and stall playback.

      Note

      PerHTMLMediaElement ready states[HTML] logic,HTMLMediaElement's readyStatechanges may trigger events on the HTMLMediaElement.

      Note

      This transition occurs because media data for the current position has been removed. Playback cannot progress until media for thecurrent playback positionis appended or the 3.15.5 Changes to selected/enabled track state.

4. If the[[buffer full flag]]equals true and this object is ready to accept more bytes, then set the[[buffer full flag]]to false.

This algorithm is run to free up space in thisSourceBufferwhen new data is appended.

1. Letnew dataequal the data that is about to be appended to this SourceBuffer.
   Issue 289:Editorial? Coded Frame eviction algorithm needs to note that "buffer full flag" may be updated immediately based on |new data|

   Need to recognize step here that implementationsMAYdecide to set [[buffer full flag]]true here if it predicts that processing new datain addition to any existing bytes in[[input buffer]] would exceed the capacity of theSourceBuffer.Such a step enables more proactive push-back from implementations before acceptingnew datawhich would overflow resources, for example. In practice, at least one implementation already does this.

2. If the[[buffer full flag]]equals false, then abort these steps.
3. Letremoval rangesequal a list of presentation time ranges that can be evicted from the presentation to make room for thenew data.
   Note

   ImplementationsMAYuse different methods for selectingremoval rangesso web applicationsSHOULD NOTdepend on a specific behavior. The web application can use thebufferedattribute to observe whether portions of the buffered data have been evicted.

4. For each range inremoval ranges,run thecoded frame removalalgorithm with startandendequal to the removal range start and end timestamp respectively.

Follow these steps when thecoded frame processingalgorithm needs to generate a splice frame for two overlapping audiocoded frames:

1. Lettrack bufferbe thetrack bufferthat will contain the splice.
2. Letnew coded framebe the newcoded frame,that is being added totrack buffer,which triggered the need for a splice.
3. Letpresentation timestampbe thepresentation timestampfornew coded frame.
4. Letdecode timestampbe the decode timestamp fornew coded frame.
5. Letframe durationbe thecoded frame durationofnew coded frame.
6. Letoverlapped framebe thecoded frameintrack bufferwith a presentation intervalthat containspresentation timestamp.
7. Updatepresentation timestampanddecode timestampto the nearest audio sample timestamp based on sample rate of the audio inoverlapped frame.If a timestamp is equidistant from both audio sample timestamps, then use the higher timestamp (e.g., floor(x * sample_rate + 0.5) / sample_rate).
   Note

   For example, given the following values:

   • Thepresentation timestampofoverlapped frameequals 10.
   • The sample rate ofoverlapped frameequals 8000 Hz
   • presentation timestampequals 10.01255
   • decode timestampequals 10.01255

   presentation timestampanddecode timestampare updated to 10.0125 since 10.01255 is closer to 10 + 100/8000 (10.0125) than 10 + 101/8000 (10.012625)

8. If the user agent does not support crossfading then run the following steps:
   1. Removeoverlapped framefromtrack buffer.
   2. Add a silence frame totrack bufferwith the following properties:
      • Thepresentation timestampset to theoverlapped framepresentation timestamp.
      • Thedecode timestampset to theoverlapped framedecode timestamp.
      • Thecoded frame durationset to difference betweenpresentation timestampand theoverlapped framepresentation timestamp.
      Note

      Some implementationsMAYapply fades to/from silence to coded frames on either side of the inserted silence to make the transition less jarring.

   3. Return to caller without providing a splice frame.
      Note

      This is intended to allownew coded frameto be added to thetrack buffer as ifoverlapped framehad not been in thetrack bufferto begin with.

9. Letframe end timestampequal the sum ofpresentation timestampand frame duration.
10. Letsplice end timestampequal the sum ofpresentation timestampand the splice duration of 5 milliseconds.
11. Letfade out coded framesequaloverlapped frameas well as any additional frames intrack bufferthat have apresentation timestampgreater than presentation timestampand less thansplice end timestamp.
12. Remove all the frames included infade out coded framesfromtrack buffer.
13. Return a splice frame with the following properties:
    • Thepresentation timestampset to theoverlapped framepresentation timestamp.
    • Thedecode timestampset to theoverlapped framedecode timestamp.
    • Thecoded frame durationset to difference betweenframe end timestamp and theoverlapped framepresentation timestamp.
    • The fade out coded frames equalsfade out coded frames.
    • The fade in coded frame equalsnew coded frame.
      Note

      If thenew coded frameis less than 5 milliseconds in duration, then coded frames that are appended after thenew coded framewill be needed to properly render the splice.

    • The splice timestamp equalspresentation timestamp.
    Note

    See theaudio splice renderingalgorithm for details on how this splice frame is rendered.

The following steps are run when a spliced frame, generated by theaudio splice framealgorithm, needs to be rendered by the media element:

1. Letfade out coded framesbe thecoded framesthat are faded out during the splice.
2. Letfade in coded framesbe thecoded framesthat are faded in during the splice.
3. Letpresentation timestampbe thepresentation timestampof the first coded frame infade out coded frames.
4. Letend timestampbe the sum of thepresentation timestampand the coded frame durationof the last frame infade in coded frames.
5. Letsplice timestampbe thepresentation timestampwhere the splice starts. This corresponds with thepresentation timestampof the first frame in fade in coded frames.
6. Letsplice end timestampequalsplice timestampplus five milliseconds.
7. Letfade out samplesbe the samples generated by decodingfade out coded frames.
8. Trimfade out samplesso that it only contains samples betweenpresentation timestampandsplice end timestamp.
9. Letfade in samplesbe the samples generated by decodingfade in coded frames.
10. Iffade out samplesandfade in samplesdo not have a common sample rate and channel layout, then convertfade out samplesandfade in samplesto a common sample rate and channel layout.
11. Letoutput samplesbe a buffer to hold the output samples.
12. Apply a linear gain fade out with a starting gain of 1 and an ending gain of 0 to the samples betweensplice timestampandsplice end timestampinfade out samples.
13. Apply a linear gain fade in with a starting gain of 0 and an ending gain of 1 to the samples betweensplice timestampandsplice end timestampinfade in samples.
14. Copy samples betweenpresentation timestamptosplice timestampfromfade out samplesintooutput samples.
15. For each sample betweensplice timestampandsplice end timestamp,compute the sum of a sample fromfade out samplesand the corresponding sample infade in samplesand store the result inoutput samples.
16. Copy samples betweensplice end timestamptoend timestampfromfade in samplesintooutput samples.
17. Renderoutput samples.

Note

Here is a graphical representation of this algorithm.

Follow these steps when thecoded frame processingalgorithm needs to generate a splice frame for two overlapping timed textcoded frames:

1. Lettrack bufferbe thetrack bufferthat will contain the splice.
2. Letnew coded framebe the newcoded frame,that is being added totrack buffer,which triggered the need for a splice.
3. Letpresentation timestampbe thepresentation timestampfornew coded frame
4. Letdecode timestampbe the decode timestamp fornew coded frame.
5. Letframe durationbe thecoded frame durationofnew coded frame.
6. Letframe end timestampequal the sum ofpresentation timestampand frame duration.
7. Letfirst overlapped framebe thecoded frameintrack bufferwith a presentation intervalthat containspresentation timestamp.
8. Letoverlapped presentation timestampbe thepresentation timestampof thefirst overlapped frame.
9. Letoverlapped framesequalfirst overlapped frameas well as any additional frames intrack bufferthat have apresentation timestampgreater than presentation timestampand less thanframe end timestamp.
10. Remove all the frames included inoverlapped framesfromtrack buffer.
11. Update thecoded frame durationof thefirst overlapped frameto presentation timestampminusoverlapped presentation timestamp.
12. Addfirst overlapped frameto thetrack buffer.
13. Return to caller without providing a splice frame.
    Note

    This is intended to allownew coded frameto be added to thetrack bufferas if it hadn't overlapped any frames intrack bufferto begin with.

The following steps are run periodically, whenever theSourceBuffer Monitoringalgorithm is scheduled to run.

Havingenough managed data to ensure uninterrupted playbackis an implementation defined condition where the user agent determines that it currently has enough data to play the presentation without stalling for a meaningful period of time. This condition is constantly evaluated to determine when to transition the value of streaming.These transitions indicate when the user agent believes it has enough data buffered or it needs more data respectively.

Beingable to retrieve and buffer data in an efficient wayis an implementation defined condition where the user agent determines that it can fetch new data in an energy efficient manner while able to achieve the desired memory usage.

1. Run theMediaSourceSourceBuffer Monitoringalgorithm.
2. Letcan play uninterrupted and efficientlybe a flag that is true if the bufferedattribute contains aTimeRangesthat includes the current playback position andenough managed data to ensure uninterrupted playbackand is able to retrieve and buffer data in an efficient way

   Ifcan play uninterrupted and efficientlyis not equal to streaming,queue an element taskon themedia element that runs the following steps:

   1. Setthisstreamingattribute tocan play uninterrupted and efficiently.
   2. Ifcan play uninterrupted and efficientlyis falsefire an eventcalled startstreamingat theManagedMediaSource.
   3. Otherwise,fire an eventcalledendstreamingat the ManagedMediaSource.

1. For eachbufferinthis'ssourceBuffers:

   1. Run thebuffer'smemory cleanupalgorithm.

The following steps are run at the completion of all operations to the ManagedSourceBufferbufferthat would cause abuffer's bufferedto change. That is onceappendBuffer(), remove()ormemory cleanupalgorithm have completed.

1. Letprevious buffered rangesequal the bufferedattribute before the changes occurred.
2. Letnew buffered rangesequal the newbuffered TimeRanges.
3. Letaddedequal theprevious buffered rangessubtracted fromnew buffered ranges.
4. Letremovedequal thenew buffered rangessubtracted fromprevious buffered ranges.
5. LeteventInitDictbe a newBufferedChangeEventInitdictionary initialized with addedas itsaddedRangesandremovedas its removedRanges
6. Queue a tasktofire an eventnamedbufferedchangeatbufferusing the BufferedChangeEventinterface, initialized witheventInitDict.

1. Ifthisis not inthis'sManagedMediaSourceparent activeSourceBuffers:

   1. Run thecoded frame removalalgorithm with start set to 0, end set to positive infinity, and abort these steps.

2. Letremoval rangesequal a list of presentation time ranges that can be evicted from the presentation to ensure uninterrupted playback from currentTimeuntil such presentation could be retrieved again.
   Note

   Implementations can use different strategies for selectingremoval rangesso web applications shouldn't depend on a specific behavior. The web application would listen to thebufferedchangeevent to observe whether portions of the buffered data have been evicted.

3. For each range inremoval ranges,run thecoded frame removalalgorithm with startandendequal to the removal range start and end timestamp respectively.