Guidelines for Implementation: DASH-IF Interoperability Points

1. Document editing notes

Documentation: https://dashif.org/DocumentAuthoring/

Example document repository: https://dashif.org/DocumentAuthoring/

Live discussion in #document-authoring on Slack.

2. Chapter 1

Placeholder text. This document will eventually contain IOP v5.

3. Timing and addressing

This chapter describes an interoperable view of DASH presentation timing and segment addressing. This interpretation is considerably narrower than afforded by [MPEGDASH], constraining services to a specific set of reasonably flexible behaviors that are highly interoperable with modern client platforms. Services conforming to this document SHALL use this timing model.

The presentation manifest or MPD defines the MPD timeline which serves as the baseline for all scheduling decisions made during DASH presentation playback.

The playback of a static MPD SHALL NOT depend on the mapping of the MPD timeline to real time. A client MAY play any part of the presentation at any time.

The MPD timeline of a dynamic MPD SHALL have a fixed mapping to real time, with each point on the timeline corresponding to a point in real time. Clients MAY introduce an additional offset with respect to real time to the extent allowed by the time shift signaling in the MPD.

Note: In addition to mapping the content to real time, a dynamic MPD can be updated during the presentation. Updates may add new periods and remove or modify existing ones, though some restrictions apply. See §3.8.5 MPD updates.

The zero point in the MPD timeline of a dynamic MPD SHALL be mapped to the point in real time indicated by MPD@availabilityStartTime. This value SHALL NOT change between MPD updates.

The ultimate purpose of the MPD is to enable the client to obtain media samples for playback. The following data structures are most relevant to locating and scheduling the samples:

1. The MPD consists of consecutive periods which map data onto the MPD timeline.

2. Each period contains of one or more representations, each of which provides media samples inside a sequence of media segments.

3. Representations within a period are grouped in adaptation sets, which associate related representations and decorate them with metadata.

The chapters below explore these relationships in detail.

3.1. Periods

An MPD SHALL define an ordered list of one or more consecutive periods. A period is both a time span on the MPD timeline and a definition of the data to be presented during this time span. Period timing is relative to the zero point of the MPD timeline.

Common reasons for defining multiple periods are:

• Assembling a presentation from multiple self-contained pieces of content.

• Inserting ads in the middle of existing content and/or replacing spans of existing content with ads.

• Adding/removing certain representations as the nature of the content changes (e.g. a new title starts with a different set of offered languages).

• Updating period-scoped metadata (e.g. codec configuration or DRM signaling).

Periods are self-contained - a service SHALL NOT require a client to know the contents of another period in order to correctly present a period. Knowledge of the contents of different periods MAY be used by a client to achieve seamless period transitions, especially when working with period-connected representations.

All periods SHALL be consecutive and non-overlapping. A period MAY have a duration of zero.

Note: A period with a duration of zero might, for example, be the result of ad-insertion logic deciding not to insert any ad.

<MPD xmlns="urn:mpeg:dash:schema:mpd:2011" type="static">
  <Period>
    ...
  </Period>
  <Period start="PT20S" duration="PT20S">
    ...
  </Period>
</MPD>

In a static MPD, the first period SHALL start at the zero point of the MPD timeline. In a dynamic MPD, the first period SHALL start at or after the zero point of the MPD timeline.

In a static MPD, the last period SHALL have a Period@duration. In a dynamic MPD, the last period MAY have a Period@duration, in which case it is considered to have a fixed duration. If without Period@duration, the last period in a dynamic MPD SHALL be considered to have an unlimited duration.

Note: In a dynamic MPD, a period with an unlimited duration may be converted to fixed-duration by an MPD update. Periods in a dynamic MPD may also be shortened or removed entirely under certain conditions. See §3.8.5 MPD updates.

MPD@mediaPresentationDuration MAY be present. If present, it SHALL accurately match the duration between the zero point on the MPD timeline and the end of the last period. Clients SHALL calculate the total duration of a static MPD by adding up the durations of each period and SHALL NOT rely on the presence of MPD@mediaPresentationDuration.

Note: This calculation is necessary because the durations of XLink periods can only be known after the XLink is resolved. Therefore it is impossible to always determine the total MPD duration on the service side as only the client is guaranteed to have access to all the required knowledge.

3.2. Representations

A representation is a sequence of segment references and a description of the samples within the referenced media segments. Each representation belongs to exactly one adaptation set and to exactly one period, although a representation may be connected with a representation in another period.

Each segment reference addresses a media segment that corresponds to a specific time span on the sample timeline. Each media segment contains samples for a specific time span on the sample timeline.

Note: Simple addressing allows the actual time span of samples within a media segment to deviate from the nominal time span described in the MPD. All timing-related clauses in this document refer to the nominal timing described in the MPD unless otherwise noted.

The exact mechanism used to define segment references depends on the addressing mode used by the representation. All representations in the same adaptation set SHALL use the same addressing mode.

In a static MPD, a representation SHALL contain enough segment references to cover the entire time span of the period.

In a dynamic MPD, a representation SHALL contain enough segment references to cover the time span of the period that intersects with the time shift buffer.

Note: Media segments only become available when their end point is within the availability window. It is a valid situation that a media segment is required to be referenced but is not yet available.

A dynamic MPD must remain valid for its entire validity duration after publishing. In other words, a dynamic MPD SHALL supply enough segment references to allow the time shift buffer to extend to now + MPD@minimumUpdatePeriod, where now is the current time according to the synchronized clock.

An unnecessary segment reference is one that is not defined as required by this chapter.

In a static MPD, a representation SHALL NOT contain unnecessary segment references, except when using indexed addressing in which case such segment references MAY be present.

In a dynamic MPD, a representation SHALL NOT contain unnecessary segment references except when any of the following applies, in which case an unnecessary segment reference MAY be present:

1. The segment reference is for future content and will eventually become necessary.

2. The segment reference is defined via indexed addressing.

3. The segment reference is defined by an <S> element that defines multiple references using S@r, some of which are necessary.

4. Removal of the segment reference is not allowed by content removal constraints.

There SHALL NOT be gaps or overlapping media segments in a representation.

Clients SHALL NOT present any samples from media segments that are entirely outside the period, even if such media segments are referenced.

Media segment start/end points MAY be unaligned with period start/end points except when using simple addressing, which requires the first media segment start point to match the period start point.

Note: Despite constraints on first media segment alignment, simple addressing allows the media samples within the first media segment to be unaligned with the period start point, allowing for synchronization of different reprensetations.

If a media segment overlaps a period boundary, clients SHOULD NOT present the samples that lie outside the period and SHOULD present the samples that lie either partially or entirely within the period.

Note: In the end, which samples are presented is entirely up to the client. It may sometimes be impractical to present media segments only partially, depending on the capabilties of the client platform, the type of media samples involved and any dependencies between samples.

3.3. Sample timeline

The samples within a representation exist on a linear sample timeline defined by the encoder that created the samples. One or more sample timelines are mapped onto the MPD timeline by metadata stored in or referenced by the MPD.

Note: A sample timeline is linear - encoders are expected to use an appropriate timescale and sufficiently large timestamp fields to avoid any wrap-around. If wrap-around does occur, a new period must be started in order to establish a new sample timeline.

The sample timeline is formed after applying any [ISOBMFF] edit lists.

A sample timeline SHALL be shared by all representations in the same adaptation set. Representations in different adaptation sets MAY use different sample timelines.

The sample timeline is measured in timescale units defined as a number of units per second. This value SHALL be present in the MPD as SegmentTemplate@timescale or SegmentBase@timescale (depending on the addressing mode).

Note: While optional in [MPEGDASH], the presence of the @timescale attribute is required by the interoperable timing model because the default value of 1 is unlikely to match any real-world content and is far more likely to indicate an unintentional content authoring error.

The point on the sample timeline indicated by @presentationTimeOffset SHALL be equivalent to the period start point on the MPD timeline. The value is provided by SegmentTemplate@presentationTimeOffset or SegmentBase@presentationTimeOffset, depending on the addressing mode, and has a default value of 0 timescale units.

Note: To transform a sample timeline position SampleTime to an MPD timeline position, use the formula MpdTime = Period@start + (SampleTime - @presentationTimeOffset) / @timescale.

@presentationTimeOffset SHALL NOT change between updates of a dynamic MPD.

3.4. Media segments

A media segment is an HTTP-addressable data structure that contains one or more media samples.

Media segments SHALL contain one or more consecutive media samples. Consecutive media segments in the same representation SHALL contain consecutive media samples.

Media segments SHALL contain the media samples that exactly match the time span on the sample timeline that is assigned to the media segment by the MPD, except when using simple addressing in which case a certain amount of inaccuracy MAY be present as defined in §3.5.3.1 Inaccuracy in media segment timing when using simple addressing.

Even when using simple addressing, the media segment that starts at or overlaps the period start point SHALL contain a media sample that starts at or overlaps the period start point and the media segment that ends at or overlaps the period end point SHALL contain a media sample that ends at or overlaps the period end point.

3.5. Segment addressing modes

This section defines the addressing modes that can be used for referencing media segments and initialization segments in interopreable DASH presentations.

Addressing modes not defined in this chapter SHALL NOT be used by DASH services. Clients SHOULD support all addressing modes defined in this chapter.

You SHOULD choose the addressing mode based on the nature of the content:

Content generated on the fly

Use explicit addressing.

Content generated in advance of publishing

Use explicit addressing or indexed addressing.

A service MAY use simple addressing which enables the packager logic to be very simple. This simplicity comes at a cost of reduced applicability to multi-period scenarios and reduced client compatibility.

All representations in the same adaptation set SHALL use the same addressing mode. Representations in different adaptation sets MAY use different addressing modes. Period-connected representations SHALL use the same addressing mode in every period.

3.5.1. Indexed addressing

A representation that uses indexed addressing consists of an [ISOBMFF] track file containing an index segment, an initialization segment and a sequence of media segments.

Note: This addressing mode is sometimes called "SegmentBase" in other documents.

Clauses in section only apply to representations that use indexed addressing.

Note: [MPEGDASH] makes a distinction between "segment" (HTTP-addressable entity) and "subsegment" (byte range of an HTTP-addressable entity). This document does not make such a distinction and has no concept of subsegments. Usage here matches the definition of CMAF Segment.

The MPD defines the byte range of the track file that contains the index segment. The index segment informs the client of all the media segments that exist, the time spans they cover on the sample timeline and their byte ranges.

Multiple representations SHALL NOT be stored in the same track file.

At least one Representation/BaseURL element SHALL be present in the MPD and SHALL contain a reference to the track file.

The SegmentBase@indexRange attribute SHALL be present in the MPD and SHALL reference the byte range of the segment index in the track file. The value SHALL be formatted as a byte-range-spec as defined in [RFC7233], referencing a single range of bytes.

The SegmentBase@timescale attribute SHALL be present and its value SHALL match the value of the timescale field in the index segment.

The AdaptationSet@subsegmentStartsWithSAP attribute SHALL be present in the MPD and SHALL have a value of 1 or 2, depending on the sample structure of the media segments.

We need to clarify how to determine the right value. #235

The SegmentBase/Initialization@range attribute SHALL reference the byte range of the initialization segment in the track file. The value SHALL be formatted as a byte-range-spec as defined in [RFC7233], referencing a single range of bytes. The Initialization@sourceURL attribute SHALL NOT be used.

<MPD xmlns="urn:mpeg:dash:schema:mpd:2011">
  <Period>
    <AdaptationSet subsegmentStartsWithSAP="1">
      <Representation>
        <BaseURL>showreel_audio_dashinit.mp4</BaseURL>
        <SegmentBase timescale="48000" presentationTimeOffset="8100" indexRange="848-999">
          <Initialization range="0-847"/>
        </SegmentBase>
      </Representation>
    </AdaptationSet>
  </Period>
</MPD>

3.5.1.1. Structure of the index segment

The index segment SHALL consist of a single Segment Index Box (sidx) as defined by [ISOBMFF]. The field layout is as follows:

aligned(8) class SegmentIndexBox extends FullBox('sidx', version, 0) {
  unsigned int(32) reference_ID;
  unsigned int(32) timescale;

  if (version==0) {
    unsigned int(32) earliest_presentation_time;
    unsigned int(32) first_offset;
  }
  else {
    unsigned int(64) earliest_presentation_time;
    unsigned int(64) first_offset;
  }

  unsigned int(16) reserved = 0;
  unsigned int(16) reference_count;

  for (i = 1; i <= reference_count; i++)
  {
    bit (1)              reference_type;
    unsigned int(31)     referenced_size;
    unsigned int(32)     subsegment_duration;
    bit(1)               starts_with_SAP;
    unsigned int(3)      SAP_type;
    unsigned int(28)     SAP_delta_time;
  }
}

The values of the fields SHALL be as follows:

Note: This matches the definitiuon in [ISOBMFF] but is simply re-worded into an [MPEGDASH] context to better guide implementations in choosing the correct values.

reference_ID

The track_ID of the track that contains the data of this representation.

timescale

Same as the timescale field of the Media Header Box and same as the SegmentBase@timescale attribute in the MPD.

earliest_presentation_time

The start timestamp of the first media segment on the sample timeline, in timescale units.

first_offset

Distance from the anchor point to the first media segment.

reference_count

Total number of media segments referenced by the index segment.

reference_type

0

referenced_size

Size of the media segment in bytes. Media segments are assumed to be consecutive, so this is also the distance to the start of the next media segment.

subsegment_duration

Duration of the media segment in timescale units.

starts_with_SAP

1

SAP_type

Either 1 or 2, depending on the sample structure in the media segment.

SAP_delta_time

0

We need to clarify how to determine the right value for SAP_type. #235

3.5.1.2. Moving the period start point (indexed addressing)

When splitting periods in two or performing other types of editorial timing adjustments, a service might want to start a period at a point after the "natural" start point of the representations within.

For representations that use indexed addressing, perform the following adjustments to set a new period start point:

1. Update SegmentBase@presentationTimeOffset to indicate the desired start point on the sample timeline.

2. Update Period@duration to match the new duration.

3.5.2. Explicit addressing

A representation that uses explicit addressing consists of a set of media segments accessed via URLs constructed using a template defined in the MPD, with the exact time span covered by each media segment described in the MPD.

Note: This addressing mode is sometimes called "SegmentTemplate with SegmentTimeline" in other documents.

Clauses in section only apply to representations that use explicit addressing.

The MPD SHALL contain a SegmentTemplate/SegmentTimeline element that contains a set of segment references which satisfies the requirements defined in this document. The references exist as a sequence of S elements, each of which references one or more media segments with start time S@t and duration S@d timescale units on the sample timeline. The SegmentTemplate@duration attribute SHALL NOT be present.

To enable concise reference definitions, an S element may be a repeating segment reference that indicates a number of repeated consecutive media segments with the same duration. The value of S@r SHALL indicate the number of additional consecutive media segments that exist.

Note: Only additional references are counted, so S@r=5 indicates a total of 6 consecutive media segments with the same duration.

The start time of a media segment SHALL be calculated from the start time and duration of the previous media segment if not specified by S@t. There SHALL NOT be any gaps or overlap between media segments.

The value of S@r SHALL be nonnegative, except for the last S element which MAY have a negative value in S@r, indicating that the repeated references continue indefinitely up to a media segment that either ends at or overlaps the period end point.

Updates to a dynamic MPD MAY add more S elements, remove expired S elements, increment SegmentTemplate@startNumber, add the S@t attribute to the first S element or increase the value of S@r on the last S element but SHALL NOT otherwise modify existing S elements.

The SegmentTemplate@media attribute SHALL contain the URL template for referencing media segments, using the $Time$ or $Number$ template variable to unique identify media segments. The SegmentTemplate@initialization attribute SHALL contain the URL template for referencing initialization segments.

If using $Number$ addressing, the number of the first segment reference SHALL be defined by SegmentTemplate@startNumber (default value 1). The S@n attribute SHALL NOT be used - segment numbers form a continuous sequence starting with SegmentTemplate@startNumber.

<MPD xmlns="urn:mpeg:dash:schema:mpd:2011">
  <Period duration="PT900S">
    <AdaptationSet>
      <Representation>
        <SegmentTemplate timescale="1000" presentationTimeOffset="900"
            media="video/$Time$.m4s" initialization="video/init.mp4">
          <SegmentTimeline>
            <S t="900" d="4001" r="224" />
          </SegmentTimeline>
        </SegmentTemplate>
      </Representation>
    </AdaptationSet>
  </Period>
</MPD>

<MPD xmlns="urn:mpeg:dash:schema:mpd:2011">
  <Period>
    <AdaptationSet>
      <Representation>
        <SegmentTemplate timescale="1000" presentationTimeOffset="810"
            media="video/$Time$.m4s" initialization="video/init.mp4">
          <SegmentTimeline>
            <S t="120" d="8520"/>
            <S d="8640"/>
            <S d="8600"/>
            <S d="8680"/>
            <S d="9360" r="1"/>
            <S d="8480"/>
            <S d="9080"/>
            <S d="6440"/>
            <S d="10000"/>
            <S d="8360"/>
          </SegmentTimeline>
        </SegmentTemplate>
      </Representation>
    </AdaptationSet>
  </Period>
</MPD>

3.5.2.1. Moving the period start point (explicit addressing)

When splitting periods in two or performing other types of editorial timing adjustments, a service might want to start a period at a point after the "natural" start point of the representations within.

For representations that use explicit addressing, perform the following adjustments to set a new period start point:

Note: See §3.2 Representations and §3.8.5.2 Removing content from the MPD to understand the constraints that apply to segment reference removal.

3.5.3. Simple addressing

Once we have a specific @earliestPresentationTime proposal submitted to MPEG we need to update this section to match. See #245.

A representation that uses simple addressing consists of a set of media segments accessed via URLs constructed using a template defined in the MPD, with the nominal time span covered by each media segment described in the MPD.

Simple addressing defines the nominal time span of each media segment in the MPD. The true time span covered by samples within the media segment can be slightly different than the nominal time span. See §3.5.3.1 Inaccuracy in media segment timing when using simple addressing.

Note: This addressing mode is sometimes called "SegmentTemplate without SegmentTimeline" in other documents.

Clauses in section only apply to representations that use simple addressing.

The SegmentTemplate@duration attribute SHALL define the nominal duration of a media segment in timescale units.

The set of segment references SHALL consist of the first media segment starting exactly at the period start point and all other media segments following in a consecutive series of equal time spans of SegmentTemplate@duration timescale units, ending with a media segment that ends at or overlaps the period end time.

The SegmentTemplate@media attribute SHALL contain the URL template for referencing media segments, using either the $Time$ or $Number$ template variable to uniquely identify media segments. The SegmentTemplate@initialization attribute SHALL contain the URL template for referencing initialization segments.

If using $Number$ addressing, the number of the first segment reference SHALL be defined by SegmentTemplate@startNumber (default value 1).

<MPD xmlns="urn:mpeg:dash:schema:mpd:2011">
  <Period duration="PT900S">
    <AdaptationSet>
      <Representation>
        <SegmentTemplate timescale="1000" presentationTimeOffset="900"
            media="video/$Number$.m4s" initialization="video/init.mp4"
            duration="4001" startNumber="800" />
      </Representation>
    </AdaptationSet>
  </Period>
</MPD>

3.5.3.1. Inaccuracy in media segment timing when using simple addressing

When using simple addressing, the samples contained in a media segment MAY cover a different time span on the sample timeline than what is indicated by the nominal timing in the MPD, as long as no constraints defined in this document are violated by this deviation.

The allowed deviation is defined as the maximum offset between the edges of the nominal time span (as defined by the MPD) and the edges of the true time span (as defined by the contents of the media segment). The deviation is evaluated separately for each edge.

This allowed deviation does not relax any requirements that do not explicitly define an exception. For example, periods must still be covered with samples for their entire duration, which constrains the flexibility allowed for the first and last media segment in a period.

The deviation SHALL be no more than 50% of the nominal media segment duration and MAY be in either direction.

Note: This results in a maximum true duration of 200% (+50% outward extension on both edges) and a minimum segment duration of 1 sample (-50% inward from both edges would result in 0 but empty segments are not allowed).

Allowing inaccurate timing is intended to enable reasoning on the sample timeline using average values for media segment timing. If the addressing data says that a media segment contains 4 seconds of data on average, a client can predict with reasonable accuracy which samples are found in which media segments, while at the same time the service is not required to publish per-segment timing data in the MPD. It is expected that the content is packaged with this contraint in mind (i.e. every segment cannot be inaccurate in the same direction - a shorter segment now implies a longer segment in the future to make up for it).

3.5.3.2. Moving the period start point (simple addressing)

When splitting periods in two or performing other types of editorial timing adjustments, a service might want to start a period at a point after the "natural" start point of the representations within.

Simple addressing is challenging to use in such scenarios. You SHOULD convert simple addressing representations to use explicit addressing before adjusting the period start point or splitting a period. See §3.5.3.3 Converting simple addressing to explicit addressing.

The rest of this chapter provides instructions for situations where you choose not to convert to explicit addressing.

To move the period start point, for representations that use simple addressing:

• Every simple addressing representation in the period must contain a media segment that starts at the new period start point.

• Media segments starting at the new period start point must contain a sample that starts at or overlaps the new period start point.

Note: If you are splitting a period, also keep in mind the requirements on period end point sample alignment for the period that remains before the split point.

Finding a suitable new start point that conforms to the above requirements can be very difficult. If inaccurate timing is used, it may be altogether impossible. This is a limitation of simple addressing.

Having ensured conformance to the above requirements for the new period start point, perform the following adjustments:

3.5.3.3. Converting simple addressing to explicit addressing

It may sometimes be desirable to convert a presentation from simple addressing to explicit addressing. This chapter provides an algorithm to do this.

Simple addressing allows for inaccuracy in media segment timing. No inaccuracy is allowed by explicit addressing. The mechanism of conversion described here only applies when there is no inaccuracy. If the nominal time spans in original the MPD differ from the true time spans of the media segments, re-package the content from scratch using explicit addressing instead of converting.

To perform the conversion, execute the following steps:

<MPD xmlns="urn:mpeg:dash:schema:mpd:2011">
  <Period duration="PT900S">
    <AdaptationSet>
      <Representation>
        <SegmentTemplate timescale="1000" presentationTimeOffset="900"
            media="video/$Number$.m4s" initialization="video/init.mp4"
            duration="4001" startNumber="800" />
      </Representation>
    </AdaptationSet>
  </Period>
</MPD>

<MPD xmlns="urn:mpeg:dash:schema:mpd:2011">
  <Period duration="PT900S">
    <AdaptationSet>
      <Representation>
        <SegmentTemplate timescale="1000" presentationTimeOffset="900"
            media="video/$Number$.m4s" initialization="video/init.mp4"
            startNumber="800">
          <SegmentTimeline>
            <S t="900" d="4001" r="224" />
          </SegmentTimeline>
        </SegmentTemplate>
      </Representation>
    </AdaptationSet>
  </Period>
</MPD>

3.5.4. Expanding URL template variables

Expansion rules for URL template variables such as $Time$ and $Number are defined by [MPEGDASH].

The set of string formatting suffixes used SHALL be restricted to %0[width]d.

Note: The string format suffixes are not intended for general-purpose string formatting. Restricting it to only this single suffix enables the functionality to be implemented without a string formatting library.

3.6. Segment alignment

Media segments are said to be aligned if the start/end points of all media segments on the sample timeline are equal in all representations that belong to the same adaptation set.

Media segments SHALL be aligned. When using simple addressing or explicit addressing, this means AdaptationSet@segmentAlignment=true in the MPD. When using indexed addressing, this means AdaptationSet@subsegmentAlignment=true in the MPD.

Equivalent aligned media segments in different representations SHALL contain samples for the same true time span, even if using simple addressing with inaccurate media segment timing.

3.7. Period connectivity

In certain circumstances content may be offered such that a representation is technically compatible with the content of a representation in a previous period. Such representations are period-connected.

Initialization segments of period-connected representations SHALL be functionally equivalent (i.e. the initialization segment from any period-connected representation can be used to initialize playback of any period-connected representation).

Note: Connectivity is generally achieved by using the same encoder to encode the content of multiple periods using the same settings. Keep in mind, however, that decryption is also a part of the client media pipeline - it is not only the codec parameters that are configured by the initialization segment.

Such content SHOULD be signaled in the MPD as period-connected. This is expected to help clients ensure seamless playback across period transitions. Any subset of the representations in a period MAY be period-connected with their counterparts in a future or past period. Period connectivity MAY be chained across any number of periods.

An MPD MAY contain unrelated periods between periods that contain period-connected representations.

The sample timelines of period-connected representations MAY be mutually discontinuous (e.g. due to skipping some content, encoder clock wrap-around or editorial decisions).

The following signaling SHALL be used to identify period-connected representations across two periods:

• Representation@id is equal.

• AdaptationSet@id is equal.

• The adaptation set in the second period has a supplemental property with:

  • @shemeIdUri set to urn:mpeg:dash:period-connectivity:2015.

  • @value set to the Period@id of the first period.

Note: Not all representations in an adaptation set need to be period-connected. For example, if a new period is introduced to add a representation that contains a new video quality level, all other representations will likely be connected but not the one that was added.

As a period may start and/or end in the middle of a media segment, the same media segment MAY simultaneously exist in two period-connected representations, with one part of it scheduled for playback during the first period and the other part during the second period. This is likely to be the case when no sample timeline discontinuity is introduced by the transition.

Clients SHOULD NOT present a media segment twice when it occurs on both sides of a period transition in a period-connected representation.

Clients SHOULD ensure seamless playback of period-connected representations in consecutive periods.

Note: The exact mechanism that ensures seamless playback depends on client capabilities and will be implementation-specific. Any shared media segment overlapping the period boundary may need to be detected and deduplicated to avoid presenting it twice.

3.7.1. Period continuity

In addition to period connectivity, [MPEGDASH] defines period continuity, which is a special case of period connectivity where the two samples on the boundary between the connected representations are consecutive on the same sample timeline.

Note: The above can only be true if the sample boundary exactly matches the period boundary.

Period continuity MAY be signaled in the MPD when the above condition is met, in which case period connectivity SHALL NOT be simultaneously signaled on the same representation. Continuity implies connectivity.

The signaling of period continuity is the same as for period connectivity, except that the value to use for @schemeIdUri is urn:mpeg:dash:period-continuity:2015.

Clients MAY take advantage of any platform-specific optimizations for seamless playback that knowledge of period continuity enables; beyond that, clients SHALL treat continuity the same as connectivity.

3.8. Dynamic MPDs

This section only applies to dynamic MPDs.

Dynamic MPDs have two main factors that differentiate them from static MPDs:

1. Dynamic MPDs may change over time, with clients retrieving new snapshots of the MPD when the validity duration of the previous snapshot expires.

2. Playback of a dynamic MPD is synchronized to a real time clock (with some amount of client-chosen time shift allowed).

A dynamic MPD must conform to the constraints in this document not only at its moment of initial publishing but through the entire validity duration of the MPD (as defined by MPD@minimumUpdatePeriod).

3.8.1. Real time clock synchronization

It is critical for dynamic MPDs to synchronize the clocks of the service and the client. The time indicated by the clock does not necessarily need to match some universal standard as long as the two are mutually synchronized.

A dynamic MPD SHALL include at least one UTCTiming element that defines a clock synchronization mechanism. If multiple UTCTiming elements are listed, their order determines the order of preference.

A client presenting a dynamic MPD SHALL synchronize its local clock according to the UTCTiming elements in the MPD and SHALL emit a warning or error to application developers when clock synchronization fails, no UTCTiming elements are defined or none of the referenced clock synchronization mechanisms are supported by the client.

Note: The use of a "default time source" is not compatible with the interoperable timing model. The mechanism of time synchronization must always be explicitly defined in the MPD by every service and interoperable clients cannot assume a default time source.

The set of time synchronization mechanisms SHALL be restricted to the following schemes defined in [MPEGDASH]:

• urn:mpeg:dash:utc:http-xsdate:2014

• urn:mpeg:dash:utc:http-iso:2014

• urn:mpeg:dash:utc:http-ntp:2014

• urn:mpeg:dash:utc:ntp:2014

• urn:mpeg:dash:utc:http-head:2014

• urn:mpeg:dash:utc:direct:2014

3.8.2. Availability

A segment is available if an HTTP request to acquire it can be successfully performed to completion by a client. In a dynamic MPD, new media segments continuously become available and stop being available with the passage of time.

An availability window is a time span on the MPD timeline that determines which media segments are available. Each representation has its own availability window.

The availability window is calculated as follows:

In the absence of MPD@timeShiftBufferDepth the start of the availability window is MPD@availabilityStartTime.

A service SHALL ensure that all media segments that have their end point inside or at the end of the availability window are available.

It is the responsibility of the service to ensure that media segments are available to clients when they are described as available by the MPD. To define proper availability timing, one must consider the entire publishing flow through the CDN, not only when the origin server makes data available to the CDN.

Clients MAY attempt to acquire any available media segment. Clients SHALL NOT attempt to acquire media segments that are not available.

Clients SHOULD NOT assume that media segments described by the MPD as available are always available at the correct moment in time and SHOULD implement appropriate retry behavior.

3.8.3. Time shift buffer

The time shift buffer is a time span on the MPD timeline that defines a baseline for content that a client can present at the current moment in time according to the synchronized clock (now).

The following additional factors further constrain the set of media segments that can be presented at the current time:

1. §3.8.2 Availability - not every media segment in the time shift buffer is guaranteed to be available.

2. §3.8.4 Presentation delay - the service may define a delay that forbids the use of a section of the time shift buffer.

The time shift buffer extends from now - MPD@timeShiftBufferDepth to now.

In the absence of MPD@timeShiftBufferDepth the start of the time shift buffer is MPD@availabilityStartTime.

Clients MAY present samples from media segments that overlap the time shift buffer, assuming no other constraints forbid it. Clients SHALL NOT present samples from media segments that are entirely outside the time shift buffer (whether in the past or the future).

The start of the time shift buffer MAY be before the start of the first period.

A dynamic MPD SHALL contain a period that ends at or overlaps the end point of the time shift buffer, except when reaching the end of live content in which case the last period MAY end before the end of the time shift buffer.

Clients SHALL NOT allow seeking into regions of the time shift buffer that are not covered by periods.

3.8.4. Presentation delay

There is a natural conflict between the availability window and the time shift buffer. It is legal for a client to present media segments as soon as they overlap the time shift buffer, yet such media segments might not yet be available.

Furthermore, the delay between media segments entering the time shift buffer and becoming available might be different for different representations that use different segment durations. This difference may also change over time if a representation does not use a constant segment duration.

Clients SHALL calculate a suitable presentation delay to ensure that the media segments it schedules for playback are available and that there is sufficient time to download them once they become available. In essence, the presentation delay decreases the time shift buffer, creating an effective time shift buffer with a reduced duration.

Services MAY define the MPD@suggestedPresentationDelay attribute to provide a suggested presentation delay. Clients SHOULD use MPD@suggestedPresentationDelay when provided, ignoring the calculated value.

Note: As different clients might use different algorithms for calculating the presentation delay, providing MPD@suggestedPresentationDelay enables services to roughly synchronize the playback start position of clients.

The effective time shift buffer is the time span from the start of the time shift buffer to now - PresentationDelay.

Clients SHALL constrain seeking to the effective time shift buffer. Clients SHALL NOT attempt to present media segments that fall entirely outside the effective time shift buffer.

3.8.5. MPD updates

Dynamic MPDs MAY change over time. The nature of the change is not restricted unless such a restriction is explicitly defined.

Some common reasons to make changes in dynamic MPDs:

• Adding new segment references to an existing period.

• Adding new periods.

• Converting unlimited-duration periods to fixed-duration periods by adding Period@duration.

• Removing segment references and/or periods that have fallen out of the time shift buffer.

• Shortening an existing period when changes in content scheduling take place.

• Removing MPD@minimumUpdatePeriod to signal that MPD will no longer be updated (even though it may remain a dynamic MPD).

• Converting the MPD to a static MPD to signal that a live service has become available on-demand as a recording.

The following basic restrictions are defined here for MPD updates:

• MPD@id SHALL NOT change.

• MPD@availabilityStartTime SHALL NOT change.

• Period@id SHALL NOT change.

• Period@start SHALL NOT change.

• Period@duration SHALL NOT change except when explicitly allowed by other statements in this document.

• The adaptation sets present in a period (i.e. the set of AdaptationSet@id values) SHALL NOT change.

• The representations present in an adaptation set (i.e. the set of Representation@id values) SHALL NOT change.

• The functional behavior of a representation (identified by a matching Representation@id value) SHALL NOT change, neither in terms of metadata-driven behavior (including metadata inherited from adaptation set level) nor in terms of media segment timing. In particular:

  • SegmentTemplate@presentationTimeOffset SHALL NOT change.

  • SegmentBase@presentationTimeOffset SHALL NOT change.

Additional restrictions on MPD updates are defined by other parts of this document.

Clients SHOULD use @id to track period, adaptation set and representation identity across MPD updates.

The presence or absence of MPD@minimumUpdatePeriod SHALL be used by a service to signal whether the MPD might be updated (with presence indicating potential for future updates). The value of this field indicates the validity duration of the present snapshot of the MPD, starting from the moment it was retrieved.

Clients SHALL process state changes that occur during the MPD validity duration. For example new media segments will become available over time if they are referenced by the MPD and old ones become unavailable, even without an MPD update.

Note: A missing MPD@minimumUpdatePeriod attribute indicates an infinite validinity period (the MPD will never be updated). The value 0 indicates that the MPD has no validity after the moment it was retrieved. In such a situation, the client will have to acquire a new MPD whenever it wants to make new media segments available (no "natural" state changes will occur).

In practice, clients will also require some time to download and process an MPD update - a service SHOULD NOT assume perfect update timing. Conversely, a client SHOULD NOT assume that it can get all updates in time (it may already be attempting to buffer some media segments that were removed by an MPD update).

In addition to signaling that clients are expected to poll for regular MPD updates, a service MAY publish in-band events to schedule MPD updates at moments of its choosing.

3.8.5.1. Adding content to the MPD

There are two mechanisms for adding content:

• Additional segment references MAY be added to the last period.

• Additional periods MAY be added to the end of the MPD.

Multiple content adding mechanisms MAY be combined in a single MPD update.

Segment references SHALL NOT be added to any period other than the last period. An MPD update MAY combine adding segment references to the last period with adding of new periods.

Note: The duration of the last period cannot change as a result of adding segment references. A live service will generally use a period with an unlimited duration to continuously add new references.

When using simple addressing or explicit addressing, it is possible for a period to define an infinite sequence of segment references that extends to the end of the period (e.g. using SegmentTemplate@duration or r="-1"). Such self-extending reference sequences SHALL be considered equivalent to explicitly defined reference sequences that extend to the end of the period and clients MAY obtain new references from such sequences even between MPD updates.

An MPD update that adds content MAY be combined with an MPD update that removes content.

3.8.5.2. Removing content from the MPD

Removal of content is only allowed if the content to be removed is not yet available to clients and will not become available until clients receive the MPD update. See §3.8.2 Availability.

Let EarliestRemovalPoint be availability window end + MPD@minimumUpdatePeriod.

Note: As each representation has its own availability window, so does each representation have its own EarliestRemovalPoint.

Media segments that overlap or end before EarliestRemovalPoint might be considered by clients to be available at the time the MPD update is processed and therefore cannot be removed.

An MPD update removing content MAY remove any segment references to media segments that start after EarliestRemovalPoint at the time the update is published but SHALL NOT remove any other segment references.

The following mechanisms exist removing content:

• The last period MAY change from unlimited duration to fixed duration.

• The duration of the last period MAY be shortened.

• One or more periods MAY be removed entirely from the end of the MPD.

Multiple content removal mechanisms MAY be combined in a single MPD update.

The above mechanisms can be used in any extent as long as no constraints defined in this document are violated. In particular, consider the constraints on invalidating segment references defined in this chapter.

Clients SHALL NOT fail catastrophically if an MPD update removes already buffered data but MAY incur unexpected time shift or a visible transition at the point of removal. It is the responsibility of the service to avoid removing data that may already be in use.

In addition to editorial removal from the end of the MPD, content naturally expires due to the passage of time. Expired content also needs to be removed:

• Explicitly defined segment references (S elements) SHALL be removed when they have expired (i.e. the media segment end point has fallen out of the time shift buffer).

  • A repeating explicit segment reference (S element with @r != 0) SHALL NOT be removed until all repetitions have expired.

• Periods with their end points before the time shift buffer SHALL be removed.

Note: When using indexed addressing or simple addressing, removal of segment references from the end of the period only requires changing Period@duration.

An MPD update that removes content MAY be combined with an MPD update that adds content.

3.8.5.3. Update signaling via in-band events

Services MAY signal the MPD validity duration by embedding in-band messages into representations instead of specifying a fixed validity duration in the MPD. This allows services to trigger MPD refreshes at exactly the desired time and to avoid needless MPD refreshes.

The rest of this chapter only applies to services and clients that use in-band MPD validity signaling.

Services SHALL define MPD@minimumUpdatePeriod=0 and add an in-band event stream to every audio representation or, if no audio representations are present, to every video representation. The in-band event stream MAY also be added to other representations. The in-band event stream SHALL be identical in every representation where it is present.

The in-band event stream SHALL be signaled on the adaptation set level by an InbandEventStream element with @scheme_id_uri="urn:mpeg:dash:event:2012" and a @value of 1 or 3, where:

• A value of 1 indicates that in-band events only extend the MPD validity duration.

• A value of 3 indicates that in-band events also contain the updated MPD snapshot when updates occur.

Note: For a detailed definition of the mechanism and the event message data structures, see [MPEGDASH]. This chapter is merely a high level summary of the most important aspects relevant to interoperability.

Services SHALL emit in-band events as [MPEGDASH] emsg boxes to signal the validity duration using the following logic:

• Lack of an in-band MPD validity event in a media segment indicates that an MPD that was valid at the start of the media segment remains valid up to the end of the media segment (mapped to wall clock time).

• The presence of an in-band MPD validity event in a media segment indicates that the MPD with MPD@publishTime equal to the event’s publish_time field remains valid up to the event start time (mapped to wall clock time).

The in-band events used for signaling MPD validity duration SHALL have scheme_id_uri and value matching the InbandEventStream element. Clients SHALL NOT use in-band events for MPD validity update signaling if these fields on the events do not match the InbandEventStream element or if the InbandEventStream element is not present.

In-band events with value=3 SHALL provide an updated MPD in the event’s mpd field as UTF-8 encoded text without a byte order mark.

Multiple media segments MAY signal the same validity update event (identified by matching id field on event), enabling the signal to be delivered several segments in advance of the MPD expiration.

3.8.5.4. End of live content

Live services can reach a point where no more content will be produced - existing content will be played back by clients and once they reach the end, playback will cease.

Services SHALL define a fixed duration for the last period, remove the MPD@minimumUpdatePeriod attribute and cease performing MPD updates to signal that no more content will be added to the MPD. The MPD@type MAY be changed to static at this point or later if the service is to be converted to a static MPD for on-demand viewing.

3.9. XLink elements

Some XML elements in an MPD may be external to the MPD itself, delay-loaded by clients based on different triggers. This mechanism is called XLink and it enables client-side MPD composition from different sources. For the purposes of timing and addressing, it is important to ensure that the duration of each period can be accurately determined both before and after XLink resolution.

Note: XLink functionality in DASH is defined by [MPEGDASH] and [XLINK]. This document provides a high level summary of the behavior and defines interoperability requirements.

XLink elements are those in the MPD that carry the xlink:href attribute. When XLink resolution is triggered, the client will query the URL referenced by this attribute. What happens next depends on the result of this query:

Non-empty result containing a valid XML fragment

The entire XLink element is replaced with the query result. A single XLink element MAY be replaced with multiple elements of the same type.

Empty result or query failure

The XLink element remains as-is with the XLink attributes removed.

When XLink resolution is triggered depends on the value of the xlink:actuate attribute. A value of onLoad indicates resolution at MPD load-time, whereas a value of onRequest indicates resolution on-demand at the time the client wishes to use the element. The default value is onRequest.

Services SHALL publish MPDs that conform to the requirements in this document even before XLink resolution. This is necessary because the behavior in case of XLink resolution failure is to retain the element as-is.

<MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xlink="http://www.w3.org/1999/xlink" type="static">
  <Period duration="PT30S">
    ...
  </Period>
  <Period duration="PT0S" xlink:href="https://example.com/256479/clips/53473/as_period">
  </Period>
</MPD>

3.10. Forbidden techniques

Some aspects of [MPEGDASH] are not compatible with the interoperable timing model defined in this document. In the interest of clarity, they are explicitly listed here:

• The @presentationDuration attribute SHALL NOT be used.

3.11. Examples

This section is informative.

3.11.1. Offer content with imperfectly aligned tracks

It may be that for various content processing workflow reasons, some tracks have a different duration from others. For example, the audio track might start a fraction of a second before the video track and end some time before the video track ends.

You now have some choices to make in how you package these tracks into a DASH presentation that conforms to this document. Specifically, there exists the requirement that every representation must cover the entire period with media samples.

The simplest option is to define a single period that contains representations resulting from cutting the content to match the shortest common time span, thereby covering the entire period with samples. Depending on the nature of the data that is removed, this may or may not be acceptable.

If you wish to preserve track contents in their entirety, the most interoperable option is to add padding samples (e.g. silence or black frames) to all tracks to ensure that all representations have enough data to cover the entire period with samples. This may require customization of the encoding process, as the padding must match the codec configuration of the real content and might be impractical to add after the real content has already been encoded.

Another option that preserves track contents is to split the content into multiple periods that each contain a different set of representations, starting a new period whenever a track starts or ends. This enables you to ensure every representations covers its period with samples. The upside of this approach is that it can be done easily, requiring only manipulation of the MPD. The downside is that some clients may be unable to seamlessly play across every period transition.

You may wish to combine the different approaches, depending on the track, to achieve the optimal result.

Some clients are known to fail when transitioning from a period with audio and video to a period with only one of these components. You should avoid such transitions unless you have exact knowledge of the capabilities of your clients.

3.11.2. Split a period

There exist scenarios where you would wish to split a period in two. Common reasons would be:

• to insert an ad period in the middle of an existing period.

• parameters of one adaptation set change (e.g. KID or display aspect ratio), requiring a new period to update signaling.

• some adaptation sets become available or unavailable (e.g. different languages).

This example shows how an existing period can be split in a way that clients capable of seamless period-connected playback do not experience interruptions in playback among representations that are present both before and after the split.

Our starting point is a presentation with a single period that contains an audio representation with short samples and a video representation with slightly longer samples, so that media segment start points do not always overlap.

Note: Periods may be split at any point in time as long as both sides of the split remain in conformance to this document (e.g. each contains at least 1 media segment). Furthermore, period splitting does not require manipulation of the segments themselves, only manipulation of the MPD.

Let’s split this period at position 220. This split occurs during segment 3 for both representations and during sample 8 and sample 5 of the audio and video representation, respectively.

The mechanism that enables period splitting in the middle of a segment is the following:

• a media segment that overlaps a period boundary exists in both periods.

• representations that are split are signaled in the MPD as period-connected.

• a representation that is period-connected with a representation in a previous period is marked with the period connectivity descriptor.

• clients are expected to deduplicate boundary-overlapping media segments for representations on which period connectivity is signaled, if necessary for seamless playback (implementation-specific).

• clients are expected to present only the samples that are within the bounds of the current period (may be limited by client platform capabilities).

After splitting the example presentation, we arrive at the following structure.

If indexed addressing is used, both periods will reference all segments as both periods will use the same unmodified index segment. Clients are expected to ignore media segments that fall outside the period bounds.

Simple addressing has significant limitations on alignment at period start, making it unsuitable for some multi-period scenarios. See §3.5.3.2 Moving the period start point (simple addressing).

Other periods (e.g. ads) may be inserted between the two periods resulting from the split. This does not affect the addressing and timing of the two periods.

3.11.3. Change the default_KID

In encrypted content, the default_KID of a representation might need to be changed at certain points in time. Often, the changes are closely synchronized in different representations.

To perform the default_KID change, start a new period on every change, treating each representation as an independently changing element. With proper signaling, clients can perform this change seamlessly.

What about period connectivity? #238

The same pattern can also be applied to other changes in representation configuration.