| Internet-Draft | MMT for MoQ | November 2026 |
| Ramadan | Expires 6 May 2027 | [Page] |
This document specifies the use of MPEG Media Transport (MMT) as a container format for Media over QUIC (MoQ). MMT carries media over broadcast (ATSC 3.0, ARIB STD-B60), multicast, and unicast (QUIC/WebTransport) networks. This specification defines the mapping of MMT packets to MoQ objects, the "iso-ssbg1-v1" block profile under which the Forward Error Correction (FEC) mechanism for MoQ protects MMTP packets -- its source symbols, its Source FEC Payload ID trailer, its repair packet and its keyframe overlay -- the derivation of MoQ catalogs from MMTP signaling, and compatibility considerations for both ATSC 3.0 and ARIB STD-B60 systems, including the ATSC 3.0 signed region.¶
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."¶
This Internet-Draft will expire on 6 May 2027.¶
Copyright (c) 2026 IETF Trust and the persons identified as the document authors. All rights reserved.¶
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶
MPEG Media Transport (MMT) [ISO.23008-1] is a transport protocol designed for delivery of multimedia content over heterogeneous networks. MMT is supported as a transport in ATSC 3.0 broadcast television (Americas, South Korea), and is the basis for ARIB STD-B60 (Japan). MMT provides native support for:¶
This document defines how MMT-encapsulated media can be transported over MoQ [I-D.ietf-moq-transport], enabling:¶
This specification complements existing MoQ packaging formats. The MOQT Streaming Format (MSF) [I-D.ietf-moq-msf] carries media packaged with the Low Overhead Media Container (LOC) [I-D.ietf-moq-loc], and CMSF [I-D.ietf-moq-cmsf] carries media packaged in the Common Media Application Format (CMAF):¶
| Format | Container | Primary Use Case | FEC Block Profile |
|---|---|---|---|
| MSF (LOC) | WebCodecs chunks | Low-latency unicast | "moq-object-v1" |
| CMSF | CMAF fMP4 | Adaptive streaming | "moq-object-v1" |
| This spec (MMT) | MMTP + MPU (ISOBMFF) | Broadcast bridge | "iso-ssbg1-v1" |
The last column names the block profile under which the FEC mechanism of [MOQ-FEC-CORE] protects tracks of each packaging. "moq-object-v1" is the profile of [MOQ-FEC-CORE] Section 5.2: each loc or cmaf object whose Object Status is Normal occupies whole source symbols. "iso-ssbg1-v1" is the profile that Section 7.1 of this document defines: each FEC source packet (FEC Type 1) is one source symbol. Both are registered in the "MoQ FEC Block Profiles" registry of [MOQ-FEC-CORE] Section 15.3.¶
MMT packaging fits these cases:¶
CMAF packaging (CMSF) fits these cases:¶
A publisher that offers CMAF content can also carry it as an mmtp track under "iso-ssbg1-v1", where an MPU corresponds to a CMAF Fragment (Section 4.2). Such a track is also how CMAF content reaches multicast, because a multicast endpoint carries only mmtp tracks ([MOQ-MULTICAST] Section 4.1). The cmaf track and its mmtp rendition are separate FEC instances ([MOQ-FEC-CORE] Section 5.1), and symbols of one never repair the other.¶
This document is a companion of [MOQ-FEC-CORE], which defines the scheme-independent FEC mechanism for MoQ and the configuration data model that a catalog carries. [MOQ-FEC-BLOCK] specifies the Reed-Solomon and RaptorQ schemes, [MOQ-FEC-CATALOG] the catalog encoding of the configuration, and [MOQ-FEC-MULTICAST] the delivery of protected tracks over multicast, including the receiver repair policy. This document defines the "iso-ssbg1-v1" block profile, the part of that set that is specific to MMTP packets, and the ATSC 3.0 and ARIB STD-B60 interworking that relies on it. [MOQ-MULTICAST] defines multicast endpoint discovery and the multicast packet format on which the multicast delivery of mmtp tracks rests.¶
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.¶
MMTP: MMT Protocol - the packet layer of MMT (ISO 23008-1 Clause 9)¶
MPU: Media Processing Unit - a generic container for independently decodable timed or non-timed data ([ISO.23008-1] Clause 3.1.14)¶
MFU: Media Fragment Unit - a sample or subsample of timed media data, or an item of non-timed media data, within an MPU ([ISO.23008-1] Clause 9.3.1)¶
AL-FEC: Application-Layer Forward Error Correction¶
AU: Access Unit¶
SLT: Service Layer Table - ATSC 3.0 bootstrap signaling¶
MPT: MMT Package Table - signaling table containing MMT asset info¶
PA: Package Access - the signaling message that carries the tables a receiver needs to consume a Package, among them the MPT ([ISO.23008-1] Clause 10.3.2)¶
MPI: Media Presentation Information - signalling table carrying the presentation information document ([ISO.23008-1] Clause 10.3.8)¶
The FEC terms of this document -- Source Track, Repair Track, Repair Layer, FEC Instance, Source Symbol, Repair Symbol, Source Block, SS_ID, SS_Start, SSB_length, RS_ID, RSB_length, Encoding Symbol ID (ESI) and Block Profile -- are those of [MOQ-FEC-CORE] Section 2, and the RS_ID offset O_i and per-block repair symbol count P_i of layer i are those of [MOQ-FEC-CORE] Section 7.3.1. Under "iso-ssbg1-v1" the following take a form of their own, and the remaining terms are specific to this document:¶
Source Symbol: Under "iso-ssbg1-v1", one FEC source packet framed as Section 7.1 specifies.¶
Repair Track: Under "iso-ssbg1-v1", a repair track carries MMTP FEC repair packets ([ISO.23008-1] Section C.4.3; Section 7.3). The layer-0 repair track carries the FEC repair flow ([ISO.23008-1] 3.1.9) that an ISO AL-FEC message names for the source track's FEC encoded flow (3.1.7); further repair layers and a keyframe overlay are repair tracks that only the catalog declares (Section 8.3).¶
SS_ID: Under "iso-ssbg1-v1", the 32-bit sequence number of a source symbol is carried in the Source FEC Payload ID of its FEC source packet ([ISO.23008-1] Section C.5.2) and advances by one per FEC source packet (Section 7.2).¶
Source Block: Under "iso-ssbg1-v1", the ISO source symbol block ([ISO.23008-1] 3.1.34), generated from a source packet block (3.1.32).¶
FEC Block: The combination of SSB_length source symbols and P repair symbols; the ISO encoding symbol block ([ISO.23008-1] 3.1.5).¶
Keyframe Overlay: A second, independent FEC instance over a source
track whose source symbols are only the fragments of Random Access
Points, with its own source symbol count, its own repair symbols, and
its own repair track, declared with "scope": "keyframe"
(Section 7.4.1). An overlay is not a repair layer and carries no
layer index. It is defined only under "iso-ssbg1-v1".¶
Base Instance: Of a source track that a keyframe overlay protects, the instance that is not the overlay; of any other source track, its only instance.¶
Interleave Depth: When positive, the complete time span, in
milliseconds, of one FEC source/repair packet block, signaled as
interleaveDepthMs ([MOQ-FEC-CORE] Section
9.3); the catalog form of
protection_window_time in [ISO.23008-1] Sections 11.2 and C.6.3.
Zero is not a millisecond span: under "iso-ssbg1-v1" it selects the ISO
packet-count alternative (Section 7.5).¶
Object Transmission Information (OTI): Parameters required to configure a RaptorQ decoder, as defined in [RFC6330] and derived by [MOQ-FEC-BLOCK] Section 6.1.¶
MMT uses a layered architecture:¶
+---------------------------------------------+ | Application Layer | +---------------------------------------------+ | MPU (Media Processing Unit) | | +---------+---------+---------+---------+ | | | MFU 0 | MFU 1 | MFU 2 | ... | | | | (IDR) | (P) | (P) | | | | +---------+---------+---------+---------+ | +---------------------------------------------+ | MMTP (MMT Protocol) | | +-----------------------------------------+| | | MMTP Header | Payload (MPU fragment) || | | 12 bytes | Variable || | +-----------------------------------------+| +---------------------------------------------+ | Transport: UDP/IP (broadcast/multicast) | | QUIC/WebTransport (unicast) | +---------------------------------------------+¶
MMTP Header (12 bytes minimum) {
Version (2),
Packet Counter Flag (C) (1),
FEC_type (FEC) (2),
Reserved (1),
Extension Flag (X) (1),
RAP Flag (R) (1),
Reserved (2),
type (6),
Packet ID (16),
Timestamp (32),
Packet Sequence Number (32),
[Packet Counter (32)], // Present if C=1 (never: C=0)
[Header Extension (..)] // Present if X=1
}
¶
The fixed fields sum to 96 bits (three 32-bit words). Byte 0 is version(2) | packet_counter_flag(1) | FEC_type(2) | reserved(1) | extension_flag(1) | RAP_flag(1); byte 1 is reserved(2) | type(6), i.e. the packet type occupies the low six bits of the second byte ([ISO.23008-1] Clause 9.2).¶
Senders MUST set all reserved bits in the MMTP packet header to zero. Receivers MUST ignore their values after validating the header; they MUST NOT assign private semantics to reserved bits.¶
Every packet of an MMTP packet flow -- media, repair, and signaling packets alike, on every delivery path -- MUST set the packet counter flag C to 0, so no packet carries the Packet Counter field. [ISO.23008-1] Clause 9.2.3 requires all packets of an MMTP flow to share one setting of C; this profile fixes that setting at 0. The MMTP packet header is therefore 12 bytes unless X = 1 adds a header extension.¶
This document profiles MMTP version 0 (V = 00) of
ISO/IEC 23008-1:2023 as amended by Amd 1:2025. A packet whose version
is not 00 MUST be rejected before its type field is parsed: version
1 uses a different type-field layout and is not defined by this
profile. The C = 0 requirement above is also a conformance gate;
packets with C = 1 MUST be rejected. An X = 1 header extension is
parsed according to its declared length, but this document defines no
additional version or packet-counter semantics.¶
Key fields for MoQ mapping:¶
Packet ID: Maps to MoQ track within namespace. This document reserves packet_id 0 for the MMTP signaling flow: MMTP packets of packet type 0x02 carrying signaling messages, such as the PA message of a multicast endpoint flow ([MOQ-MULTICAST] Section¶
packetId the endpoint advertises for the
track ([MOQ-MULTICAST] Section
4.1). On every delivery path, a
receiver MUST NOT deliver an MMTP packet on packet_id 0 to a media
or repair decoder, including one received as the payload of an
object of a media or repair track: a receiver that processes MMTP
signaling reads it as a signaling message, and any other receiver
discards it ([MOQ-MULTICAST] Section
5 states the same rule for
every unlisted packet_id on multicast)¶
packet_id, incremented by one modulo
2^32 across the flow ([ISO.23008-1] Clause 9.2.3); carried inside
the object payload for wrap-aware MMTP-layer loss/ordering detection.
It is not the MoQ Object ID. Object IDs number the objects of a
group in source order, starting at zero in each group (Section 4.6).¶
A signaling packet (packet type 0x02, on packet_id 0) MUST carry its messages in the signaling message mode of [ISO.23008-1] Clause 9.3.4: its MMTP payload begins with the 2-byte payload header of Clause 9.3.4.2:¶
Signaling Message Payload Header {
f_i (2), // fragmentation indicator
RES (4), // reserved, set to 0
H (1), // 1: each MSG_length is 32 bits
A (1), // aggregation_flag
frag_count (8), // fragmentation counter
}
¶
When A = 1, a MSG_length field of 16 bits, or 32 bits when H = 1, precedes each aggregated message; when A = 0 no MSG_length field is present. The signaling message follows ([ISO.23008-1] Clause 10.2), beginning with its message_id. A packet that carries one complete signaling message without aggregation has f_i = 00, A = 0, and frag_count = 0, and its sender sets H = 0 (H has no effect when A = 0), so its payload header is the two bytes 0x00 0x00 and the message_id immediately follows them. A receiver ignores H when A = 0 and parses the header field by field, never by matching those two bytes.¶
An MMT stream maps to MoQ tracks as follows:¶
| MMT Component | MoQ Mapping |
|---|---|
| Package | Namespace |
| Asset, by Packet ID (video) | Track "video" |
| Asset, by Packet ID (audio) | Track "audio" |
| MPU (one MPU per group) | Group; number derived from media time (Section 4.4.1) |
| MPU data unit (FT=0, FT=1, or FT=2) | Subgroup ID |
| MMTP packet (fragment) within a data unit | Object; Object IDs count the group's objects from 0 in source order (Section 4.6) |
| AL-FEC signaling message (OPTIONAL, packet_id 0) | Object in Subgroup 0 immediately before FT=0 (Section 8) |
| AL-FEC repair | Track "video/repair" |
[I-D.ietf-moq-msf] Section 6 maps each object to a new stream. An mmtp track does not follow that rule. It maps each MPU data unit to one subgroup (Section 4.3), and the objects of a subgroup share one stream ([I-D.ietf-moq-transport] Section 2.2), so each data unit's fragments travel in order on one stream. An mmtp track keeps the Group and Object numbering of [I-D.ietf-moq-msf] Sections 6.1 and 6.2: Group numbers are unique and increase monotonically (Section 4.4.1), and Object IDs start at zero in each group and increase by one (Section 4.6).¶
Each MoQ object carries one complete MMTP packet:¶
MoQ Object Payload {
MMTP Packet Header (12+ bytes),
MMTP Payload (MPU-mode payload header,
then an FT=0, FT=1, or FT=2 data-unit fragment),
[Source FEC Payload ID (32)], // trails FEC Type=1 packets
}
¶
The object payload begins at byte zero of the preserved MMTP packet
header. It has no MoQ-private envelope, length prefix, harness prefix,
or other undocumented bytes before that header. For an FEC Type 1
packet, the Source FEC Payload ID and any profile-defined MMTP header
extensions remain in their specified wire locations. No object-layer
prefix, suffix, or private replacement for those fields is permitted.
[MOQ-FEC-CORE] Section 5.6 requires
every source object to carry the FEC Source Info Object Property (type
0x11); on an mmtp source object its value is the four-byte form of
that section, which names instance 0 and carries the network-byte-order
SS_ID of the base instance, the value of the packet's Source FEC
Payload ID (Section 7.2); the property is object metadata
and is not inserted into this payload.¶
The one exception is the OPTIONAL AL-FEC signaling object of Section 8: its payload is a complete MMTP signaling packet (packet type 0x02) on packet_id 0 with FEC Type 0. Its MMTP payload is the signaling message payload header of Section 3.1 followed by one complete AL-FEC message (f_i = 00, A = 0), rather than an MPU-mode payload header and data-unit fragment, and it carries no Source FEC Payload ID. No other object on an mmtp source track uses packet type 0x02 or packet_id 0.¶
Throughout this document, "MMTP packet" means the header-included wire unit and "MMTP payload" means the bytes that follow the packet header, per [ISO.23008-1]. Both MMTP header layers are always preserved: the packet header carries the FEC Type, RAP Flag, and timestamp, and the MPU-mode payload header (the first bytes of the MMTP payload) carries the Fragmentation Indicator and MPU sequence number that receivers and the FEC layer depend on. A publisher that wishes to deliver raw ISOBMFF fragments without MMTP encapsulation publishes the track with CMAF packaging [I-D.ietf-moq-cmsf] instead. In this profile an MPU can contain one or more ISOBMFF movie fragments: FT=1 carries one movie fragment's metadata and each FT=2 data unit carries one whole sample or one producer-authored subsample of a sample (Section 5.2.1).¶
This section defines the object layout of mfu mode
(mmtpMode: "mfu", Section 13.1).¶
Group boundaries align with MPU boundaries, and subgroup boundaries align with MPU data-unit boundaries:¶
ftyp, mmpu, and moov, and uses FI=0. When the
publisher carries the OPTIONAL AL-FEC signaling object of Section 8,
that object is in Subgroup 0 of the same group and is the object
immediately before the FT=0 object; Subgroup 0 then carries exactly
these two objects. No other object precedes FT=0, and a group
carries at most one signaling object.¶
offset order. Section 5.2 defines the only permitted
fragmentation state machine.¶
The MPU sequence number remains present in-band (in the MMTP payload header of every MPU fragment) and delimits which packets belong to the same group, but it is not the MoQ Group number. Publishers MUST NOT copy the MPU sequence number into the Group number, and subscribers MUST NOT derive media time or switching decisions from it. (The MPU sequence number is a per-Asset counter. It starts at zero for the first MPU in an Asset and increments by one for every following MPU ([ISO.23008-1] Clause 7.3.3), but it is Asset-local: independently packetized renditions need not assign the same MPU number to the same media instant. Counter-based group numbering therefore cannot satisfy the switching-set alignment requirements of Section 4.4.)¶
For an ISOBMFF source, the publisher MUST preserve initialization in
FT=0, MUST copy the authored moof and the original mdat prefix
through the byte before the first media sample byte-for-byte into FT=1,
and MUST copy the referenced sample bytes byte-for-byte and without
transformation into FT=2. A receiver can therefore reconstruct the
authored movie fragment without synthesizing box fields, sample timing,
or media bytes (Section 5.1).¶
Mapping each data unit to its own subgroup makes a complete metadata unit or sample the unit of in-order delivery, loss, and FEC. A publisher MUST send the objects of an mmtp track in subgroups, not as datagrams: a datagram object belongs to no subgroup ([I-D.ietf-moq-transport] Section 2.2) and cannot carry this mapping. Subgroups can be delivered concurrently, and a subgroup can still be lost, for example when its stream is reset on a delivery timeout ([I-D.ietf-moq-transport] Sections 5.2 and 11.3.2); the resulting gap invalidates the affected movie fragment rather than authorizing inferred metadata, timing, or cadence.¶
When multiple tracks represent alternative renditions of the same media (e.g., an adaptive bitrate (ABR) ladder), they form a switching set: a set of alternate tracks that MUST be time-aligned per [I-D.ietf-moq-msf] Section 4.2 (the same construct that [I-D.ietf-moq-cmsf] Section 3.2 applies to CMAF tracks). MoQ Group boundaries MUST be media-time-aligned across all tracks of a switching set so that subscribers can switch at group boundaries. Tracks with equal group durations MUST carry the same Group number for the same media interval. Only an intra-only mmtp track can have a shorter group duration than the other tracks, under the exception of Section 4.4.2, which also defines its Group number mapping. A cmaf member keeps equal group durations and the Group number alignment that [I-D.ietf-moq-cmsf] Section 3.2 requires. The switch is without discontinuity when the MPUs are closed groups of pictures (GOPs). On a join or switch at a Group boundary, leading pictures (Section 4.4.1) that reference the preceding Group are not decodable, and presentation of the new track begins at its RAP. If the old track's next MPU has leading pictures, neither track has decodable pictures for their presentation interval, and the receiver conceals that gap.¶
Switching-set membership is declared in the catalog via the per-track field:¶
altGroup (OPTIONAL, unsigned integer)¶
defined by [I-D.ietf-moq-msf] Section 5.2.12: an integer
identifying a group of tracks that are alternate versions of one
another. Two tracks are members of the same switching set if and only if both
carry the altGroup field and the values are equal. A track that
omits altGroup is a member of no switching set, and the
switching-set requirements of this section do not bind it. A repair
track (Section 8.2) carries no altGroup. It is not an alternate
rendition, and its Groups are FEC blocks ([MOQ-FEC-CORE] Section
7.6), not media-time-aligned
Groups. A subscriber finds the repair tracks of the rendition it
selects through the repairTrack of each of that rendition's FEC
instances and the source fields of its repair tracks
([MOQ-FEC-CATALOG] Section 3.2; the data model of
[MOQ-FEC-CORE] Section 10.2).¶
The Group number of every mmtp-packaged track MUST be computed from media time by the formula in this section. This single definition is what makes group boundaries align across the renditions of a switching set. FEC interleave timing is independent of this grouping; Section 7.5 does not derive FEC block boundaries or a number of Groups per FEC block from this formula:¶
relative_ticks = ticks - presentationAnchorTicks group_number = floor(relative_ticks / groupDurationTicks)¶
where:¶
ticks is the presentation timestamp (PTS) of the MPU's first AU in
decode order, expressed in the track's timescale (Hz, Section
4.4.2) as a signed integer. That AU is first in the MPU payload, as
ISO/IEC 23008-1 requires, and is the MPU's RAP (Section 4.3).
Leading pictures, which follow the RAP in decode order but present
before it, do not select ticks, even when they present within the
preceding MPU's Group interval.¶
groupDurationTicks is the per-track group duration expressed
in the same timescale, as a positive integer. The catalog
signals it either directly or through an exact millisecond
conversion as specified in Section 4.4.2.¶
presentationAnchorTicks is the signed timestamp in this track's
timescale that corresponds to presentation time zero. It includes
the publisher's edit-list, encoder-priming, and composition-time
decisions, and MUST be applied before the formula. A track that is
intended to be rendered with another audio or video track MUST signal
this field; a track that is not part of a synchronized presentation
MAY omit it and use zero as its local origin.¶
The publisher MUST choose presentationAnchorTicks no later than the
smallest ticks value of the MPUs it admits for the track. If an
MPU's ticks is earlier than the anchor, the publisher MUST reject
that MPU. Before the first group is published, the publisher MAY
instead revise the catalog anchor and validate the complete initial
input again; after publication, the anchor is immutable for that track
instance. It MUST NOT clamp the Group number to zero. The floor in
the formula is mathematical floor (floor division), not truncation
towards zero. With the anchor requirement, every admitted MPU has a
non-negative relative_ticks value. The requirement binds ticks
only; a leading picture of an admitted MPU can present before the
anchor.¶
Timeline origin: ticks is measured on the track's media
presentation timeline; presentationAnchorTicks maps that track to
the shared presentation timeline. For every set of audio and video
tracks intended to be rendered together, the publisher MUST choose one
presentation origin and MUST signal an anchor for each track such that¶
presentation_time = (ticks - presentationAnchorTicks) / timescale¶
has the same rational value for samples representing the same media
instant. This is a media-time relation, not an NTP or wall-clock
relation, and it is not the Timestamp field of the MMTP packet header
(Section 3.1). For
ISOBMFF-derived MPU content, the anchor is applied to the composition
timeline after edit-list and priming adjustments. All tracks of a
switching set MUST use the same presentation origin as well. A
publisher that resets the timeline (a timeline discontinuity) MUST
reset it identically, at the same media instant, across every affected
audio, video, and switching-set track. A discontinuity on an existing
track does not reset its Group number sequence.¶
This Group anchor is a MoQ scheduling and switching key only. ISO/IEC
23008-1 presentation timing remains authoritative: the presentation time
of an MPU is signaled by the presentation information (or the
mpu_timestamp_descriptor when no presentation information is present),
with any applicable elst offset, and is not rewritten by this Group
number calculation.¶
For each track, when the MPU changes, the computed Group number MUST be strictly greater than the Group number of the preceding admitted MPU. The publisher MUST NOT put two different MPUs in one Group, reuse a Group number, or alter authored timestamps to satisfy this rule. A publisher that restarts or declares a discontinuity MUST retain the last Group number for the existing track instance and resume with a greater number; numeric gaps are allowed. The publisher SHOULD signal the skipped numbers with the PRIOR_GROUP_ID_GAP Object Property ([I-D.ietf-moq-transport] Section 10.8), as [I-D.ietf-moq-msf] Section 6.1 recommends. If it cannot resume with a greater number while preserving the formula and authored timing, it MUST close that track instance and start a new track with a distinct (namespace, track name) pair before emitting the reset MPU. That new identity is the Group-number scope for the restarted stream; it MUST NOT be presented as a continuation of the old track. Group-number state is transport state; this document defines no new catalog field for it.¶
Because its Group numbers follow media time, an mmtp publisher cannot start a Group early. It MUST NOT set the DYNAMIC_GROUPS Track Property ([I-D.ietf-moq-transport] Section 10.6) to 1.¶
For two tracks A and B, a receiver can verify that timestamps describe the same presentation instant without floating-point conversion:¶
(ticksA - presentationAnchorTicksA) * timescaleB ==
(ticksB - presentationAnchorTicksB) * timescaleA
¶
The group-duration agreement rule in this section applies after the
same anchor conversion. Consequently, equal rational values of
groupDurationTicks / timescale and correctly paired
presentationAnchorTicks values are what align 30-fps video with
audio; per-track timescale and groupDurationTicks values alone do
not establish synchronization.¶
All arithmetic is in integers: relative_ticks and groupDurationTicks
are integers, floor(relative_ticks / groupDurationTicks) is integer division,
and implementations MUST NOT compute the group number in floating
point or apply any tolerance. An input held in another unit
(seconds, milliseconds, a different timescale) MUST be converted to
an exact integer tick count before the formula is applied; if that
conversion is not exact, the publisher's timescale or group-duration
choice is wrong (Section 4.4.2) -- implementations MUST NOT round to
compensate.¶
For ISOBMFF content, authored composition timestamps use the media
track's mdhd timescale. If that differs from the catalog timescale,
the publisher MUST convert the adjusted composition timestamps by the
exact ratio timescale / mdhd.timescale before applying this formula.
It MUST reject a timestamp that is not an integer in the selected
catalog clock; it MUST NOT relabel media ticks as asset-clock ticks.¶
Worked example. A switching set of two video renditions of the same
content: rendition A with timescale: 90000, rendition B with
timescale: 44100, both with presentationAnchorTicks: 0 and
groupDurationMs: 2000 in the catalog (Section 4.4.2). The sample
that is 10.5 seconds into the shared presentation timeline has an
integer timestamp in both timescales:¶
A: groupDurationTicks = 2000 * 90000 / 1000 = 180000 B: groupDurationTicks = 2000 * 44100 / 1000 = 88200 Same media instant, t = 10.5 s on the shared timeline: A: relative_ticks = 945000 -> floor(945000 / 180000) = 5 B: relative_ticks = 463050 -> floor(463050 / 88200) = 5¶
Both renditions place the instant in group 5. No intermediate value is shared between the two computations -- only the integer results agree -- which is exactly the property group-boundary switching requires.¶
Negative authored timestamps use the same rule and are not clamped. For
one 90 kHz track with groupDurationTicks: 180000, choose the shared
presentation anchor presentationAnchorTicks: -180000, which is earlier
than the ticks of the first admitted MPU:¶
MPU 0 ticks = -90000: relative_ticks = 90000 -> group 0 MPU 1 ticks = 90000: relative_ticks = 270000 -> group 1¶
The negative authored PTS therefore maps to group 0 through the shared
earlier anchor. A catalog with the same samples and
presentationAnchorTicks: 0 is invalid for this profile; the publisher
rejects MPU 0 instead of mapping it to group 0.¶
With reordered samples, the first AU in decode order selects the Group number. This example lists the authored PTS values of the first three AUs of two consecutive MPUs in decode order. MPU 0 is a closed GOP. MPU 1 is an open GOP: its RAP presents at 180000, and its two leading pictures present before it, inside Group 0's interval:¶
{
"timescale": 90000,
"presentationAnchorTicks": 0,
"groupDurationTicks": 180000,
"mpuSamplePtsInDecodeOrder": [
[0, 6000, 3000],
[180000, 174000, 177000]
],
"ticks": [0, 180000],
"groupNumbers": [0, 1]
}
¶
ticks is the first entry of each MPU's list. Selecting the minimum
presented PTS instead would give ticks 0 and 174000, both in Group 0,
which violates the strictly-greater rule above.¶
For a subscriber to apply the formula in Section 4.4.1, the catalog MUST publish the group duration for every mmtp-packaged track. The catalog publishes either the millisecond form or the exact tick form defined here:¶
groupDurationMs (REQUIRED unless groupDurationTicks is present,
positive unsigned integer, milliseconds)
¶
The timescale is the MSF timescale field ([I-D.ietf-moq-msf]
Section 5.2.21), resolved per the track's packaging value
([I-D.ietf-moq-msf] Section 5.2.4; this document adds mmtp,
Section 13.1):¶
packaging
|
Timescale source |
|---|---|
mmtp
|
The timescale field, REQUIRED for mmtp tracks (Section 13.1); absence is a catalog error, and the 90 000 asset_timescale default of [ISO.23008-1] Clause 10.3.9.3 does not apply. |
cmaf
|
The timescale field. [I-D.ietf-moq-msf] Section 5.2.21 makes it optional, so this document makes it REQUIRED for any cmaf track to which this section applies (a member of a switching set governed by Section 4.4.1); leaving it unset on such a track is a catalog error. |
loc
|
The timescale field. As for cmaf, this document makes it REQUIRED for any loc track to which this section applies; leaving it unset on such a track is a catalog error. When the publisher also sends the TIMESCALE Property ([I-D.ietf-moq-loc] Section 2.3.1.2) on such a track, its value MUST equal the field. |
(unrecognized packaging) |
1 000 000 (microseconds). This row covers track entries whose packaging value is unrecognized; the convention is fixed by this document and does not derive from [I-D.ietf-moq-transport]. A track entry without packaging is malformed ([I-D.ietf-moq-msf] Section 5.2.4) and fails catalog validation (Section 6.1, step 1). |
When groupDurationTicks is absent, conversion is
groupDurationTicks = groupDurationMs * timescale / 1000. The
catalog publisher MUST choose a groupDurationMs such that the
resulting tick count is a positive integer and the conversion is exact
(i.e. (groupDurationMs * timescale) % 1000 == 0). If this cannot be
satisfied, or if the intended cadence is not an integer number of
milliseconds, the catalog MUST publish ticks directly via the exact
tick field:¶
groupDurationTicks (OPTIONAL, positive unsigned integer)¶
If both groupDurationMs and groupDurationTicks are present,
groupDurationTicks wins and groupDurationMs is informational; it is
not used in the conversion and need not have an exact millisecond
conversion. Cases where the override is required include:¶
1/30 s = 33.333... ms): expressing this in integer ms is
impossible. If the track timescale is divisible by 30, publishers
set groupDurationTicks: timescale / 30 directly (e.g. 3000 at
90 000 Hz, 1500 at 45 000 Hz). A timescale not divisible by 30
cannot represent this cadence as an integer tick count and MUST be
changed or rejected for this profile; it MUST NOT be rounded.¶
333 x 44 100 / 1000 = 14 685.3 ticks. A publisher whose authored
cadence is the exact integral cadence of 14 685 ticks MUST publish
groupDurationTicks: 14685 (about 332.993 ms); it MUST NOT claim
that this preserves an exact 333 ms duration. A publisher that
requires exactly 333 ms MUST choose a compatible timescale or reject
that cadence.¶
For example, this catalog uses the exact 30-per-second cadence on a
44 100 Hz track. It omits groupDurationMs because 33.333... ms is
not an integer millisecond value:¶
{
"version": 1,
"tracks": [{
"name": "video-30fps",
"packaging": "mmtp",
"isLive": true,
"mmtpMode": "mfu",
"timescale": 44100,
"presentationAnchorTicks": 0,
"groupDurationTicks": 1470,
"codec": "avc1.64001f",
"framerate": 30,
"bitrate": 2000000
}]
}
¶
Switching-set agreement: every track in the same switching set MUST
publish the same effective group duration as a rational number of
seconds. Raw groupDurationTicks values need not be equal when track
timescales differ. For tracks A and B, subscribers validate exact
equality with integer arithmetic:¶
A.groupDurationTicks * B.timescale ==
B.groupDurationTicks * A.timescale
¶
One exception applies, to mmtp tracks only. An mmtp track is
intra-only when its framerate field is present and its group
duration is one frame period at that rate:¶
groupDurationTicks * framerate == timescale¶
Because framerate need not be an integer (29.97), this comparison is
made within a relative tolerance of 10^-3 of timescale; a GOP track
misses it by a factor of at least two. The tolerance only classifies
a track as intra-only. It never enters Group number arithmetic: the
remainder test below, the quotient n, and every Group number use
exact integers, and Section 4.4.1 forbids any tolerance there. Every
other track in the switching set, including every track whose
packaging is not mmtp, is a GOP track. Exact equality above is
required between all GOP tracks. An intra-only track I MAY instead publish a shorter
group duration that divides that of the GOP tracks a whole number of
times. For every GOP track G in the switching set:¶
(G.groupDurationTicks * I.timescale) %
(I.groupDurationTicks * G.timescale) == 0
¶
Since an intra-only track has a group boundary at every frame, each GOP
group boundary is also one of its group boundaries, and switching
remains group-aligned. With quotient
n = (G.groupDurationTicks * I.timescale) / (I.groupDurationTicks *
G.timescale) and a shared presentation anchor, Group g of G starts at
the same instant as Group g * n of I. This mapping is the alternate
mechanism that [I-D.ietf-moq-msf] Section 4.2 requires for
time-aligned tracks whose Group IDs differ at the same instant. A
subscriber can therefore switch
from G to I at any G group boundary. It can switch from I to G only at
an I group whose number is a multiple of n. Each intra-only track is
checked only against the GOP tracks, so switching between two
intra-only tracks of one switching set is group-aligned only at GOP
track group boundaries. The exception does not
apply to a switching set with no GOP track: intra-only tracks with no
GOP peer still require exact equality.¶
For example, an mmtp 15 fps intra-only base track with timescale: 90000 and
groupDurationTicks: 6000 may share a switching set with a 15 fps GOP
track with timescale: 1000000 and groupDurationTicks: 1000000,
because 1000000 * 90000 = 15 * (6000 * 1000000). With
groupDurationTicks: 1050000 on the GOP track the remainder is non-zero
and the catalog is invalid. Two GOP tracks with group durations of 2 s
and 1 s are invalid even though one nests in the other.¶
Subscribers SHOULD validate this at catalog load and, on disagreement, SHOULD reject the catalog and signal a catalog-validation error to the application. Subscribers that proceed despite disagreement are non-conformant for switching across the affected tracks; ABR transitions on such tracks will produce group-boundary discontinuities.¶
The group duration above describes the cadence of MMTP media groups (one MPU per group, Section 4.3). It does not describe whether an MPU contains additional keyframes after its required group-start random access point. This document defines an OPTIONAL per-track field for the cadence of all keyframes, which the repair policy of [MOQ-FEC-MULTICAST] Section 6.1 uses to classify FEC blocks:¶
keyframeIntervalMs (OPTIONAL, unsigned integer, milliseconds)¶
keyframeIntervalMs is the interval between consecutive video
keyframes, in integer milliseconds. It is distinct from
groupDurationMs: groupDurationMs is the duration of
one MPU/group, whereas keyframeIntervalMs describes every keyframe,
including any additional keyframes inside an MPU. When the only
keyframes are the required group-start RAPs and group duration is
constant, the two values are equal; additional RAPs can make the
keyframe interval shorter. The field applies to video tracks only; a
publisher that cannot determine the cadence simply omits it.¶
Its absence is not a catalog error:
subscribers MUST NOT reject a catalog solely because it is missing,
and a receiver MUST NOT substitute a locally chosen or observed
cadence for it. When present, keyframeIntervalMs MUST be a
positive integer; a zero value is a catalog error.¶
The repair policy of [MOQ-FEC-MULTICAST] Section 6.1 relies on a published keyframe interval to exclude keyframes from part of a group, so when the field is present it binds every keyframe of the track: the publisher MUST NOT emit a keyframe other than at a group start or a whole number of intervals after one. A publisher whose encoder can place keyframes off that cadence, for example at scene cuts, omits the field.¶
As with the group duration, the integer-millisecond form cannot express a keyframe cadence whose tick count is non-integer on the track's timescale. A publisher MAY publish an exact override:¶
keyframeIntervalTicks (OPTIONAL, unsigned integer)¶
expressed in the same timescale as groupDurationTicks (Section 4.4.2,
selected by the track's packaging). If both
keyframeIntervalMs and keyframeIntervalTicks are present,
keyframeIntervalTicks wins. Unlike
groupDurationTicks, this document places no switching-set agreement
requirement on the keyframe interval: it is a per-track hint,
and tracks in a switching set MAY carry different keyframe cadences.¶
Publishers signal MPU metadata inline as the complete FT=0 object in
Subgroup 0 of every group (Section 4.3), preceded in that subgroup only
by the OPTIONAL AL-FEC signaling object of Section 8. Its data unit contains the
authored initialization boxes, including ftyp, mmpu, and moov.¶
In addition, publishers SHOULD carry decoder initialization data in
the catalog, using the mechanism of the base catalog that
[I-D.ietf-moq-cmsf] Section 3.1 also uses for CMAF headers. The
data is an entry of the root initDataList array
([I-D.ietf-moq-msf] Section 5.1.7) whose type is "inline" and
whose data is the base64 encoding of an ISOBMFF initialization
segment (ftyp + moov) sufficient to initialize the decoder for
the track (including the codec configuration record, e.g. avcC or
hvcC). The track's initRef field ([I-D.ietf-moq-msf]
Section 5.2.13) holds that entry's id. This lets a subscriber
initialize its decoder at catalog load, before the first FT=0 MPU
metadata object arrives, removing one delivery round trip from the
join path (Section 6.1). A receiver ignores an initRef that names
no entry, or names an entry whose type is not "inline", and joins
without catalog initialization data.¶
Catalog initialization data is a join optimization, not a replacement
for in-band metadata: the per-group FT=0 MPU metadata remains
authoritative, and a receiver MUST adopt in-band metadata when it
differs from the catalog's initialization data. Catalog content is
untrusted input; a receiver MUST validate that the decoded data is
a well-formed ftyp + moov sequence before passing it to a
decoder, and MUST discard it otherwise.¶
Some implementations carry the same bytes in a per-track string
field, initData, which [I-D.ietf-moq-msf] does not define. A
receiver MAY accept initData on an mmtp track and validates it in
the same way. When a track carries both initRef and initData, the
receiver uses initRef.¶
The publisher numbers the objects of each group in source order, as [I-D.ietf-moq-msf] Section 6.2 requires. The first object of the group has Object ID 0. It is the AL-FEC signaling object when the group carries one (Section 8), and the FT=0 object otherwise. Each later object, in whichever subgroup, has the Object ID of the object before it plus one. An Object ID therefore identifies one object of its group, as [I-D.ietf-moq-transport] Section 2.1 requires, and the Subgroup ID still identifies the data unit (Section 4.3). The packetizer emits each data unit's packets contiguously (Section 5.2.2), and Subgroup IDs follow source order (Section 5.3.2), so ascending Object ID order within a group is also ascending Subgroup ID order. A FETCH or fill fetch stream, which sends a group's objects in Object ID order ([I-D.ietf-moq-transport] Section 3.2.1), therefore delivers them in the order that Section 4.3 requires.¶
MoQ Transport carries Object IDs as deltas on the wire ([I-D.ietf-moq-transport] Sections 11.3.1 and 11.4.1.1), and the objects of every subgroup other than Subgroup 0 have Object IDs greater than zero. A receiver MUST reconstruct each object's absolute Object ID from that encoding. It MUST NOT assume that a subgroup's first Object ID is zero, that successive objects differ by a fixed step, or that an absolute Object ID is carried explicitly.¶
Object IDs identify preserved source objects within a group; they do not replace the MMTP packet sequence, FI countdown, or active-data-unit association required by Sections 5.2 and 5.3.¶
This timed-media profile transports each authored MPU data unit without rewriting it:¶
MFU Mode (one subgroup per MPU data unit):
Group N:
[SG 0 / Obj 0: [MMTP type=0x02 id=0][SH][AL-FEC msg]] (Section 8)
SG 0 / Obj s: [MMTP][PH FT=0 FI=0][ftyp+mmpu+moov]
SG 1 / Obj s+1: [MMTP][PH FT=1 FI=0][moof+mdat prefix]
SG 2 / Obj s+2: [MMTP][PH FT=2 FI=0][DU][sample 1]
SG 3 / Obj s+3: [MMTP][PH FT=2 FI=0][DU][sample 2]
...
Fragmented FT=2 sample (Section 5.2):
Group N, SG 2:
Obj s+2: [MMTP][PH FT=2 FI=1][DU][sample part 1]
Obj s+3: [MMTP][PH FT=2 FI=2][sample part 2]
Obj s+4: [MMTP][PH FT=2 FI=3][sample part 3]
Subsample-partitioned FT=2 sample (Section 5.2.1): one sample,
two producer-authored subsample data units:
Group N:
SG 4 / Obj a..b: [MMTP][PH FT=2 FI=1][DU offset=0] ... FI=3
SG 5 / Obj b+1..c: [MMTP][PH FT=2 FI=1][DU offset=L0] ... FI=3
[MMTP] = MMTP packet header (Section 3.1): 12 bytes, since C = 0,
plus a header extension when X = 1.
[PH] = MPU-mode payload header, which carries the Fragmentation
Indicator (FI); the FI is not in the MMTP packet header.
[SH] = signaling message payload header (Section 3.1).
[...] = OPTIONAL signaling object. When present it is Object 0,
the first object of SG 0, and FT=0 is the object after it.
s = 1 when the signaling object is present, 0 when it is
absent. Object IDs count the group's objects in source
order across subgroups (Section 4.6).
¶
For conformance, an mmtp track in this document carries timed media
only. ISO/IEC 23008-1 Clause 9.3.2.3 defines the Timed Flag as
indicating whether the MPU is timed or non-timed, independently of the
FT value. Every MPU-mode packet with FT=0, FT=1, or FT=2 in this
profile MUST carry T = 1; a packet with T = 0 MUST be rejected.
Checking only FT=2 would leave metadata packets unchecked. The ISO
non-timed DU identity and four-byte header remain outside this profile;
no Item_ID-based mapping is defined here.¶
FT=0 initializes the track, FT=1 supplies the authoritative movie fragment structure and sample table, and FT=2 supplies the exact bytes of each sample. That order is a decoder contract, not a hint.¶
A single MFU can exceed the path MTU, but the ISO MPU-mode payload
header does not permit an unbounded number of fragments. The 8-bit
fragment_counter is the number of MMTP payloads containing fragments
of the same data unit that succeed the current payload; it is not a
modulo sequence number. A complete FI=0 data unit MUST carry zero.
For a fragmented data unit carried in N payloads, where 2 <= N <= 256,
the FI=1 payload MUST carry N - 1, every succeeding payload MUST
decrement the value by one, and the FI=3 payload MUST carry zero. The
counter MUST NOT wrap or repeat within a data unit.¶
The mfu mapping defined here carries exactly one data unit per MMTP
payload, so aggregation_flag MUST be zero. The upstream MMTP
packetizer produces conformant data units, and an input MMTP flow used
by this mapping MUST already satisfy the 256-payload bound for every
data unit. In this profile an FT=2 timed data unit carries one whole
sample, or exactly one producer-authored subsample of it
(Section 5.2.1). A sample whose media exceeds one data unit's
capacity MUST be partitioned into subsample data units by the origin
packetizer per Section 5.2.1, before this mapping applies -- never by
a relay and never by this mapping, which does not repacketize or
subdivide an input data unit. Such a sample MUST NOT be carried by
wrapping or extending the counter, silently truncated, or split into a
second MPU, and an access unit MUST NOT be split across multiple MPUs
([ISO.23008-1] Section 6.4). A publisher MUST reject a purported
data unit that exceeds the bound, has an invalid countdown, or uses
aggregation. It MUST NOT wrap or extend the counter, divide one data
unit's overflow between subgroups, or synthesize MoQ Object IDs to
make a non-conformant MMTP data unit appear valid.¶
Each MMTP packet maps to exactly one MoQ object, and the publisher MUST NOT reassemble MFU fragments:¶
The receiver MUST reassemble each data unit from the objects of its subgroup before media processing. After restoring serial packet order (Section 5.3.3), it applies the FI/counter state machine of Section 5.3:¶
fragment_counter countdown with
the state machine of Section 5.3.2. Any missing, repeated,
wrapped, or non-decrementing count invalidates the entire MFU.¶
The bytes concatenated are each fragment's media (data unit)
bytes only -- not everything after the MMTP packet header. Every
fragment begins with a 12-or-more-byte MMTP packet header. The
receiver MUST parse the C and X flags, packet-counter presence, and
header-extension length to locate the following 8-byte MPU-mode
payload header; it MUST NOT assume a fixed offset of 12 bytes. The
MPU-mode payload header contains, by [ISO.23008-1]
Clause 9.3.2.3 name with this document's alias in parentheses,
length (payload_length, 16), FT (fragment_type, 4), T
(timed_flag, 1), f_i (fragmentation_indicator, 2),
aggregation_flag (1), fragment_counter (8), and
MPU_sequence_number (32); both headers are stripped from every
fragment.¶
The 16-bit payload_length field MUST be encoded and interpreted
as the number of bytes following that field within the MPU payload,
as defined by [ISO.23008-1]. It therefore includes the remaining
six bytes of the 8-byte MPU-mode payload header. For FT=2 packets,
the declared lengths are:¶
FI=0 or FI=1: payload_length = 6 + DU_header_octets + media_octets FI=2 or FI=3: payload_length = 6 + media_octets¶
DU_header_octets is 14 for timed media and 4 for non-timed media,
and the right-hand side MUST NOT exceed the 16-bit maximum of 65535.
A trailing Source FEC Payload ID is outside the declared MPU payload
and MUST NOT be counted. A publisher MUST NOT encode a media-only or
DU-plus-media length convention. A receiver MUST require the
declared MPU payload to end exactly at the protected MPU boundary
before the Source FEC Payload ID, when present. It MUST reject a
declaration that is too short, cannot contain the required header
bytes, extends beyond the available MPU payload bytes, leaves
undeclared bytes before the outer trailer, or consumes that trailer.¶
The DU header ([ISO.23008-1] Figures 12 and 13: 14 bytes for
timed media -- movie_fragment_sequence_number (32),
sample_number (32), offset (32), priority (8),
dep_counter (8) -- or 4 bytes for non-timed media) follows the
payload header on an FI=0 object and on the FI=1 first fragment;
it is likewise stripped rather than concatenated into the media
stream. When the packet's FEC Type is 1, the trailing 4-byte
Source FEC Payload ID (Section 7.2) is excluded as
well.
4. FT=0 and FT=1 MUST each be unaggregated, complete FI=0 data units
with fragment_counter zero. Their payload length is six plus the
metadata octets; they have no DU header. Fragmented or empty
metadata is rejected by this profile.
5. The receiver MUST bound both in-flight data-unit memory and pending
FT=1 sample records. At either bound it MUST reclaim a slot by
evicting the least recently admitted incomplete record (an open
data unit or a parked continuation run) and admit the new one; it
rejects only when no such record exists. A record that has
already completed is never a candidate. An incomplete data
unit is discarded on declared discontinuity or unrecoverable loss
and is never emitted.¶
For very large frames where FEC cannot recover the loss, receivers SHOULD support resolution-tier fallback (subscribe to a lower-resolution track of the switching set and upscale) rather than stalling.¶
At a given packet size, one data unit carries at most 256 MMTP
payloads of media. A single intra-coded 4K or 8K sample routinely
exceeds that capacity. This section defines the only conformant way
to carry such a sample: the origin packetizer partitions it into
subsample data units, using the offset field that [ISO.23008-1]
Section 9.3.2.3 already defines in the timed-MFU DU header. The
partition is producer-authored and wire-stated; nothing about it is
inferred.¶
A subsample data unit is an ordinary FT=2 timed data unit: it has its
own DU header, its own FI chain, and its own fragment_counter
countdown bounded by 256 payloads (Section 5.2). Its identity is
(packet_id, MPU_sequence_number, movie_fragment_sequence_number,
sample_number, offset), where offset is the authored byte position
of the data unit's first media byte within the sample.¶
Origin packetizer requirements:¶
offset zero) whenever the sample fits within one data unit's
capacity. Subsample partitioning of a sample that fits is not
conformant.¶
offset zero and each subsequent one MUST carry an offset equal
to the sum of the media lengths of all preceding subsample data
units of that sample: offsets are strictly increasing, with no gap
and no overlap.¶
priority and dep_counter describe the sample and SHOULD be
identical across its subsample data units.¶
Relays and this mapping MUST NOT create, merge, or re-partition subsample data units.¶
Receiver sample assembly: a subsample-partitioned sample completes
only when every one of its subsample data units has completed under
the state machine of Section 5.3.2 and, ordered by offset, the data
units begin at zero, are byte-adjacent (each offset equals the
preceding offset plus the preceding data unit's media length), and
their total media length equals the sample's authored size from the
FT=1 record. A gap, overlap, duplicate offset, or length mismatch
invalidates the sample; the receiver MUST NOT emit a partial sample or
synthesize missing bytes. Completion is a protocol event -- every
constituent subgroup complete -- and on reliable transport a receiver
MUST NOT substitute timers, cadence, buffer occupancy, or successor
arrival for it. On lossy paths each subsample data unit completes,
possibly after FEC recovery ([MOQ-FEC-CORE] Section
9), or is discarded
as incomplete (Section 5.2, item 5), like any other data unit; no
additional sample-level timer is defined or permitted.¶
The sample's RAP flag is the RAP flag of the FI=0 or FI=1 object of its offset-0 data unit. FEC is unaffected by the partition: source symbols are MMTP packets (Section 7.1), and the packet stream's block and interleave structure does not depend on data-unit boundaries.¶
[ISO.23008-1] does not prohibit a packetizer from emitting an
unrelated payload of the same packet_id between two fragments of one
data unit. This profile does prohibit it.¶
An origin packetizer MUST emit every multi-packet unit on consecutive
packet_sequence_number values of its packet_id, in authored order,
with no other payload of that packet_id between its packets: the
fragments of a data unit in FI order, and the subsample data units of
one sample (Section 5.2.1) in ascending offset order. The sequence
numbers of the FI=1 fragment, of each FI=2 fragment, and of the FI=3
fragment MUST be successive values under
modulo-2^32 arithmetic. A data unit of a different packet_id MAY be
emitted at any point, and a further data unit of the same packet_id
MAY be emitted once that data unit's FI=3 fragment has been emitted,
subject to the sample-level contiguity above.¶
The requirement constrains the packetizer's emission order only. Loss
or reordering on the path does not violate it: the receiver restores
serial packet order before reassembly (Section 5.3.3). Repair
carried on its own flow bears a different packet_id and therefore
cannot interleave.¶
A violating data unit never completes at the receiver, which associates continuations only through active data-unit state (Section 5.3) and evicts the stranded records without a protocol error (Section 5.2, item 5). A receiver MUST NOT relax association to tolerate such a producer: a continuation admitted into the wrong data unit corrupts media that would otherwise have been discarded.¶
Object IDs are transport coordinates, not an MMTP continuation key. A fragment recovered below the MoQ object layer has no Object ID, and Object IDs from two delivery paths are not comparable. The MMTP DU header is also not a per-fragment coordinate: this profile carries it only in the FI=0 or FI=1 payload ([ISO.23008-1] Clause 9.3.2 does not specify whether FI=2 and FI=3 fragments repeat it). FI=2 and FI=3 continuations carry media bytes but no movie-fragment sequence, sample number, or sample offset.¶
Consequently a receiver MUST associate continuations through the ordered, bounded active-data-unit state defined here. It MUST NOT use an MMTP Timestamp, MoQ Object ID, signaling identifier, packet arrival time or order, inferred byte offset, recently completed identity, or a (stream, MPU sequence number) pair or fragment counter alone as a fallback association.¶
For an FT=2 first or complete fragment, [ISO.23008-1] defines a transport-independent data-unit identity in the DU header:¶
| Object class | Data-unit identity | [ISO.23008-1] |
|---|---|---|
| Timed-media MFU | packet_id, MPU_sequence_number, movie_fragment_sequence_number, sample_number, offset | Section 9.3.2.3 |
| Non-timed media MFU | packet_id, MPU_sequence_number, Item_ID | Section 9.3.2.3 |
For this timed-media profile, offset MUST be zero for a whole-sample
data unit and MUST equal the authored subsample byte position for a
subsample data unit (Section 5.2.1);
movie_fragment_sequence_number and sample_number MUST be non-zero,
and MPU_sequence_number MAY be zero (indeed, ISO/IEC 23008-1 requires
zero for the first MPU in an Asset). The packet_id MUST be
non-zero: packet_id 0 is the MMTP signaling flow (Section 3.1) and
carries no media data unit, so a receiver MUST reject an MPU-mode
packet (packet type 0x00) on packet_id 0 rather than admit it to
active state. A non-zero packet_id is validated in the scope of its
MMTP session; on multicast delivery it is validated against the
packetId values the endpoint advertises ([MOQ-MULTICAST]
Section 5). A continuation inherits the accepted first-fragment
identity from active state; it does not reconstruct or guess another
copy of the DU header.¶
Section 4.3 assigns Subgroup 0 to FT=0. The publisher assigns every
following FT=1 or FT=2 data unit the next unused Subgroup ID in source
order. All packets from one FI=1 through FI=3 chain share that ID.
An FI=2 or FI=3 packet therefore remains in the subgroup opened by the
immediately preceding FI=1 packet on the same packet_id; a new data
unit MUST NOT interleave before the chain closes (Section 5.2.2).¶
The publisher and receiver each maintain at most one active fragmented
data unit per (packet_id, MPU_sequence_number) while processing
subgroups in order:¶
fragment_counter == 0 and no active data unit. For
FT=2 it requires a valid DU header and represents the complete
sample. FT=0 and FT=1 are permitted only in this form.¶
c in 1..255, and, for
FT=2, a valid DU header. It creates active state containing the
data-unit type, timed flag, full first-fragment identity, destination
subgroup, and next_expected_counter = c - 1.¶
next_expected_counter == 0, and a wire counter of zero. It appends
the declared media bytes, closes the state, and emits the complete
data unit only after all size and authored-timing checks succeed.¶
An FI=0 or FI=1 while state is active is an overlap. An FI=2 or FI=3 without it is an orphan. A changed packet ID, MPU sequence, FT, timed flag, subgroup, skipped/repeated counter, FI=2 at zero, or FI=3 before the expected counter reaches zero is a mismatch or gap. The receiver MUST reject all of these and MUST NOT advance, replace, or consume valid state or pending sample timing on rejection. A new MPU arriving while the prior MPU has an incomplete data unit declares loss and discontinuity; the incomplete unit is discarded and never emitted.¶
A receiver admits a fragment only after its declared MPU payload
boundary, outer FEC trailer, stream scope, subgroup order, and active-DU
transition all validate. Duplicate packets may be dropped by exact
MMTP packet identity before admission, but a duplicate MUST NOT create
or advance active state. Multi-path combining and FEC recovery MUST
restore the same per-packet_id serial packet order before this state
machine, using modulo-2^32 serial-number arithmetic across
packet_sequence_number wrap within a bounded replay window. The
MMTP-session and MoQ-track scope are part of that ordering state;
neither mechanism may fabricate a missing transition.¶
Completion is transactional. For a whole-sample data unit the
receiver first validates the full FI chain and the matching unconsumed
FT=1 sample record; the reassembled FT=2 media length MUST equal that
sample's authored size. For a subsample-partitioned sample it
additionally requires every constituent subsample data unit to have
completed, with the byte-adjacency and total-length checks of
Section 5.2.1. Only then may it emit the sample and consume that
record. Every (sample, offset) data unit is admitted at most once,
and every described sample is emitted at most once. A failed check leaves every pre-existing valid state and timing record
unchanged, except that an explicitly declared discontinuity discards
the incomplete prior unit.¶
An MPU begins at a Stream Access Point: the first access unit of an MPU is a random access point (RAP) ([ISO.23008-1] Section 6.4; Section 4.3). Accordingly a receiver:¶
On a multicast or Automatic Multicast Tunneling (AMT) leg, [MOQ-FEC-MULTICAST] Section 6 specifies the receiver repair policy this implies.¶
This section defines the join procedure for subscribers. A subscriber reaches its first decodable frame without waiting for the next group boundary by joining at the start of the current group with the fill mechanism of [I-D.ietf-moq-transport] Section 3.5. The join needs no relay behavior beyond [I-D.ietf-moq-transport], and relays never parse MMTP (Section 6.2).¶
A subscriber joins a live mmtp-packaged track as follows:¶
Catalog. Subscribe to the catalog track (the track named
catalog, Section 13) as [I-D.ietf-moq-msf] Section 5 requires,
joining at the start of its current group
([I-D.ietf-moq-transport] Section 3.5), and validate the
catalog. Validation includes the per-track REQUIRED fields of
Section 13.1. A subscriber MUST NOT issue SUBSCRIBE for a track
whose catalog entry fails validation. It also applies the
switching-set agreement check of Section 4.4.2 and handles a
disagreement as that section specifies.¶
Decoder initialization. If the catalog provides initialization data for the track (Section 4.5) and that data validates, initialize the decoder from it immediately. In all cases the subscriber MUST accept the authoritative FT=0 object in Subgroup 0 before admitting this group's FT=1 or FT=2 data units. An AL-FEC signaling object that precedes FT=0 in Subgroup 0 (Section 8) is not decoder initialization and does not satisfy this step.¶
Subscribe at a group boundary. Join at the start of the current group, as [I-D.ietf-moq-transport] Section 3.5 describes. Send SUBSCRIBE with a LOCATION_FILTER of type 0x05 (Next Object) and a FILL_PARAMETERS parameter whose LOCATION_FILTER has type 0x01 (Relative Start) and StartGroup = 1 ([I-D.ietf-moq-transport] Sections 3.4, 9.20.9, and 9.20.15). The publisher sends the current group, from its start up to the Largest Object, on a fill fetch stream, and every later object on the subscription, so each object arrives once. The group start carries FT=0 initialization (Subgroup 0), the FT=1 metadata governing the first samples, and the RAP-bearing first FT=2 subgroup (Section 4.3); a subscription that starts mid-group without a fill yields objects that cannot be decoded until the next group. A data unit that was still open at the join arrives partly on the fill fetch stream and partly on its subgroup stream, and the receiver combines the two before reassembly (Section 5.2). If the fill fails or leaves a gap, the subscriber waits for the next group, which the subscription delivers from its start. A subscriber that does not want the current group sends SUBSCRIBE with a LOCATION_FILTER of type 0x01 and StartGroup = 0 instead. A NEW_GROUP_REQUEST ([I-D.ietf-moq-transport] Section 9.20.19) cannot shorten that wait (Section 4.4.1).¶
Repair track. If the subscriber wants FEC protection and the catalog signals FEC for the track, it SHOULD subscribe to the repair track (Section 8.2) at the same time as the source track, so that the first FEC block spanning the join point is repairable. Repair-track subscription is selective and optional ([MOQ-FEC-CORE] Section 7.2).¶
Presentation gate. A subscriber MUST NOT submit media to the decoder until it has (a) accepted the authoritative FT=0 and FT=1 data units and (b) a completely reassembled FT=2 sample whose RAP flag is set (Section 5.2). Objects received before that point are buffered or discarded according to the receiver's reassembly bounds.¶
Loss and discontinuity. If a RAP MFU is discarded under unrepairable loss (Section 5.2), the subscriber MUST treat the track as discontinuous and MUST NOT resume presentation before the next completely reassembled RAP MFU (normally the start of the next group). The same holds when the group's FT=0 data unit, or the FT=1 data unit governing its RAP, is lost, since the RAP then cannot be admitted (Section 4.3). On a multicast or AMT leg, loss that spares those three data units does not make the track discontinuous; [MOQ-FEC-MULTICAST] Section 6 specifies its handling.¶
This section is informative. A relay serves the fill fetch stream of Section 6.1 from its cache, and fetches from upstream the objects it does not hold ([I-D.ietf-moq-transport] Sections 3.2.3 and 3.4). A relay that caches the current group of each track it serves can therefore start a late joiner without an upstream round trip. The join needs no mmtp-specific relay behavior. A fill fetch stream sends a group's objects in Object ID order, which is source order on an mmtp track (Section 4.6), so it delivers Subgroup 0 and each metadata subgroup before the FT=2 subgroups that depend on them, and the relay does not parse MMTP.¶
Retention of groups older than the current group (for example to serve FETCH-based catch-up) is permitted but out of scope for this document.¶
MMT's AL-FEC framework supports multiple FEC schemes, with the MMT FEC codes specified in [ISO.23008-10] and parameters signaled via MMTP signaling messages (Section 8.3); neither the framework nor ATSC 3.0, which uses it for MMTP sessions ([ATSC-A331] Section 8.1.2.5), requires a particular code, so the codes are specified by [MOQ-FEC-BLOCK]: "reed-solomon", the Reed-Solomon code of [ISO.23008-10] Clause 6 (its Section 5), is the mandatory-to-implement scheme, and RaptorQ [RFC6330] (its Section 6) is OPTIONAL ([MOQ-FEC-BLOCK] Section 10). FEC repair uses the model defined in [MOQ-FEC-CORE] (its Sections 3.1 and 7), under the block profile that Section 7.1 defines for mmtp tracks:¶
Source Track: video +-- Objects: MMTP packets with FEC Type=0 or 1 (source) Repair Track: video/repair +-- Objects: MMTP packets with FEC Type=2 (repair)¶
This section defines that profile, "iso-ssbg1-v1": its source symbols, the Source FEC Payload ID trailer, the MMTP form of the repair packet, the keyframe overlay and the interleave window. It states where the profile deviates from [MOQ-FEC-CORE], whose remaining rules apply unchanged ([MOQ-FEC-CORE] Section 5.2).¶
"iso-ssbg1-v1" is the block profile ([MOQ-FEC-CORE] Section
5.2) of this document:
the ISO/IEC 23008-1 Annex C profile, all of whose rules a receiver
applies together. It admits the mmtp packaging
(Section 15). A source symbol is the ssbg_mode1 framing of
one FEC source packet (below). A block holds SSB_length source symbols,
with sourceSymbols as the upper bound on SSB_length ([MOQ-FEC-CORE]
Section 5.4), carried in band as
maximum_k_for_repair_flow (Section 8.3). SS_IDs advance by one per
FEC source packet, blocks are consecutive, and a source symbol is placed
in its block by range membership ([MOQ-FEC-CORE] Section
5.5). SS_ID arithmetic is modulo 2^32,
and a block's repair objects are grouped under its 62-bit extended
SS_Start ([MOQ-FEC-CORE] Section
7.6). An in-band AL-FEC message
that describes such a track carries the coding byte 0x14 (Section 8.3).¶
The rules of [MOQ-FEC-CORE] apply to a track under this profile with the following deviations, which [MOQ-FEC-CORE] Section 5.2 asks a companion profile to state:¶
The unit of protection is the MMTP packet, not the MoQ object: a source symbol is formed from one FEC source packet as this section specifies, and the Protected Object Unit of [MOQ-FEC-CORE] Section 5.2 is not used. A source object of an mmtp track carries one source symbol, the FEC source packet that is its payload, so SS_IDs advance by one per source object.¶
A source packet carries its SS_ID in a 4-byte trailer, the Source FEC Payload ID (Section 7.2), which is the canonical carrier. The FEC Source Info Object Property of the object that carries the packet restates it in the 4-byte form, which names instance 0 ([MOQ-FEC-CORE] Section 5.6); the Property is object metadata, not part of the protected packet, and is not recovered with it. When an object carries a source symbol but no FEC Source Info Property, a receiver that parses the packet MAY take the SS_ID from the trailer and feed the symbol to the base decoder (Section 7.2), a deviation from [MOQ-FEC-CORE] Section 5.6, under which such an object is not a source object; a receiver that does not parse the packet has no SS_ID for it and feeds it to no decoder.¶
A repair object is one complete MMTP repair packet of 12 + 13 + T bytes (Section 7.3), not the 13 + T byte payload of [MOQ-FEC-CORE] Section 7.5.¶
The keyframe overlay (Section 7.4) is this profile's form of an overlapping instance ([MOQ-FEC-CORE] Section 8.4), with two differences from that section and from [MOQ-FEC-CORE] Section 5.1: a RAP packet's overlay SS_ID travels in an MMTP header extension, not in a second entry of the FEC Source Info Property, and the Property keeps the 4-byte form, so a receiver learns a packet's overlay membership and overlay SS_ID from that extension, which lies inside the protected packet (Section 7.4.2), rather than from the Property alone; and because a recovered packet carries its overlay SS_ID but not its base SS_ID, a base recovery places the packet in the overlay instance, whereas an overlay recovery counts toward the base instance only where the receiver has determined the packet's base SS_ID independently (Section 7.1).¶
interleaveDepthMs admits the value zero, with the ISO packet-count
meaning of Section 7.5, under the permission of [MOQ-FEC-CORE]
Section 10.1 for a block profile to give zero
a meaning of its own.¶
The symbol carriage of [MOQ-FEC-CORE] Section 6.2, and with it the symbol track and the FEC Unit Start Property, does not apply to an mmtp track: its source symbols reach a receiver inside the track's own objects or as multicast datagrams ([MOQ-MULTICAST] Section 5), each carrying one whole FEC source packet.¶
Per [ISO.23008-1] Section C.3.3 (signalled per C.6.3, Table C.5), the FEC block structure is described by the source symbol block generation (SSBG) mode (ssbg_mode). This profile uses ssbg_mode1, with the one-stage FEC coding structure and FEC payload ID mode 0 ([ISO.23008-1] Clause C.1).¶
ssbg_mode1: One MMTP source packet = one source symbol, which records the packet's length. Each MoQ object or multicast UDP datagram carries exactly one FEC symbol. Repair symbols are exactly T bytes on the wire; a source packet MAY be shorter than T on the wire -- the length field and the zero padding that complete its T-byte source symbol are applied by the encoder and decoder as part of the source symbol construction below and are not transmitted. This is the natural model for MMTP because packets are sized to fit in UDP datagrams and no fragmentation or reassembly is needed at the FEC layer. Every source symbol is exactly T bytes, so a block of SSB_length source symbols is exactly SSB_length x T bytes -- the property the derived Transfer Length of [MOQ-FEC-BLOCK] Section 6.1 relies on. This is ssbg_mode1 of [ISO.23008-1], not ssbg_mode0, which carries no length and requires every packet of the flow to be exactly one symbol long.¶
Source symbol construction: The protected packet of an FEC source packet is the complete MMTP packet -- the MMTP packet header, including any header extension, followed by the MMTP payload -- before its 4-byte Source FEC Payload ID is appended ([ISO.23008-1] Section C.4.2; Section 7.2). Its T-byte source symbol is ([ISO.23008-1] Section C.3.3):¶
Source Symbol {
Length (16), # protected packet length, octets
Protected Packet (..), # Length octets
Padding (..), # zero octets, up to T in total
}
¶
The Length field is in network byte order and every Padding octet is zero, so the largest protected packet a symbol can carry is T - 2 octets, and T MUST be at least 14: a shorter symbol carries no protected packet of the 12-octet MMTP packet header (Section 3.1). (Under "raptorq", whose T is a multiple of 8, the least T is therefore 16, [MOQ-FEC-BLOCK] Section 6.1.) A publisher MUST size every FEC source packet so that its protected packet does not exceed T - 2 octets. A receiver that observes a source packet whose protected packet exceeds T - 2 octets MUST NOT feed it to the decoder -- a truncated symbol would corrupt recovery for the entire block -- but MAY still deliver the packet as media. On a RAP packet under a keyframe overlay (Section 7.4), the base and the overlay instance form the symbol from the same protected packet, so the two protect one identical T-byte symbol.¶
Recovery yields the block's SSB_length source symbols. From each, the receiver takes the Length octets that follow the Length field as the recovered protected packet, and delivers that packet to the MMTP parser. A receiver MUST NOT deliver a recovered symbol whose Length is less than 12, the size in bytes of the MMTP packet header (Section 3.1), or points past the end of the symbol, or whose Padding octets are not all zero: each means that the symbol is not the one the publisher encoded.¶
A recovered packet carries no Source FEC Payload ID, of either instance: the trailer is appended after FEC encoding and is not part of the protected packet. A receiver MUST NOT re-append a trailer before delivering the recovered packet to the MMTP parser, and MUST NOT expect the recovered bytes to end in a block coordinate. This holds identically for a symbol recovered by a keyframe overlay, which the receiver delivers as media on the strength of the overlay recovery alone. The 0x8B01 header extension of a RAP packet under an overlay (Section 7.4.3) is part of the protected packet, by contrast, and is recovered with it.¶
The packet's base SS_ID was in the trailer, and this document defines no mapping from an overlay coordinate to a base one: a receiver MUST NOT place an overlay-recovered symbol into a base source block unless it has determined that packet's base SS_ID independently. Absent that, an overlay recovery is usable as media but MUST NOT be counted toward base block recovery, and in particular MUST NOT be treated as reducing the number of erasures the base decoder still has to solve. The other direction needs no such determination, because a RAP packet recovered by the base instance still carries its overlay SS_ID (Section 7.4.3).¶
SS_ID: Every FEC source packet carries a Source FEC Payload ID (Section 7.2), and a receiver that parses the packet takes each source symbol's SS_ID from that field on every path that delivers the packet with its trailer. The SS_ID is also carried alongside the packet where the packaging says so: on the MoQ path every object carrying a source symbol carries it in the FEC Source Info Object Property ([MOQ-FEC-CORE] Section 5.6), under the same value and the same definition, whether or not that object also delivers the trailer. A receiver that does not parse the FEC source packet -- a relay, or a MoQ receiver with no MMTP parser -- takes the SS_ID from there. The SS_ID advances by one per FEC source packet, modulo 2^32, in the order the publisher emits the packets ([ISO.23008-1] Section C.5.2), and the blocks of an FEC instance are consecutive: each starts at the SS_ID that follows the last source symbol of its predecessor ([MOQ-FEC-CORE] Section 5.4). Every transport path yields the same SS_ID for the same packet, because each path carries the same MMTP packet, so receivers MAY combine symbols from any transport path toward the same FEC block recovery. A loc or cmaf track reaches multicast only through a separate mmtp rendition track ([MOQ-MULTICAST] Section 4.1), whose symbols belong to another FEC instance.¶
OTI and sub-blocks: MoQ carries no RaptorQ Object Transmission Information (OTI, [RFC6330] Section 3.3) in-session: receivers derive it per block from the instance's configuration and the block's SSB_length per [MOQ-FEC-BLOCK] Section 6.1. The AL-FEC message of an MMTP session (Section 8.3) carries none either. When a source block is too large to decode in the decoder's working memory, RFC 6330 permits dividing each block into N sub-blocks (the Number of Sub-Blocks field of the OTI). The derived OTI fixes N = 1, so catalog-signaled sessions do not use sub-blocks; the AL-FEC message of [ISO.23008-1] carries no OTI, and the split source blocks of its Clause C.2.2 are the two-stage coding structure, which this profile does not use. A publisher using MMTP packaging under this profile therefore uses ssbg_mode1 without sub-blocks, and never ssbg_mode2.¶
Exactly one 4-byte Source FEC Payload ID (SS_ID; ffsrpts_flag = 0, so no FFSRP_TS, [ISO.23008-1] Section C.5.1) is appended at the end of each MMTP FEC source packet -- following the MMTP payload, outside the MMTP packet header -- with or without a keyframe overlay (Section 7.4.3). The trailer lies outside the declared MPU payload (Section 5.2) and outside the protected packet (Section 7.1).¶
For MoQ-only receivers (no A3SA verification), the trailing 4-byte Source FEC Payload ID remains the canonical identifier by which a source packet is placed in its block ([MOQ-FEC-CORE] Section 5.5). QUIC transport encryption provides equivalent integrity protection. Section 14.1 states why the ATSC 3.0 signing mechanism does not authenticate the trailer, and what a receiver that requires authenticated block assignment does instead.¶
Whatever the receiver's verification posture, the FEC Source Info Object Property ([MOQ-FEC-CORE] Section 5.6) carries the same value for a receiver that does not parse the packet; where an object delivers both and they disagree, the trailer is the one that is canonical, because it is the field the FEC encoder emitted and the one the receiver locates the symbol by. The Property restates it at the MoQ layer.¶
On an mmtp source object the Property's value is the 4-byte form of [MOQ-FEC-CORE] Section 5.6, which names instance 0: the sequence number of the source symbol the object carries, the same value and the same definition that the Source FEC Payload ID trailer of the FEC source packet carries. The Property is that trailer's MoQ-layer carrier: one identifier, one definition, whichever path the symbol travels. Both carry the same value and a publisher MUST make them identical, on every source object, because the object payload is the whole FEC source packet, trailer included. The Property is wanted there as well: it places the symbol at the MoQ layer, reachable by a relay and by a receiver that does not parse MMTP, whereas the trailer is the field the FEC encoder emitted. A packet the decoder recovers carries neither (Section 7.1): the Property is not recovered with the object, and the trailer is appended after FEC encoding, so a receiver that recovers a packet determines its SS_ID from the block coordinate that recovered it, within that FEC instance's SS_ID space; an overlay coordinate does not yield a base SS_ID.¶
Where a keyframe overlay applies (Section 7.4), an object carrying a Random Access Point fragment is a source symbol of two FEC instances, whose SS_ID spaces are disjoint, so "the source symbol the object carries" names two SS_IDs rather than one. The Property carries the base instance's SS_ID -- the value the Source FEC Payload ID trailer carries -- and a publisher MUST NOT place an overlay SS_ID in it. No second Property entry conveys the overlay SS_ID: it travels in the 0x8B01 MMTP header extension, which lies inside the MMTP packet header and is therefore part of the protected packet of both instances (Section 7.4.3), so it rides inside the object's payload on this path as on any other. A receiver takes an overlay SS_ID from that extension and from nothing else, and in particular MUST NOT infer one from the Property.¶
An object that carries a source symbol but no FEC Source Info Property violates the publisher requirement of [MOQ-FEC-CORE] Section 5.6. A receiver that does not parse the FEC source packet has no SS_ID for it -- at the MoQ layer the Property is the only carrier -- so it MUST NOT feed such an object to the base decoder and MUST NOT count it in K or in any base block, and MAY still deliver it as media. A receiver that parses the packet MAY take the SS_ID from the trailer instead, since the packaging delivers it; this is a deviation from that section, which has no trailer to fall back on. An overlay SS_ID is unaffected either way: it travels in the 0x8B01 MMTP header extension (Section 7.4.3), which the Property does not carry, and that section alone governs when the overlay decoder may be fed.¶
Each object on a repair track is one complete MMTP repair packet (packet type 0x03, FEC Type 2) carrying exactly one repair symbol. Its layout is the ISO 23008-1 AL-FEC repair packet form -- the MMTP packet header followed by the 13-byte Repair FEC Payload ID and the repair symbol data ([ISO.23008-1] Sections C.4.3 and C.5.3; ssbg_mode1, one-stage FEC with N = 1, and ffsrpts_flag = 0, so no FFSRP_TS, Section C.5.1) -- so the same repair packet is valid on MoQ, multicast UDP, and native broadcast paths without translation, and matches the repair-object depiction of Section 7. The layout below reproduces [ISO.23008-1] Section C.5.3 for convenience. The Repair FEC Payload ID carries the fields of the repair object of [MOQ-FEC-CORE] Section 7.5 with the semantics that section gives them -- SS_Start, RSB_length, RS_ID and SSB_length -- and this section defines only the MMTP packet header in front of them:¶
Repair Packet {
MMTP Packet Header (96), # packet type 0x03, FEC Type 2
Repair FEC Payload ID {
SS_Start (32),
RSB_length (24),
RS_ID (24),
SSB_length (24),
},
Repair Symbol Data (..), # exactly one symbol, T bytes
}
¶
MMTP Packet Header: The standard 12-byte MMTP packet header (Section
3.1) with packet type 0x03 (repair) and FEC Type 2; its packet_id
routes the repair flow. Like every packet of its MMTP packet flow,
source and signaling packets included, a repair packet has the packet
counter flag C = 0 (Section 3.1; [ISO.23008-1] Clause 9.2.3). A
repair packet MUST NOT carry a header extension (X = 0). Because C = 0
and X = 0, the Repair FEC Payload ID always begins at byte offset 12.¶
A repair object under this profile is therefore 12 + 13 + T bytes long, not the 13 + T bytes that [MOQ-FEC-CORE] Section 7.5 requires of a repair object under "moq-object-v1"; a receiver MUST discard a repair object of an mmtp track whose payload is not exactly 12 + 13 + T bytes. It reads the Repair FEC Payload ID at byte offset 12, and every other rule of that section applies to the fields: a receiver routes the packet to the decoder of the instance that the configuration relates to the repair track it arrived on, and discards a packet whose RS_ID is not less than RSB_length or that fails a bound of the instance's scheme ([MOQ-FEC-BLOCK] Section 7).¶
A source track MAY additionally be protected by a keyframe overlay: a
second, independent FEC instance whose source symbols are only the
fragments of Random Access Points of that source track. It is this
profile's form of an overlapping instance ([MOQ-FEC-CORE] Section
8.4). The overlay is defined only for a
source track under "iso-ssbg1-v1", because its membership and its
overlay SS_ID rely on the MMTP RAP_flag and header extension
(Section 7.4.2 and Section 7.4.3), which a loc or
cmaf object does not have; a track under "moq-object-v1" has no overlay,
and a catalog that declares one for such a track fails the rule of
Section 7.4.1.¶
Losing part of a Random Access Point makes every dependent frame undecodable until the next one; the overlay protects RAP fragments across a block of its own, which can be wider than the base window, without changing the base instance's latency.¶
An overlay is a separate FEC instance over the same source track. It
has its own sourceSymbols (K_overlay) and its own repairSymbols,
declared on its own instance entry with its own repair track
([MOQ-FEC-CATALOG] Section 3.1), and it shares
the source track's scheme, blockProfile and symbolSize. It does
not share the base instance's K, its block partition, its SS_ID space,
or its RS_ID space. Because each source packet carries one base SS_ID,
a different K or interleave window needs a separate rendition; the
keyframe overlay is a separate FEC instance, not a layer.¶
The overlay's instance object carries one member that this document defines, under the rule of [MOQ-FEC-CATALOG] Section 3.4 that lets a block profile's document define further keys for the tracks it admits, and further exceptions among them:¶
scope (string, OPTIONAL): When present, MUST be "keyframe", and
declares this instance to be the source track's keyframe overlay. It is
an unknown-key exception (Section 7.4.4). An overlay is
defined only for a source track under "iso-ssbg1-v1"
(Section 7.4).¶
Every per-block quantity of [MOQ-FEC-CORE] -- K, the total repair
count P, SS_ID, ESI, RS_ID, SS_Start, RSB_length, SSB_length, the
short-block rules, the wrap epoch and repair group IDs, and the bounds
of its Section 10.3 -- is defined within one FEC instance
(its Section 5.1). For the base instance, K is the
source track's sourceSymbols and P is the sum of repairSymbols over
its published layers ([MOQ-FEC-CORE] Section
7.3.1); for an overlay, K and P are the
sourceSymbols and repairSymbols of its own instance. A repair
packet belongs to the instance of the track it arrived on
(Section 7.4.2).¶
An overlay is not a repair layer. It carries no repairLayer, it is
not a member of the layer index chain, and it is excluded from the
contiguity and priority monotonicity rules of [MOQ-FEC-CORE] Section
10.3, which order the base layer chain only. Its
priority is independent within 192 to 255 ([MOQ-FEC-CORE] Section
10.3, rule 10, and Section
11), and MAY be numerically lower than a
base enhancement layer's so that a keyframe repair symbol is delivered
before an ordinary one.¶
Overlay ESIs never collide with base ESIs because the two are different instances decoded by different decoders; a receiver does not feed a repair symbol of one instance to the decoder of the other, nor reuse a source symbol's coordinate in one instance as its coordinate in the other ([MOQ-FEC-CORE] Section 5.1). A RAP source packet is a source symbol of both instances, at an unrelated coordinate in each (Section 7.4.2).¶
A source track has at most one overlay, a wire constraint explained in
Section 7.4.3, and none when its blockProfile is
"moq-object-v1"; the overlay instance's repair track carries no
repairLayer. A receiver MUST validate this before starting FEC for an
overlay instance of the source track, and on any violation MUST fail
closed as [MOQ-FEC-CORE] Section 10.3 requires for its
own rules: it reports the offending catalog path, does not start FEC for
any overlay instance of that source track, and MUST NOT repair the
catalog by ignoring a track. An overlay is declared only by the
catalog, never in an in-band AL-FEC message, and only where no FEC
source or repair packet carries FFSRP_TS (Section 7.4.3).¶
Under "reed-solomon" the overlay has its own bound: K_overlay +
P_overlay is at most 255 ([MOQ-FEC-BLOCK] Section
7), whatever the base instance uses, and its rows
and columns of A are numbered within the overlay instance
([MOQ-FEC-BLOCK] Section 5.2). An overlay with
K_overlay 256 and P_overlay 32 has an RS_ID range of 0 to 31, an ESI
range in a full overlay block of 256 to 287, and packets carrying
RSB_length 32; it is valid only under "raptorq", since 256 + 32 = 288
exceeds 255. Under "reed-solomon", an overlay whose instance declares
sourceSymbols 192 and repairSymbols 32 is valid, since 192 + 32 =
224.¶
The overlay's source symbols are the fragments of the source track that
belong to a Random Access Point, in transmission order. A publisher
that declares an overlay MUST set RAP_flag on a source packet of the
track, MUST include the packet in the overlay instance, and MUST carry
the packet's overlay SS_ID in its 0x8B01 header extension
(Section 7.4.3), if and only if that packet's payload
belongs to a keyframe MFU covered by the overlay.¶
Overlay source symbols are numbered consecutively within the overlay
instance, in the order the publisher emits them: the overlay SS_ID is 0
at the first overlay source symbol of the stream and advances by one per
overlay source symbol, modulo 2^32 ([MOQ-FEC-CORE] Section
7.6). It is the value the RAP
source packet carries in its 0x8B01 header extension
(Section 7.4.3). It lives in the overlay instance's own
SS_ID space, independent of the base instance's, and the two IDs on one
packet are unrelated numbers. An overlay source block holds at most
K_overlay overlay source symbols, and overlay blocks, including short
ones, are formed, located and wrapped exactly as base blocks are
([MOQ-FEC-CORE] Sections 5.4,
5.5 and
7.6), with K_overlay as the upper
bound on their SSB_length. Because a RAP is a small fraction of the
media, an overlay block spans many base blocks; the overlay's recovery
latency is therefore its own block span and is unrelated to
interleaveDepthMs.¶
A receiver knows that a source packet is an overlay source symbol
because the packet carries the 0x8B01 header extension
(Section 7.4.3). It knows that a repair packet belongs to
the overlay because it arrives on the repair track that the catalog
declares for the overlay instance of this source track, and that track
has its own packet_id. The instance is therefore fixed by the track
the packet arrived on; no field of the repair packet distinguishes the
two, and a receiver does not infer the instance from SS_Start,
RSB_length, or SSB_length, whose values are only interpretable once the
instance is known ([MOQ-FEC-CORE] Section
7.5).¶
When the catalog declares a keyframe overlay for a source track (Section 7.4), a source packet of that track carrying a Random Access Point fragment is a source symbol of two FEC instances. It carries its base SS_ID in the one Source FEC Payload ID trailer that every FEC source packet carries (Section 7.2), and its overlay SS_ID (Section 7.4.2) in an MMTP header extension ([ISO.23008-1] Clause 9.2.3) of type 0x8B01:¶
Source Packet with Overlay {
MMTP Packet Header {
..., # X = 1, RAP_flag = 1
Header Extension {
Type (16) = 0x8B01,
Length (16) = 4,
Overlay SS_ID (32),
},
},
MMTP Payload (..),
Source FEC Payload ID (32), # base instance SS_ID
}
¶
A publisher that declares an overlay MUST carry this extension as the
header extension, with X = 1, of exactly the source packets that it
includes in the overlay instance, that is the packets on which it sets
RAP_flag (Section 7.4.2), and of no other packet of the track.
An MMTP packet carries at most one header extension ([ISO.23008-1]
Clause 9.2.3), so the RAP packets of a track under an overlay carry no
other MMTP header extension, and a publisher that needs another header
extension on those packets cannot use the overlay for that track.
[ISO.23008-1] specifies no particular header extension and no registry
of header extension types; this document uses the type 0x8B01 for the
overlay SS_ID (Section 15).¶
The trailer is never more than one Source FEC Payload ID. On a packet with or without the extension, the last four bytes are the base SS_ID, as [ISO.23008-1] Section C.4.2 places it, so a receiver that does not implement overlays locates the trailer and the payload as on any other packet, stepping over the extension by its Length field. The extension lies inside the MMTP packet header, so it is part of the protected packet of both instances (Section 7.1), which protect one identical source symbol. The permission of [ISO.23008-1] Clause 9.2.3 to discard a header extension therefore extends only to a receiver's own processing: a node that forwards the packet MUST NOT remove or alter the extension. A RAP packet recovered by either instance carries its overlay SS_ID, so a receiver MAY place a RAP packet that the base instance recovered into the overlay block that the extension locates; Section 7.1 states why the converse requires the packet's base SS_ID.¶
A receiver that implements overlays takes a packet's overlay SS_ID from
this extension and from nothing else. It MUST NOT feed the overlay
decoder a source packet that lacks the extension, whatever its
RAP_flag, or one whose extension has a Length other than 4, and MUST
NOT infer an overlay SS_ID from RAP_flag, from the packet's position,
or from its base SS_ID.¶
A source track has at most one overlay (Section 7.4.1) because a source packet has one header extension, and so carries one overlay SS_ID; relaxing that bound would be a revision of this document and a change to the source packet layout, not a catalog-only change.¶
The overlay is declared only by the catalog: Section 8.3 forbids describing the overlay's repair flow in an AL-FEC message, which identifies an FEC source flow by packet_id ([ISO.23008-1] Clause C.6.3) and cannot describe the RAP subset that the overlay protects, so a receiver that consumes the flow without the catalog uses the base instance alone. An overlay is defined only without FFSRP_TS: a publisher MUST NOT declare an overlay for a source track whose FEC source or repair packets carry FFSRP_TS (ffsrpts_flag = 1, [ISO.23008-1] Section C.5.1), and the extension carries an SS_ID alone.¶
Section 14.1 applies to the overlay SS_ID as to the base one: A3SA signing authenticates neither, and a receiver that requires authenticated block assignment treats each as that section requires.¶
A receiver that implements overlays and has joined the overlay repair track decodes the two instances independently and MAY use the recovered RAP fragments even when the base instance failed to recover the block that carried them.¶
The overlay is the publisher-side half of keyframe protection. The receiver-side half is the repair policy of [MOQ-FEC-MULTICAST] Section 6, under which a receiver gives a keyframe-bearing base block that in-band decoding leaves unrecovered priority unicast repair; its Section 6.6 states how the two combine.¶
A receiver that does not implement overlays MUST fail closed on a
catalog that declares one for a track it intends to consume, that is on
an FEC instance carrying the unknown scope key, and MUST NOT ignore
the key and consume the source track anyway. This is an unknown-key
exception that this document defines (Section 7.4.1) under
[MOQ-FEC-CATALOG] Section 3.4, which lets a block
profile's document define further exceptions: a receiver that ignored
the key could not tell the overlay's repair track from a repair layer of
the base instance, and would feed the base decoder repair symbols of
another FEC instance (Section 7.4.2).¶
This mechanism is not backward compatible with receivers that predate it, for the same reason. A publisher that must serve such receivers publishes the overlay on a separate source track, or does not publish it.¶
Interleaving and typical interleave windows (interleaveDepthMs) follow
[MOQ-FEC-CORE] Section 9.3. A positive
interleaveDepthMs is the elapsed source/repair send-span bound for
each actual FEC block; it does not select FEC block boundaries, derive a
number of groups per block, or require a fixed relationship between an
FEC block and a media Group. A typical window for ARIB STD-B60 content
is 2000 ms. The receiver's block-abandonment timeout -- how long it
waits for a block's source and repair symbols before declaring the block
unrecoverable -- follows the ISO receiver-buffer model [ISO.23008-1]
Section 11.2.¶
[MOQ-FEC-CORE] Section 10.1 admits a positive
interleaveDepthMs and lets a block profile give the value zero a
meaning of its own. Under this profile the field MAY also be zero. A
value of zero has the ISO meaning of using protection_window_size
instead of a time window ([ISO.23008-1] Clause C.6.3); this catalog
profile carries no packet-count window. A catalog-only receiver
therefore MUST NOT treat zero as a zero-duration span or use it to
derive a block count, and has no elapsed-time FEC deadline from this
field. If an applicable ISO packet-count window is accepted in-band, it
governs separate count-based processing; it does not supply a
millisecond deadline, it is not converted to milliseconds without an
independently defined time mapping, and the receiver MUST NOT assume a
media cadence to convert it. A publisher uses a zero value only with
that ISO protection_window_size alternative. An absent field supplies
no catalog-signaled span; neither an absent nor a zero value supplies an
elapsed-time deadline. Without an elapsed-time deadline, open-block
state is bounded only by the local policy of [MOQ-FEC-CORE] Section
9.4, and on a multicast or AMT leg the receiver
performs no block classification or unicast repair
([MOQ-FEC-MULTICAST] Section 6) and
keeps its repair-layer selection fixed (its Section
4).¶
FEC parameters for mmtp-packaged tracks are signaled in the catalog
fec field, whose encoding [MOQ-FEC-CATALOG] defines (its Section
3) for the configuration data model of
[MOQ-FEC-CORE] Section 10, which is
authoritative for MoQ receivers (that section). A MoQ receiver requires
no in-session FEC signaling: the catalog carries every FEC parameter it
needs. This document does not redefine the catalog fields but specifies
MMT-specific considerations for their use.¶
A publisher MAY also carry the MMTP AL-FEC signaling message of
Section 8.3 on a MoQ mmtp source track, as an object whose payload
is an MMTP signaling packet on packet_id 0 (Section 3.1). When
present, it is published in Subgroup 0 of an MPU group as the object
immediately before, and in the same subgroup as, that group's FT=0
object, and a group carries at most one such object (Sections 4.2
and 4.3). On a MoQ
track such an object is a redundant in-band copy of catalog state: the
message MUST NOT signal FEC parameters that contradict the track's
catalog fec instances, and where the two disagree the catalog governs,
as [MOQ-FEC-CORE] Section 10 requires
of any in-band signaling. The object carries no media data
unit and no repair symbol, so a MoQ receiver MUST NOT deliver it to
a media or repair decoder (Section 3.1); a receiver that does not
process MMTP signaling discards it.¶
The signaling object is not an FEC source symbol. Its MMTP packet
has FEC Type 0 and no Source FEC Payload ID; it is not protected by
any repair symbol, is not counted in any FEC block or in
sourceSymbols (K), and has no SS_ID and no Encoding Symbol ID.
Because it occupies an Object ID (Object ID 0 of its group,
Section 4.6), Object IDs on an mmtp source track are not positions
in the FEC block: a receiver MUST take each source symbol's SS_ID from
the Source FEC Payload ID that trails its FEC Type 1 packet
(Section 7.2), never from its Object ID, and places the
symbol in its block by that SS_ID ([MOQ-FEC-CORE] Section
5.5).¶
A MoQ track carries only the signaling packets its publisher chose to place on that track, so the packet_sequence_number values it exhibits on packet_id 0 need not be contiguous. A receiver MUST NOT infer loss from a packet_sequence_number gap on packet_id 0 of a MoQ track, and MUST NOT defer any packet, including a later packet on packet_id 0, behind such a gap.¶
When used with MMT packaging, the instance parameters of [MOQ-FEC-CORE] Section 10.1, as the catalog encodes them ([MOQ-FEC-CATALOG] Section 3.1), map as follows:¶
scheme: the scheme identifier ([MOQ-FEC-BLOCK] Section¶
fec_code_id_for_repair_flow (Section
8.3)¶
blockProfile: "iso-ssbg1-v1" (Section 7.1; registered in [MOQ-FEC-CORE] Section 15.3), the ISO 23008-1 Annex C block profile, for which an AL-FEC message carries the coding byte 0x14 (Section 8.3); REQUIRED on every FEC instance, with no default¶
sourceSymbols: The largest number of MMTP packets (source symbols) that an FEC block may hold; each block holds SSB_length of them ([MOQ-FEC-CORE] Section 5.4). On the MMT path the canonical FEC source symbol is one whole MMTP packet, before its Source FEC Payload ID, framed with its length (Section 7.1)¶
interleaveDepthMs: When positive, the elapsed source/repair send
span bound of each actual FEC block, as defined by [MOQ-FEC-CORE]
Sections 10.1 and
9.3; Section 7.5 defines absent and
zero values under this profile. It is independent of
groupDurationMs and does not imply a block count or block boundary
from MoQ Groups. When ingesting broadcast content, use the ISO
protection-window signaling or a known source/repair send-span bound.
A source-frame count alone does not bound the actual packet send span.¶
See [MOQ-FEC-CATALOG] for the complete catalog field definitions, [MOQ-FEC-CORE] Section 15.2 for the scheme registry, and [MOQ-FEC-BLOCK] for the scheme bounds and the OTI derivation (its Sections 7 and 6.1).¶
For FEC block alignment and timing to be deterministic, an mmtp publisher MUST, in addition to the encoder constraints of [MOQ-FEC-CORE] Section 5.3:¶
Signal the catalog-authored group duration and frame rate in the MoQ
catalog: publish a positive duration as groupDurationMs when its
integer-millisecond form is exact, or as the exact positive
groupDurationTicks; when both are present, groupDurationTicks
wins and groupDurationMs is informational. Signal the keyframe
cadence in keyframeIntervalMs (or keyframeIntervalTicks), and the
frame rate in the framerate field (Sections 4.4.2, 4.4.3 and 13.1).
The ISO 23008-1 AL-FEC message has no field for either value
([ISO.23008-1] Table C.3), so an encoder signals neither there, and
a receiver takes neither from it.¶
If frame rate or GOP size changes mid-session, update the MoQ catalog and signal a manifest discontinuity for DASH/HLS.¶
When interleaveDepthMs is positive, choose it so that the actual
elapsed send span from the first source or repair packet of every
block to its last does not exceed the signaled value. A zero value
is usable only with the ISO protection_window_size alternative of
Section 7.5. The value is not derived from a group count, frame
count, or per-symbol interval.¶
A source track's repair tracks are the track that each of its FEC
instances names in repairTrack and the tracks that name it in source
([MOQ-FEC-CATALOG] Section 3.2; [MOQ-FEC-CORE]
Section 10.2);
[MOQ-FEC-CATALOG] Section 6 gives their
conventional names. Subscription to them follows [MOQ-FEC-CORE]
Section 7.2, and their priorities its Section
11.¶
For SSM and Any-Source Multicast (ASM) delivery there is no bidirectional signaling channel. FEC parameters reach multicast receivers via:¶
MoQ Catalog Extension (authoritative for MoQ receivers,
[MOQ-FEC-CORE] Section 10):
out-of-band delivery in the catalog fec field (see the worked
catalog of [MOQ-FEC-CATALOG] Appendix A).¶
MMTP AL-FEC Signaling (message_id=0x0203): in-band delivery of the AL-FEC message of [ISO.23008-1] Clause C.6, with Table C.3 as replaced by [ISO.23008-1.AMD1], for ISO/IEC 23008-1 receivers that consume the flow without the catalog, such as native broadcast (ATSC 3.0, ARIB STD-B60) receivers. A publisher that sends this message MUST carry it in an MMTP signaling packet (packet type 0x02) on packet_id 0, the MMTP signaling flow, after the signaling message payload header of [ISO.23008-1] Clause 9.3.4.2 (Section 3.1); on multicast delivery it shares the multicast group of the media it describes, and on MoQ delivery it MAY be carried as an object of an mmtp source track under the rules of Section 8. No track is advertised on packet_id 0 ([MOQ-MULTICAST] Section 4.1), and a receiver does not deliver these packets to a media or repair decoder (Section 3.1; [MOQ-MULTICAST] Section 5)¶
Each FEC source flow has one FEC encoded flow ([ISO.23008-1]
Clause C.6.3), so the message describes a source track's flow as one
entry of the message's fec_flow_descriptor naming exactly one repair
flow, the track's layer-0 repair track ([MOQ-FEC-CORE] Section
7.3). For a track whose instance's blockProfile is
"iso-ssbg1-v1", that entry carries:¶
number_of_assets of 1, the one Asset it lists being the source
track's, the Asset whose packet_id is the track's: an FEC source
flow of [ISO.23008-1] may carry the packets of several Assets
(its Clause C.1), where an FEC instance of [MOQ-FEC-CORE]
protects one source track (its Section 5.1);¶
fec_coding_structure 0001
(one-stage), ssbg_mode 01, ffsrpts_flag 0 and
fec_payload_id_mode 0 (Section 7.1);¶
fec_code_id_for_repair_flow, the code point that
[ISO.23008-10] Table 1 assigns to the code that the instance's
scheme names ([MOQ-FEC-BLOCK] Section
3): 3, the "6330 code" (RaptorQ), for
"raptorq", and 1, the Reed-Solomon code, for "reed-solomon";¶
repair_flow_id, the packet_id of the layer-0 repair
packets. [ISO.23008-1.AMD1] makes this field 16 bits in the
one-stage branch, where the 2023 edition of Table C.3 gives it 8
bits and so cannot name a repair packet_id above 255;¶
length_of_repair_symbol, the instance's symbolSize (T), the
byte length of every repair symbol and, under ssbg_mode1, of every
source symbol ([MOQ-FEC-CORE] Section 10.1;
Section 7.1 and Section 7.3), and as its
num_of_repair_symbol_per_packet 1, the one repair symbol each repair
packet carries (Section 7.3; [MOQ-FEC-CORE] Section
7.5), a field whose semantics
[ISO.23008-1] Clause C.6.3 gives for ssbg_mode2 alone;¶
maximum_k_for_repair_flow, the instance's sourceSymbols:
like that field, it bounds each block's SSB_length and is not the size
of any particular block ([MOQ-FEC-CORE] Section
5.4); and¶
maximum_p_for_repair_flow, the RSB_length that every
repair packet of the track's FEC instance carries: the total
repair count over the published layers, which is the instance's
repairSymbols while layer 0 is the only published layer
([MOQ-FEC-CORE] Section 7.3.1). A
publisher that appends a layer ([MOQ-FEC-CORE] Section
7.3.2) raises that total and sends a new
version of the message carrying it only after the last repair packet
generated under the previous total, of whichever block: a block's
RSB_length is the total in force when its repair symbols were
generated ([MOQ-FEC-CORE] Section 7.5), so
every repair packet
sent after the new version carries a value no lower than the
field it names. The path keeps no order between the message, on
packet_id 0, and the repair flow (Section 5.2.2), so a receiver
of both can take a packet generated under the previous total
after the new version; the blocks generated under the raised
total follow every block generated under the previous one in the
serial order of their SS_Start ([MOQ-FEC-CORE] Section
7.6), as its Section
7.3.2 requires of the publisher, which the
conversion rule of Section 13.2 relies on instead.¶
A receiver that holds layer 0 under the AL-FEC message alone, without
the catalog (Section 10), applies the fail-closed check of
[MOQ-FEC-CORE] Section 7.3.2 with
maximum_p_for_repair_flow standing for its O_0 + P_0, by that check's
anchored clause alone: layer 0 has no clause of its own track, its O_0
being 0 and its symbols' coordinates depending on no published count, so
a packet of the flow generated under a previous total, which the message
can precede (above), fails it closed on nothing.¶
Layered repair is an extension of [ISO.23008-1] that an ISO AL-FEC
message cannot express: each FEC source flow has one FEC encoded flow,
whose one-stage coding structure carries a single repair_flow_id
([ISO.23008-1] Clause C.6.3), so only layer 0 maps to an ISO repair
flow. A repair layer is also unrelated to the layers of ISO layer-aware
FEC (LA-FEC, Clause C.2.3). A publisher MUST NOT describe a repair
layer above layer 0, or a keyframe overlay, in the message, whether in
that entry or in an entry of its own: the one-stage coding structure
names one repair flow per FEC encoded flow, the overlay protects a RAP
subset that no packet_id identifies, and only the catalog declares
either ([MOQ-FEC-CORE] Section 7.3;
Section 7.4 and Section 7.4.3).¶
MMT content can be delivered via IP multicast (SSM, AMT) and TreeDN [RFC9706]. Platform-specific delivery paths, the multicast endpoint catalog extension, and TreeDN/AMT integration are defined in [MOQ-MULTICAST], and the multicast packet format in its Section 5.¶
This specification is designed for interoperability with the Application-Layer FEC of ATSC A/331 [ATSC-A331] MMTP sessions. [ATSC-A331] applies AL-FEC to an MMTP session, when it applies it at all, through the AL-FEC framework of [ISO.23008-1] Annex C and requires no FEC code of its own (its Section 8.1.2.5): the framework carries whichever code the AL-FEC message names by its [ISO.23008-10] code point. The codes are those of [MOQ-FEC-BLOCK], which specifies two, "reed-solomon" (code point 1, mandatory to implement) and "raptorq" (code point 3, OPTIONAL) (its Sections 3 and 10); Section 8.3 gives the AL-FEC message a publisher sends for a track of either, and an MMTP session whose AL-FEC message names any other code is outside this document and its companions until the "MoQ FEC Schemes" registry of [MOQ-FEC-CORE] Section 15.2 lists that code. A publisher that ingests an A/331 MMTP session takes the FEC parameters of its tracks from that message and the asset locations of the MP table (Section 13.2), and a receiver that consumes the session without the catalog takes them from the message alone, holding the one repair flow as layer 0 under the block total the message bounds, as Section 8.3 specifies. An MMTP session carries no [RFC6330] OTI: the derived OTI of [MOQ-FEC-BLOCK] Section 6.1 fixes N = 1, and sub-blocks are never used (Section 7.1). ROUTE delivery, whose repair protocol (Annex A.4 of [ATSC-A331]) is RaptorQ-based, is outside the scope of this document.¶
ARIB STD-B60 [ARIB-B60] (Japan's MMT-based broadcasting standard) uses the same ISO 23008-1 foundation as ATSC 3.0. Publishers SHOULD preserve the original FEC parameters when ingesting ARIB STD-B60 content.¶
ARIB STD-B60 uses the MMTP Timestamp of Section 3.1: the UTC send time of the packet in NTP short format.¶
Clients SHOULD attempt transports in the preference order that [MOQ-MULTICAST] Section 3 defines for native and browser clients. AL-FEC (Section 7) is the loss recovery on SSM/AMT paths, which have no retransmission.¶
The MoQ catalog indicates MMT packaging and multicast endpoints.¶
The catalog of this suite is an MSF catalog: the root object and its
fields (version, generatedAt, tracks, ...), the requirement that
parsers ignore fields they do not understand, and the base per-track
fields are those of [I-D.ietf-moq-msf] Section 5. The catalog is
delivered as a MoQ track whose case-sensitive Track Name is catalog
([I-D.ietf-moq-msf] Section 5). The documents of this suite extend
the base in the same manner as [I-D.ietf-moq-cmsf] extends it: this
document adds the packaging value mmtp and the mmtp track fields
(Section 13.1), [MOQ-FEC-CATALOG] adds the per-track fec array of
FEC instances and the fec-repair and fec-source packaging values
(its Sections 3 and
3.3), and [MOQ-MULTICAST] adds the root-level
multicast member. Every catalog field this suite uses is defined
either by [I-D.ietf-moq-msf] or locally by a document of this suite
(with cmaf tracks per [I-D.ietf-moq-cmsf]); no field is defined by
any other catalog specification.¶
This document defines a single packaging value:¶
packaging: "mmtp"¶
carried in the per-track packaging field of the catalog
([I-D.ietf-moq-msf] Section 5.2.4). The value extends the
allowed packaging values of that specification in the same manner
as the cmaf value of [I-D.ietf-moq-cmsf] Section 3.5.1; see
Section 15 for the registry request. Object payloads of an
mmtp-packaged track are whole MMTP packets per Section 4.2.¶
Per-track catalog fields for mmtp packaging:¶
| Field | Status | Type | Description |
|---|---|---|---|
packaging
|
REQUIRED | String | MUST be "mmtp" |
mmtpMode
|
REQUIRED | String | MUST be "mfu" (Section 5); defined by this document |
timescale
|
REQUIRED | Number | Media timescale in Hz ([I-D.ietf-moq-msf] Section 5.2.21, profiled REQUIRED here; Section 4.4.2) |
presentationAnchorTicks
|
REQUIRED for synchronized audio/video tracks | Signed integer | Track timestamp corresponding to shared presentation time zero (Section 4.4.1); defined by this document |
groupDurationMs
|
Conditional | Number | REQUIRED without the tick form; positive integer ms (Section 4.4.2); informational when ticks are present |
groupDurationTicks
|
OPTIONAL | Number | Positive exact group duration in track ticks; authoritative when present (Section 4.4.2); defined by this document |
altGroup
|
OPTIONAL | Number | Switching-set membership key ([I-D.ietf-moq-msf] Section 5.2.12, profiled in Section 4.4) |
keyframeIntervalMs
|
OPTIONAL | Number | Keyframe/GOP cadence, integer ms; video only (Section 4.4.3); defined by this document |
keyframeIntervalTicks
|
OPTIONAL | Number | Integer-tick override for the keyframe interval (Section 4.4.3); defined by this document |
initRef
|
OPTIONAL | String |
id of the root initDataList entry, of type "inline", that holds the base64 ftyp+moov init segment ([I-D.ietf-moq-msf] Sections 5.1.7 and 5.2.13, profiled in Section 4.5) |
fec
|
OPTIONAL | Array | FEC instances, encoded per [MOQ-FEC-CATALOG] Section 3 (Section 8) |
All other base track fields of [I-D.ietf-moq-msf] Section 5.2
apply unchanged. In particular name and the rendition
parameters -- codec ([I-D.ietf-moq-msf] Section 5.2.18; required
for media tracks per that section), framerate, width, height,
samplerate, channelConfig, bitrate, and lang -- are flat,
top-level fields of the track object.¶
mmtpMode selects the object layout and MUST be "mfu". It delivers
one subgroup per FT=0, FT=1, or FT=2 data unit as defined in Sections
4.3 and 5. A subscriber MUST reject a track whose
mmtpMode is absent or has any other value; the object layout cannot
be inferred safely from received objects.¶
Example, with the data value shortened:¶
{
"version": 1,
"tracks": [{
"name": "video",
"packaging": "mmtp",
"isLive": true,
"mmtpMode": "mfu",
"timescale": 90000,
"presentationAnchorTicks": 0,
"groupDurationMs": 1000,
"codec": "avc1.640028",
"width": 1920,
"height": 1080,
"framerate": 30,
"bitrate": 6000000,
"initRef": "video-init"
}],
"initDataList": [{
"id": "video-init",
"type": "inline",
"data": "AAAAIGZ0eXBpc29t..."
}]
}
¶
This section is informative: it documents how a converter that ingests an MMTP broadcast service, an ATSC 3.0 MMTP session ([ATSC-A331] Section 8.1.2.1) or an ARIB STD-B60 service, derives a MoQ catalog from the MMT signaling, and imposes no interoperability requirements on MoQ publishers or subscribers. A converter's output catalog is subject to the normative catalog requirements of Section 13.1 like any other catalog. The MP table gives the location of each Asset ([ISO.23008-1] Clause 10.3.9) and the AL-FEC message the FEC parameters of each protected flow ([ISO.23008-1] Annex C.6; Section 8.3 gives the fields a MoQ publisher sets, which a converter reads back).¶
The multicast field in the output uses the multicast endpoint
format of [MOQ-MULTICAST] Section
4.1. Conversion rules:¶
multicast.endpoints[] entry (sourceAddress, groupAddress,
port), and the Assets it carries -> that endpoint's tracks[]¶
packetId is the Asset's packet_id,
unique within the session ([ATSC-A331] Section 8.1.2.1.1) and
carried by every packet of the flow, so a converter cannot renumber
it. [MOQ-MULTICAST] Section 4.1
requires each packetId value to
denote one sub-flow across all endpoints of a sending entity and
reserves 0 for the signaling flow, so a converter publishes
sessions whose packet_ids collide under different namespaces, and
never advertises a track on packet_id 0¶
scheme from
fec_code_id_for_repair_flow, "reed-solomon" for code point 1 and
"raptorq" for 3 ([MOQ-FEC-BLOCK] Section
3), symbolSize (T) from
length_of_repair_symbol, sourceSymbols (K) from
maximum_k_for_repair_flow, repairSymbols (P) from
maximum_p_for_repair_flow, and a repair track, named in
repairTrack, whose packetId is the packet_id that repair_flow_id
names (Section 8.3). The converter accepts an
entry whose number_of_assets is 1, whose coding byte is 0x14 and
whose fec_code_id_for_repair_flow is 1 or 3: a source flow of
several Assets ([ISO.23008-1] Clause C.1) is one block sequence
over the packets of several packet_ids, which no FEC instance of
[MOQ-FEC-CORE] protects (its Section 5.1 protects
one source track), a flow under any other coding structure,
ssbg_mode, FFSRP_TS or FEC payload ID mode is not one that
"iso-ssbg1-v1" describes (Section 7.1), a session whose message
names any other code is outside this document and its companions until
the "MoQ FEC Schemes" registry of [MOQ-FEC-CORE] Section
15.2 lists that code
(Section 10), and the conversion of such a flow fails. The
one
repair flow the
one-stage entry names is the track's layer 0, and
maximum_p_for_repair_flow is its P while layer 0 is the only
published layer, as in a one-stage session of [ISO.23008-1],
whose one repair flow carries the block's whole repair count.
That field bounds the repair count of a block ([ISO.23008-1]
Clause C.6.3: the maximum allowed number of repair symbols in a
repair symbol block) where "iso-ssbg1-v1" carries P as the RSB_length
of every repair packet (Section 7.3; [MOQ-FEC-CORE] Section
7.5), on
which every receiver holding the highest layer the converter lists
(below) fails closed when a packet of that layer, where it is
above 0, or of a block at or after (in the serial order of
[MOQ-FEC-CORE] Section 7.6) one
of which it has received a repair
packet carrying at least the value, carries less than the
O_i + P_i its catalog derives for that layer, the field itself
(its Section 7.3.2), so
the converter, which receives the flow's repair packets before it
lists the repair track (below), lists it only once it has received
a repair packet of the flow carrying an RSB_length of at least the
field of the version it holds, and fails the conversion of a flow
whose repair packets carry an RSB_length below that field on a
block at or after, in the serial order of their SS_Start
([MOQ-FEC-CORE] Section 7.6),
one of which it has received a repair packet carrying at least the
field: such a flow is not one
"iso-ssbg1-v1" describes. A block before that one whose repair
packets carry less was generated under the previous total, after
whose last repair packet a publisher that appends a layer sends
the raised field (Section 8.3) but which the path, keeping no
order between the message and the repair flow (Section 5.2.2),
can deliver after it; and a flow no block of which carries at
least the field is one the converter never lists. Such a packet
of an earlier block, below the repairSymbols the catalog gives
layer 0 (below) where the layers the converter lists above 0 count
less than the appended one, fails a receiver of the catalog closed
on nothing: the clause of [MOQ-FEC-CORE] Section
7.3.2 for a packet of a layer's own track
applies to layers above 0 alone, layer 0's
coordinates depending on no count, and its anchored clause, which
anchors a held layer on a block of which the receiver has received
a repair packet carrying at least that layer's O_i + P_i, for
layer 0 its repairSymbols (the field where the converter lists
none above 0), treats such a packet as the conforming packet it
is, every block generated under a total below a layer's O_i + P_i
preceding, in the serial order of its Section
7.6, every block generated under
one at or above it (its Section 7.3.2, which
has
a publisher apply a raised total from a block onward in that order
and lower none), on the repair track the
converter publishes as on the session's own endpoint, which
carries the sender's packets and from which the converter
withholds nothing.
An RSB_length above the
field is the append case of the next sentence. A
MoQ publisher that appended a layer ([MOQ-FEC-CORE] Section
7.3.2) sends the raised total in that field
(Section 8.3) while the named flow carries only the P_0 symbols of
each block ([MOQ-FEC-CORE] Section 7.3.1),
and neither the message nor the MP table describes
the layers above, so the converter sets repairSymbols to
maximum_p_for_repair_flow less the repairSymbols of every
layer above 0 that it lists, and lists a layer it learns from
provisioning only where provisioning places it, with every other
layer it lists, above every layer it does not list, the topmost
layers of the publisher's set: layer 0's repairSymbols is then
the publisher's total below the lowest layer it lists above 0 (the
whole field where it lists none) and each listed layer's O_i
([MOQ-FEC-CORE] Section 7.3.1) is the
publisher's, while listing a
layer above which lies one the converter does not list would give
it an O_i that is not the publisher's, which no check detects, and
an O_i + P_i that its own-track packets generated before the
append of the layer it does not list fall below (its Section
7.3.2). One that lists none publishes the
whole field, which
overstates layer 0's count, and the loss tolerance a receiver
derives from it, by the layers it does not list. The converter
keeps the repairSymbols it first
published for a track, and those of every layer it has listed,
for as long as it publishes the track: a later version of the
message that raises maximum_p_for_repair_flow changes nothing in
its catalog by itself, since [MOQ-FEC-CORE] Section
7.3.2 forbids
changing the repairSymbols of a published layer while any
receiver may still be consuming the source track, and says that a
receiver reading the smaller bound excludes only symbols of layers
it has not joined. A layer above 0 that the converter learns from
provisioning after it published the track, one that provisioning
places next above the highest layer the converter lists, the
publisher having appended it after the converter read the field
with no layer the converter has not learned below it, so that its
O_i, the total of every layer below it ([MOQ-FEC-CORE] Section
7.3.1), is the total the catalog derives for
the highest layer
listed before it, layer 0's repairSymbols plus those of every
layer above 0 listed before it, and the field of the version it
holds has risen by at least that layer's repairSymbols, it
appends at the next index (its Section
7.3.2): its symbols' RS_IDs
run upward from that total, and its O_i + P_i is at most the
field, which every block generated under the raised total carries.
The field's rise alone places no layer, a field risen by more than
a layer's repairSymbols counting a layer the converter has not
learned, which can lie below the learned one. A layer that
provisioning does not so place, or whose symbols the field the
converter first read already counted, it lists only by retiring
the source encoding and publishing the source track and its repair
tracks as a separate rendition that lists it, since appending it
would give it an O_i that is not the publisher's, which no check
detects, or an O_i + P_i above the wire RSB_length, on which every
receiver joining it fails closed (that section), while lowering
layer 0's repairSymbols to fit it is the change that section
forbids. A
later version that lowers maximum_p_for_repair_flow below the
total its catalog derives for the highest layer it lists, layer
0's repairSymbols plus those of every layer above 0 it lists, is
a change that section lets no publisher make, on which a receiver
holding the highest layer it lists fails closed on the first
packet of that layer, where it is above 0, or of a block at or
after, in that serial order, one of which it has received a repair
packet carrying at least that total, whose RSB_length falls below
the O_i + P_i its catalog derives for that layer, which is that
total, and one that no such packet reaches, the check anchored on
no block it received or, holding that layer alone, reached by no
repair packet after the change, learns of the change only at the
next catalog refresh (that section), so the
converter
retires the source encoding and
publishes the source track and its repair tracks as a separate
rendition carrying the lowered value, as that section has a
publisher that must retire a layer do; a repair packet whose
RSB_length falls below that total once the track is published, on
a block at or after, in that serial order, one of which it has
received a repair packet carrying at least that total, is the same
violation and has the converter retire the source encoding in the
same way, the separate rendition listing the repair track as the
first listing did, once a repair packet of the flow received after
that one carries at least the field of the version it then holds,
as the violating packet, below the field of the version held when
it arrived, did not. The
conversion
of a flow fails where the FEC instance it would yield fails a rule of
[MOQ-FEC-CORE] Sections 10.1 and
10.3, or of [MOQ-FEC-CATALOG] Section
5, on which a receiver fails closed: a
maximum_p_for_repair_flow of 0 names no repair flow a catalog can
publish ([MOQ-FEC-CORE] Section 10.1
requires a repairSymbols of at least 1), a
maximum_k_for_repair_flow of 0 names no block (a sourceSymbols of
at least 1), and a length_of_repair_symbol below 14 names a symbol
that protects no MMTP packet under the T - 2 construction of
Section 7.1, whose protected packet begins with the 12-octet MMTP
packet header of Section 3.1 (a symbolSize of at least 14 under
"iso-ssbg1-v1"); an entry that names code point 3 with a
length_of_repair_symbol that is not a multiple of 8 names a T no
"raptorq" track can carry ([MOQ-FEC-BLOCK] Section
6.1 fixes the Symbol Alignment at 8, which an MMTP
session, carrying no OTI, does not impose on its symbol length), one
with maximum_k_for_repair_flow above 56403 fails the RaptorQ bound
of [MOQ-FEC-BLOCK] Section 7, and one with
maximum_k_for_repair_flow plus maximum_p_for_repair_flow above
2^24 names an ESI no instance can produce ([MOQ-FEC-CORE] Section
10.3 bounds K + totalP - 1 by 2^24 - 1, a sum the two
24-bit fields can exceed); and an entry that names code point 1 with
maximum_k_for_repair_flow plus maximum_p_for_repair_flow above 255
names a block no "reed-solomon" instance can hold (the Reed-Solomon
bound of [MOQ-FEC-BLOCK] Section 7, which
bounds K plus the repair count of every layer). The
conversion also fails for a flow whose repair_flow_id is 0, the
signaling flow (Section 3.1), the flow's own packet_id, or the
packet_id of another track of the session, a value that
[MOQ-MULTICAST] Section 4.1
requires to denote one sub-flow (a
repair flow carries a packet_id of its own, [ISO.23008-1]
Clause C.6.3), and for the flows of any two entries that name one
repair_flow_id or list one Asset: two source tracks naming one
repair track, or two base instances for one track, are shapes no FEC
instance of [MOQ-FEC-CORE] has, a repair packet belonging to the
instance that the configuration relates to the track it arrived on
(its Sections 5.1 and
7.5). The
message names the repair flow
by packet_id without a location, and the MP table does not list it
([MOQ-MULTICAST] Section 5),
so the converter lists the repair
track in the tracks[] of each endpoint on whose session it
receives packets of that packet_id, which is where a receiver
without the catalog finds them; a repair-only endpoint
([MOQ-FEC-MULTICAST] Section 4) is one
the converter learns from provisioning, not from the signaling¶
interleaveDepthMs, the ISO protection-window send
span from
the first source or repair packet to the last: set only when the
AL-FEC message's protection_window_time or an independently known
source/repair send schedule supplies that span, as a positive
integer millisecond upper bound that covers the complete span,
never rounded down; when the span is unknown, or the message
signals zero (the size-based case), the converter omits it¶
isLive = true ([I-D.ietf-moq-msf]
Section 5.2.7) on each media track and on its repair track, which
the converter publishes while it ingests the service¶
presentationAnchorTicks; a converter preserves one
common presentation origin across audio and video, and does not
emit zero merely because the source offset was unavailable¶
bitrate ([I-D.ietf-moq-msf] Section 5.2.22) from whatever
signaling carries it, or from another source, or the conversion of
that track fails¶
Flows sharing one (sourceAddress, groupAddress, port) tuple collapse
into a single endpoint whose tracks[] array lists them all. MMT
signaling carries no networkSource ([MOQ-MULTICAST] Section
4.2).
A converter that adds an AMT network source advertises "driad"
discovery only for a source address whose reverse DNS zone carries an
AMT relay record, and otherwise names a relay service hostname in
relay with "dns" discovery ([MOQ-MULTICAST] Section
4.2.1).¶
timescale, the group-duration field and mmtpMode are required
fields (Section 13.1). The converter takes timescale from
asset_timescale in the MP table when asset_timescale_flag is 1,
and uses 90 000 Hz when that flag is 0, as specified by
[ISO.23008-1] Clause 10.3.9.3. For each timed
Asset, it obtains the presentation time of each MPU's first AU in
presentation order from the PI document when PI is present;
ISO/IEC 23008-1 Clause 10.5.2.1 says that the MPU timestamp
descriptor is ignored in that case. Otherwise it decodes each
descriptor's 64-bit NTP mpu_presentation_time value
([ISO.23008-1] Clause 10.5.2) and does not treat the encoded NTP
integer as a media tick count. It converts those presentation
instants through the selected asset/media clock, retaining the
authored media timing when NTP fractional quantization does not land
on an integer media tick. That instant is not ticks
(Section 4.4.1) when the MPU has leading pictures: ticks is that
instant plus the difference between the authored composition times
(Section 5.1) of the MPU's first AU in decode order and its
earliest-presented AU, converted to the selected timescale as
Section 4.4.1 requires.¶
The converter derives the group duration from the authored MPU/media
cadence and validates it against those ticks values. It may emit
groupDurationMs only when the conversion to the selected timescale
is an exact positive integer tick count; when the cadence is exact
only in ticks it emits groupDurationTicks.
The fixed-cadence profile cannot represent a source whose MPU cadence
is not one constant positive duration, so conversion fails rather
than rounding or inventing a cadence. It sets mmtpMode itself.
The AL-FEC message does not name blockProfile, which is required on
every FEC instance ([MOQ-FEC-CORE] Section
10.1); the converter sets it to "iso-ssbg1-v1",
the profile whose entry carries the coding byte 0x14 that the converter
accepts (Section 8.3; Section 7.1).¶
MMT content protection uses Common Encryption (CENC) [ISO.23001-7] which is preserved through MoQ transport. The MMTP packet and payload headers are not encrypted by CENC, so receivers, and MMTP gateways outside MoQ, can inspect packet type, sequence, and fragmentation metadata without accessing media sample content. A MoQ relay treats the object payload, MMTP headers included, as opaque ([I-D.ietf-moq-transport] Section 7.8) and never parses MMTP (Section 6.2).¶
Per [ATSC-A360] Section 5.2.2.5, the ATSC 3.0 signing mechanism, whose certificates the ATSC 3.0 Security Authority (A3SA) administers, covers signaling messages and MA3 messages (packet type 0x2) but excludes asset packets (type 0x00 MPU and type 0x03 repair).¶
Neither [ISO.23008-1] nor [ATSC-A331] carries the SS_ID inside the
A3SA signed region. In particular the MMT hint sample does not: neither
its ISO syntax ([ISO.23008-1] Clause 8.3.2) nor the ATSC
MMTHSampleATSC3 syntax ([ATSC-A331] Section 7.2.4.2.2.3) has a
Source FEC Payload ID field, and hint samples travel as MPU-mode data
units on the Asset's own packet_id ([ATSC-A331] Section 7.2.4.2.3),
that is in asset packets, which A3SA does not sign. This document
therefore defines no A3SA-authenticated FEC block assignment: block
assignment taken from the trailing Source FEC Payload ID
(Section 7.2) of an A3SA verified packet is
unauthenticated. A receiver that requires authenticated FEC block
assignment MUST obtain it from an integrity mechanism, other than A3SA
signing, whose coverage includes the complete MMTP packet, the trailer
included. Version 2 of the bc-provenance profile ([MOQ-MULTICAST]
Section 7.2) is such a mechanism: its
digest covers the whole UDP payload as transmitted, the Source FEC
Payload ID included. The same holds for the overlay SS_ID
(Section 7.4.3): A3SA signing authenticates neither, and a
receiver that requires authenticated block assignment treats each as
this section requires.¶
An attacker who can modify the Source FEC Payload ID without detection could redirect source symbols to incorrect FEC blocks, causing recovery failures or corrupted output. On multicast networks without A3SA verification, receivers SHOULD cross-check the SS_ID against the MMTP packet_sequence_number to detect obvious tampering (e.g., an SS_ID that advances further than the packet_sequence_number of the same packet_id, whereas SS_ID advances by one per FEC source packet, Section 7.1).¶
Multicast-specific security considerations (source authentication, replay protection, AMT relay trust) are defined in [MOQ-MULTICAST] Section 7.¶
This document defines the catalog packaging value "mmtp"
(Section 13.1), extending the allowed packaging values of the MOQT
Streaming Format [I-D.ietf-moq-msf] Section 5.2.4 in the same
manner as [I-D.ietf-moq-cmsf]. If a registry of packaging
values is established for that format, this document requests
registration of:¶
| Value | Description | Reference |
|---|---|---|
| "mmtp" | MMTP packets carrying MPU/MFU payloads | This document |
This document requests the registration of the block profile "iso-ssbg1-v1" (Section 7.1) in the "MoQ FEC Block Profiles" registry that [MOQ-FEC-CORE] Section 15.3 creates:¶
| Block Profile | Admitted Packaging | Reference |
|---|---|---|
| "iso-ssbg1-v1" | mmtp | This document, Section 7.1 |
The header extension type 0x8B01 of Section 7.4.3 is an MMTP value, not a MoQ one; [ISO.23008-1] Clause 9.2.3 defines no registry of MMTP header extension types, so its use requires no IANA action.¶
This document requests no MoQ message type registrations. FEC signaling uses no MoQ control message ([MOQ-FEC-CORE] Section 15.1); no document in this suite requests a control-message codepoint.¶
This appendix is informative. It is to be removed before publication as an RFC.¶
The FEC draft [I-D.ramadan-moq-fec] was split into a core mechanism [MOQ-FEC-CORE] and companion documents [MOQ-FEC-BLOCK], [MOQ-FEC-CATALOG] and [MOQ-FEC-MULTICAST]. The MMTP block profile moved into this document, and every citation of the FEC draft was re-pointed. The section numbers in the -00 entry below are those of -01.¶
Moved in from draft-ramadan-moq-fec-01: the "iso-ssbg1-v1" block
profile definition (-01 Sections 5.1, 7.3, 7.4 and 8.3, as the new
Section 7.1), the Source FEC Payload ID trailer and its relation to
the FEC Source Info Property (-01 Sections 5.3 and 8.5, as Section
7.2), the MMTP form of the repair packet (-01 Section 7.1, as Section
7.3), the keyframe overlay and its 0x8B01 header extension (-01
Sections 2, 5.1.1, 5.1.2, 6.3.1, 6.3.4 rule 10, 6.4 to 6.4.3, 7.6.6
and 8.5.1, as Section 7.4), the ISO packet-count meaning of
interleaveDepthMs zero (-01 Sections 2, 5.1, 8.4 and 9, into Section
7.5, which was Section 7.1), the mmtp encoder constraints (-01 Section
8.4, into Section 8.1), the ISO single-repair-flow paragraph and the
catalog-less layer-0 check (-01 Sections 6.3 and 6.3.9, into Section
8.3), ATSC 3.0 compatibility (-01 Section 12, as the new Section 10),
the A3SA signed region (-01 Sections 8.5, 8.5.1 and 14.5, as the new
Section 14.1), the "iso-ssbg1-v1" registration and the 0x8B01 note
(-01 Sections 15.1 and 15.4, into Section 15), the MMTP capacity
arithmetic (-01 Appendix B, as Appendix B), the mmtp rendition of CMAF
content (-01 Section 13, 3893-3900, into Section 1.1; the diagram of
-01 Section 13 depicts direct protection of a cmaf track under
"moq-object-v1" and is not carried here) and the FEC terms specific to
this profile (-01 Section 2, into Section 2). Section 7.2 OTI
Signaling of -01 is folded into Section 7.1.¶
Section 7.1 states where the profile deviates from [MOQ-FEC-CORE]:
the unit of protection is the MMTP packet, the trailer carries the
SS_ID and the FEC Source Info Property restates it in the 4-byte form
naming instance 0 (and a receiver that parses the packet can place a
symbol by the trailer when the Property is absent, Section 7.2), the
repair object is an MMTP packet of 12 + 13 + T bytes rather than 13 +
T, the keyframe overlay is the profile's form of an overlapping
instance with its overlay SS_ID in the 0x8B01 extension rather than in
a Property entry, so that overlay membership is read from the
extension inside the protected packet rather than from the Property
alone and an overlay recovery counts toward the base instance only
with an independently determined base SS_ID, interleaveDepthMs
admits zero, and symbol carriage does not apply.¶
Duplicates collapsed to one normative copy: the in-band precedence rule (Section 8 cites [MOQ-FEC-CORE]), the coding-byte paragraph (Section 8.3), the prohibition on describing the overlay in the AL-FEC message (Section 8.3; Section 7.4.3 keeps the rationale and the FFSRP_TS prohibition), and the catalog-less layer-0 anchored check (Section 8.3; Sections 10 and 13.2 cite it). Three further -01 passages that fall to this document were already stated by draft-ramadan-moq-mmt-01 Sections 8 and 8.3 and are collapsed into that text rather than moved: the -01 Section 5.2 sentences naming the AL-FEC message as the in-band FEC signaling of the mmtp packaging for a receiver that consumes the flow without the catalog (Section 8.3; Section 8 makes the message optional on a MoQ track), the -01 Section 8.3 paragraph on the signaling object that carries no source symbol, no SS_ID and no ESI and is not counted in K (Section 8), and the -01 Appendix A.1 sentence that no in-session message is required and that a track may carry no AL-FEC signaling object (Section 8). Their profile-neutral sentences are kept by [MOQ-FEC-CORE] Sections 5.3 and 10 and Appendix E.¶
Every citation of the FEC draft is re-pointed: the mechanism to
[MOQ-FEC-CORE], the schemes, their bounds, the OTI derivation and
the mandatory-to-implement designation to [MOQ-FEC-BLOCK], the
catalog encoding, the worked catalog, the naming convention and the
unknown-key rules to [MOQ-FEC-CATALOG], and the receiver repair
policy, block classification and repair-only endpoints to
[MOQ-FEC-MULTICAST]; the profile, the trailer, the repair packet and
the overlay are cited within this document. Catalog field names
follow the data model of [MOQ-FEC-CORE]: algorithm is scheme,
depends is source, the per-track fec field is an array of FEC
instances, and the "none" scheme is no longer used (a track without
instances declares no FEC). [RFC6330] is now an informative
reference, since RaptorQ is specified by [MOQ-FEC-BLOCK].
[ATSC-A360] is a new informative reference and [MOQ-MULTICAST] is
cited at its -02 revision.¶
Section 1.2 names the document set, Section 2 lists the FEC terms this document uses and points at [MOQ-FEC-CORE] for the rest, and the Abstract and Section 1 say that this document defines the block profile, the overlay and the ATSC 3.0 interworking. Sections 10 to 14 of -01 are now Sections 11 to 15.¶
The worked catalogs carry "version": 1, the catalog version of
[MOQ-FEC-CATALOG], in place of the "draft-01" string of -01.¶
Sentences that are new rather than moved, listed verbatim with the -01 text they replace where one exists:¶
This specification defines the mapping of MMT packets to MoQ objects, the "iso-ssbg1-v1" block profile under which the Forward Error Correction (FEC) mechanism for MoQ protects MMTP packets -- its source symbols, its Source FEC Payload ID trailer, its repair packet and its keyframe overlay -- the derivation of MoQ catalogs from MMTP signaling, and compatibility considerations for both ATSC 3.0 and ARIB STD-B60 systems, including the ATSC 3.0 signed region.¶
FEC: MMT's AL-FEC integrates with MoQ FEC repair tracks per [MOQ-FEC-CORE], under the "iso-ssbg1-v1" block profile that this document defines (Section 7)¶
Broadcast interworking: ATSC 3.0 A/331 AL-FEC sessions and the ATSC 3.0 signed region (Sections 10 and 14.1), and ARIB STD-B60 (Section 11)¶
The last column names the block profile under which the FEC mechanism of [MOQ-FEC-CORE] protects tracks of each packaging.¶
"moq-object-v1" is the profile of [MOQ-FEC-CORE] Section 5.2: each loc or cmaf object whose Object Status is Normal occupies whole source symbols.¶
"iso-ssbg1-v1" is the profile that Section 7.1 of this document defines: each FEC source packet (FEC Type 1) is one source symbol.¶
Both are registered in the "MoQ FEC Block Profiles" registry of [MOQ-FEC-CORE] Section 15.3.¶
A publisher that offers CMAF content can also carry it as an mmtp track under "iso-ssbg1-v1", where an MPU corresponds to a CMAF Fragment (Section 4.2). [reworded from -01 13 3893-3895: "Alternatively, the publisher carries the media as an mmtp track under iso-ssbg1-v1, where an MPU corresponds to a CMAF Fragment (see [MOQ-MMT])"; the alternative it followed, direct protection under "moq-object-v1", is [MOQ-FEC-CORE] Section 13.2, and the pointer to this document became a pointer to Section 4.2; the two sentences that follow are carried verbatim from -01 13 3895-3900 with the Section 4.1 citation written as a token]¶
Section 1.2, all four sentences: This document is a companion of [MOQ-FEC-CORE], which defines the scheme-independent FEC mechanism for MoQ and the configuration data model that a catalog carries. [MOQ-FEC-BLOCK] specifies the Reed-Solomon and RaptorQ schemes, [MOQ-FEC-CATALOG] the catalog encoding of the configuration, and [MOQ-FEC-MULTICAST] the delivery of protected tracks over multicast, including the receiver repair policy. This document defines the "iso-ssbg1-v1" block profile, the part of that set that is specific to MMTP packets, and the ATSC 3.0 and ARIB STD-B60 interworking that relies on it. [MOQ-MULTICAST] defines multicast endpoint discovery and the multicast packet format on which the multicast delivery of mmtp tracks rests.¶
The FEC terms of this document -- Source Track, Repair Track, Repair Layer, FEC Instance, Source Symbol, Repair Symbol, Source Block, SS_ID, SS_Start, SSB_length, RS_ID, RSB_length, Encoding Symbol ID (ESI) and Block Profile -- are those of [MOQ-FEC-CORE] Section 2, and the RS_ID offset O_i and per-block repair symbol count P_i of layer i are those of [MOQ-FEC-CORE] Section 7.3.1.¶
Under "iso-ssbg1-v1" the following take a form of their own, and the remaining terms are specific to this document:¶
Base Instance: Of a source track that a keyframe overlay protects, the instance that is not the overlay; of any other source track, its only instance. (New definition; -01 used the term without defining it. [MOQ-FEC-MULTICAST] Section 2 carries the same text.)¶
[MOQ-FEC-CORE] Section 5.6 requires every source object to carry the FEC Source Info Object Property (type 0x11); on an mmtp source object its value is the four-byte form of that section, which names instance 0 and carries the network-byte-order SS_ID of the base instance, the value of the packet's Source FEC Payload ID (Section 7.2); the property is object metadata and is not inserted into this payload.¶
A subscriber finds the repair tracks of the rendition it selects through the repairTrack of each of that rendition's FEC instances and the source fields of its repair tracks ([MOQ-FEC-CATALOG] Section 3.2; the data model of [MOQ-FEC-CORE] Section 10.2).¶
This section defines that profile, "iso-ssbg1-v1": its source symbols, the Source FEC Payload ID trailer, the MMTP form of the repair packet, the keyframe overlay and the interleave window.¶
It states where the profile deviates from [MOQ-FEC-CORE], whose remaining rules apply unchanged ([MOQ-FEC-CORE] Section 5.2).¶
The rules of [MOQ-FEC-CORE] apply to a track under this profile with the following deviations, which [MOQ-FEC-CORE] Section 5.2 asks a companion profile to state:¶
The unit of protection is the MMTP packet, not the MoQ object: a source symbol is formed from one FEC source packet as this section specifies, and the Protected Object Unit of [MOQ-FEC-CORE] Section 5.2 is not used.¶
A source object of an mmtp track carries one source symbol, the FEC source packet that is its payload, so SS_IDs advance by one per source object.¶
A source packet carries its SS_ID in a 4-byte trailer, the Source FEC Payload ID (Section 7.2), which is the canonical carrier.¶
The FEC Source Info Object Property of the object that carries the packet restates it in the 4-byte form, which names instance 0 ([MOQ-FEC-CORE] Section 5.6); the Property is object metadata, not part of the protected packet, and is not recovered with it.¶
When an object carries a source symbol but no FEC Source Info Property, a receiver that parses the packet MAY take the SS_ID from the trailer and feed the symbol to the base decoder (Section 7.2), a deviation from [MOQ-FEC-CORE] Section 5.6, under which such an object is not a source object; a receiver that does not parse the packet has no SS_ID for it and feeds it to no decoder. [new; restates the -01 5.3 permission (1103-1104) as a deviation in the Section 7.1 list]¶
A repair object is one complete MMTP repair packet of 12 + 13 + T bytes (Section 7.3), not the 13 + T byte payload of [MOQ-FEC-CORE] Section 7.5.¶
The keyframe overlay (Section 7.4) is this profile's form of an overlapping instance ([MOQ-FEC-CORE] Section 8.4), with two differences from that section and from [MOQ-FEC-CORE] Section 5.1: a RAP packet's overlay SS_ID travels in an MMTP header extension, not in a second entry of the FEC Source Info Property, and the Property keeps the 4-byte form, so a receiver learns a packet's overlay membership and overlay SS_ID from that extension, which lies inside the protected packet (Section 7.4.2), rather than from the Property alone; and because a recovered packet carries its overlay SS_ID but not its base SS_ID, a base recovery places the packet in the overlay instance, whereas an overlay recovery counts toward the base instance only where the receiver has determined the packet's base SS_ID independently (Section 7.1). [the clauses from "with two differences" on are new, restating Sections 7.1, 7.2 and 7.4.2 against the core's instance-membership rule]¶
interleaveDepthMs admits the value zero, with the ISO packet-count meaning of Section 7.5, under the permission of [MOQ-FEC-CORE] Section 10.1 for a block profile to give zero a meaning of its own.¶
The symbol carriage of [MOQ-FEC-CORE] Section 6.2, and with it the symbol track and the FEC Unit Start Property, does not apply to an mmtp track: its source symbols reach a receiver inside the track's own objects or as multicast datagrams ([MOQ-MULTICAST] Section 5), each carrying one whole FEC source packet.¶
This profile uses ssbg_mode1, with the one-stage FEC coding structure and FEC payload ID mode 0 ([ISO.23008-1] Clause C.1). [reworded from -01 7.3: "The iso-ssbg1-v1 block profile (Section 5.1) uses ssbg_mode1 ..."; the sentence "This subsection specifies the source symbols of iso-ssbg1-v1; Section 7.6 specifies those of moq-object-v1" is dropped]¶
(Under "raptorq", whose T is a multiple of 8, the least T is therefore 16, [MOQ-FEC-BLOCK] Section 6.1.) [carried from -01 5.1 688-691, reduced to the iso-ssbg1-v1 consequence]¶
OTI and sub-blocks: MoQ carries no RaptorQ Object Transmission Information (OTI, [RFC6330] Section 3.3) in-session: receivers derive it per block from the instance's configuration and the block's SSB_length per [MOQ-FEC-BLOCK] Section 6.1. [-01 mmt 7.2, "from the catalog fec fields" reworded]¶
The trailer lies outside the declared MPU payload (Section 5.2) and outside the protected packet (Section 7.1).¶
Section 14.1 states why the ATSC 3.0 signing mechanism does not authenticate the trailer, and what a receiver that requires authenticated block assignment does instead.¶
On an mmtp source object the Property's value is the 4-byte form of [MOQ-FEC-CORE] Section 5.6, which names instance 0: the sequence number of the source symbol the object carries, the same value and the same definition that the Source FEC Payload ID trailer of the FEC source packet carries. [-01 5.3 1034-1039 reworded for the 4-byte form]¶
Both carry the same value and a publisher MUST make them identical, on every source object, because the object payload is the whole FEC source packet, trailer included. [-01 5.3 1046-1049, two sentences merged]¶
No second Property entry conveys the overlay SS_ID: ... [-01 5.3 1072 "No second Property conveys" reworded for the list form of the core]¶
A receiver that parses the packet MAY take the SS_ID from the trailer instead, since the packaging delivers it; this is a deviation from that section, which has no trailer to fall back on. [-01 5.3 1103-1104 plus the deviation clause]¶
The Repair FEC Payload ID carries the fields of the repair object of [MOQ-FEC-CORE] Section 7.5 with the semantics that section gives them -- SS_Start, RSB_length, RS_ID and SSB_length -- and this section defines only the MMTP packet header in front of them:¶
A repair object under this profile is therefore 12 + 13 + T bytes long, not the 13 + T bytes that [MOQ-FEC-CORE] Section 7.5 requires of a repair object under "moq-object-v1"; a receiver MUST discard a repair object of an mmtp track whose payload is not exactly 12 + 13 + T bytes.¶
It reads the Repair FEC Payload ID at byte offset 12, and every other rule of that section applies to the fields: a receiver routes the packet to the decoder of the instance that the configuration relates to the repair track it arrived on, and discards a packet whose RS_ID is not less than RSB_length or that fails a bound of the instance's scheme ([MOQ-FEC-BLOCK] Section 7).¶
It is this profile's form of an overlapping instance ([MOQ-FEC-CORE] Section 8.4).¶
a track under "moq-object-v1" has no overlay, and a catalog that declares one for such a track fails the rule of Section 7.4.1. [-01 7.6.6 2629-2634 condensed]¶
An overlay is a separate FEC instance over the same source track. It has its own sourceSymbols (K_overlay) and its own repairSymbols, declared on its own instance entry with its own repair track ([MOQ-FEC-CATALOG] Section 3.1), and it shares the source track's scheme, blockProfile and symbolSize. [-01 6.4.1: "declared on its own repair track" and "algorithm" renamed]¶
Its priority is independent within 192 to 255 ([MOQ-FEC-CORE] Section 10.3, rule 10, and Section 11), and MAY be numerically lower than a base enhancement layer's so that a keyframe repair symbol is delivered before an ordinary one. [-01 6.4.1: the range is now the core's validation rule 10 and is cited]¶
Overlay ESIs never collide with base ESIs because the two are different instances decoded by different decoders; a receiver does not feed a repair symbol of one instance to the decoder of the other, nor reuse a source symbol's coordinate in one instance as its coordinate in the other ([MOQ-FEC-CORE] Section 5.1). [-01 6.4.1: the two MUST NOTs are the rule of [MOQ-FEC-CORE] Section 5.1, now cited rather than restated; one normative copy]¶
A source track has at most one overlay, a wire constraint explained in Section 7.4.3, and none when its blockProfile is "moq-object-v1"; the overlay instance's repair track carries no repairLayer. [-01 6.3.4 rule 10 and 6.4.1 merged]¶
A receiver MUST validate this before starting FEC for an overlay instance of the source track, and on any violation MUST fail closed as [MOQ-FEC-CORE] Section 10.3 requires for its own rules: it reports the offending catalog path, does not start FEC for any overlay instance of that source track, and MUST NOT repair the catalog by ignoring a track. [-01 6.3.4 preamble 1349-1354 restated for rule 10; fail closed is per instance as [MOQ-FEC-CORE] Section 10.3 scopes it, not per source track as -01 had]¶
An overlay with K_overlay 256 and P_overlay 32 has an RS_ID range of 0 to 31, an ESI range in a full overlay block of 256 to 287, and packets carrying RSB_length 32; it is valid only under "raptorq", since 256 + 32 = 288 exceeds 255. [-01 5.1.2 930-938 and 6.4.1 1813-1815 merged, no longer tied to the -01 worked catalog]¶
It knows that a repair packet belongs to the overlay because it arrives on the repair track that the catalog declares for the overlay instance of this source track, and that track has its own packet_id. [-01 6.4.2: "declares with scope: keyframe" reworded]¶
The instance is therefore fixed by the track the packet arrived on; no field of the repair packet distinguishes the two, and a receiver does not infer the instance from SS_Start, RSB_length, or SSB_length, whose values are only interpretable once the instance is known ([MOQ-FEC-CORE] Section 7.5). [-01 6.4.2: the MUST NOT is the rule of [MOQ-FEC-CORE] Section 7.5, applied to the MMTP repair packet by Section 7.3; now cited rather than restated]¶
The overlay is declared only by the catalog: Section 8.3 forbids describing the overlay's repair flow in an AL-FEC message, which identifies an FEC source flow by packet_id ([ISO.23008-1] Clause C.6.3) and cannot describe the RAP subset that the overlay protects, so a receiver that consumes the flow without the catalog uses the base instance alone. [-01 8.5.1 3041-3046: the MUST NOT is collapsed into the one copy in Section 8.3]¶
The receiver-side half is the repair policy of [MOQ-FEC-MULTICAST] Section 6, under which a receiver gives a keyframe-bearing base block that in-band decoding leaves unrecovered priority unicast repair; its Section 6.6 states how the two combine. [-01 6.4.3: the Section 11.4 and 11.4.6 pointers re-pointed]¶
A receiver that does not implement overlays MUST fail closed on a catalog that declares one for a track it intends to consume, that is on an FEC instance carrying the unknown scope key, and MUST NOT ignore the key and consume the source track anyway. [-01 6.4.3: "on a fec-repair track carrying" became "on an FEC instance carrying"]¶
scope (string, OPTIONAL): When present, MUST be "keyframe", and declares this instance to be the source track's keyframe overlay. It is an unknown-key exception (Section 7.4.4). An overlay is defined only for a source track under "iso-ssbg1-v1" (Section 7.4). [-01 5.1.1 816-820: "declares this track" became "declares this instance", the member moved from the repair-track entry to the overlay's instance object, and "one of the unknown-key exceptions below" became an exception this document defines under [MOQ-FEC-CATALOG] Section 3.4; the -01 overlay sourceSymbols member (821-824) is now the instance's ordinary sourceSymbols, Section 7.4.1]¶
The overlay's instance object carries one member that this document defines, under the rule of [MOQ-FEC-CATALOG] Section 3.4 that lets a block profile's document define further keys for the tracks it admits, and further exceptions among them. [new bridging sentence, Section 7.4.1]¶
This is an unknown-key exception that this document defines (Section 7.4.1) under [MOQ-FEC-CATALOG] Section 3.4, which lets a block profile's document define further exceptions. [-01 6.4.3: "one of the unknown-key exceptions of Section 5.1.1" re-pointed, since the catalog binding no longer lists scope]¶
[MOQ-FEC-CORE] Section 10.1 admits a positive interleaveDepthMs and lets a block profile give the value zero a meaning of its own. Under this profile the field MAY also be zero.¶
A publisher uses a zero value only with that ISO protection_window_size alternative. [-01 8.4 item 4, 2923-2925 reworded]¶
An absent field supplies no catalog-signaled span; neither an absent nor a zero value supplies an elapsed-time deadline. [-01 5.1 723 and mmt-01 7.1 1497-1498 merged]¶
FEC parameters for mmtp-packaged tracks are signaled in the catalog fec field, whose encoding [MOQ-FEC-CATALOG] defines (its Section¶
the message MUST NOT signal FEC parameters that contradict the track's catalog fec instances, and where the two disagree the catalog governs, as [MOQ-FEC-CORE] Section 10 requires of any in-band signaling.¶
When used with MMT packaging, the instance parameters of [MOQ-FEC-CORE] Section 10.1, as the catalog encodes them ([MOQ-FEC-CATALOG] Section 3.1), map as follows:¶
scheme: the scheme identifier ([MOQ-FEC-BLOCK] Section¶
blockProfile: ... REQUIRED on every FEC instance, with no default [replaces "REQUIRED when algorithm is not none"]¶
See [MOQ-FEC-CATALOG] for the complete catalog field definitions, [MOQ-FEC-CORE] Section 15.2 for the scheme registry, and [MOQ-FEC-BLOCK] for the scheme bounds and the OTI derivation (its Sections 7 and 6.1).¶
For FEC block alignment and timing to be deterministic, an mmtp publisher MUST, in addition to the encoder constraints of [MOQ-FEC-CORE] Section 5.3: [-01 8.4 preamble, scoped to mmtp]¶
A zero value is usable only with the ISO protection_window_size alternative of Section 7.5. [-01 8.4 item 4]¶
A source track's repair tracks are the track that each of its FEC instances names in repairTrack and the tracks that name it in source ([MOQ-FEC-CATALOG] Section 3.2; [MOQ-FEC-CORE] Section 10.2); [MOQ-FEC-CATALOG] Section 6 gives their conventional names.¶
A receiver that holds layer 0 under the AL-FEC message alone, without the catalog (Section 10), applies the fail-closed check of [MOQ-FEC-CORE] Section 7.3.2 with maximum_p_for_repair_flow standing for its O_0 + P_0, by that check's anchored clause alone: layer 0 has no clause of its own track, its O_0 being 0 and its symbols' coordinates depending on no published count, so a packet of the flow generated under a previous total, which the message can precede (above), fails it closed on nothing. [-01 6.3.9 1631-1642 and 1713-1715 and -01 12 3844-3851 merged into one copy]¶
Layered repair is an extension of [ISO.23008-1] that an ISO AL-FEC message cannot express: each FEC source flow has one FEC encoded flow, whose one-stage coding structure carries a single repair_flow_id ([ISO.23008-1] Clause C.6.3), so only layer 0 maps to an ISO repair flow. [-01 6.3 1206-1211, "Under iso-ssbg1-v1" dropped as implied]¶
The codes are those of [MOQ-FEC-BLOCK], which specifies two, "reed-solomon" (code point 1, mandatory to implement) and "raptorq" (code point 3, OPTIONAL) (its Sections 3 and 10); ... outside this document and its companions until the "MoQ FEC Schemes" registry of [MOQ-FEC-CORE] Section 15.2 lists that code. [-01 12 3836-3843: "this document's to specify" re-pointed to the block-codes draft]¶
... and a receiver that consumes the session without the catalog takes them from the message alone, holding the one repair flow as layer 0 under the block total the message bounds, as Section 8.3 specifies. [-01 12 3847-3851: the check itself now lives in Section 8.3]¶
[MOQ-FEC-CATALOG] adds the per-track fec array of FEC instances and the fec-repair and fec-source packaging values (its Sections 3 and 3.3)¶
fec | OPTIONAL | Array | FEC instances, encoded per [MOQ-FEC-CATALOG] Section 3 (Section 8)¶
the AL-FEC message's entry for a flow -> the track's FEC instance (instance 0, encoded as [MOQ-FEC-CATALOG] Section 3.1 specifies): scheme from fec_code_id_for_repair_flow, "reed-solomon" for code point 1 and "raptorq" for 3 ([MOQ-FEC-BLOCK] Section 3), ...¶
a session whose message names any other code is outside this document and its companions until the "MoQ FEC Schemes" registry of [MOQ-FEC-CORE] Section 15.2 lists that code (Section 10)¶
The conversion of a flow fails where the FEC instance it would yield fails a rule of [MOQ-FEC-CORE] Sections 10.1 and 10.3, or of [MOQ-FEC-CATALOG] Section 5, on which a receiver fails closed: ... fails the RaptorQ bound of [MOQ-FEC-BLOCK] Section 7 ... ([MOQ-FEC-CORE] Section 10.3 bounds K + totalP - 1 by 2^24 - 1 ...) ... (the Reed-Solomon bound of [MOQ-FEC-BLOCK] Section 7, which bounds K plus the repair count of every layer). [rules 5, 6 and 12 of -01 6.3.4 re-pointed]¶
two source tracks naming one repair track, or two base instances for one track, are shapes no FEC instance of [MOQ-FEC-CORE] has, a repair packet belonging to the instance that the configuration relates to the track it arrived on (its Sections 5.1 and 7.5).¶
The AL-FEC message does not name blockProfile, which is required on every FEC instance ([MOQ-FEC-CORE] Section 10.1); the converter sets it to "iso-ssbg1-v1", the profile whose entry carries the coding byte 0x14 that the converter accepts (Section 8.3; Section 7.1).¶
The same holds for the overlay SS_ID (Section 7.4.3): A3SA signing authenticates neither, and a receiver that requires authenticated block assignment treats each as this section requires. [-01 8.5.1 3055-3057 moved here]¶
An attacker who can modify the Source FEC Payload ID without detection could redirect source symbols to incorrect FEC blocks, causing recovery failures or corrupted output. [-01 14.5: "it" made explicit after the first sentence of 14.5 was merged into the paragraph above]¶
This document requests the registration of the block profile "iso-ssbg1-v1" (Section 7.1) in the "MoQ FEC Block Profiles" registry that [MOQ-FEC-CORE] Section 15.3 creates:¶
Appendix B, lead sentence, parameter block and the sentence after the block: This appendix is informative. It carries the MMTP capacity arithmetic of the -01 FEC draft's catalog example, for two mmtp tracks with the FEC parameters shown. The block lists each track's scheme, K, P, T, groupDurationMs, interleaveDepthMs and bitrate in place of the -01 catalog JSON (-01 Appendix B 4415-4497, now in [MOQ-FEC-CATALOG]). The video instance is valid only under "raptorq", since 625 + 125 = 750 exceeds the "reed-solomon" bound of 255 ([MOQ-FEC-BLOCK] Section 7); the audio instance, with 32 + 8 = 40, is within that bound under either scheme.¶
Appendix B, two sentences carried from -01 Appendix B with their
keyword lowercased because the appendix is informative: "Every
protected MMTP packet still must fit within T - 2
(Section 7.1)" (-01: "still MUST fit within T - 2"; the rule
is Section 7.1) and "The publisher must choose the actual
source-packet count and block boundaries ..." (-01: "The publisher
MUST choose ...").¶
Appendix B, parameter block: the video and audio lines restate the K, P, T, groupDurationMs, interleaveDepthMs and bitrate values of the -01 Appendix B catalog, which this document does not carry.¶
altGroup, so it is not a
member of a switching set. In -00 it carried the altGroup of the
track it protects, although its Groups are FEC blocks, not
media-time-aligned Groups.¶
initDataList and initRef
fields of [I-D.ietf-moq-msf] in place of the per-track initData
field, which the current revisions of [I-D.ietf-moq-msf] and
[I-D.ietf-moq-cmsf] do not define. A receiver may still accept
initData from -00 implementations (Sections 4.5, 6.1, 12, and 12.1).¶
timescale field of cmaf tracks
is attributed to [I-D.ietf-moq-msf] Section 5.2.21, not to
[I-D.ietf-moq-cmsf]. A loc switching-set member, like a cmaf one,
carries a catalog timescale, which a TIMESCALE Property on the track
must equal, so [I-D.ietf-moq-loc] is now a normative reference. A
track entry without packaging is malformed.¶
fec_code_id_for_repair_flow the [ISO.23008-10] Table 1 code point
of the track's algorithm: 3 for "raptorq" and 1 for "reed-solomon".
Sections 7 and 8.1 follow [I-D.ramadan-moq-fec], which now fully
specifies "reed-solomon" and makes it the mandatory-to-implement
scheme. RaptorQ is optional there. Section 7 states that neither the
MMT AL-FEC framework nor [ATSC-A331], which uses it for MMTP
sessions, requires a particular code, so the codes are those
[I-D.ramadan-moq-fec] specifies. [ISO.23008-10] is now a normative
reference. The message's length_of_repair_symbol carries the
track's symbolSize, which Section 12.2 reads back, and
num_of_repair_symbol_per_packet is 1; its
maximum_p_for_repair_flow carries the RSB_length of the track's FEC
instance, the block total, which Section 12.2 reads back as
repairSymbols less the layers above 0 it lists, listing a layer it
learns from provisioning only where provisioning places it, with every
other listed layer, above every layer it does not list, keeping the
value it first published for a track, and those of the layers it has
listed, across later versions of the message that raise the field,
retiring the source encoding under one that lowers the field below the
total its catalog derives for the highest layer it lists
([I-D.ramadan-moq-fec] Section 6.3.9), appending at the next index a
layer it learns from provisioning after it published the track only
where provisioning places it next above the highest layer it lists
with no unlearned layer below it, the field's rise alone placing none,
and listing one provisioning does not so place, or whose symbols the
field it first read already counted, only by retiring the source
encoding and publishing a separate rendition. Section 12.2 states
that the field is a bound in [ISO.23008-1] where "iso-ssbg1-v1"
carries P as every repair packet's RSB_length, lists the repair track
only once a repair packet of the flow carries at least the field, and
fails the conversion of a flow whose repair packets carry an
RSB_length below the field on a block at or after, in the serial order
of [I-D.ramadan-moq-fec] Section 7.5, one received carrying at least
the field, since a publisher that appends a layer sends the new
version only after the last repair packet generated under the previous
total (Section 8.3) while the path keeps no order between the message
and the repair flow (Section 5.2.2); once the track is published such
a packet, below the total its catalog derives for the highest layer it
lists, retires the source encoding, the separate rendition listing the
repair track again as the first listing did; a previous-total packet
of an earlier block fails a receiver of the catalog closed on nothing,
the own-track clause of [I-D.ramadan-moq-fec] Section 6.3.9 applying
to layers above 0 alone. Section 8.3 sets the entry's
number_of_assets to 1, the one Asset being the source track's.
Section 12.2 accepts only an entry with a number_of_assets of 1, the
coding byte 0x14 and code point 1 or 3, and fails the conversion of a
flow whose fec object would fail a catalog rule of
[I-D.ramadan-moq-fec] Sections 5.1 and 6.3.4 (a maximum K or maximum
P of 0, a symbol length below 14, a RaptorQ symbol length that is not
a multiple of 8, a RaptorQ maximum K above 56403 or maximum K plus
maximum P above 2^24, a Reed-Solomon maximum K plus maximum P above
255), of one whose repair flow is packet_id 0, the flow's own
packet_id or the packet_id of another track of the session, and of the
flows of two entries that name one repair flow or list one Asset.¶
fecOTI attribute and says the AL-FEC message carries no OTI,
and the Section 8.1 algorithm rule reads the code point from
fec_code_id_for_repair_flow alone, with no ROUTE-ingest case.¶
This appendix is informative. It carries the MMTP capacity arithmetic of the catalog example of draft-ramadan-moq-fec-01 ([I-D.ramadan-moq-fec]), Appendix B of that revision, as a historical example: its parameters are the ones shown here, and are not those of the worked catalog of [MOQ-FEC-CATALOG].¶
video: scheme raptorq, K 625, P 125, T 1000,
groupDurationMs 1000, interleaveDepthMs 1000,
bitrate 4990000 bit/s
audio: scheme raptorq, K 32, P 8, T 512,
groupDurationMs 1000, interleaveDepthMs 1000,
bitrate 128000 bit/s
¶
The video instance is valid only under "raptorq", since 625 + 125 = 750 exceeds the "reed-solomon" bound of 255 ([MOQ-FEC-BLOCK] Section 7); the audio instance, with 32 + 8 = 40, is within that bound under either scheme.¶
Parameter consistency: both tracks use one-second MMTP groups (one MPU per MoQ Group), and each has a 1000 ms source/repair send-span budget. The group duration does not determine the FEC partition. A worked capacity check must include the MMTP and MPU-mode headers, not only the media bitrate. With no MMTP header extension, a timed FT=2 packet has the following media capacity, depending on whether it starts a data unit (Section 5.2):¶
FI=0 or FI=1: T - 2 (length) - 12 (MMTP) - 8 (MPU) - 14 (DU) = T - 36 FI=2 or FI=3: T - 2 (length) - 12 (MMTP) - 8 (MPU) = T - 22¶
The source packet also carries a 4-byte Source FEC Payload ID on the
wire, outside the protected packet and therefore outside these symbol
budgets (Section 7.2). FT=0/FT=1 metadata consumes
source packets but does not carry sample bytes, and any MMTP extension
consumes additional space. Every protected MMTP packet still must fit
within T - 2 (Section 7.1). Using the larger continuation
capacity gives a media ceiling, not a packetization plan:¶
Video: 4,990,000 bit/s x 1.0 s / 8 = 623,750 bytes of media per
one-second group. K x (T - 22) = 625 x 978 = 611,250 bytes is an
upper bound on the media capacity of one full block. The media
therefore needs at least ceil(623,750 / 978) = 638 source packets
and at least two K=625 blocks; MMTP metadata, fragmentation, and
authored packet boundaries can require more. Repair overhead is P /
K = 125 / 625 = 20%.¶
Audio: 128,000 bit/s x 1.0 s / 8 = 16,000 bytes of media per group;
the upper-bound capacity of one full block is K x (T - 22) = 32 x 490
= 15,680 bytes. The media therefore needs at least
ceil(16,000 / 490) = 33 source packets and at least two K=32 blocks;
packetization and metadata can require more. Repair overhead is
P / K = 8 / 32 = 25%.¶
The publisher must choose the actual source-packet count and block
boundaries so that every packet, including FT=0 and FT=1 metadata, fits
the selected symbol size and every block's repair fields describe the
actual (SS_Start, SSB_length) partition. These calculations are
capacity examples, not a rule that one FEC block equals one Group.¶
Portions of this document were drafted with the assistance of AI tools. The author reviewed, edited and takes responsibility for all of its content.¶