Carriage of ID3 Timed Metadata in the Common Media Application Format (CMAF)

1. Introduction

HTTP Live Streaming (HLS) [HLS] supports the inclusion of timed metadata in ID3 format [ID3] in various container formats, as described in [TM-HLS].

A large ecosystem has built up around carrying timed ID3 metadata in HLS for applications such as ad delivery & audience measurement. There are many benefits to adopting CMAF for HLS media delivery, but without a specification for carrying ID3 as sparse timed metadata in CMAF, deployment by companies in this ecosystem is blocked.

This specification describes how such ID3 metadata can be carried as timed metadata in a CMAF-compatible fragmented MP4 (fMP4) stream [CMAF] as used by the HLS protocol.

CMAF-compatible fragmented MP4 can also be used in DASH. The elements defined in this specification may also be used with DASH.

2. Timed Metadata in a CMAF-compatible stream

2.1. Overview

Timed Metadata in a CMAF-compatible stream is signaled via one or more Event Message boxes (emsg) [CMAF] per segment.

Event messages with the scheme specified in this document will identify boxes that carry ID3v2 metadata [ID3].

2.2. ID3 Metadata in an Event Message Box

2.2.1. Introduction

One or more Event Message boxes (emsg) [CMAF] can be included per segment. Version 1 of the Event Message box [DASH] must be used.

2.2.2. Syntax

For convenience, the follow box definition is reproduced from [DASH], section 5.10.3.3.3.

aligned(8) class DASHEventMessageBox extends FullBox('emsg', version, flags = 0)
{
  if (version==0) {
    string              scheme_id_uri;
    string              value;
    unsigned int(32)    timescale;
    unsigned int(32)    presentation_time_delta;
    unsigned int(32)    event_duration;
    unsigned int(32)    id;
  } else if (version==1) {
    unsigned int(32)    timescale;
    unsigned int(64)    presentation_time;
    unsigned int(32)    event_duration;
    unsigned int(32)    id;
    string              scheme_id_uri;
    string              value;
  }
  unsigned int(8) message_data[];
}

2.2.3. Semantics

scheme_id_uri MUST be set to https://aomedia.org/emsg/ID3 to identify ID3v2 metadata [ID3].

value may either be an absolute or relative user-specified URI which defines the semantics of the id field. Any relative URI is considered to be relative to the scheme_id_uri.

message_data MUST contain complete ID3 version 2.4 data [ID3].

In general, ID3 don’t carry a duration and in those cases the event_duration field should be set to 0xFFFFFFFF. If in a particular case, the ID3 message carries a duration, it should be reflected in the event_duration field.

The presentation_time must be within the time interval of the fragment.

The id field is not restricted in this version of the specification.

2.3. Signaling

Files compliant to this specification should signal it using the brand aid3 as part of the list compatible brands in the file type box. Manifest formats using files compliant to this specification may signal these files using the following URN: urn:aomedia:cmaf:id3.