ARIB STD-T64-C.S0050-B v1.0. 3GPP2 File Formats for Multimedia Services

Transcription

1 ARIB STD-T-C.S000-B v.0 GPP File Formats for Multimedia Services Refer to "Industrial Property Rights (IPR)" in the preface of ARIB STD-T for Related Industrial Property Rights. Refer to "Notice" in the preface of ARIB STD-T for Copyrights

2 Original Specification This standard, ARIB STD-T-C.S000-B v.0, was prepared by GPP-WG of Association of Radio Industries and Businesses (ARIB) based upon the GPP specification, C.S000-B v.0. Modification to the original specification None. Notes None.

3 GPP C.S000-B Version.0 Date: May 00 GPP File Formats for Multimedia Services COPYRIGHT NOTICE GPP and its Organizational Partners claim copyright in this document and individual Organizational Partners may copyright and issue documents or standards publications in individual Organizational Partner's name based on this document. Requests for reproduction of this document should be directed to the GPP Secretariat at Requests to reproduce individual Organizational Partner's documents should be directed to that Organizational Partner. See for more information

4 No Text 0 0 0

5 PREFACE This document describes the file formats to be used in GPP Multimedia services. GPP File Formats for Multimedia Services

6 Contents Contents... List of Figures... List of Tables... Scope References...0 Abbreviations... Introduction... GPP File Format.g.... Conformance..... File identification.. Registration of codecs.. Interpretation of GPP file format.. Limitation of the ISO base media file format. Codec Registration..... Overview.. Sample Description Box. Video..... MPEG- Video.. H... H./AVC. Audio and Speech..... MPEG- AAC and HE AAC.. AMR.. EVRC.. EVRC-B.. EVRC-WB.. K (QCELP).. SMV.. VMR-WB. Timed Text Format.... Asset Information.... Encryption.... Video-Buffer... Presentation and Layout Support (SMIL).... Media Synchronization and Presentation Format..... Document Conformance.. User Agent Conformance GPP File Formats for Multimedia Services

7 GPP SMIL Language Profile definition.. Content Model 0 File Format for K Speech.QCP... Compact Multimedia Format.cmf.... Description of CMF Content.... Formal Syntax of CMF Content.... Tables..... TimeBase.. Pitch Bend.. Fine Pitch Bend. Acceptable Profiles for CMF file format..... Talking Picture Messaging.. Audio-only Profile.. Picture Ringers.. Animated Ringers. CMF Conformance Guidelines..... AAC Requirements.. Subchunk Requirements.. MIDI Requirements.. MIP Requirements.. Wave Packet Requirements.. "dls-bank-change" event.. ADPCM Requirements.. Cue and Jump Points.. Recycle Requirements. File Extension and MIME type for Media presentation... Annex A File formats: difference with GPP (Informative)... Annex A. Relations between ISO, GPP, and GPP file format... Annex A. Differences between GPP and GPP... Annex A. Usage of GPP branding... Annex A. Relationship of GPP and GPP Profiles...0 Annex B Guideline for File Format Usage (Informative)... Annex B. MSS (Multimedia Streaming Service)... Annex B. Server storage for RTP streaming... Annex B. Transmission format for pseudo-streaming... Annex B. MMS Annex B. File download and play back... Annex C SMIL Profile Differences Between GPP and GPP (Informative)... Annex C. Additional functionality... Annex C. Interoperability between GPP and GPP SMIL... GPP File Formats for Multimedia Services

8 0 Annex D GPP SMIL Authoring Guidelines (Informative)... Annex D. General... Annex D. BasicLinking... Annex D. BasicLayout... Annex D. EventTiming... Annex D. AccessKeyTiming... Annex D. MultiArcTiming... Annex D. BasicAnimation... Annex D. MediaParam... Annex D. MetaInformation... Annex E Additional Specification for the System Component Test Attribute (Normative)...0 Annex E. General...0 Annex E. Definition of Attribute Encoding...0 Annex E. Behavior of a GPP SMIL Player...0 Annex F Description of CMF to SMIL Conversion (Informative)... Annex F. Conversion Mechanics... GPP File Formats for Multimedia Services

9 0 List of Figures Figure -: ISO File Format Box Structure Hierarchy... Figure -: EVRC Frame byte alignment... Figure -: EVRC-B Frame byte alignment... Figure -: EVRC-WB Frame byte alignment... Figure -: K (QCELP) Frame byte alignment... Figure -: SMV frame byte alignment... Figure -: VMR-WB Frame byte alignment... Figure A : File formats in ISO... Figure A : GPP file format... Figure A : GPP file format... Figure B : Hinted Presentation for Streaming (Reprint from ISO/IEC -)... Figure B : Basic sequence of pseudo-streaming... Figure B : Fragmented movie file format.... GPP File Formats for Multimedia Services

10 0 0 0 List of Tables Table -: The File-Type Box... Table -: SampleEntry fields... 0 Table -: EVRCSampleEntry fields... Table -: The EVRCSpecificBox fields for EVRCSampleEntry... Table -: EVRCDecSpecStruc... Table - : EVRCSampleEntry fields... Table - : he EVRCBSpecificBox fields for EVRCBSampleEntry... Table -: EVRCBDecSpecStruc... Table -: EVRCWBSampleEntry fields... Table -0: The EVRCWBSpecificBox fields for EVRCWBSampleEntry... Table -: EVRCWBDecSpecStruc... Table -: QCELPSampleEntry fields... Table -: The QCELPSpecificBox fields for QCELPSampleEntry... Table -: QCELPDecSpecStruc... Table -: Mapping table... Table -: SMVSampleEntry fields... Table -: The SMVSpecificBox fields for SMVSampleEntry... Table -: SMV DECSpecStruc... Table -: VMRSampleEntry fields... Table -0: The VMRSpecificBox fields for VMRSampleEntry... Table -: VMRDecSpecStruc... Table -: VMR mode_set bit field assignments... Table -: The GAD Information box... Table - Additional formats for encrypted media tracks... Table - GPP SMIL MIME types and attributes... Table -: Content model for the GPP SMIL profile... Table -: TimeBase Values... Table -: Pitch Bend Range values... Table -: Fine PITCH bend range values... Table -: Allowed formats for each media type... Table A--: Brand usage in G files: = defined support... Table A--: Relationship of GPP and GPP profiles... 0 Table C--: Name value pairs for MediaParam module that are additional to GPP. GPP File Formats for Multimedia Services

11 Scope The objective is to define and standardize a set of common file formats to be used in multimedia services (such as Multimedia Streaming Service (MSS) and Multimedia Messaging Service (MMS)) and to provide interoperability with existing G and the Internet multimedia services to the greatest extent possible. The specific media types and descriptions to be covered include: video, audio, images, graphics, high fidelity audio as well as presentation layout and synchronization. GPP File Formats for Multimedia Services

12 References. Void.. Void.. ISO/IEC -:00 Information technology Coding of audiovisual-objects Part : ISO base media file format.. ISO/IEC -:00, Information technology Coding of audiovisual-objects Part : MP file format.. GPP TS., "Transparent end-to-end packet switched streaming service (PSS); GPP file format (GP) (Release )".. ISO/IEC -:00: Information Technology Coding of Audio- Visual Objects Part : Visual.. ITU-T Recommendation H.: Video Coding for Low Bit rate Communication. GPP TS.0 "Adaptive Multi-Rate (AMR) Speech Codec; General Description. GPP TS.: AMR Speech Codec, wideband; General Description. 0. GPP C.S00-0-: Enhanced Variable Rate Codec, Speech Service Option for Wideband Spread Spectrum Digital Systems. IETF RFC : RTP Payload Format for PureVoice(tm) Audio.. IETF RFC : RTP Payload Format for Enhanced Variable Rate Codecs (EVRC) and Selectable Mode Vocoders (SMV), June 00.. IETF RFC : RTP payload format and file storage format for the Adaptive Multi-Rate (AMR) Adaptive Multi-Rate Wideband (AMR-WB) audio codecs, March 00.. GPP C.S000-0: High Rate Speech Service Option for Wideband Spread Spectrum Communications Systems. GPP C.S000-0 Version.0 Selectable Mode Vocoder Service Option for Wideband Spread Spectrum Communication Systems.. GPP TS.0: Adaptive Multi-Rate (AMR) speech codec frame structure. GPP TS.0: AMR Wideband Speech Codec; Frame Structure. WC Recommendation: "Synchronized Multimedia Integration Language (SMIL.0)", ITU-T Recommendation H. (annex X): "Annex X: Profiles and levels definition". 0. IETF RFC : Augmented BNF for Syntax Specification: ABNF.. IETF RFC : The QCP File Format and Media Types for Speech Data.. Unicode Standard Annex #: Unicode Newline guidelines, by Mark Davis. An integral part of The Unicode Standard, Version... "Standard MIDI Files.0", RP-00, "The Complete MIDI.0 Detailed GPP File Formats for Multimedia Services 0

13 0 Specification, Document Version." The MIDI Manufacturers Association, Los Angeles, CA, USA, February.. ITU-T Recommendation T.: "Information technology; Digital compression and coding of continuous-tone still images: Requirements and guidelines".. IETF RFC 0: "PNG (Portable Networks Graphics) Specification version.0".. Technical note TN0: Understanding the Differences between Apple and Windows IMA-ADPCM Compressed Sound Files, Developer Connection, Windows BMP us/gdi/bitmaps_jhv.asp. Windows Multimedia, Resource Interchange File Format/WAVE us/multimed/htm/_win_resource_interchange_file_format_services.as p, us/multimed/htm/_win_about_waveform_audio.asp Complete MIDI.0 Detailed Specification, v., (second edition), MIDI Manufacturers Association (MMA), General MIDI Specification, v., MIDI Manufacturer Association (MMA), September 00.. GPP C.S000-0 GPP File Formats for Multimedia Services.. IETF RFC Real-Time Transport Protocol (RTP) Payload Format for the Variable-Rate Multimode Wideband (VMR-WB) Audio Codec.. GPP C.S00-A v.0 Source-Controlled Variable-Rate Multimode Wideband Speech Codec (VMR-WB), Service Options and for Spread Spectrum Systems, April 00.. ISO/IEC -:00: Information technology Coding of audio - visual objects Part : Advanced Video Coding (AVC) file format.. GPP TS. The GPP File Format.. GPP TS. GPP Timed Text.. ISO/IEC -0:00: Information technology Coding of audiovisual objects Part 0: Advanced Video Coding.. GPP C.S00-0 Multimedia Streaming Service for Spread Spectrum Systems.. ISO/IEC -:00, "Information technology - Coding of audio-visual objects - Part : Audio". 0. ISO/IEC -:00/Amd :00, "Bandwidth Extension".. ISO/IEC -:00/Amd :00/Cor :00.. IETF RFC : "The Secure Real-time Transport Protocol".. GPP TS.0 Universal Geographical Area Description.. Scalable polyphony MIDI specification, Version.0, RP-0, The MIDI GPP File Formats for Multimedia Services

14 0 Manufacturers Association, Los Angeles, CA, USA, 00.. WC Recommendation: "Scalable Vector Graphics (SVG) Tiny. Specification", DLS/XMF Format for Mobile Devices, MIDI Manufacturers Association, Los Angeles, CA, Sept IETF RFC : "MIME Type Registrations for GPP Multimedia files".. WC Recommendation: REC-CSS-0: "Cascading Style Sheets", level CSS Specification -May-. CSS/. GPP TS. "GPP SMIL Language Profile". 0. WC Recommendation: "Cascading Style Sheets, Level, CSS Specification". GPP C.S00-B: Enhanced Variable rate Codec Speech Service option and for wideband spread spectrum Digital Systems.. GPP C.S00-C: Enhanced Variable rate Codec Speech Service option, and 0 for wideband spread spectrum Digital Systems.. IETF RFC : Enhancements to RTP Payload Formats for EVRC Family Codecs, January 00 GPP File Formats for Multimedia Services

15 Abbreviations For the purpose of this document, the following abbreviations apply: G Third Generation system GP File Format for GPP Multimedia Services GPP Third Generation Partnership Project GPP Third Generation Partnership Project AAC Advanced Audio Coding ABNF Augmented BNF ADPCM Adaptive Differential Pulse Code Modulation AMR Adaptive Multi-Rate AMR-WB Adaptive Multi-Rate Wideband AVC Advanced Video Coding BNF Backus-Naur Form BMP Bit Map Picture CMF Compact Multimedia Format CSS Cascading Style Sheets, level DLS Downloadable Sound(s) EVRC Enhanced Variable Rate Codec FFMS File Formats for Multimedia Services GAD Geographical Area Description HE AAC High Efficiency AAC HRD Hypothetical Reference Decoder HTML Hyper Text Markup Language HTTP Hypertext Transfer Protocol IETF Internet Engineering Task Force IMA International Multimedia Association IP Internet Protocol ISO International Standards Organization ITU-T International Telecommunication Union - Telecommunication Sector JPEG Joint Photographic Experts Group LED Light Emitting Diode MIDI Musical Instrument Digital Interface MIME Multipurpose Internet Mail Extensions MIP Maximum Instantaneous Polyphony MMA MIDI Manufacturers Association GPP File Formats for Multimedia Services

16 MMS MP MPEG MSS PDA PNG QCELP RFC RIFF RTCP RTP SBR SDP SMIL SRTP SMV TCP TOC URI VMR Multimedia Messaging Service MPEG- File Format Motion Picture Experts Group Multimedia Streaming Service Personal Digital Assistant Portable Network Graphics Qualcomm Code Excited Linear Prediction Request for Comments Resource Interchange File Format Real-Time Control Protocol Real-time Transport Protocol Spectral Band Replication Session Description Protocol Synchronized Multimedia Integration Language Secure Realtime Transport Protocol Selectable Mode Vocoder Transport Control Protocol Table of Contents Uniform Resource Identifier Variable-Rate Multimode [Wideband Vocoder] GPP File Formats for Multimedia Services

17 0 Introduction The purpose of this standard is to define a set of file formats to be used with GPP multimedia services. Among these file formats is a new format designated as the GPP file format or.g file format. It is the recommended format to use and can contain multiple media types (such as, video, audio, and timed text). Also included in this release are a presentation and layout description language and file format,.smi, and a compact multimedia file format,.cmf. These file formats can be used for, but not limited to: - multimedia content downloading to a terminal, - multimedia file generation and uploading from the originating terminal, - multimedia content exchange between MMS and/or MSS servers, - multimedia content storing to a server, and - multimedia message exchange with other industry system. This document does not specify how this file format is used in specific services. In other words, it should be expressly mandated, if necessary, in other service specifications whether to use this specification. GPP File Formats for Multimedia Services

18 0 GPP File Format.g The purpose of this section is to define the GPP file format for multimedia services. This file format is based on the ISO base media file format [ ]. Also, it adopts the methodology defined in [] to integrate necessary structures for inclusion of non-iso codecs such as H. [], AMR [], AMR-WB [], and extends this approach to include GPP specific codecs: EVRC [0], SMV [], VMR-WB [], and K Speech (QCELP) []. Currently the GPP file format also defines extensions for: AVC file format [..], Asset information [.], Encryption [.], Video buffer information [.].. Conformance The GPP file format, used in the specification for timed media (such as video, audio, and timed-text), is structurally based on the ISO base media file format defined in [ ]. However, the conformance statement for GPP files is defined in the present document by addressing file identification (file extension, brand identifier and MIME type definition) and registration of codecs. 0 NOTE: Future releases may expand the conformance statement for GPP files to include more codecs or functionalities by defining new boxes. Boxes of unknown type in the GPP file shall be ignored. 0.. File identification GPP multimedia files can be identified using several mechanisms. When stored in traditional computer file systems, these files should be given the file extension.g (readers should allow mixed case for the alphabetic characters). The following MIME types should be used: video/gpp (for visual or audio/visual content, where visual includes both video and timed text) and audio/gpp (for purely audio content). [ ] A file-type box in Table -, as defined in the ISO Base Media File Format [], shall be present in conforming files. The file type box ftyp shall occur before any variable-length box (e.g. movie, free space, media data). Only a fixed-size box such as a file signature, if required, may precede it. The brand identifier for this specification is 'gc'. This brand identifier shall occur in the compatible brands list, and may also be the primary brand. Readers should check the compatible brands list for the identifiers they recognize, and not rely on the file having a particular primary brand, for maximum compatibility. Files may be compatible with more than one brand, and have a 'best use' other than this specification, yet still be compatible with this specification. GPP File Formats for Multimedia Services

19 0 0 0 Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() 'ftyp' Brand Unsigned int() The major or best use of this file MinorVersion Unsigned int() CompatibleBrands Unsigned int() A list of brands, to end of the Box Table -: The File-Type Box Brand: Identifies the best use of this file. The brand should match the file extension. For files with extension.g and conforming to release 0 of this specification, the brand shall be ga. For files with extension.g and conforming to release A of this specification, the brand shall be gb. For files with extension.g and conforming to release B of this specification, the brand shall be gc. MinorVersion: This identifies the minor version of the brand. Files with brand 'gx', where x is an alphabetic character, shall have a corresponding release X.Y.Z such that X = when x = 'a'; X = when x = 'b'; and so on. A conforming minor version value for releasex.y.z uses the byte aligned and right adjusted value of release X* + y*+z. CompatibleBrands: A list of brand identifiers (to the end of the Box). 'gc' shall be a member of this list. ga and gb shall also be members of this list if the file is in conformance with release 0 and/or release A of this specification. The brand compatibility list shall include major brands 'gp', 'gp' and/or 'gp' as described in [] when the file content meets the conditions described therein. Brands shall not be placed in the compatibility list if playback is not possible given the applicable methods. See Annex A. for additional information... Registration of codecs In GPP files, AVC video, MPEG- video, MPEG- AAC audio streams, and other ISO codec streams, as well as non-iso media streams such as AMR narrow-band speech, AMR WB speech, EVRC speech, H. video, K speech, SMV speech, VMR-WB speech, and timed text, can be included as described in this specification... Interpretation of GPP file format All index numbers used in the GPP file format start with the value one rather than zero, in particular first-chunk in Sample to chunk box, sample-number in Sync sample box and shadowed-sample-number, sync-sample-number in Shadow sync sample box... Limitation of the ISO base media file format The following limitation to the ISO base media file format [] shall apply to a GPP file: A GPP file shall be self-contained, i.e., there shall not be references to external media data from inside the GPP file. GPP File Formats for Multimedia Services

20 . Codec Registration.. Overview The purpose of this section is to give some background information about the Sample Description Box in the ISO base media file format [ ]. The following sections define the necessary structures for integration of video, audio, speech, and timed text in a GPP file. This specification provides details for support of codecs defined within GPP. Support for codecs not defined by GPP is provided using external references... Sample Description Box In an ISO file, Sample Description Box gives detailed information about the coding type 0 used, and any initialization information needed for that coding. The Sample Description Box can be found in the ISO file format Box Structure Hierarchy shown in Figure -. Movie Box Track Box Media Box Media Information Box Sample Table Box Sample Description Box Figure -: ISO File Format Box Structure Hierarchy The Sample Description Box can have one or more Sample Entry Boxes. Valid Sample Entry Boxes already defined for ISO [0] and MP [0] include MPAudioSampleEntry, MPVisualSampleEntry, and HintSampleEntry. GPP File Formats for Multimedia Services

21 0 In addition, the Sample Entry Box for H. video shall be HSampleEntry. The Sample Entry Box for AMR and AMR-WB speech shall be AMRSampleEntry. The Sample Entry Box for EVRC speech shall be EVRCSampleEntry. The Sample Entry Box for K (QCELP) speech shall be QCELPSampleEntry or MPAudioSampleEntry. (Note: for K speech a g file parser shall be able to read both storage methods.) The Sample Entry Box for SMV speech shall be SMVSampleEntry. The Sample Entry Box for VMR-WB speech shall be VMRSampleEntry. The Sample Entry Box for timed text shall be TextSampleEntry. The Sample Entry Box for AVC shall be AVCSampleEntry. The Sample Entry Box for EVRC-B speech shall be EVRCBSampleEntry. The Sample Entry Box for EVRC-WB speech shall be EVRCWBSampleEntry. GPP File Formats for Multimedia Services

22 0 The format of SampleEntry and its fields are explained as follows: SampleEntry ::= MPVisualSampleEntry MPAudioSampleEntry HSampleEntry AVCSampleEntry AMRSampleEntry EVRCSampleEntry EVRCBSampleEntry EVRCWBSampleEntry QCELPSampleEntry SMVSampleEntry TextSampleEntry VMRSampleEntry HintSampleEntry Field Type Details Value MPVisualSampleEntry Entry type for visual samples defined in section.. of the present document. MPAudioSampleEntry Entry type for audio samples defined in section.. of the present document. HSampleEntry Entry type for H. visual samples defined in section.. of the present document. AVCSampleEntry Entry type for AVC samples defined in section.. of the present document. AMRSampleEntry Entry type for AMR and AMR-WB speech samples defined in section.. of the present document. EVRCSampleEntry Entry type for EVRC speech samples defined in section.. of the present document. EVRCBSampleEntry Entry type for EVRC-B speech samples defined in section.. of the present document. EVRCWBSampleEntry Entry type for EVRC-WB speech samples defined in section.. of the present document. QCELPSampleEntry Entry type for k (QCELP) speech samples defined in section 0 of the present document. SMVSampleEntry Entry type for SMV speech samples defined in section.. of the present document. TextSampleEntry Entry type for timed text samples defined in section. of the present document. VMRSampleEntry Entry type for VMR-WB speech samples defined in section.. of the present document. HintSampleEntry Entry type for hint track samples defined in the ISO specification [0]. Table -: SampleEntry fields GPP File Formats for Multimedia Services 0

23 . Video This section describes Sample Entries for video... MPEG- Video If MPEG- Video [] is supported then it shall be supported using the MPVisualSampleEntry Box as described in [ ]. NOTE: Throughout this document MPEG- Visual is referred to as MPEG- video, which should be taken to mean encoding of natural (pixel based) video using MPEG- Visual methods... H. If H. Video [] is supported then it shall be supported using the HSampleEntry 0 Box as described in [ ]... H./AVC If MPEG- AVC Video [] is supported then it shall be supported using the AVCSampleEntry Box as described in [ ].. Audio and Speech This section describes Sample Entries for audio and speech. 0.. MPEG- AAC and HE AAC If MPEG- AAC Profile or MPEG- HE AAC Profile [][0][] is supported then it shall be supported using the MPAudioSampleEntry Box as described in [ ]. When HE AAC is stored in the GPP file format, implicit signaling of SBR [][0][] shall not be used... AMR If AMR [] or AMR-WB [] speech is supported then they shall be supported using the AMRSampleEntry Box as described in [ ]. 0.. EVRC EVRC speech data shall be stored inside of a media track in such a way that is described in Section of []. The magic number shall not be included. The codec data frames are stored in a consecutive order with a single TOC entry field as a prefix per each of data frame, where the TOC field is extended to one octet by setting the four most significant bits of the octet to zero, as illustrated in the following figure. <-- Octet --> <-- Octet --> <-- --> <-- Octet N --> FR Type One EVRC speech data frame GPP File Formats for Multimedia Services

24 Figure -: EVRC Frame byte alignment EVRCSampleEntry Box For EVRC, the Box type of the EVRCSampleEntry Box shall be 'sevc'. The EVRCSampleEntry Box is defined as follows: EVRCSampleEntry ::= BoxHeader Reserved_ Data-reference-index Reserved_ Reserved_ Reserved_ Reserved_ TimeScale Reserved_ EVRCSpecificBox Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() 'sevc Reserved_ Unsigned int() [] 0 Data-referenceindex Unsigned int() Index to a data reference that to use to retrieve the sample data. Data references are stored in data reference Boxs. Reserved_ unsigned int() [] 0 Reserved_ unsigned int() Reserved_ unsigned int() Reserved_ unsigned int() 0 TimeScale Unsigned int() Copied from media header Box of this media Reserved_ unsigned int() 0 EVRCSpecificBox Information specific to the decoder. Table -: EVRCSampleEntry fields If one compares the MPAudioSampleEntry Box to the EVRCSampleEntry Box the main difference is in the replacement of the ESDBox, which is specific to MPEG- systems, with a box suitable for EVRC speech. The EVRCSpecificBox field structure is described in section EVRCSpecificBox field for EVRCSampleEntry Box The EVRCSpecificBox fields for EVRC shall be as defined in Table -. The EVRCSpecificBox for the EVRCSampleEntry Box shall always be included if the GPP file contains EVRC media. GPP File Formats for Multimedia Services

25 0 0 0 Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() devc DecSpecificInfo EVRCDecSpecStruc Structure which holds the EVRC Specific information Table -: The EVRCSpecificBox fields for EVRCSampleEntry BoxHeader Size and Type: indicates the size and type of the EVRC decoder-specific Box. The type shall be devc. DecSpecificInfo: the structure where the EVRC stream specific information resides. The EVRCDecSpecStruc is defined as follows: Field Type Details Value vendor Unsigned int() decoder_version Unsigned int() frames_per_sample Unsigned int() Table -: EVRCDecSpecStruc The definitions of EVRCDecSpecStruc members are as follows: vendor: four character code of the manufacturer of the codec, e.g. 'VXYZ'. The vendor field gives information about the vendor whose codec is used to create the encoded data. It is an informative field which may be used by the decoding end. If a manufacturer already has a four character code it should be used in this field. Otherwise, a vendor may create a four character code which best expresses the vendor s name. This field may be ignored. decoder_version: version of the vendor s decoder which can decode the encoded stream in the best (i.e. optimal) way. This field is closely associated with the vendor field. It may be used advantageously by vendors, which have optimal encoder-decoder version pairs. The value shall be set to 0 if the decoder version has no importance for the vendor. This field may be ignored.frames_per_sample: defines the number of frames to be considered as 'one sample' inside the MP file. This number shall be greater than 0 and should be carefully chosen since the access unit is decided depending on the value defined by this field. For example, a value of means each frame is treated as one sample. A value of 0 means that 0 frames (of duration 0 msec each) are aggregated and treated as one sample. It must be noted that, in this case, one sample duration is 0 (msec/frame) x 0 (frame) = 00 msec. For the last sample of the stream, the number of frames can be smaller than frames_per_sample, if the number of remaining frames is smaller than frames_per_sample... EVRC-B EVRC-B speech data shall be stored inside of a media track in such a way that is described in Section of [ ]. The magic number shall not be included. The codec data frames are stored in a consecutive order with a single TOC entry field as a prefix per each of data frame, where the TOC field is extended to one octet by setting the four most significant bits of the octet to zero, as illustrated in the following figure. <-- Octet --> <-- Octet --> <-- --> <-- Octet N --> GPP File Formats for Multimedia Services

26 FR Type One EVRC-B speech data frame EVRCBSampleEntry Box Figure -: EVRC-B Frame byte alignment For EVRC-B, the Box type of the EVRCBSampleEntry Box shall be 'secb'. The EVRCBSampleEntry Box is defined as follows: EVRCBSampleEntry ::= BoxHeader Reserved_ Data-reference-index Reserved_ Reserved_ Reserved_ Reserved_ TimeScale Reserved_ EVRCBSpecificBox Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() 'secb Reserved_ Unsigned int() [] 0 Data-referenceindex Unsigned int() Index to a data reference that to use to retrieve the sample data. Data references are stored in data reference Boxs. Reserved_ unsigned int() [] 0 Reserved_ unsigned int() Reserved_ unsigned int() Reserved_ unsigned int() 0 TimeScale Unsigned int() Copied from media header Box of this media Reserved_ unsigned int() 0 EVRCBSpecificBox Information specific to the decoder. Table - : EVRCBSampleEntry fields If one compares the MPAudioSampleEntry Box to the EVRCBSampleEntry Box the main difference is in the replacement of the ESDBox, which is specific to MPEG- systems, with a box suitable for EVRC-B speech. The EVRCBSpecificBox field structure is described in section EVRCBSpecificBox field for EVRCBSampleEntry Box The EVRCBSpecificBox fields for EVRC-B shall be as defined in Table -. The EVRCBSpecificBox for the EVRCBSampleEntry Box shall always be included if the GPP file contains EVRC-B media. GPP File Formats for Multimedia Services

27 0 0 0 Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() decb DecSpecificInfo EVRCBDecSpecStruc Structure which holds the EVRC- B Specific information Table - : he EVRCBSpecificBox fields for EVRCBSampleEntry BoxHeader Size and Type: indicates the size and type of the EVRC-B decoder-specific Box. The type shall be decb. DecSpecificInfo: the structure where the EVRC-B stream specific information resides. The EVRCBDecSpecStruc is defined as follows: Field Type Details Value Vendor Unsigned int() decoder_version Unsigned int() frames_per_sample Unsigned int() Table -: EVRCBDecSpecStruc The definitions of EVRCBDecSpecStruc members are as follows: vendor: four character code of the manufacturer of the codec, e.g. 'VXYZ'. The vendor field gives information about the vendor whose codec is used to create the encoded data. It is an informative field which may be used by the decoding end. If a manufacturer already has a four character code it should be used in this field. Otherwise, a vendor may create a four character code which best expresses the vendor s name. This field may be ignored. decoder_version: version of the vendor s decoder which can decode the encoded stream in the best (i.e. optimal) way. This field is closely associated with the vendor field. It may be used advantageously by vendors, which have optimal encoder-decoder version pairs. The value shall be set to 0 if the decoder version has no importance for the vendor. This field may be ignored. frames_per_sample: defines the number of frames to be considered as 'one sample' inside the MP file. This number shall be greater than 0 and should be carefully chosen since the access unit is decided depending on the value defined by this field. For example, a value of means each frame is treated as one sample. A value of 0 means that 0 frames (of duration 0 msec each) are aggregated and treated as one sample. It must be noted that, in this case, one sample duration is 0 (msec/frame) x 0 (frame) = 00 msec. For the last sample of the stream, the number of frames can be smaller than frames_per_sample, if the number of remaining frames is smaller than frames_per_sample... EVRC-WB EVRC-WB speech data shall be stored inside of a media track as described in this section. The EVRC-WB encoded speech data frames are stored in a consecutive order with a single TOC entry field as a prefix per each of data frame, where the TOC field is extended to one octet by setting the four most significant bits of the octet to zero, as illustrated in the following figure. GPP File Formats for Multimedia Services

28 0 0 <-- Octet --> <-- Octet --> <-- --> <-- Octet N--> FR Type One EVRC-WB speech data frame EVRCWBSampleEntry Box Figure -: EVRC-WB Frame byte alignment For EVRC-WB, the Box type of the EVRCWBSampleEntry Box shall be 'secw'. The EVRCWBSampleEntry Box is defined as follows: EVRCWBSampleEntry ::= BoxHeader Reserved_ Data-reference-index Reserved_ Reserved_ Reserved_ Reserved_ TimeScale Reserved_ EVRCWBSpecificBox Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() 'secw Reserved_ Unsigned int() 0 [] Data-reference-index Unsigned int() Index to a data reference that to use to retrieve the sample data. Data references are stored in data reference Boxs. Reserved_ unsigned int() 0 [] Reserved_ unsigned int() Reserved_ unsigned int() Reserved_ unsigned int() 0 TimeScale Unsigned int() Copied from media header Box of this media Reserved_ unsigned int() 0 EVRCWBSpecificBox Information specific to the decoder. Table -: EVRCWBSampleEntry fields If one compares the MPAudioSampleEntry Box to the EVRCWBSampleEntry Box the main difference is in the replacement of the ESDBox, which is specific to MPEG- systems, with a box suitable for EVRC-WB speech. The EVRCWBSpecificBox field structure is described in section EVRCWBSpecificBox field for EVRCWBSampleEntry Box The EVRCWBSpecificBox fields for EVRC-WB shall be as defined in Table -0. The EVRCWBSpecificBox for the EVRCWBSampleEntry Box shall always be included if the GPP file contains EVRC-WB media. GPP File Formats for Multimedia Services

29 0 0 0 Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() decw DecSpecificInfo EVRCWBDecSpecStr uc Structure which holds the EVRC- WB Specific information Table -0: The EVRCWBSpecificBox fields for EVRCWBSampleEntry BoxHeader Size and Type: indicates the size and type of the EVRC-WB decoderspecific Box. The type shall be decw. DecSpecificInfo: the structure where the EVRC-WB stream specific information resides. The EVRCWBDecSpecStruc is defined as follows: Field Type Details Value vendor Unsigned int() decoder_version Unsigned int() frames_per_sample Unsigned int() Table -: EVRCWBDecSpecStruc The definitions of EVRCWBDecSpecStruc members are as follows: vendor: four character code of the manufacturer of the codec, e.g. 'VXYZ'. The vendor field gives information about the vendor whose codec is used to create the encoded data. It is an informative field which may be used by the decoding end. If a manufacturer already has a four character code it should be used in this field. Otherwise, a vendor may create a four character code which best expresses the vendor s name. This field may be ignored. decoder_version: version of the vendor s decoder which can decode the encoded stream in the best (i.e. optimal) way. This field is closely associated with the vendor field. It may be used advantageously by vendors, which have optimal encoder-decoder version pairs. The value shall be set to 0 if the decoder version has no importance for the vendor. This field may be ignored. frames_per_sample: defines the number of frames to be considered as 'one sample' inside the MP file. This number shall be greater than 0 and should be carefully chosen since the access unit is decided depending on the value defined by this field. For example, a value of means each frame is treated as one sample. A value of 0 means that 0 frames (of duration 0 msec each) are aggregated and treated as one sample. It must be noted that, in this case, one sample duration is 0 (msec/frame) x 0 (frame) = 00 msec. For the last sample of the stream, the number of frames can be smaller than frames_per_sample, if the number of remaining frames is smaller than frames_per_sample... K (QCELP) (Note: for K speech a g file parser shall be able to read both the QCELPSampleEntry and the MPAudioSampleEntry storage methods.) K speech data shall be stored inside of a media track in the same way for codec data frame format as described in Section. of []. Each codec data frame is zero-padded to become of multiple of octets and the frames are stored in a consecutive order, as GPP File Formats for Multimedia Services

30 0 0 illustrated in the following figure. Here z is the stuffing bit used to keep byte alignment; its value is 0. <-- Octet --> <-- Octet --> <-- --> <-- Octet N --> Rate One K speech data frame z z QCELPSampleEntry Box Figure -: K (QCELP) Frame byte alignment For K, the box type of the QCELPSampleEntry Box shall be 'sqcp'. The QCELPSampleEntry Box is defined as follows: QCELPSampleEntry::= BoxHeader Reserved_ Data-reference-index Reserved_ Reserved_ Reserved_ Reserved_ TimeScale Reserved_ QCELPSpecificBox Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() 'sqcp Reserved_ Unsigned int() [] 0 Data-reference-index Unsigned int() Index to a data reference that to use to retrieve the sample data. Data references are stored in data reference Boxs. Reserved_ Const unsigned int() [] 0 Reserved_ Const unsigned int() Reserved_ Const unsigned int() Reserved_ Const unsigned int() 0 TimeScale Unsigned int() Copied from media header Box of this media Reserved_ Const unsigned int() 0 QCELPSpecificBox Information specific to the decoder. Table -: QCELPSampleEntry fields If one compares the MPAudioSampleEntry Box to the QCELPSampleEntry Box the main difference is in the replacement of the ESDBox, which is specific to MPEG- systems, with a box suitable for k. The QCELPSpecificBox field structure is described in Section... GPP File Formats for Multimedia Services

31 QCELPSpecificBox field for QCELPSampleEntry Box The QCELPSpecificBox fields for K speech shall be as defined in Table -. The QCELPSpecificBox for the QCELPSampleEntry Box shall always be included if the GPP file contains K speech media. Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() dqcp DecSpecificInfo QCELPDecSpecStruc Structure which holds the K (QCELP) speech specific information Table -: The QCELPSpecificBox fields for QCELPSampleEntry BoxHeader Size and Type: indicate the size and type of the k decoder-specific Box. The type shall be dqcp. DecSpecificInfo: the structure where the K speech stream specific information resides. The QCELPDecSpecStruc is defined as follows: Field Type Details Value vendor Unsigned int() decoder_version Unsigned int() frames_per_sample Unsigned int() Table -: QCELPDecSpecStruc The definitions of QCELPDecSpecStruc members are as follows: vendor: four character code of the manufacturer of the codec, e.g. 'VXYZ'. The vendor field gives information about the vendor whose codec is used to create the encoded data. It is an informative field, which may be used by the decoding end. If a manufacturer already has a four character code, it is recommended that it uses the same code should be used in this field. Otherwise, a vendor may create a four character code which best expresses the vendor s name. Else, it is recommended that the manufacturer creates a four character code which best addresses the manufacturer s name. This field may be safely ignored. decoder_version: version of the vendor s decoder which can decode the encoded stream in the best (i.e. optimal) way. This field is closely associated with the vendor field. It may be used advantageously by the vendors, which have optimal encoderdecoder version pairs. The value shall be set to 0 if the decoder version has no importance for the vendor. This field may be safely ignored. frames_per_sample: defines the number of frames to be considered as 'one sample' inside the file. This number shall be greater than 0 and should be carefully chosen since the access unit is decided depending on the value defined by this field. A value of means each frame is treated as one sample. A value of 0 means that 0 frames (of duration 0 msec each) are aggregated and treated as one sample. It must be noted that, in this case, one sample duration is 0 (msec/frame) x 0 (frame) = 00 msec. For the last sample of the stream, the number of frames can be smaller than frames_per_sample, if the number of remaining frames is smaller than frames_per_sample. GPP File Formats for Multimedia Services

32 K (QCELP) Support in MPAudioSampleEntry Box (Note: for K speech a g file parser shall be able to read both the QCELPSampleEntry and the MPAudioSampleEntry storage methods.) For storage of K speech media, MPAudioSampleEntry also can be used. K speech data shall be stored inside of a media track in the same way as described in [] When storing a K speech bitstream in a GPP file, the handler-type field within the HandlerAtom shall be set to soun to indicate media of type AudioStream, and the SampleEntry Box type shall be 'mpa' and the same Box described in Section.. is used. For inclusion of K speech media in MPAudioSampleEntry, the stream type specific information is in the ESDBox structure. The K speech codec is to be signaled by new value from the User Public area of objecttypeindication within the DecoderConfigDescriptor structure. objecttypeindication = 0xE ; The QCELPDecoderSpecificInfo in ABNF [0] format is specified as QCELPDecoderSpecificInfo = QLCM fmt The above ABNF rule indicates that QCELPDecoderSpecificInfo is the same as the header for K vocoder in.qcp file format as described in [], but without RIFF, riffsize, or anything after fmt. In addition, if the size of packets is completely constant, i.e. fixed rate encoding, the following rules apply to the definition of fmt (see variable-rate referenced by codec-info). num-rates = 0 rate-map = 0 major =... Mapping of QCELPSampleEntry Box and K Support in MPAudioSampleEntry Box Variables in QCELPSampleEntry Box and DecoderSpecificInfo in MPAudioSampleEntry Box for K is translated as described in the following table. GPP File Formats for Multimedia Services 0

33 QCELPSampleEntry Vendor decoder_version framespersample (fps) fps = spb / (sps * 0.0) N.A. N.A. N.A. N.A... SMV MPAudioSampleEntry The first bytes of Name field The fifth byte of Name field N.A. AvgBitsPerSec (abps) Calculated based on duration field in Track Header Box and data size in Sample Size Box. bytesperblock (bpb) Calculated according to the equation: abps = bpb * (bits/byte) / (0.0 * fps) sampleperblock (spb) spb = sps * 0.0 * fps numofrates and bytesperpacket When fixed rate encoding is used, all fields should be set 0x00. When variable rate encoding is used, example in packet definition in []. Table -: Mapping table 0 0 SMV speech data shall be stored inside of a media track in such a way that is described in Section of [ ]. The magic number is not included. The codec data frames are stored in a consecutive order with a single TOC entry field as a prefix per each of data frame, where the TOC field is extended to one octet by setting the four most significant bits of the octet to zero, as illustrated in the following figure. <-- Octet --> <-- Octet --> <-- --> <-- Octet N --> FR Type One SMV speech data frame SMVSampleEntry Box Figure -: SMV frame byte alignment For SMV speech, the box type of the SMVSampleEntry Box shall be 'ssmv'. The SMVSampleEntry Box is defined as follows: SMVSampleEntry ::= BoxHeader Reserved_ Data-reference-index Reserved_ GPP File Formats for Multimedia Services

34 0 0 Reserved_ Reserved_ Reserved_ TimeScale Reserved_ SMVSpecificBox Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() 'ssmv Reserved_ Unsigned int() [] 0 Data-reference-index Unsigned int() Index to a data reference that to use to retrieve the sample data. Data references are stored in data reference Boxs. Reserved_ Const unsigned int() [] 0 Reserved_ Const unsigned int() Reserved_ Const unsigned int() Reserved_ Const unsigned int() 0 TimeScale Unsigned int() Copied from media header Box of this media Reserved_ Const unsigned int() 0 SMVSpecificBox Information specific to the decoder. Table -: SMVSampleEntry fields If one compares the AudioSampleEntry Box for the SMVSampleEntry Box the main difference is in the replacement of the ESDBox, which is specific to MPEG- systems, with a box suitable for SMV. The SMVSpecificBox field structure is described in section SMVSpecificBox field for SMVSampleEntry Box The SMVSpecificBox fields for SMV shall be as defined in Table -. The SMVSpecificBox for the SMVSampleEntry Box shall always be included if the MP file contains SMV media. Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() dsmv DecSpecificInfo SMVDecSpecStruc Structure which holds SMV Specific information Table -: The SMVSpecificBox fields for SMVSampleEntry BoxHeader Size and Type: indicate the size and type of the SMV decoder-specific Box. The type shall be dsmv. DecSpecificInfo: the structure where the SMV stream specific information resides. The SMVDecSpecStruc is defined as follows: Field Type Details Value Vendor Unsigned int() decoder_version Unsigned int() Frames_per_sample Unsigned int() Table -: SMV DECSpecStruc The definitions of SMVDecSpecStruc members are as follows: GPP File Formats for Multimedia Services

35 vendor: four character code of the manufacturer of the codec, e.g. 'VXYZ'. The vendor field gives information about the vendor whose codec is used to create the encoded data. It is an informative field, which may be used by the decoding end. If a manufacturer already has a four character code, it is recommended that it uses the same code should be used in this field. Otherwise, a vendor may create a four character code which best expresses the vendor s name. Else, it is recommended that the manufacturer creates a four character code which best addresses the manufacturer s name. This field may be safely ignored. decoder_version: version of the vendor s decoder which can decode the encoded stream in the best (i.e. optimal) way. This field is closely associated with the vendor field. It may be used advantageously by the vendors, which have optimal encoderdecoder version pairs. The value shall be set to 0 if the decoder version has no importance for the vendor. This field may be safely ignored. frames_per_sample: defines the number of frames to be considered as 'one sample' inside the MP file. This number shall be greater than 0 and should be carefully chosen since the access unit is decided depending on the value defined by this field. For example, a value of means each frame is treated as one sample. A value of 0 means that 0 frames (of duration 0 msec each) are aggregated and treated as one sample. It must be noted that, in this case, one sample duration is 0 (msec/frame) x 0 (frame) = 00 msec. For the last sample of the stream, the number of frames can be smaller than frames_per_sample, if the number of remaining frames is smaller than frames_per_sample... VMR-WB VMR-WB speech data are stored in the stream according to the VMR-WB storage file format (see Section. in [ ]). The codec data frames are stored in a consecutive order with a single TOC entry field as a prefix per each of data frame, as illustrated in the following figure. <-- Octet --> <-- Octet --> <-- --> <-- Octet N --> P FT Q P P One VMR-WB speech data frame Figure.-: VMR-WB Frame byte alignment The FT field (Frame Types) and the Q bit (Frame Quality Indicator) are defined in section.. of []. The P bits are padding bits and shall be set to VMRSampleEntry Box For VMR-WB speech the box type of the VMRSampleEntry Box shall be 'svmr'. The VMRSampleEntry Box is defined as follows: VMRSampleEntry ::= BoxHeader Reserved_ GPP File Formats for Multimedia Services

36 0 0 Data-reference-index Reserved_ Reserved_ Reserved_ Reserved_ TimeScale Reserved_ VMRSpecificBox: Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() ' svmr Reserved_ Unsigned int() [] 0 Data-referenceindex Unsigned int() Index to a data reference that to use to retrieve the sample data. Data references are stored in data reference Boxs. Reserved_ Const unsigned int() [] 0 Reserved_ Const unsigned int() Reserved_ Const unsigned int() Reserved_ Const unsigned int() 0 TimeScale Unsigned int() Copied from media header Box of this media Reserved_ Const unsigned int() 0 VMRSpecificBox Information specific to the decoder. Table -: VMRSampleEntry fields If one compares the MPAudioSampleEntry Box to the VMRSampleEntry Box the main difference is in the replacement of the ESDBox, which is specific to MPEG- systems, with a box suitable for VMR-WB. The VMRSpecificBox field structure is described in section VMRSpecificBox field for VMRSampleEntry Box The VMRSpecificBox fields for VMR-WB speech shall be as defined in Table -0. The VMRSpecificBox for the VMRSampleEntry Box shall always be included if the GPP file contains VMR-WB media. Field Type Details Value BoxHeader.Size Unsigned int() BoxHeader.Type Unsigned int() dvmr DecSpecificInfo VMRDecSpecStruc Structure which holds the VMR- WB Specific information Table -0: The VMRSpecificBox fields for VMRSampleEntry BoxHeader Size and Type: indicate the size and type of the VMR decoder-specific Box. The type shall be dvmr. DecSpecificInfo: the structure where the VMR-WB stream specific information resides. The VMRDecSpecStruc is defined as follows: GPP File Formats for Multimedia Services