Internet Engineering Task Force Audio-Video Transport Working Group INTERNET-DRAFT H. Schulzrinne draft-ietf-avt-profile-03.txt AT&T Bell Laboratories October 20, 1993 Expires: 12/31/93 Sample Profile and Encodings for the Use of RTP for Audio and Video Conferences with Minimal Control Status of this Memo This document is an Internet Draft. Internet Drafts are working documents of the Internet Engineering Task Force (IETF), its Areas, and its Working Groups. Note that other groups may also distribute working documents as Internet Drafts. Internet Drafts are draft documents valid for a maximum of six months. Internet Drafts may be updated, replaced, or obsoleted by other documents at any time. It is not appropriate to use Internet Drafts as reference material or to cite them other than as a ``working draft'' or ``work in progress.'' Please check the I-D abstract listing contained in each Internet Draft directory to learn the current status of this or any other Internet Draft. Distribution of this document is unlimited. Abstract This note describes a profile for the use of the real-time transport protocol (RTP) and the associated control protocol, RTCP, within audio and video multiparticipant conferences with minimal control. It provides interpretations of generic fields within the RTP specification suitable for audio and video conferences. In particular, this document defines a set of default mappings from format index to encodings. The document also describes how audio and video data may be carried within RTP. It defines a set of standard encodings and their names when used within RTP. However, the definitions are independent of the particular transport mechanism used. The descriptions provide pointers to reference implementations and the detailed standards. This document is meant as an aid for implementors of audio, video and other real-time multimedia applications. INTERNET-DRAFT draft-ietf-avt-profile-03.txt October 20, 1993 Contents 1 Introduction 2 2 Demultiplexing 3 3 Audio 3 3.1 Encoding-independent recommendations . . . . . . . . . . . . . . . 3 3.2 Recommended Audio Encodings. . . . . . . . . . . . . . . . . . . . 4 3.3 The RTCP FMT Option for Audio. . . . . . . . . . . . . . . . . . . 6 3.4 Port Assignment. . . . . . . . . . . . . . . . . . . . . . . . . . 7 4 Video 8 4.1 The RTCP FMT Option for Video. . . . . . . . . . . . . . . . . . . 9 4.2 Port Assignment. . . . . . . . . . . . . . . . . . . . . . . . . . 9 5 Miscellaneous 10 6 Address of Author 10 1 Introduction This profile defines aspects of RTP left unspecified in the RTP protocol definition (RFC TBD). This profile is intended for the use within audio and video conferences with minimal session control. In particular, no support for the negotiation of parameters or membership control is provided. Other profiles may make different choices for the items specified here. The profile specifies the use of RTP over unicast and multicast UDP as well as ST-II. For unicast UDP and ST-II, references to multicast addresses are to be ignored. The use of this profile is indicated by the use of a media-specific well-known port number. The profile may also be used with other port numbers. For example, the use of a particular session announcement tool could imply use of this profile. H. Schulzrinne Expires 12/31/93 [Page 2] internet-dRAFT draft-ietf-avt-profile-03.txt October 20, 1993 2 Demultiplexing For applications which choose to share a single network destination address and port for both audio and video, the default channel identifier for audio is 0 and for video is 1. In that case, the port number for audio is used. This combination should only be used when it is known that all receiving applications can properly demultiplex audio and video. 3 Audio 3.1 Encoding-independent recommendations The following recommendations are default operating parameters. Ap- plications should be prepared to handle other values. The ranges given are meant to give guidance to application writers, allowing a set of applications conforming to these guidelines to interoperate without additional negotiation. These guidelines are not intended to restrict operating parameters for applications that can negotiate a set of interoperable parameters, e.g., through a conference control protocol. For packetized audio, the default packetization interval should have a duration of 20 ms, unless otherwise noted in Table 1. The packetization interval determines the minimum end-to-end delay; longer packets introduce less header overhead but higher delay and make packet loss more noticeable. For non-interactive applications such as lectures or links with severe bandwidth constraints, a higher packetization delay may be appropriate. For frame-based encodings (marked as F in the table 1 below) such as LPC, CELP and GSM, the sender may choose to combine several frame intervals into a single message. The receiver can tell the number of frames contained in a message since the frame duration is defined as part of the encoding. If multiple channels are used, the left channel information always precedes the right-channel information. For more than two channels, the convention followed by the AIFF-C audio interchange format should be followed. (The AIFF-C specification is available by anonymous ftp at ftp.sgi.com in the file sgi/aiff-c.9.26.91.ps.) For two-channel stereo, the sequence is left, right; for three channels, left, right, center; for quadrophonic systems, front left, front right, rear left, rear right; for four-channel systems, left, center, right, and surround sound; for six-channel systems left, left center, center, right, right center and surround sound. The sampling frequency should be drawn from the set: 8, 11.025, 16, 22.05, 44.1 and 48 kHz. H. Schulzrinne Expires 12/31/93 [Page 3] INTERNET-DRAFT draft-ietf-avt-profile-03.txt October 20, 1993 3.2 Recommended Audio Encodings The table 1 shows the names, types (sample vs. frame oriented), per-channel bit rates and default sampling frequencies of recommended encodings. The list is partially drawn from the document "Recommended practices for enhancing digital audio compatibility in multimedia systems", published by the Interactive Multimedia Assocation, Version 3.00, Oct. 1992 (referenced as [IMA]). The names are for identification only; they correspond to the names used within the Real-Time Transport Protocol (RTP). Other applications may choose different namings. Note that the L16 encoding may be used with different sampling rates. The CCITT changed its name in 1993 to ITU-T; to limit confusion, both old and new name are used. name nom. sampling rate type frame description kHz kb/s S/F ms _________________________________________________________________________ L16 48 768 S 16-bit linear, 2's complement L16 44.1 705.6 S L16 22.05 352.8 S L16 11.025 176.4 S G722 16 64 S CCITT/ITU-T subband ADPCM PCMU 8 64 S CCITT/ITU-T mu-law PCM PCMA 8 64 S CCITT/ITU-T A-law PCM G721 8 32 S CCITT/ITU-T ADPCM IDVI 8 32 S Intel/DVI ADPCM [IMA] G723 8 24 S CCITT/ITU-T ADPCM GSM 8 13 F 20 RTE/LTP GSM 06.10 1016 8 4.8 F 30 CELP _________________________________________________________________________ Table 1: Audio encodings For multi-octet encodings, octets are transmitted in network byte order (i.e., most significant octet first). A detailed description of the encodings is given below. The names shown (L16, PCMU, etc.) are limited to four characters and suitable to be used for identification in protocols such as RTP (RFC TBD). L16: denotes uncompressed audio data, using 16-bit signed representation with 65535 equally divided steps between minimum and maximum signal level, ranging from -32768 to 32767. The value is represented in two's complement notation. PCMU: specified in CCITT/ITU-T recommendation G.711. Audio data is encoded as eight bits per sample, after companding. Code to convert between linear and mu-law companded data is available in the IMA document. PCMA: specified in CCITT/ITU-T recommendation G.711. Audio data is encoded H. Schulzrinne Expires 12/31/93 [Page 4] INTERNET-DRAFT draft-ietf-avt-profile-03.txt October 20, 1993 as eight bits per sample, after companding. Code to convert between linear and A-law companded data is available in the IMA document. G721 through G729: specified in the corresponding CCITT/ITU-T recommenda- tions. Reference implementations for G.721 and G.723 are available as part of the CCITT/ITU-T Software Tool Library (STL) from the ITU General Secretariat, Sales Service, Place du Nations, CH-1211 Geneve 20, Switzerland. The library is covered by a license and is available for anonymous ftp on gaia.cs.umass.edu, file pub/ccitt/ccitt_tools.tar.Z. GSM: (group speciale mobile) denotes the European GSM 06.10 provisional standard for full-rate speech transcoding, prI-ETS 300 036, which is based on RPE/LTP (residual pulse excitation/long term prediction) coding at a rate of 13 kb/s. A reference implementation was written by Carsten Borman and Jutta Degener (TU Berlin, Germany) and is available for anonymous ftp from tub.cs.tu-berlin.de, directory tub/tubmik. 1016: uses code-excited linear prediction (CELP) and is specified in Federal Standard FED-STD 1016, published by the Office of Technology and Standards, Washington, DC 20305-2010. The U. S. DoD's Federal-Standard-1016 based 4800 bps code excited linear prediction voice coder version 3.2 (CELP 3.2) Fortran and C simulation source codes are available for worldwide distribution at no charge (on DOS diskettes, but configured to compile on Sun SPARC stations) from: Bob Fenichel, National Communications System, Washington, D.C. 20305, phone +1-703-692-2124, fax +1-703-746-4960. Example input and processed speech files, a technical information bulletin, and the official standard "Federal Standard 1016, Telecom- munications: Analog to Digital Conversion of Radio Voice by 4,800 bit/second Code Excited Linear Prediction (CELP)" are included at no charge. According to Vincent Cate (Carnegie Mellon), the distribution is also available for anonymous ftp at furmint.nectar.cs.cmu.edu (128.2.209.111) in directory celp.audio.compression. The following articles describes the Federal-Standard-1016 4.8-kbps CELP coder: Campbell, Joseph P. Jr., Thomas E. Tremain and Vanoy C. Welch, "The Proposed Federal Standard 1016 4800 bps Voice Coder: CELP," Speech Technology Magazine, April/May 1990, p. 58-64. Campbell, Joseph P. Jr., Thomas E. Tremain and Vanoy C. Welch, "The Federal Standard 1016 4800 bps CELP Voice Coder," Digital Signal Processing, Academic Press, 1991, Vol. 1, No. 3, p. 145-155. Campbell, Joseph P. Jr., Thomas E. Tremain and Vanoy C. Welch, "The DoD 4.8 kbps Standard (Proposed Federal Standard 1016)," in Advances in Speech Coding, ed. Atal, Cuperman and Gersho, Kluwer Academic H. Schulzrinne Expires 12/31/93 [Page 5] INTERNET-DRAFT draft-ietf-avt-profile-03.txt October 20, 1993 Publishers, 1991, Chapter 12, p. 121-133. Campbell, Joseph P. Jr., Thomas E. Tremain and Vanoy C. Welch, "The Proposed Federal Standard 1016 4800 bps Voice Coder: CELP," Speech Technology Magazine, April/May 1990, p. 58-64. Copies of the FS-1016 document are available for $2.50 each from: GSA Rm 6654 7th & D St SW Washington, D.C. 20407 1-202-708-9205 DVI: is specified in the "Recommended Practices for Enhancing Digital Audio Compatibility in Multimedia Systems", published by the Interactive Multimedia Association (IMA), Annapolis, MD. The document also contains reference implementations for mu-law to 16-bit, ADPCM and sample rate conversions. For sample-based encodings, a receiver should accept packets representing between 0 and 200 ms of audio data.(1) Receivers should be prepared to accept multi-channel audio, but may choose to only play a single channel. All block-oriented audio codecs should be able to encode and decode several consecutive blocks within a single packet. Since the frame size for the block-oriented codecs is given, there is no need to use a separate designation for the same encoding, but with different number of blocks per packet. 3.3 The RTCP FMT Option for Audio Unless specified with the FMT option, the mapping between the format field in an RTP packet and audio encodings, sampling rates and channel counts is specified by Tables 2. Format values of 31 and below cannot be redefined by FMT options. In other words, only values of 32 and above are valid in the format field within an FMT option. The receiver is expected to discard RTP packets containing media data with unknown format field values. Sites are expected to keep the mapping between format and encoding constant, so that lost packets containing FMT options do not lead the receiver to misinterpret media data. Additional standard encodings may be registered with the Internet Assigned ------------------------------ 1. This restriction allows reasonable buffer sizing for the receiver. H. Schulzrinne Expires 12/31/93 [Page 6] INTERNET-DRAFT draft-ietf-avt-profile-03.txt October 20, 1993 Numbers Authority (IANA). The format name is intended to describe the format in an unambiguous way; it is interpreted as a sequence of four ASCII characters, with uppercase and lowercase characters treated as distinct. Format names beginning with the letter 'X' are reserved for experimental use and not subject to registration. These experimental encodings may be mapped to format values 32 and above using the FMT option. Additional standard mappings to format values of 31 and below may also be registered with IANA. Registered assignments are published periodically in the Assigned Numbers RFC. Within the FMT option, the format name is followed by a field containing a channel count and a sample rate field, measured in samples per second.(2) A channel count of zero is considered invalid. A packetization interval of 20 ms or a multiple thereof is suggested as it leads to integral sample counts for all common sampling rates. 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |F| FMT | length |0|0| format | reserved | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | name of format | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | channels | sampling rate (Hz) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ... encoding specific parameters ... +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Figure 1: FMT option for audio encodings 3.4 Port Assignment ST-II SAP and UDP port 5005 is the default destination for multicast real-time audio data carried by RTP for this profile. A fixed port number is useful as it is less likely than a randomly chosen port number to be already in use by another application at one or more of the intended destination hosts. Also, fixed port numbers allow traffic statistics to be collected and may simplify firewall implementations. A single fixed port number requires that hosts allow several processes to use a single UDP port with different multicast addresses. (The particular port number was chosen to lie in the range above 5000 to accomodate port number ------------------------------ 2. Fractional samples per second was considered excessive as the typical crystal accuraccy of 100 ppm translates into about one Hz or more of sampling rate inaccuracy. H. Schulzrinne Expires 12/31/93 [Page 7] INTERNET-DRAFT draft-ietf-avt-profile-03.txt October 20, 1993 index encoding sampling rate channels name (kHz) __________________________________________ 0 PCMU 8 1 1 1016 8 1 2 G721 8 1 3 GSM 8 1 4 G723 8 1 5 IDVI 8 1 10 L16 44.1 2 __________________________________________ Table 2: Standard audio encodings allocation practice within the Unix operating system, where port numbers below 1024 can only be used by privileged processes and port numbers between 1024 and 5000 are automatically assigned by the operating system.) Unicast connections may use the this or a set of mutually agreed-upon port numbers. 4 Video The following video encodings are currently defined, with their abbreviated names used for identification: CPV: This encoding, "Compressed Packet Video" is implemented by Concept, Bolter, and ViewPoint Systems video codecs. JPEG: The encoding is specified in ISO Standards DIS 10918-1 and DIS 10918-2. The data is formatted according to the JFIF (JPEG File Interchange Format) defined by C-Cube Microsystems. H261: The encoding is specified in CCITT/ITU-T standard H.261. The packetization and RTP-specific properties are described in RFC TBD. nv: The encoding is implemented in the program 'nv' developed at Xerox PARC by Ron Frederick. CUSM: The encoding is implemented in the program CU-SeeMe developed at Cornell University by Dick Cogger, Scott Brim, Tim Dorcey and John Lynn. PicW: The encoding is implemented in the program PictureWindow developed at Bolt, Beranek and Newman (BBN). H. Schulzrinne Expires 12/31/93 [Page 8] INTERNET-DRAFT draft-ietf-avt-profile-03.txt October 20, 1993 4.1 The RTCP FMT Option for Video Unless specified with the RTCP FMT option, the mapping between the format field in an RTP packet and the video encoding is specified by Tables 3. The second paragraph of Section 3.3 applies for video as well. Within the video FMT option, a one-octet numeric version identifier further describes the encoding. Unless otherwise defined, the version identifier has the value zero. 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |F| FMT | length |0|0| format | reserved | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | name of format | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | version | encoding-specific parameters | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ... encoding-specific parameters ... +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Figure 2: FMT option for video encodings number name ______________ 26 JPEG 27 CUSM 28 nv 29 PicW 30 Bolt 31 H261 Table 3: Format values for standard video encodings 4.2 Port Assignment ST-II SAP and UDP port 5006 is the default destination for multicast real-time video data carried by RTP for this profile. The remainder of section 3.4 applies. H. Schulzrinne Expires 12/31/93 [Page 9] INTERNET-DRAFT draft-ietf-avt-profile-03.txt October 20, 1993 5 Miscellaneous RTCP messages should be sent periodically, with a period varying randomly around a set mean to avoid synchronized bursts of RTCP packets. (For example, the time between messages could vary uniformly between one half and 1.5 times the mean.) The average period between transmissions determines the additional network load due to RTCP packets and also determines how long it will take a new arrival to discover the identities of the other conference participants. The average period should be chosen such that no more than a small fraction (say, 1%) of the media bandwidth is consumed by RTCP messages from all sources, with a minimum period of a few seconds. By scaling the message frequency with the (slowly increasing) number of observed participants, a new conference participant will quickly inform all other participants of its arrival and then slow its announcement rate. 6 Address of Author Henning Schulzrinne AT&T Bell Laboratories MH 2A244 600 Mountain Avenue Murray Hill, NJ 07974-0636 telephone: +1 908 582 2262 facsimile: +1 908 582 5809 electronic mail: hgs@research.att.com H. Schulzrinne Expires 12/31/93 [Page 10]