1
2
3
4
5
6
7Network Working Group                                     H. Schulzrinne
8Request for Comments: 3551                           Columbia University
9Obsoletes: 1890                                                S. Casner
10Category: Standards Track                                  Packet Design
11                                                               July 2003
12
13
14              RTP Profile for Audio and Video Conferences
15                          with Minimal Control
16
17Status of this Memo
18
19   This document specifies an Internet standards track protocol for the
20   Internet community, and requests discussion and suggestions for
21   improvements.  Please refer to the current edition of the "Internet
22   Official Protocol Standards" (STD 1) for the standardization state
23   and status of this protocol.  Distribution of this memo is unlimited.
24
25Copyright Notice
26
27   Copyright (C) The Internet Society (2003).  All Rights Reserved.
28
29Abstract
30
31   This document describes a profile called "RTP/AVP" for the use of the
32   real-time transport protocol (RTP), version 2, and the associated
33   control protocol, RTCP, within audio and video multiparticipant
34   conferences with minimal control.  It provides interpretations of
35   generic fields within the RTP specification suitable for audio and
36   video conferences.  In particular, this document defines a set of
37   default mappings from payload type numbers to encodings.
38
39   This document also describes how audio and video data may be carried
40   within RTP.  It defines a set of standard encodings and their names
41   when used within RTP.  The descriptions provide pointers to reference
42   implementations and the detailed standards.  This document is meant
43   as an aid for implementors of audio, video and other real-time
44   multimedia applications.
45
46   This memorandum obsoletes RFC 1890.  It is mostly backwards-
47   compatible except for functions removed because two interoperable
48   implementations were not found.  The additions to RFC 1890 codify
49   existing practice in the use of payload formats under this profile
50   and include new payload formats defined since RFC 1890 was published.
51
52
53
54
55
56
57
58Schulzrinne & Casner        Standards Track                     [Page 1]
59
60RFC 3551                    RTP A/V Profile                    July 2003
61
62
63Table of Contents
64
65   1.  Introduction .................................................  3
66       1.1  Terminology .............................................  3
67   2.  RTP and RTCP Packet Forms and Protocol Behavior ..............  4
68   3.  Registering Additional Encodings .............................  6
69   4.  Audio ........................................................  8
70       4.1  Encoding-Independent Rules ..............................  8
71       4.2  Operating Recommendations ...............................  9
72       4.3  Guidelines for Sample-Based Audio Encodings ............. 10
73       4.4  Guidelines for Frame-Based Audio Encodings .............. 11
74       4.5  Audio Encodings ......................................... 12
75            4.5.1   DVI4 ............................................ 13
76            4.5.2   G722 ............................................ 14
77            4.5.3   G723 ............................................ 14
78            4.5.4   G726-40, G726-32, G726-24, and G726-16 .......... 18
79            4.5.5   G728 ............................................ 19
80            4.5.6   G729 ............................................ 20
81            4.5.7   G729D and G729E ................................. 22
82            4.5.8   GSM ............................................. 24
83            4.5.9   GSM-EFR ......................................... 27
84            4.5.10  L8 .............................................. 27
85            4.5.11  L16 ............................................. 27
86            4.5.12  LPC ............................................. 27
87            4.5.13  MPA ............................................. 28
88            4.5.14  PCMA and PCMU ................................... 28
89            4.5.15  QCELP ........................................... 28
90            4.5.16  RED ............................................. 29
91            4.5.17  VDVI ............................................ 29
92   5.  Video ........................................................ 30
93       5.1  CelB .................................................... 30
94       5.2  JPEG .................................................... 30
95       5.3  H261 .................................................... 30
96       5.4  H263 .................................................... 31
97       5.5  H263-1998 ............................................... 31
98       5.6  MPV ..................................................... 31
99       5.7  MP2T .................................................... 31
100       5.8  nv ...................................................... 32
101   6.  Payload Type Definitions ..................................... 32
102   7.  RTP over TCP and Similar Byte Stream Protocols ............... 34
103   8.  Port Assignment .............................................. 34
104   9.  Changes from RFC 1890 ........................................ 35
105   10. Security Considerations ...................................... 38
106   11. IANA Considerations .......................................... 39
107   12. References ................................................... 39
108       12.1 Normative References .................................... 39
109       12.2 Informative References .................................. 39
110   13. Current Locations of Related Resources ....................... 41
111
112
113
114Schulzrinne & Casner        Standards Track                     [Page 2]
115
116RFC 3551                    RTP A/V Profile                    July 2003
117
118
119   14. Acknowledgments .............................................. 42
120   15. Intellectual Property Rights Statement ....................... 43
121   16. Authors' Addresses ........................................... 43
122   17. Full Copyright Statement ..................................... 44
123
1241. Introduction
125
126   This profile defines aspects of RTP left unspecified in the RTP
127   Version 2 protocol definition (RFC 3550) [1].  This profile is
128   intended for the use within audio and video conferences with minimal
129   session control.  In particular, no support for the negotiation of
130   parameters or membership control is provided.  The profile is
131   expected to be useful in sessions where no negotiation or membership
132   control are used (e.g., using the static payload types and the
133   membership indications provided by RTCP), but this profile may also
134   be useful in conjunction with a higher-level control protocol.
135
136   Use of this profile may be implicit in the use of the appropriate
137   applications; there may be no explicit indication by port number,
138   protocol identifier or the like.  Applications such as session
139   directories may use the name for this profile specified in Section
140   11.
141
142   Other profiles may make different choices for the items specified
143   here.
144
145   This document also defines a set of encodings and payload formats for
146   audio and video.  These payload format descriptions are included here
147   only as a matter of convenience since they are too small to warrant
148   separate documents.  Use of these payload formats is NOT REQUIRED to
149   use this profile.  Only the binding of some of the payload formats to
150   static payload type numbers in Tables 4 and 5 is normative.
151
1521.1 Terminology
153
154   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
155   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
156   document are to be interpreted as described in RFC 2119 [2] and
157   indicate requirement levels for implementations compliant with this
158   RTP profile.
159
160   This document defines the term media type as dividing encodings of
161   audio and video content into three classes: audio, video and
162   audio/video (interleaved).
163
164
165
166
167
168
169
170Schulzrinne & Casner        Standards Track                     [Page 3]
171
172RFC 3551                    RTP A/V Profile                    July 2003
173
174
1752. RTP and RTCP Packet Forms and Protocol Behavior
176
177   The section "RTP Profiles and Payload Format Specifications" of RFC
178   3550 enumerates a number of items that can be specified or modified
179   in a profile.  This section addresses these items.  Generally, this
180   profile follows the default and/or recommended aspects of the RTP
181   specification.
182
183   RTP data header: The standard format of the fixed RTP data
184      header is used (one marker bit).
185
186   Payload types: Static payload types are defined in Section 6.
187
188   RTP data header additions: No additional fixed fields are
189      appended to the RTP data header.
190
191   RTP data header extensions: No RTP header extensions are
192      defined, but applications operating under this profile MAY use
193      such extensions.  Thus, applications SHOULD NOT assume that the
194      RTP header X bit is always zero and SHOULD be prepared to ignore
195      the header extension.  If a header extension is defined in the
196      future, that definition MUST specify the contents of the first 16
197      bits in such a way that multiple different extensions can be
198      identified.
199
200   RTCP packet types: No additional RTCP packet types are defined
201      by this profile specification.
202
203   RTCP report interval: The suggested constants are to be used for
204      the RTCP report interval calculation.  Sessions operating under
205      this profile MAY specify a separate parameter for the RTCP traffic
206      bandwidth rather than using the default fraction of the session
207      bandwidth.  The RTCP traffic bandwidth MAY be divided into two
208      separate session parameters for those participants which are
209      active data senders and those which are not.  Following the
210      recommendation in the RTP specification [1] that 1/4 of the RTCP
211      bandwidth be dedicated to data senders, the RECOMMENDED default
212      values for these two parameters would be 1.25% and 3.75%,
213      respectively.  For a particular session, the RTCP bandwidth for
214      non-data-senders MAY be set to zero when operating on
215      unidirectional links or for sessions that don't require feedback
216      on the quality of reception.  The RTCP bandwidth for data senders
217      SHOULD be kept non-zero so that sender reports can still be sent
218      for inter-media synchronization and to identify the source by
219      CNAME.  The means by which the one or two session parameters for
220      RTCP bandwidth are specified is beyond the scope of this memo.
221
222
223
224
225
226Schulzrinne & Casner        Standards Track                     [Page 4]
227
228RFC 3551                    RTP A/V Profile                    July 2003
229
230
231   SR/RR extension: No extension section is defined for the RTCP SR
232      or RR packet.
233
234   SDES use: Applications MAY use any of the SDES items described
235      in the RTP specification.  While CNAME information MUST be sent
236      every reporting interval, other items SHOULD only be sent every
237      third reporting interval, with NAME sent seven out of eight times
238      within that slot and the remaining SDES items cyclically taking up
239      the eighth slot, as defined in Section 6.2.2 of the RTP
240      specification.  In other words, NAME is sent in RTCP packets 1, 4,
241      7, 10, 13, 16, 19, while, say, EMAIL is used in RTCP packet 22.
242
243   Security: The RTP default security services are also the default
244      under this profile.
245
246   String-to-key mapping: No mapping is specified by this profile.
247
248   Congestion: RTP and this profile may be used in the context of
249      enhanced network service, for example, through Integrated Services
250      (RFC 1633) [4] or Differentiated Services (RFC 2475) [5], or they
251      may be used with best effort service.
252
253      If enhanced service is being used, RTP receivers SHOULD monitor
254      packet loss to ensure that the service that was requested is
255      actually being delivered.  If it is not, then they SHOULD assume
256      that they are receiving best-effort service and behave
257      accordingly.
258
259      If best-effort service is being used, RTP receivers SHOULD monitor
260      packet loss to ensure that the packet loss rate is within
261      acceptable parameters.  Packet loss is considered acceptable if a
262      TCP flow across the same network path and experiencing the same
263      network conditions would achieve an average throughput, measured
264      on a reasonable timescale, that is not less than the RTP flow is
265      achieving.  This condition can be satisfied by implementing
266      congestion control mechanisms to adapt the transmission rate (or
267      the number of layers subscribed for a layered multicast session),
268      or by arranging for a receiver to leave the session if the loss
269      rate is unacceptably high.
270
271      The comparison to TCP cannot be specified exactly, but is intended
272      as an "order-of-magnitude" comparison in timescale and throughput.
273      The timescale on which TCP throughput is measured is the round-
274      trip time of the connection.  In essence, this requirement states
275      that it is not acceptable to deploy an application (using RTP or
276      any other transport protocol) on the best-effort Internet which
277      consumes bandwidth arbitrarily and does not compete fairly with
278      TCP within an order of magnitude.
279
280
281
282Schulzrinne & Casner        Standards Track                     [Page 5]
283
284RFC 3551                    RTP A/V Profile                    July 2003
285
286
287   Underlying protocol: The profile specifies the use of RTP over
288      unicast and multicast UDP as well as TCP.  (This does not preclude
289      the use of these definitions when RTP is carried by other lower-
290      layer protocols.)
291
292   Transport mapping: The standard mapping of RTP and RTCP to
293      transport-level addresses is used.
294
295   Encapsulation: This profile leaves to applications the
296      specification of RTP encapsulation in protocols other than UDP.
297
2983.  Registering Additional Encodings
299
300   This profile lists a set of encodings, each of which is comprised of
301   a particular media data compression or representation plus a payload
302   format for encapsulation within RTP.  Some of those payload formats
303   are specified here, while others are specified in separate RFCs.  It
304   is expected that additional encodings beyond the set listed here will
305   be created in the future and specified in additional payload format
306   RFCs.
307
308   This profile also assigns to each encoding a short name which MAY be
309   used by higher-level control protocols, such as the Session
310   Description Protocol (SDP), RFC 2327 [6], to identify encodings
311   selected for a particular RTP session.
312
313   In some contexts it may be useful to refer to these encodings in the
314   form of a MIME content-type.  To facilitate this, RFC 3555 [7]
315   provides registrations for all of the encodings names listed here as
316   MIME subtype names under the "audio" and "video" MIME types through
317   the MIME registration procedure as specified in RFC 2048 [8].
318
319   Any additional encodings specified for use under this profile (or
320   others) may also be assigned names registered as MIME subtypes with
321   the Internet Assigned Numbers Authority (IANA).  This registry
322   provides a means to insure that the names assigned to the additional
323   encodings are kept unique.  RFC 3555 specifies the information that
324   is required for the registration of RTP encodings.
325
326   In addition to assigning names to encodings, this profile also
327   assigns static RTP payload type numbers to some of them.  However,
328   the payload type number space is relatively small and cannot
329   accommodate assignments for all existing and future encodings.
330   During the early stages of RTP development, it was necessary to use
331   statically assigned payload types because no other mechanism had been
332   specified to bind encodings to payload types.  It was anticipated
333   that non-RTP means beyond the scope of this memo (such as directory
334   services or invitation protocols) would be specified to establish a
335
336
337
338Schulzrinne & Casner        Standards Track                     [Page 6]
339
340RFC 3551                    RTP A/V Profile                    July 2003
341
342
343   dynamic mapping between a payload type and an encoding.  Now,
344   mechanisms for defining dynamic payload type bindings have been
345   specified in the Session Description Protocol (SDP) and in other
346   protocols such as ITU-T Recommendation H.323/H.245.  These mechanisms
347   associate the registered name of the encoding/payload format, along
348   with any additional required parameters, such as the RTP timestamp
349   clock rate and number of channels, with a payload type number.  This
350   association is effective only for the duration of the RTP session in
351   which the dynamic payload type binding is made.  This association
352   applies only to the RTP session for which it is made, thus the
353   numbers can be re-used for different encodings in different sessions
354   so the number space limitation is avoided.
355
356   This profile reserves payload type numbers in the range 96-127
357   exclusively for dynamic assignment.  Applications SHOULD first use
358   values in this range for dynamic payload types.  Those applications
359   which need to define more than 32 dynamic payload types MAY bind
360   codes below 96, in which case it is RECOMMENDED that unassigned
361   payload type numbers be used first.  However, the statically assigned
362   payload types are default bindings and MAY be dynamically bound to
363   new encodings if needed.  Redefining payload types below 96 may cause
364   incorrect operation if an attempt is made to join a session without
365   obtaining session description information that defines the dynamic
366   payload types.
367
368   Dynamic payload types SHOULD NOT be used without a well-defined
369   mechanism to indicate the mapping.  Systems that expect to
370   interoperate with others operating under this profile SHOULD NOT make
371   their own assignments of proprietary encodings to particular, fixed
372   payload types.
373
374   This specification establishes the policy that no additional static
375   payload types will be assigned beyond the ones defined in this
376   document.  Establishing this policy avoids the problem of trying to
377   create a set of criteria for accepting static assignments and
378   encourages the implementation and deployment of the dynamic payload
379   type mechanisms.
380
381   The final set of static payload type assignments is provided in
382   Tables 4 and 5.
383
384
385
386
387
388
389
390
391
392
393
394Schulzrinne & Casner        Standards Track                     [Page 7]
395
396RFC 3551                    RTP A/V Profile                    July 2003
397
398
3994.  Audio
400
4014.1  Encoding-Independent Rules
402
403   Since the ability to suppress silence is one of the primary
404   motivations for using packets to transmit voice, the RTP header
405   carries both a sequence number and a timestamp to allow a receiver to
406   distinguish between lost packets and periods of time when no data was
407   transmitted.  Discontiguous transmission (silence suppression) MAY be
408   used with any audio payload format.  Receivers MUST assume that
409   senders may suppress silence unless this is restricted by signaling
410   specified elsewhere.  (Even if the transmitter does not suppress
411   silence, the receiver should be prepared to handle periods when no
412   data is present since packets may be lost.)
413
414   Some payload formats (see Sections 4.5.3 and 4.5.6) define a "silence
415   insertion descriptor" or "comfort noise" frame to specify parameters
416   for artificial noise that may be generated during a period of silence
417   to approximate the background noise at the source.  For other payload
418   formats, a generic Comfort Noise (CN) payload format is specified in
419   RFC 3389 [9].  When the CN payload format is used with another
420   payload format, different values in the RTP payload type field
421   distinguish comfort-noise packets from those of the selected payload
422   format.
423
424   For applications which send either no packets or occasional comfort-
425   noise packets during silence, the first packet of a talkspurt, that
426   is, the first packet after a silence period during which packets have
427   not been transmitted contiguously, SHOULD be distinguished by setting
428   the marker bit in the RTP data header to one.  The marker bit in all
429   other packets is zero.  The beginning of a talkspurt MAY be used to
430   adjust the playout delay to reflect changing network delays.
431   Applications without silence suppression MUST set the marker bit to
432   zero.
433
434   The RTP clock rate used for generating the RTP timestamp is
435   independent of the number of channels and the encoding; it usually
436   equals the number of sampling periods per second.  For N-channel
437   encodings, each sampling period (say, 1/8,000 of a second) generates
438   N samples.  (This terminology is standard, but somewhat confusing, as
439   the total number of samples generated per second is then the sampling
440   rate times the channel count.)
441
442   If multiple audio channels are used, channels are numbered left-to-
443   right, starting at one.  In RTP audio packets, information from
444   lower-numbered channels precedes that from higher-numbered channels.
445
446
447
448
449
450Schulzrinne & Casner        Standards Track                     [Page 8]
451
452RFC 3551                    RTP A/V Profile                    July 2003
453
454
455   For more than two channels, the convention followed by the AIFF-C
456   audio interchange format SHOULD be followed [3], using the following
457   notation, unless some other convention is specified for a particular
458   encoding or payload format:
459
460      l  left
461      r  right
462      c  center
463      S  surround
464      F  front
465      R  rear
466
467      channels  description  channel
468                                1     2   3   4   5   6
469      _________________________________________________
470      2         stereo          l     r
471      3                         l     r   c
472      4                         l     c   r   S
473      5                        Fl     Fr  Fc  Sl  Sr
474      6                         l     lc  c   r   rc  S
475
476         Note: RFC 1890 defined two conventions for the ordering of four
477         audio channels.  Since the ordering is indicated implicitly by
478         the number of channels, this was ambiguous.  In this revision,
479         the order described as "quadrophonic" has been eliminated to
480         remove the ambiguity.  This choice was based on the observation
481         that quadrophonic consumer audio format did not become popular
482         whereas surround-sound subsequently has.
483
484   Samples for all channels belonging to a single sampling instant MUST
485   be within the same packet.  The interleaving of samples from
486   different channels depends on the encoding.  General guidelines are
487   given in Section 4.3 and 4.4.
488
489   The sampling frequency SHOULD be drawn from the set:  8,000, 11,025,
490   16,000, 22,050, 24,000, 32,000, 44,100 and 48,000 Hz.  (Older Apple
491   Macintosh computers had a native sample rate of 22,254.54 Hz, which
492   can be converted to 22,050 with acceptable quality by dropping 4
493   samples in a 20 ms frame.)  However, most audio encodings are defined
494   for a more restricted set of sampling frequencies.  Receivers SHOULD
495   be prepared to accept multi-channel audio, but MAY choose to only
496   play a single channel.
497
4984.2  Operating Recommendations
499
500   The following recommendations are default operating parameters.
501   Applications SHOULD be prepared to handle other values.  The ranges
502   given are meant to give guidance to application writers, allowing a
503
504
505
506Schulzrinne & Casner        Standards Track                     [Page 9]
507
508RFC 3551                    RTP A/V Profile                    July 2003
509
510
511   set of applications conforming to these guidelines to interoperate
512   without additional negotiation.  These guidelines are not intended to
513   restrict operating parameters for applications that can negotiate a
514   set of interoperable parameters, e.g., through a conference control
515   protocol.
516
517   For packetized audio, the default packetization interval SHOULD have
518   a duration of 20 ms or one frame, whichever is longer, unless
519   otherwise noted in Table 1 (column "ms/packet").  The packetization
520   interval determines the minimum end-to-end delay; longer packets
521   introduce less header overhead but higher delay and make packet loss
522   more noticeable.  For non-interactive applications such as lectures
523   or for links with severe bandwidth constraints, a higher
524   packetization delay MAY be used.  A receiver SHOULD accept packets
525   representing between 0 and 200 ms of audio data.  (For framed audio
526   encodings, a receiver SHOULD accept packets with a number of frames
527   equal to 200 ms divided by the frame duration, rounded up.)  This
528   restriction allows reasonable buffer sizing for the receiver.
529
5304.3  Guidelines for Sample-Based Audio Encodings
531
532   In sample-based encodings, each audio sample is represented by a
533   fixed number of bits.  Within the compressed audio data, codes for
534   individual samples may span octet boundaries.  An RTP audio packet
535   may contain any number of audio samples, subject to the constraint
536   that the number of bits per sample times the number of samples per
537   packet yields an integral octet count.  Fractional encodings produce
538   less than one octet per sample.
539
540   The duration of an audio packet is determined by the number of
541   samples in the packet.
542
543   For sample-based encodings producing one or more octets per sample,
544   samples from different channels sampled at the same sampling instant
545   SHOULD be packed in consecutive octets.  For example, for a two-
546   channel encoding, the octet sequence is (left channel, first sample),
547   (right channel, first sample), (left channel, second sample), (right
548   channel, second sample), ....  For multi-octet encodings, octets
549   SHOULD be transmitted in network byte order (i.e., most significant
550   octet first).
551
552   The packing of sample-based encodings producing less than one octet
553   per sample is encoding-specific.
554
555   The RTP timestamp reflects the instant at which the first sample in
556   the packet was sampled, that is, the oldest information in the
557   packet.
558
559
560
561
562Schulzrinne & Casner        Standards Track                    [Page 10]
563
564RFC 3551                    RTP A/V Profile                    July 2003
565
566
5674.4  Guidelines for Frame-Based Audio Encodings
568
569   Frame-based encodings encode a fixed-length block of audio into
570   another block of compressed data, typically also of fixed length.
571   For frame-based encodings, the sender MAY choose to combine several
572   such frames into a single RTP packet.  The receiver can tell the
573   number of frames contained in an RTP packet, if all the frames have
574   the same length, by dividing the RTP payload length by the audio
575   frame size which is defined as part of the encoding.  This does not
576   work when carrying frames of different sizes unless the frame sizes
577   are relatively prime.  If not, the frames MUST indicate their size.
578
579   For frame-based codecs, the channel order is defined for the whole
580   block.  That is, for two-channel audio, right and left samples SHOULD
581   be coded independently, with the encoded frame for the left channel
582   preceding that for the right channel.
583
584   All frame-oriented audio codecs SHOULD be able to encode and decode
585   several consecutive frames within a single packet.  Since the frame
586   size for the frame-oriented codecs is given, there is no need to use
587   a separate designation for the same encoding, but with different
588   number of frames per packet.
589
590   RTP packets SHALL contain a whole number of frames, with frames
591   inserted according to age within a packet, so that the oldest frame
592   (to be played first) occurs immediately after the RTP packet header.
593   The RTP timestamp reflects the instant at which the first sample in
594   the first frame was sampled, that is, the oldest information in the
595   packet.
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618Schulzrinne & Casner        Standards Track                    [Page 11]
619
620RFC 3551                    RTP A/V Profile                    July 2003
621
622
6234.5 Audio Encodings
624
625   name of                              sampling              default
626   encoding  sample/frame  bits/sample      rate  ms/frame  ms/packet
627   __________________________________________________________________
628   DVI4      sample        4                var.                   20
629   G722      sample        8              16,000                   20
630   G723      frame         N/A             8,000        30         30
631   G726-40   sample        5               8,000                   20
632   G726-32   sample        4               8,000                   20
633   G726-24   sample        3               8,000                   20
634   G726-16   sample        2               8,000                   20
635   G728      frame         N/A             8,000       2.5         20
636   G729      frame         N/A             8,000        10         20
637   G729D     frame         N/A             8,000        10         20
638   G729E     frame         N/A             8,000        10         20
639   GSM       frame         N/A             8,000        20         20
640   GSM-EFR   frame         N/A             8,000        20         20
641   L8        sample        8                var.                   20
642   L16       sample        16               var.                   20
643   LPC       frame         N/A             8,000        20         20
644   MPA       frame         N/A              var.      var.
645   PCMA      sample        8                var.                   20
646   PCMU      sample        8                var.                   20
647   QCELP     frame         N/A             8,000        20         20
648   VDVI      sample        var.             var.                   20
649
650   Table 1: Properties of Audio Encodings (N/A: not applicable; var.:
651            variable)
652
653   The characteristics of the audio encodings described in this document
654   are shown in Table 1; they are listed in order of their payload type
655   in Table 4.  While most audio codecs are only specified for a fixed
656   sampling rate, some sample-based algorithms (indicated by an entry of
657   "var." in the sampling rate column of Table 1) may be used with
658   different sampling rates, resulting in different coded bit rates.
659   When used with a sampling rate other than that for which a static
660   payload type is defined, non-RTP means beyond the scope of this memo
661   MUST be used to define a dynamic payload type and MUST indicate the
662   selected RTP timestamp clock rate, which is usually the same as the
663   sampling rate for audio.
664
665
666
667
668
669
670
671
672
673
674Schulzrinne & Casner        Standards Track                    [Page 12]
675
676RFC 3551                    RTP A/V Profile                    July 2003
677
678
6794.5.1 DVI4
680
681   DVI4 uses an adaptive delta pulse code modulation (ADPCM) encoding
682   scheme that was specified by the Interactive Multimedia Association
683   (IMA) as the "IMA ADPCM wave type".  However, the encoding defined
684   here as DVI4 differs in three respects from the IMA specification:
685
686   o  The RTP DVI4 header contains the predicted value rather than the
687      first sample value contained the IMA ADPCM block header.
688
689   o  IMA ADPCM blocks contain an odd number of samples, since the first
690      sample of a block is contained just in the header (uncompressed),
691      followed by an even number of compressed samples.  DVI4 has an
692      even number of compressed samples only, using the `predict' word
693      from the header to decode the first sample.
694
695   o  For DVI4, the 4-bit samples are packed with the first sample in
696      the four most significant bits and the second sample in the four
697      least significant bits.  In the IMA ADPCM codec, the samples are
698      packed in the opposite order.
699
700   Each packet contains a single DVI block.  This profile only defines
701   the 4-bit-per-sample version, while IMA also specified a 3-bit-per-
702   sample encoding.
703
704   The "header" word for each channel has the following structure:
705
706      int16  predict;  /* predicted value of first sample
707                          from the previous block (L16 format) */
708      u_int8 index;    /* current index into stepsize table */
709      u_int8 reserved; /* set to zero by sender, ignored by receiver */
710
711   Each octet following the header contains two 4-bit samples, thus the
712   number of samples per packet MUST be even because there is no means
713   to indicate a partially filled last octet.
714
715   Packing of samples for multiple channels is for further study.
716
717   The IMA ADPCM algorithm was described in the document IMA Recommended
718   Practices for Enhancing Digital Audio Compatibility in Multimedia
719   Systems (version 3.0).  However, the Interactive Multimedia
720   Association ceased operations in 1997.  Resources for an archived
721   copy of that document and a software implementation of the RTP DVI4
722   encoding are listed in Section 13.
723
724
725
726
727
728
729
730Schulzrinne & Casner        Standards Track                    [Page 13]
731
732RFC 3551                    RTP A/V Profile                    July 2003
733
734
7354.5.2 G722
736
737   G722 is specified in ITU-T Recommendation G.722, "7 kHz audio-coding
738   within 64 kbit/s".  The G.722 encoder produces a stream of octets,
739   each of which SHALL be octet-aligned in an RTP packet.  The first bit
740   transmitted in the G.722 octet, which is the most significant bit of
741   the higher sub-band sample, SHALL correspond to the most significant
742   bit of the octet in the RTP packet.
743
744   Even though the actual sampling rate for G.722 audio is 16,000 Hz,
745   the RTP clock rate for the G722 payload format is 8,000 Hz because
746   that value was erroneously assigned in RFC 1890 and must remain
747   unchanged for backward compatibility.  The octet rate or sample-pair
748   rate is 8,000 Hz.
749
7504.5.3 G723
751
752   G723 is specified in ITU Recommendation G.723.1, "Dual-rate speech
753   coder for multimedia communications transmitting at 5.3 and 6.3
754   kbit/s".  The G.723.1 5.3/6.3 kbit/s codec was defined by the ITU-T
755   as a mandatory codec for ITU-T H.324 GSTN videophone terminal
756   applications.  The algorithm has a floating point specification in
757   Annex B to G.723.1, a silence compression algorithm in Annex A to
758   G.723.1 and a scalable channel coding scheme for wireless
759   applications in G.723.1 Annex C.
760
761   This Recommendation specifies a coded representation that can be used
762   for compressing the speech signal component of multi-media services
763   at a very low bit rate.  Audio is encoded in 30 ms frames, with an
764   additional delay of 7.5 ms due to look-ahead.  A G.723.1 frame can be
765   one of three sizes:  24 octets (6.3 kb/s frame), 20 octets (5.3 kb/s
766   frame), or 4 octets.  These 4-octet frames are called SID frames
767   (Silence Insertion Descriptor) and are used to specify comfort noise
768   parameters.  There is no restriction on how 4, 20, and 24 octet
769   frames are intermixed.  The least significant two bits of the first
770   octet in the frame determine the frame size and codec type:
771
772         bits  content                      octets/frame
773         00    high-rate speech (6.3 kb/s)            24
774         01    low-rate speech  (5.3 kb/s)            20
775         10    SID frame                               4
776         11    reserved
777
778
779
780
781
782
783
784
785
786Schulzrinne & Casner        Standards Track                    [Page 14]
787
788RFC 3551                    RTP A/V Profile                    July 2003
789
790
791   It is possible to switch between the two rates at any 30 ms frame
792   boundary.  Both (5.3 kb/s and 6.3 kb/s) rates are a mandatory part of
793   the encoder and decoder.  Receivers MUST accept both data rates and
794   MUST accept SID frames unless restriction of these capabilities has
795   been signaled.  The MIME registration for G723 in RFC 3555 [7]
796   specifies parameters that MAY be used with MIME or SDP to restrict to
797   a single data rate or to restrict the use of SID frames.  This coder
798   was optimized to represent speech with near-toll quality at the above
799   rates using a limited amount of complexity.
800
801   The packing of the encoded bit stream into octets and the
802   transmission order of the octets is specified in Rec. G.723.1 and is
803   the same as that produced by the G.723 C code reference
804   implementation.  For the 6.3 kb/s data rate, this packing is
805   illustrated as follows, where the header (HDR) bits are always "0 0"
806   as shown in Fig. 1 to indicate operation at 6.3 kb/s, and the Z bit
807   is always set to zero.  The diagrams show the bit packing in "network
808   byte order", also known as big-endian order.  The bits of each 32-bit
809   word are numbered 0 to 31, with the most significant bit on the left
810   and numbered 0.  The octets (bytes) of each word are transmitted most
811   significant octet first.  The bits of each data field are numbered in
812   the order of the bit stream representation of the encoding (least
813   significant bit first).  The vertical bars indicate the boundaries
814   between field fragments.
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842Schulzrinne & Casner        Standards Track                    [Page 15]
843
844RFC 3551                    RTP A/V Profile                    July 2003
845
846
847    0                   1                   2                   3
848    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
849   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
850   |    LPC    |HDR|      LPC      |      LPC      |    ACL0   |LPC|
851   |           |   |               |               |           |   |
852   |0 0 0 0 0 0|0 0|1 1 1 1 0 0 0 0|2 2 1 1 1 1 1 1|0 0 0 0 0 0|2 2|
853   |5 4 3 2 1 0|   |3 2 1 0 9 8 7 6|1 0 9 8 7 6 5 4|5 4 3 2 1 0|3 2|
854   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
855   |  ACL2   |ACL|A| GAIN0 |ACL|ACL|    GAIN0      |    GAIN1      |
856   |         | 1 |C|       | 3 | 2 |               |               |
857   |0 0 0 0 0|0 0|0|0 0 0 0|0 0|0 0|1 1 0 0 0 0 0 0|0 0 0 0 0 0 0 0|
858   |4 3 2 1 0|1 0|6|3 2 1 0|1 0|6 5|1 0 9 8 7 6 5 4|7 6 5 4 3 2 1 0|
859   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
860   | GAIN2 | GAIN1 |     GAIN2     |     GAIN3     | GRID  | GAIN3 |
861   |       |       |               |               |       |       |
862   |0 0 0 0|1 1 0 0|1 1 0 0 0 0 0 0|0 0 0 0 0 0 0 0|0 0 0 0|1 1 0 0|
863   |3 2 1 0|1 0 9 8|1 0 9 8 7 6 5 4|7 6 5 4 3 2 1 0|3 2 1 0|1 0 9 8|
864   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
865   |   MSBPOS    |Z|POS|  MSBPOS   |     POS0      |POS|   POS0    |
866   |             | | 0 |           |               | 1 |           |
867   |0 0 0 0 0 0 0|0|0 0|1 1 1 0 0 0|0 0 0 0 0 0 0 0|0 0|1 1 1 1 1 1|
868   |6 5 4 3 2 1 0| |1 0|2 1 0 9 8 7|9 8 7 6 5 4 3 2|1 0|5 4 3 2 1 0|
869   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
870   |     POS1      | POS2  | POS1  |     POS2      | POS3  | POS2  |
871   |               |       |       |               |       |       |
872   |0 0 0 0 0 0 0 0|0 0 0 0|1 1 1 1|1 1 0 0 0 0 0 0|0 0 0 0|1 1 1 1|
873   |9 8 7 6 5 4 3 2|3 2 1 0|3 2 1 0|1 0 9 8 7 6 5 4|3 2 1 0|5 4 3 2|
874   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
875   |     POS3      |   PSIG0   |POS|PSIG2|  PSIG1  |  PSIG3  |PSIG2|
876   |               |           | 3 |     |         |         |     |
877   |1 1 0 0 0 0 0 0|0 0 0 0 0 0|1 1|0 0 0|0 0 0 0 0|0 0 0 0 0|0 0 0|
878   |1 0 9 8 7 6 5 4|5 4 3 2 1 0|3 2|2 1 0|4 3 2 1 0|4 3 2 1 0|5 4 3|
879   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
880
881                  Figure 1: G.723 (6.3 kb/s) bit packing
882
883   For the 5.3 kb/s data rate, the header (HDR) bits are always "0 1",
884   as shown in Fig. 2, to indicate operation at 5.3 kb/s.
885
886
887
888
889
890
891
892
893
894
895
896
897
898Schulzrinne & Casner        Standards Track                    [Page 16]
899
900RFC 3551                    RTP A/V Profile                    July 2003
901
902
903    0                   1                   2                   3
904    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
905   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
906   |    LPC    |HDR|      LPC      |      LPC      |   ACL0    |LPC|
907   |           |   |               |               |           |   |
908   |0 0 0 0 0 0|0 1|1 1 1 1 0 0 0 0|2 2 1 1 1 1 1 1|0 0 0 0 0 0|2 2|
909   |5 4 3 2 1 0|   |3 2 1 0 9 8 7 6|1 0 9 8 7 6 5 4|5 4 3 2 1 0|3 2|
910   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
911   |  ACL2   |ACL|A| GAIN0 |ACL|ACL|     GAIN0     |     GAIN1     |
912   |         | 1 |C|       | 3 | 2 |               |               |
913   |0 0 0 0 0|0 0|0|0 0 0 0|0 0|0 0|1 1 0 0 0 0 0 0|0 0 0 0 0 0 0 0|
914   |4 3 2 1 0|1 0|6|3 2 1 0|1 0|6 5|1 0 9 8 7 6 5 4|7 6 5 4 3 2 1 0|
915   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
916   | GAIN2 | GAIN1 |     GAIN2     |    GAIN3      | GRID  | GAIN3 |
917   |       |       |               |               |       |       |
918   |0 0 0 0|1 1 0 0|1 1 0 0 0 0 0 0|0 0 0 0 0 0 0 0|0 0 0 0|1 1 0 0|
919   |3 2 1 0|1 0 9 8|1 0 9 8 7 6 5 4|7 6 5 4 3 2 1 0|4 3 2 1|1 0 9 8|
920   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
921   |     POS0      | POS1  | POS0  |     POS1      |     POS2      |
922   |               |       |       |               |               |
923   |0 0 0 0 0 0 0 0|0 0 0 0|1 1 0 0|1 1 0 0 0 0 0 0|0 0 0 0 0 0 0 0|
924   |7 6 5 4 3 2 1 0|3 2 1 0|1 0 9 8|1 0 9 8 7 6 5 4|7 6 5 4 3 2 1 0|
925   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
926   | POS3  | POS2  |     POS3      | PSIG1 | PSIG0 | PSIG3 | PSIG2 |
927   |       |       |               |       |       |       |       |
928   |0 0 0 0|1 1 0 0|1 1 0 0 0 0 0 0|0 0 0 0|0 0 0 0|0 0 0 0|0 0 0 0|
929   |3 2 1 0|1 0 9 8|1 0 9 8 7 6 5 4|3 2 1 0|3 2 1 0|3 2 1 0|3 2 1 0|
930   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
931
932                  Figure 2: G.723 (5.3 kb/s) bit packing
933
934   The packing of G.723.1 SID (silence) frames, which are indicated by
935   the header (HDR) bits having the pattern "1 0", is depicted in Fig.
936   3.
937
938    0                   1                   2                   3
939    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
940   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
941   |    LPC    |HDR|      LPC      |      LPC      |   GAIN    |LPC|
942   |           |   |               |               |           |   |
943   |0 0 0 0 0 0|1 0|1 1 1 1 0 0 0 0|2 2 1 1 1 1 1 1|0 0 0 0 0 0|2 2|
944   |5 4 3 2 1 0|   |3 2 1 0 9 8 7 6|1 0 9 8 7 6 5 4|5 4 3 2 1 0|3 2|
945   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
946
947                   Figure 3: G.723 SID mode bit packing
948
949
950
951
952
953
954Schulzrinne & Casner        Standards Track                    [Page 17]
955
956RFC 3551                    RTP A/V Profile                    July 2003
957
958
9594.5.4  G726-40, G726-32, G726-24, and G726-16
960
961   ITU-T Recommendation G.726 describes, among others, the algorithm
962   recommended for conversion of a single 64 kbit/s A-law or mu-law PCM
963   channel encoded at 8,000 samples/sec to and from a 40, 32, 24, or 16
964   kbit/s channel.  The conversion is applied to the PCM stream using an
965   Adaptive Differential Pulse Code Modulation (ADPCM) transcoding
966   technique.  The ADPCM representation consists of a series of
967   codewords with a one-to-one correspondence to the samples in the PCM
968   stream.  The G726 data rates of 40, 32, 24, and 16 kbit/s have
969   codewords of 5, 4, 3, and 2 bits, respectively.
970
971   The 16 and 24 kbit/s encodings do not provide toll quality speech.
972   They are designed for used in overloaded Digital Circuit
973   Multiplication Equipment (DCME).  ITU-T G.726 recommends that the 16
974   and 24 kbit/s encodings should be alternated with higher data rate
975   encodings to provide an average sample size of between 3.5 and 3.7
976   bits per sample.
977
978   The encodings of G.726 are here denoted as G726-40, G726-32, G726-24,
979   and G726-16.  Prior to 1990, G721 described the 32 kbit/s ADPCM
980   encoding, and G723 described the 40, 32, and 16 kbit/s encodings.
981   Thus, G726-32 designates the same algorithm as G721 in RFC 1890.
982
983   A stream of G726 codewords contains no information on the encoding
984   being used, therefore transitions between G726 encoding types are not
985   permitted within a sequence of packed codewords.  Applications MUST
986   determine the encoding type of packed codewords from the RTP payload
987   identifier.
988
989   No payload-specific header information SHALL be included as part of
990   the audio data.  A stream of G726 codewords MUST be packed into
991   octets as follows:  the first codeword is placed into the first octet
992   such that the least significant bit of the codeword aligns with the
993   least significant bit in the octet, the second codeword is then
994   packed so that its least significant bit coincides with the least
995   significant unoccupied bit in the octet.  When a complete codeword
996   cannot be placed into an octet, the bits overlapping the octet
997   boundary are placed into the least significant bits of the next
998   octet.  Packing MUST end with a completely packed final octet.  The
999   number of codewords packed will therefore be a multiple of 8, 2, 8,
1000   and 4 for G726-40, G726-32, G726-24, and G726-16, respectively.  An
1001   example of the packing scheme for G726-32 codewords is as shown,
1002   where bit 7 is the least significant bit of the first octet, and bit
1003   A3 is the least significant bit of the first codeword:
1004
1005
1006
1007
1008
1009
1010Schulzrinne & Casner        Standards Track                    [Page 18]
1011
1012RFC 3551                    RTP A/V Profile                    July 2003
1013
1014
1015          0                   1
1016          0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
1017         +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-
1018         |B B B B|A A A A|D D D D|C C C C| ...
1019         |0 1 2 3|0 1 2 3|0 1 2 3|0 1 2 3|
1020         +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-
1021
1022   An example of the packing scheme for G726-24 codewords follows, where
1023   again bit 7 is the least significant bit of the first octet, and bit
1024   A2 is the least significant bit of the first codeword:
1025
1026          0                   1                   2
1027          0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3
1028         +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-
1029         |C C|B B B|A A A|F|E E E|D D D|C|H H H|G G G|F F| ...
1030         |1 2|0 1 2|0 1 2|2|0 1 2|0 1 2|0|0 1 2|0 1 2|0 1|
1031         +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-
1032
1033   Note that the "little-endian" direction in which samples are packed
1034   into octets in the G726-16, -24, -32 and -40 payload formats
1035   specified here is consistent with ITU-T Recommendation X.420, but is
1036   the opposite of what is specified in ITU-T Recommendation I.366.2
1037   Annex E for ATM AAL2 transport.  A second set of RTP payload formats
1038   matching the packetization of I.366.2 Annex E and identified by MIME
1039   subtypes AAL2-G726-16, -24, -32 and -40 will be specified in a
1040   separate document.
1041
10424.5.5 G728
1043
1044   G728 is specified in ITU-T Recommendation G.728, "Coding of speech at
1045   16 kbit/s using low-delay code excited linear prediction".
1046
1047   A G.278 encoder translates 5 consecutive audio samples into a 10-bit
1048   codebook index, resulting in a bit rate of 16 kb/s for audio sampled
1049   at 8,000 samples per second.  The group of five consecutive samples
1050   is called a vector.  Four consecutive vectors, labeled V1 to V4
1051   (where V1 is to be played first by the receiver), build one G.728
1052   frame.  The four vectors of 40 bits are packed into 5 octets, labeled
1053   B1 through B5.  B1 SHALL be placed first in the RTP packet.
1054
1055   Referring to the figure below, the principle for bit order is
1056   "maintenance of bit significance".  Bits from an older vector are
1057   more significant than bits from newer vectors.  The MSB of the frame
1058   goes to the MSB of B1 and the LSB of the frame goes to LSB of B5.
1059
1060
1061
1062
1063
1064
1065
1066Schulzrinne & Casner        Standards Track                    [Page 19]
1067
1068RFC 3551                    RTP A/V Profile                    July 2003
1069
1070
1071                   1         2         3        3
1072         0         0         0         0        9
1073         ++++++++++++++++++++++++++++++++++++++++
1074         <---V1---><---V2---><---V3---><---V4---> vectors
1075         <--B1--><--B2--><--B3--><--B4--><--B5--> octets
1076         <------------- frame 1 ---------------->
1077
1078   In particular, B1 contains the eight most significant bits of V1,
1079   with the MSB of V1 being the MSB of B1.  B2 contains the two least
1080   significant bits of V1, the more significant of the two in its MSB,
1081   and the six most significant bits of V2.  B1 SHALL be placed first in
1082   the RTP packet and B5 last.
1083
10844.5.6 G729
1085
1086   G729 is specified in ITU-T Recommendation G.729, "Coding of speech at
1087   8 kbit/s using conjugate structure-algebraic code excited linear
1088   prediction (CS-ACELP)".  A reduced-complexity version of the G.729
1089   algorithm is specified in Annex A to Rec. G.729.  The speech coding
1090   algorithms in the main body of G.729 and in G.729 Annex A are fully
1091   interoperable with each other, so there is no need to further
1092   distinguish between them.  An implementation that signals or accepts
1093   use of G729 payload format may implement either G.729 or G.729A
1094   unless restricted by additional signaling specified elsewhere related
1095   specifically to the encoding rather than the payload format.  The
1096   G.729 and G.729 Annex A codecs were optimized to represent speech
1097   with high quality, where G.729 Annex A trades some speech quality for
1098   an approximate 50% complexity reduction [10].  See the next Section
1099   (4.5.7) for other data rates added in later G.729 Annexes.  For all
1100   data rates, the sampling frequency (and RTP timestamp clock rate) is
1101   8,000 Hz.
1102
1103   A voice activity detector (VAD) and comfort noise generator (CNG)
1104   algorithm in Annex B of G.729 is RECOMMENDED for digital simultaneous
1105   voice and data applications and can be used in conjunction with G.729
1106   or G.729 Annex A.  A G.729 or G.729 Annex A frame contains 10 octets,
1107   while the G.729 Annex B comfort noise frame occupies 2 octets.
1108   Receivers MUST accept comfort noise frames if restriction of their
1109   use has not been signaled.  The MIME registration for G729 in RFC
1110   3555 [7] specifies a parameter that MAY be used with MIME or SDP to
1111   restrict the use of comfort noise frames.
1112
1113   A G729 RTP packet may consist of zero or more G.729 or G.729 Annex A
1114   frames, followed by zero or one G.729 Annex B frames.  The presence
1115   of a comfort noise frame can be deduced from the length of the RTP
1116   payload.  The default packetization interval is 20 ms (two frames),
1117   but in some situations it may be desirable to send 10 ms packets.  An
1118
1119
1120
1121
1122Schulzrinne & Casner        Standards Track                    [Page 20]
1123
1124RFC 3551                    RTP A/V Profile                    July 2003
1125
1126
1127   example would be a transition from speech to comfort noise in the
1128   first 10 ms of the packet.  For some applications, a longer
1129   packetization interval may be required to reduce the packet rate.
1130
1131       0                   1                   2                   3
1132       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
1133      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1134      |L|      L1     |    L2   |    L3   |       P1      |P|    C1   |
1135      |0|             |         |         |               |0|         |
1136      | |0 1 2 3 4 5 6|0 1 2 3 4|0 1 2 3 4|0 1 2 3 4 5 6 7| |0 1 2 3 4|
1137      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1138      |       C1      |  S1   | GA1 |  GB1  |    P2   |      C2       |
1139      |          1 1 1|       |     |       |         |               |
1140      |5 6 7 8 9 0 1 2|0 1 2 3|0 1 2|0 1 2 3|0 1 2 3 4|0 1 2 3 4 5 6 7|
1141      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1142      |   C2    |  S2   | GA2 |  GB2  |
1143      |    1 1 1|       |     |       |
1144      |8 9 0 1 2|0 1 2 3|0 1 2|0 1 2 3|
1145      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1146
1147                    Figure 4: G.729 and G.729A bit packing
1148
1149   The transmitted parameters of a G.729/G.729A 10-ms frame, consisting
1150   of 80 bits, are defined in Recommendation G.729, Table 8/G.729.  The
1151   mapping of the these parameters is given below in Fig. 4.  The
1152   diagrams show the bit packing in "network byte order", also known as
1153   big-endian order.  The bits of each 32-bit word are numbered 0 to 31,
1154   with the most significant bit on the left and numbered 0.  The octets
1155   (bytes) of each word are transmitted most significant octet first.
1156   The bits of each data field are numbered in the order as produced by
1157   the G.729 C code reference implementation.
1158
1159   The packing of the G.729 Annex B comfort noise frame is shown in Fig.
1160   5.
1161
1162          0                   1
1163          0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
1164         +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1165         |L|  LSF1   |  LSF2 |   GAIN  |R|
1166         |S|         |       |         |E|
1167         |F|         |       |         |S|
1168         |0|0 1 2 3 4|0 1 2 3|0 1 2 3 4|V|    RESV = Reserved (zero)
1169         +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1170
1171                       Figure 5: G.729 Annex B bit packing
1172
1173
1174
1175
1176
1177
1178Schulzrinne & Casner        Standards Track                    [Page 21]
1179
1180RFC 3551                    RTP A/V Profile                    July 2003
1181
1182
11834.5.7 G729D and G729E
1184
1185   Annexes D and E to ITU-T Recommendation G.729 provide additional data
1186   rates.  Because the data rate is not signaled in the bitstream, the
1187   different data rates are given distinct RTP encoding names which are
1188   mapped to distinct payload type numbers.  G729D indicates a 6.4
1189   kbit/s coding mode (G.729 Annex D, for momentary reduction in channel
1190   capacity), while G729E indicates an 11.8 kbit/s mode (G.729 Annex E,
1191   for improved performance with a wide range of narrow-band input
1192   signals, e.g., music and background noise).  Annex E has two
1193   operating modes, backward adaptive and forward adaptive, which are
1194   signaled by the first two bits in each frame (the most significant
1195   two bits of the first octet).
1196
1197   The voice activity detector (VAD) and comfort noise generator (CNG)
1198   algorithm specified in Annex B of G.729 may be used with Annex D and
1199   Annex E frames in addition to G.729 and G.729 Annex A frames.  The
1200   algorithm details for the operation of Annexes D and E with the Annex
1201   B CNG are specified in G.729 Annexes F and G.  Note that Annexes F
1202   and G do not introduce any new encodings.  Receivers MUST accept
1203   comfort noise frames if restriction of their use has not been
1204   signaled.  The MIME registrations for G729D and G729E in RFC 3555 [7]
1205   specify a parameter that MAY be used with MIME or SDP to restrict the
1206   use of comfort noise frames.
1207
1208   For G729D, an RTP packet may consist of zero or more G.729 Annex D
1209   frames, followed by zero or one G.729 Annex B frame.  Similarly, for
1210   G729E, an RTP packet may consist of zero or more G.729 Annex E
1211   frames, followed by zero or one G.729 Annex B frame.  The presence of
1212   a comfort noise frame can be deduced from the length of the RTP
1213   payload.
1214
1215   A single RTP packet must contain frames of only one data rate,
1216   optionally followed by one comfort noise frame.  The data rate may be
1217   changed from packet to packet by changing the payload type number.
1218   G.729 Annexes D, E and H describe what the encoding and decoding
1219   algorithms must do to accommodate a change in data rate.
1220
1221   For G729D, the bits of a G.729 Annex D frame are formatted as shown
1222   below in Fig. 6 (cf.  Table D.1/G.729).  The frame length is 64 bits.
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234Schulzrinne & Casner        Standards Track                    [Page 22]
1235
1236RFC 3551                    RTP A/V Profile                    July 2003
1237
1238
1239       0                   1                   2                   3
1240       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
1241      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1242      |L|      L1     |    L2   |    L3   |        P1     |     C1    |
1243      |0|             |         |         |               |           |
1244      | |0 1 2 3 4 5 6|0 1 2 3 4|0 1 2 3 4|0 1 2 3 4 5 6 7|0 1 2 3 4 5|
1245      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1246      | C1  |S1 | GA1 | GB1 |  P2   |        C2       |S2 | GA2 | GB2 |
1247      |     |   |     |     |       |                 |   |     |     |
1248      |6 7 8|0 1|0 1 2|0 1 2|0 1 2 3|0 1 2 3 4 5 6 7 8|0 1|0 1 2|0 1 2|
1249      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1250
1251                     Figure 6: G.729 Annex D bit packing
1252
1253   The net bit rate for the G.729 Annex E algorithm is 11.8 kbit/s and a
1254   total of 118 bits are used.  Two bits are appended as "don't care"
1255   bits to complete an integer number of octets for the frame.  For
1256   G729E, the bits of a data frame are formatted as shown in the next
1257   two diagrams (cf. Table E.1/G.729).  The fields for the G729E forward
1258   adaptive mode are packed as shown in Fig. 7.
1259
1260       0                   1                   2                   3
1261       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
1262      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1263      |0 0|L|      L1     |    L2   |    L3   |        P1     |P| C0_1|
1264      |   |0|             |         |         |               |0|     |
1265      |   | |0 1 2 3 4 5 6|0 1 2 3 4|0 1 2 3 4|0 1 2 3 4 5 6 7| |0 1 2|
1266      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1267      |       |   C1_1      |     C2_1    |   C3_1      |    C4_1     |
1268      |       |             |             |             |             |
1269      |3 4 5 6|0 1 2 3 4 5 6|0 1 2 3 4 5 6|0 1 2 3 4 5 6|0 1 2 3 4 5 6|
1270      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1271      | GA1 |  GB1  |    P2   |   C0_2      |     C1_2    |   C2_2    |
1272      |     |       |         |             |             |           |
1273      |0 1 2|0 1 2 3|0 1 2 3 4|0 1 2 3 4 5 6|0 1 2 3 4 5 6|0 1 2 3 4 5|
1274      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1275      | |    C3_2     |     C4_2    | GA2 | GB2   |DC |
1276      | |             |             |     |       |   |
1277      |6|0 1 2 3 4 5 6|0 1 2 3 4 5 6|0 1 2|0 1 2 3|0 1|
1278      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1279
1280         Figure 7: G.729 Annex E (forward adaptive mode) bit packing
1281
1282   The fields for the G729E backward adaptive mode are packed as shown
1283   in Fig. 8.
1284
1285
1286
1287
1288
1289
1290Schulzrinne & Casner        Standards Track                    [Page 23]
1291
1292RFC 3551                    RTP A/V Profile                    July 2003
1293
1294
1295       0                   1                   2                   3
1296       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
1297      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1298      |1 1|       P1      |P|       C0_1              |     C1_1      |
1299      |   |               |0|                    1 1 1|               |
1300      |   |0 1 2 3 4 5 6 7|0|0 1 2 3 4 5 6 7 8 9 0 1 2|0 1 2 3 4 5 6 7|
1301      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1302      |   |  C2_1       | C3_1        | C4_1        |GA1  | GB1   |P2 |
1303      |   |             |             |             |     |       |   |
1304      |8 9|0 1 2 3 4 5 6|0 1 2 3 4 5 6|0 1 2 3 4 5 6|0 1 2|0 1 2 3|0 1|
1305      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1306      |     |          C0_2           |       C1_2        |    C2_2   |
1307      |     |                    1 1 1|                   |           |
1308      |2 3 4|0 1 2 3 4 5 6 7 8 9 0 1 2|0 1 2 3 4 5 6 7 8 9|0 1 2 3 4 5|
1309      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1310      | |    C3_2     |     C4_2    | GA2 | GB2   |DC |
1311      | |             |             |     |       |   |
1312      |6|0 1 2 3 4 5 6|0 1 2 3 4 5 6|0 1 2|0 1 2 3|0 1|
1313      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
1314
1315         Figure 8: G.729 Annex E (backward adaptive mode) bit packing
1316
13174.5.8 GSM
1318
1319   GSM (Group Speciale Mobile) denotes the European GSM 06.10 standard
1320   for full-rate speech transcoding, ETS 300 961, which is based on
1321   RPE/LTP (residual pulse excitation/long term prediction) coding at a
1322   rate of 13 kb/s [11,12,13].  The text of the standard can be obtained
1323   from:
1324
1325   ETSI (European Telecommunications Standards Institute)
1326   ETSI Secretariat: B.P.152
1327   F-06561 Valbonne Cedex
1328   France
1329   Phone: +33 92 94 42 00
1330   Fax:   +33 93 65 47 16
1331
1332   Blocks of 160 audio samples are compressed into 33 octets, for an
1333   effective data rate of 13,200 b/s.
1334
13354.5.8.1  General Packaging Issues
1336
1337   The GSM standard (ETS 300 961) specifies the bit stream produced by
1338   the codec, but does not specify how these bits should be packed for
1339   transmission.  The packetization specified here has subsequently been
1340   adopted in ETSI Technical Specification TS 101 318.  Some software
1341   implementations of the GSM codec use a different packing than that
1342   specified here.
1343
1344
1345
1346Schulzrinne & Casner        Standards Track                    [Page 24]
1347
1348RFC 3551                    RTP A/V Profile                    July 2003
1349
1350
1351               field  field name  bits  field  field name  bits
1352               ________________________________________________
1353               1      LARc[0]     6     39     xmc[22]     3
1354               2      LARc[1]     6     40     xmc[23]     3
1355               3      LARc[2]     5     41     xmc[24]     3
1356               4      LARc[3]     5     42     xmc[25]     3
1357               5      LARc[4]     4     43     Nc[2]       7
1358               6      LARc[5]     4     44     bc[2]       2
1359               7      LARc[6]     3     45     Mc[2]       2
1360               8      LARc[7]     3     46     xmaxc[2]    6
1361               9      Nc[0]       7     47     xmc[26]     3
1362               10     bc[0]       2     48     xmc[27]     3
1363               11     Mc[0]       2     49     xmc[28]     3
1364               12     xmaxc[0]    6     50     xmc[29]     3
1365               13     xmc[0]      3     51     xmc[30]     3
1366               14     xmc[1]      3     52     xmc[31]     3
1367               15     xmc[2]      3     53     xmc[32]     3
1368               16     xmc[3]      3     54     xmc[33]     3
1369               17     xmc[4]      3     55     xmc[34]     3
1370               18     xmc[5]      3     56     xmc[35]     3
1371               19     xmc[6]      3     57     xmc[36]     3
1372               20     xmc[7]      3     58     xmc[37]     3
1373               21     xmc[8]      3     59     xmc[38]     3
1374               22     xmc[9]      3     60     Nc[3]       7
1375               23     xmc[10]     3     61     bc[3]       2
1376               24     xmc[11]     3     62     Mc[3]       2
1377               25     xmc[12]     3     63     xmaxc[3]    6
1378               26     Nc[1]       7     64     xmc[39]     3
1379               27     bc[1]       2     65     xmc[40]     3
1380               28     Mc[1]       2     66     xmc[41]     3
1381               29     xmaxc[1]    6     67     xmc[42]     3
1382               30     xmc[13]     3     68     xmc[43]     3
1383               31     xmc[14]     3     69     xmc[44]     3
1384               32     xmc[15]     3     70     xmc[45]     3
1385               33     xmc[16]     3     71     xmc[46]     3
1386               34     xmc[17]     3     72     xmc[47]     3
1387               35     xmc[18]     3     73     xmc[48]     3
1388               36     xmc[19]     3     74     xmc[49]     3
1389               37     xmc[20]     3     75     xmc[50]     3
1390               38     xmc[21]     3     76     xmc[51]     3
1391
1392                      Table 2: Ordering of GSM variables
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402Schulzrinne & Casner        Standards Track                    [Page 25]
1403
1404RFC 3551                    RTP A/V Profile                    July 2003
1405
1406
1407   Octet  Bit 0   Bit 1   Bit 2   Bit 3   Bit 4   Bit 5   Bit 6   Bit 7
1408   _____________________________________________________________________
1409       0    1       1       0       1    LARc0.0 LARc0.1 LARc0.2 LARc0.3
1410       1 LARc0.4 LARc0.5 LARc1.0 LARc1.1 LARc1.2 LARc1.3 LARc1.4 LARc1.5
1411       2 LARc2.0 LARc2.1 LARc2.2 LARc2.3 LARc2.4 LARc3.0 LARc3.1 LARc3.2
1412       3 LARc3.3 LARc3.4 LARc4.0 LARc4.1 LARc4.2 LARc4.3 LARc5.0 LARc5.1
1413       4 LARc5.2 LARc5.3 LARc6.0 LARc6.1 LARc6.2 LARc7.0 LARc7.1 LARc7.2
1414       5  Nc0.0   Nc0.1   Nc0.2   Nc0.3   Nc0.4   Nc0.5   Nc0.6  bc0.0
1415       6  bc0.1   Mc0.0   Mc0.1  xmaxc00 xmaxc01 xmaxc02 xmaxc03 xmaxc04
1416       7 xmaxc05 xmc0.0  xmc0.1  xmc0.2  xmc1.0  xmc1.1  xmc1.2  xmc2.0
1417       8 xmc2.1  xmc2.2  xmc3.0  xmc3.1  xmc3.2  xmc4.0  xmc4.1  xmc4.2
1418       9 xmc5.0  xmc5.1  xmc5.2  xmc6.0  xmc6.1  xmc6.2  xmc7.0  xmc7.1
1419      10 xmc7.2  xmc8.0  xmc8.1  xmc8.2  xmc9.0  xmc9.1  xmc9.2  xmc10.0
1420      11 xmc10.1 xmc10.2 xmc11.0 xmc11.1 xmc11.2 xmc12.0 xmc12.1 xcm12.2
1421      12  Nc1.0   Nc1.1   Nc1.2   Nc1.3   Nc1.4   Nc1.5   Nc1.6   bc1.0
1422      13  bc1.1   Mc1.0   Mc1.1  xmaxc10 xmaxc11 xmaxc12 xmaxc13 xmaxc14
1423      14 xmax15  xmc13.0 xmc13.1 xmc13.2 xmc14.0 xmc14.1 xmc14.2 xmc15.0
1424      15 xmc15.1 xmc15.2 xmc16.0 xmc16.1 xmc16.2 xmc17.0 xmc17.1 xmc17.2
1425      16 xmc18.0 xmc18.1 xmc18.2 xmc19.0 xmc19.1 xmc19.2 xmc20.0 xmc20.1
1426      17 xmc20.2 xmc21.0 xmc21.1 xmc21.2 xmc22.0 xmc22.1 xmc22.2 xmc23.0
1427      18 xmc23.1 xmc23.2 xmc24.0 xmc24.1 xmc24.2 xmc25.0 xmc25.1 xmc25.2
1428      19  Nc2.0   Nc2.1   Nc2.2   Nc2.3   Nc2.4   Nc2.5   Nc2.6   bc2.0
1429      20  bc2.1   Mc2.0   Mc2.1  xmaxc20 xmaxc21 xmaxc22 xmaxc23 xmaxc24
1430      21 xmaxc25 xmc26.0 xmc26.1 xmc26.2 xmc27.0 xmc27.1 xmc27.2 xmc28.0
1431      22 xmc28.1 xmc28.2 xmc29.0 xmc29.1 xmc29.2 xmc30.0 xmc30.1 xmc30.2
1432      23 xmc31.0 xmc31.1 xmc31.2 xmc32.0 xmc32.1 xmc32.2 xmc33.0 xmc33.1
1433      24 xmc33.2 xmc34.0 xmc34.1 xmc34.2 xmc35.0 xmc35.1 xmc35.2 xmc36.0
1434      25 Xmc36.1 xmc36.2 xmc37.0 xmc37.1 xmc37.2 xmc38.0 xmc38.1 xmc38.2
1435      26  Nc3.0   Nc3.1   Nc3.2   Nc3.3   Nc3.4   Nc3.5   Nc3.6   bc3.0
1436      27  bc3.1   Mc3.0   Mc3.1  xmaxc30 xmaxc31 xmaxc32 xmaxc33 xmaxc34
1437      28 xmaxc35 xmc39.0 xmc39.1 xmc39.2 xmc40.0 xmc40.1 xmc40.2 xmc41.0
1438      29 xmc41.1 xmc41.2 xmc42.0 xmc42.1 xmc42.2 xmc43.0 xmc43.1 xmc43.2
1439      30 xmc44.0 xmc44.1 xmc44.2 xmc45.0 xmc45.1 xmc45.2 xmc46.0 xmc46.1
1440      31 xmc46.2 xmc47.0 xmc47.1 xmc47.2 xmc48.0 xmc48.1 xmc48.2 xmc49.0
1441      32 xmc49.1 xmc49.2 xmc50.0 xmc50.1 xmc50.2 xmc51.0 xmc51.1 xmc51.2
1442
1443                        Table 3: GSM payload format
1444
1445   In the GSM packing used by RTP, the bits SHALL be packed beginning
1446   from the most significant bit.  Every 160 sample GSM frame is coded
1447   into one 33 octet (264 bit) buffer.  Every such buffer begins with a
1448   4 bit signature (0xD), followed by the MSB encoding of the fields of
1449   the frame.  The first octet thus contains 1101 in the 4 most
1450   significant bits (0-3) and the 4 most significant bits of F1 (0-3) in
1451   the 4 least significant bits (4-7).  The second octet contains the 2
1452   least significant bits of F1 in bits 0-1, and F2 in bits 2-7, and so
1453   on.  The order of the fields in the frame is described in Table 2.
1454
1455
1456
1457
1458Schulzrinne & Casner        Standards Track                    [Page 26]
1459
1460RFC 3551                    RTP A/V Profile                    July 2003
1461
1462
14634.5.8.2   GSM Variable Names and Numbers
1464
1465   In the RTP encoding we have the bit pattern described in Table 3,
1466   where F.i signifies the ith bit of the field F, bit 0 is the most
1467   significant bit, and the bits of every octet are numbered from 0 to 7
1468   from most to least significant.
1469
14704.5.9 GSM-EFR
1471
1472   GSM-EFR denotes GSM 06.60 enhanced full rate speech transcoding,
1473   specified in ETS 300 726 which is available from ETSI at the address
1474   given in Section 4.5.8.  This codec has a frame length of 244 bits.
1475   For transmission in RTP, each codec frame is packed into a 31 octet
1476   (248 bit) buffer beginning with a 4-bit signature 0xC in a manner
1477   similar to that specified here for the original GSM 06.10 codec.  The
1478   packing is specified in ETSI Technical Specification TS 101 318.
1479
14804.5.10 L8
1481
1482   L8 denotes linear audio data samples, using 8-bits of precision with
1483   an offset of 128, that is, the most negative signal is encoded as
1484   zero.
1485
14864.5.11 L16
1487
1488   L16 denotes uncompressed audio data samples, using 16-bit signed
1489   representation with 65,535 equally divided steps between minimum and
1490   maximum signal level, ranging from -32,768 to 32,767.  The value is
1491   represented in two's complement notation and transmitted in network
1492   byte order (most significant byte first).
1493
1494   The MIME registration for L16 in RFC 3555 [7] specifies parameters
1495   that MAY be used with MIME or SDP to indicate that analog pre-
1496   emphasis was applied to the signal before quantization or to indicate
1497   that a multiple-channel audio stream follows a different channel
1498   ordering convention than is specified in Section 4.1.
1499
15004.5.12 LPC
1501
1502   LPC designates an experimental linear predictive encoding contributed
1503   by Ron Frederick, which is based on an implementation written by Ron
1504   Zuckerman posted to the Usenet group comp.dsp on June 26, 1992.  The
1505   codec generates 14 octets for every frame.  The framesize is set to
1506   20 ms, resulting in a bit rate of 5,600 b/s.
1507
1508
1509
1510
1511
1512
1513
1514Schulzrinne & Casner        Standards Track                    [Page 27]
1515
1516RFC 3551                    RTP A/V Profile                    July 2003
1517
1518
15194.5.13 MPA
1520
1521   MPA denotes MPEG-1 or MPEG-2 audio encapsulated as elementary
1522   streams.  The encoding is defined in ISO standards ISO/IEC 11172-3
1523   and 13818-3.  The encapsulation is specified in RFC 2250 [14].
1524
1525   The encoding may be at any of three levels of complexity, called
1526   Layer I, II and III.  The selected layer as well as the sampling rate
1527   and channel count are indicated in the payload.  The RTP timestamp
1528   clock rate is always 90,000, independent of the sampling rate.
1529   MPEG-1 audio supports sampling rates of 32, 44.1, and 48 kHz (ISO/IEC
1530   11172-3, section 1.1; "Scope").  MPEG-2 supports sampling rates of
1531   16, 22.05 and 24 kHz.  The number of samples per frame is fixed, but
1532   the frame size will vary with the sampling rate and bit rate.
1533
1534   The MIME registration for MPA in RFC 3555 [7] specifies parameters
1535   that MAY be used with MIME or SDP to restrict the selection of layer,
1536   channel count, sampling rate, and bit rate.
1537
15384.5.14 PCMA and PCMU
1539
1540   PCMA and PCMU are specified in ITU-T Recommendation G.711.  Audio
1541   data is encoded as eight bits per sample, after logarithmic scaling.
1542   PCMU denotes mu-law scaling, PCMA A-law scaling.  A detailed
1543   description is given by Jayant and Noll [15].  Each G.711 octet SHALL
1544   be octet-aligned in an RTP packet.  The sign bit of each G.711 octet
1545   SHALL correspond to the most significant bit of the octet in the RTP
1546   packet (i.e., assuming the G.711 samples are handled as octets on the
1547   host machine, the sign bit SHALL be the most significant bit of the
1548   octet as defined by the host machine format).  The 56 kb/s and 48
1549   kb/s modes of G.711 are not applicable to RTP, since PCMA and PCMU
1550   MUST always be transmitted as 8-bit samples.
1551
1552   See Section 4.1 regarding silence suppression.
1553
15544.5.15 QCELP
1555
1556   The Electronic Industries Association (EIA) & Telecommunications
1557   Industry Association (TIA) standard IS-733, "TR45: High Rate Speech
1558   Service Option for Wideband Spread Spectrum Communications Systems",
1559   defines the QCELP audio compression algorithm for use in wireless
1560   CDMA applications.  The QCELP CODEC compresses each 20 milliseconds
1561   of 8,000 Hz, 16-bit sampled input speech into one of four different
1562   size output frames:  Rate 1 (266 bits), Rate 1/2 (124 bits), Rate 1/4
1563   (54 bits) or Rate 1/8 (20 bits).  For typical speech patterns, this
1564   results in an average output of 6.8 kb/s for normal mode and 4.7 kb/s
1565   for reduced rate mode.  The packetization of the QCELP audio codec is
1566   described in [16].
1567
1568
1569
1570Schulzrinne & Casner        Standards Track                    [Page 28]
1571
1572RFC 3551                    RTP A/V Profile                    July 2003
1573
1574
15754.5.16 RED
1576
1577   The redundant audio payload format "RED" is specified by RFC 2198
1578   [17].  It defines a means by which multiple redundant copies of an
1579   audio packet may be transmitted in a single RTP stream.  Each packet
1580   in such a stream contains, in addition to the audio data for that
1581   packetization interval, a (more heavily compressed) copy of the data
1582   from a previous packetization interval.  This allows an approximation
1583   of the data from lost packets to be recovered upon decoding of a
1584   subsequent packet, giving much improved sound quality when compared
1585   with silence substitution for lost packets.
1586
15874.5.17 VDVI
1588
1589   VDVI is a variable-rate version of DVI4, yielding speech bit rates of
1590   between 10 and 25 kb/s.  It is specified for single-channel operation
1591   only.  Samples are packed into octets starting at the most-
1592   significant bit.  The last octet is padded with 1 bits if the last
1593   sample does not fill the last octet.  This padding is distinct from
1594   the valid codewords.  The receiver needs to detect the padding
1595   because there is no explicit count of samples in the packet.
1596
1597   It uses the following encoding:
1598
1599            DVI4 codeword  VDVI bit pattern
1600            _______________________________
1601                        0  00
1602                        1  010
1603                        2  1100
1604                        3  11100
1605                        4  111100
1606                        5  1111100
1607                        6  11111100
1608                        7  11111110
1609                        8  10
1610                        9  011
1611                       10  1101
1612                       11  11101
1613                       12  111101
1614                       13  1111101
1615                       14  11111101
1616                       15  11111111
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626Schulzrinne & Casner        Standards Track                    [Page 29]
1627
1628RFC 3551                    RTP A/V Profile                    July 2003
1629
1630
16315.  Video
1632
1633   The following sections describe the video encodings that are defined
1634   in this memo and give their abbreviated names used for
1635   identification.  These video encodings and their payload types are
1636   listed in Table 5.
1637
1638   All of these video encodings use an RTP timestamp frequency of 90,000
1639   Hz, the same as the MPEG presentation time stamp frequency.  This
1640   frequency yields exact integer timestamp increments for the typical
1641   24 (HDTV), 25 (PAL), and 29.97 (NTSC) and 30 Hz (HDTV) frame rates
1642   and 50, 59.94 and 60 Hz field rates.  While 90 kHz is the RECOMMENDED
1643   rate for future video encodings used within this profile, other rates
1644   MAY be used.  However, it is not sufficient to use the video frame
1645   rate (typically between 15 and 30 Hz) because that does not provide
1646   adequate resolution for typical synchronization requirements when
1647   calculating the RTP timestamp corresponding to the NTP timestamp in
1648   an RTCP SR packet.  The timestamp resolution MUST also be sufficient
1649   for the jitter estimate contained in the receiver reports.
1650
1651   For most of these video encodings, the RTP timestamp encodes the
1652   sampling instant of the video image contained in the RTP data packet.
1653   If a video image occupies more than one packet, the timestamp is the
1654   same on all of those packets.  Packets from different video images
1655   are distinguished by their different timestamps.
1656
1657   Most of these video encodings also specify that the marker bit of the
1658   RTP header SHOULD be set to one in the last packet of a video frame
1659   and otherwise set to zero.  Thus, it is not necessary to wait for a
1660   following packet with a different timestamp to detect that a new
1661   frame should be displayed.
1662
16635.1  CelB
1664
1665   The CELL-B encoding is a proprietary encoding proposed by Sun
1666   Microsystems.  The byte stream format is described in RFC 2029 [18].
1667
16685.2 JPEG
1669
1670   The encoding is specified in ISO Standards 10918-1 and 10918-2.  The
1671   RTP payload format is as specified in RFC 2435 [19].
1672
16735.3 H261
1674
1675   The encoding is specified in ITU-T Recommendation H.261, "Video codec
1676   for audiovisual services at p x 64 kbit/s".  The packetization and
1677   RTP-specific properties are described in RFC 2032 [20].
1678
1679
1680
1681
1682Schulzrinne & Casner        Standards Track                    [Page 30]
1683
1684RFC 3551                    RTP A/V Profile                    July 2003
1685
1686
16875.4 H263
1688
1689   The encoding is specified in the 1996 version of ITU-T Recommendation
1690   H.263, "Video coding for low bit rate communication".  The
1691   packetization and RTP-specific properties are described in RFC 2190
1692   [21].  The H263-1998 payload format is RECOMMENDED over this one for
1693   use by new implementations.
1694
16955.5 H263-1998
1696
1697   The encoding is specified in the 1998 version of ITU-T Recommendation
1698   H.263, "Video coding for low bit rate communication".  The
1699   packetization and RTP-specific properties are described in RFC 2429
1700   [22].  Because the 1998 version of H.263 is a superset of the 1996
1701   syntax, this payload format can also be used with the 1996 version of
1702   H.263, and is RECOMMENDED for this use by new implementations.  This
1703   payload format does not replace RFC 2190, which continues to be used
1704   by existing implementations, and may be required for backward
1705   compatibility in new implementations.  Implementations using the new
1706   features of the 1998 version of H.263 MUST use the payload format
1707   described in RFC 2429.
1708
17095.6 MPV
1710
1711   MPV designates the use of MPEG-1 and MPEG-2 video encoding elementary
1712   streams as specified in ISO Standards ISO/IEC 11172 and 13818-2,
1713   respectively.  The RTP payload format is as specified in RFC 2250
1714   [14], Section 3.
1715
1716   The MIME registration for MPV in RFC 3555 [7] specifies a parameter
1717   that MAY be used with MIME or SDP to restrict the selection of the
1718   type of MPEG video.
1719
17205.7 MP2T
1721
1722   MP2T designates the use of MPEG-2 transport streams, for either audio
1723   or video.  The RTP payload format is described in RFC 2250 [14],
1724   Section 2.
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738Schulzrinne & Casner        Standards Track                    [Page 31]
1739
1740RFC 3551                    RTP A/V Profile                    July 2003
1741
1742
17435.8 nv
1744
1745   The encoding is implemented in the program `nv', version 4, developed
1746   at Xerox PARC by Ron Frederick.  Further information is available
1747   from the author:
1748
1749   Ron Frederick
1750   Blue Coat Systems Inc.
1751   650 Almanor Avenue
1752   Sunnyvale, CA 94085
1753   United States
1754   EMail: ronf@bluecoat.com
1755
17566.  Payload Type Definitions
1757
1758   Tables 4 and 5 define this profile's static payload type values for
1759   the PT field of the RTP data header.  In addition, payload type
1760   values in the range 96-127 MAY be defined dynamically through a
1761   conference control protocol, which is beyond the scope of this
1762   document.  For example, a session directory could specify that for a
1763   given session, payload type 96 indicates PCMU encoding, 8,000 Hz
1764   sampling rate, 2 channels.  Entries in Tables 4 and 5 with payload
1765   type "dyn" have no static payload type assigned and are only used
1766   with a dynamic payload type.  Payload type 2 was assigned to G721 in
1767   RFC 1890 and to its equivalent successor G726-32 in draft versions of
1768   this specification, but its use is now deprecated and that static
1769   payload type is marked reserved due to conflicting use for the
1770   payload formats G726-32 and AAL2-G726-32 (see Section 4.5.4).
1771   Payload type 13 indicates the Comfort Noise (CN) payload format
1772   specified in RFC 3389 [9].  Payload type 19 is marked "reserved"
1773   because some draft versions of this specification assigned that
1774   number to an earlier version of the comfort noise payload format.
1775   The payload type range 72-76 is marked "reserved" so that RTCP and
1776   RTP packets can be reliably distinguished (see Section "Summary of
1777   Protocol Constants" of the RTP protocol specification).
1778
1779   The payload types currently defined in this profile are assigned to
1780   exactly one of three categories or media types:  audio only, video
1781   only and those combining audio and video.  The media types are marked
1782   in Tables 4 and 5 as "A", "V" and "AV", respectively.  Payload types
1783   of different media types SHALL NOT be interleaved or multiplexed
1784   within a single RTP session, but multiple RTP sessions MAY be used in
1785   parallel to send multiple media types.  An RTP source MAY change
1786   payload types within the same media type during a session.  See the
1787   section "Multiplexing RTP Sessions" of RFC 3550 for additional
1788   explanation.
1789
1790
1791
1792
1793
1794Schulzrinne & Casner        Standards Track                    [Page 32]
1795
1796RFC 3551                    RTP A/V Profile                    July 2003
1797
1798
1799               PT   encoding    media type  clock rate   channels
1800                    name                    (Hz)
1801               ___________________________________________________
1802               0    PCMU        A            8,000       1
1803               1    reserved    A
1804               2    reserved    A
1805               3    GSM         A            8,000       1
1806               4    G723        A            8,000       1
1807               5    DVI4        A            8,000       1
1808               6    DVI4        A           16,000       1
1809               7    LPC         A            8,000       1
1810               8    PCMA        A            8,000       1
1811               9    G722        A            8,000       1
1812               10   L16         A           44,100       2
1813               11   L16         A           44,100       1
1814               12   QCELP       A            8,000       1
1815               13   CN          A            8,000       1
1816               14   MPA         A           90,000       (see text)
1817               15   G728        A            8,000       1
1818               16   DVI4        A           11,025       1
1819               17   DVI4        A           22,050       1
1820               18   G729        A            8,000       1
1821               19   reserved    A
1822               20   unassigned  A
1823               21   unassigned  A
1824               22   unassigned  A
1825               23   unassigned  A
1826               dyn  G726-40     A            8,000       1
1827               dyn  G726-32     A            8,000       1
1828               dyn  G726-24     A            8,000       1
1829               dyn  G726-16     A            8,000       1
1830               dyn  G729D       A            8,000       1
1831               dyn  G729E       A            8,000       1
1832               dyn  GSM-EFR     A            8,000       1
1833               dyn  L8          A            var.        var.
1834               dyn  RED         A                        (see text)
1835               dyn  VDVI        A            var.        1
1836
1837               Table 4: Payload types (PT) for audio encodings
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850Schulzrinne & Casner        Standards Track                    [Page 33]
1851
1852RFC 3551                    RTP A/V Profile                    July 2003
1853
1854
1855               PT      encoding    media type  clock rate
1856                       name                    (Hz)
1857               _____________________________________________
1858               24      unassigned  V
1859               25      CelB        V           90,000
1860               26      JPEG        V           90,000
1861               27      unassigned  V
1862               28      nv          V           90,000
1863               29      unassigned  V
1864               30      unassigned  V
1865               31      H261        V           90,000
1866               32      MPV         V           90,000
1867               33      MP2T        AV          90,000
1868               34      H263        V           90,000
1869               35-71   unassigned  ?
1870               72-76   reserved    N/A         N/A
1871               77-95   unassigned  ?
1872               96-127  dynamic     ?
1873               dyn     H263-1998   V           90,000
1874
1875               Table 5: Payload types (PT) for video and combined
1876                        encodings
1877
1878   Session participants agree through mechanisms beyond the scope of
1879   this specification on the set of payload types allowed in a given
1880   session.  This set MAY, for example, be defined by the capabilities
1881   of the applications used, negotiated by a conference control protocol
1882   or established by agreement between the human participants.
1883
1884   Audio applications operating under this profile SHOULD, at a minimum,
1885   be able to send and/or receive payload types 0 (PCMU) and 5 (DVI4).
1886   This allows interoperability without format negotiation and ensures
1887   successful negotiation with a conference control protocol.
1888
18897.  RTP over TCP and Similar Byte Stream Protocols
1890
1891   Under special circumstances, it may be necessary to carry RTP in
1892   protocols offering a byte stream abstraction, such as TCP, possibly
1893   multiplexed with other data.  The application MUST define its own
1894   method of delineating RTP and RTCP packets (RTSP [23] provides an
1895   example of such an encapsulation specification).
1896
18978.  Port Assignment
1898
1899   As specified in the RTP protocol definition, RTP data SHOULD be
1900   carried on an even UDP port number and the corresponding RTCP packets
1901   SHOULD be carried on the next higher (odd) port number.
1902
1903
1904
1905
1906Schulzrinne & Casner        Standards Track                    [Page 34]
1907
1908RFC 3551                    RTP A/V Profile                    July 2003
1909
1910
1911   Applications operating under this profile MAY use any such UDP port
1912   pair.  For example, the port pair MAY be allocated randomly by a
1913   session management program.  A single fixed port number pair cannot
1914   be required because multiple applications using this profile are
1915   likely to run on the same host, and there are some operating systems
1916   that do not allow multiple processes to use the same UDP port with
1917   different multicast addresses.
1918
1919   However, port numbers 5004 and 5005 have been registered for use with
1920   this profile for those applications that choose to use them as the
1921   default pair.  Applications that operate under multiple profiles MAY
1922   use this port pair as an indication to select this profile if they
1923   are not subject to the constraint of the previous paragraph.
1924   Applications need not have a default and MAY require that the port
1925   pair be explicitly specified.  The particular port numbers were
1926   chosen to lie in the range above 5000 to accommodate port number
1927   allocation practice within some versions of the Unix operating
1928   system, where port numbers below 1024 can only be used by privileged
1929   processes and port numbers between 1024 and 5000 are automatically
1930   assigned by the operating system.
1931
19329.  Changes from RFC 1890
1933
1934   This RFC revises RFC 1890.  It is mostly backwards-compatible with
1935   RFC 1890 except for functions removed because two interoperable
1936   implementations were not found.  The additions to RFC 1890 codify
1937   existing practice in the use of payload formats under this profile.
1938   Since this profile may be used without using any of the payload
1939   formats listed here, the addition of new payload formats in this
1940   revision does not affect backwards compatibility.  The changes are
1941   listed below, categorized into functional and non-functional changes.
1942
1943   Functional changes:
1944
1945   o  Section 11, "IANA Considerations" was added to specify the
1946      registration of the name for this profile.  That appendix also
1947      references a new Section 3 "Registering Additional Encodings"
1948      which establishes a policy that no additional registration of
1949      static payload types for this profile will be made beyond those
1950      added in this revision and included in Tables 4 and 5.  Instead,
1951      additional encoding names may be registered as MIME subtypes for
1952      binding to dynamic payload types.  Non-normative references were
1953      added to RFC 3555 [7] where MIME subtypes for all the listed
1954      payload formats are registered, some with optional parameters for
1955      use of the payload formats.
1956
1957
1958
1959
1960
1961
1962Schulzrinne & Casner        Standards Track                    [Page 35]
1963
1964RFC 3551                    RTP A/V Profile                    July 2003
1965
1966
1967   o  Static payload types 4, 16, 17 and 34 were added to incorporate
1968      IANA registrations made since the publication of RFC 1890, along
1969      with the corresponding payload format descriptions for G723 and
1970      H263.
1971
1972   o  Following working group discussion, static payload types 12 and 18
1973      were added along with the corresponding payload format
1974      descriptions for QCELP and G729.  Static payload type 13 was
1975      assigned to the Comfort Noise (CN) payload format defined in RFC
1976      3389.  Payload type 19 was marked reserved because it had been
1977      temporarily allocated to an earlier version of Comfort Noise
1978      present in some draft revisions of this document.
1979
1980   o  The payload format for G721 was renamed to G726-32 following the
1981      ITU-T renumbering, and the payload format description for G726 was
1982      expanded to include the -16, -24 and -40 data rates.  Because of
1983      confusion regarding draft revisions of this document, some
1984      implementations of these G726 payload formats packed samples into
1985      octets starting with the most significant bit rather than the
1986      least significant bit as specified here.  To partially resolve
1987      this incompatibility, new payload formats named AAL2-G726-16, -24,
1988      -32 and -40 will be specified in a separate document (see note in
1989      Section 4.5.4), and use of static payload type 2 is deprecated as
1990      explained in Section 6.
1991
1992   o  Payload formats G729D and G729E were added following the ITU-T
1993      addition of Annexes D and E to Recommendation G.729.  Listings
1994      were added for payload formats GSM-EFR, RED, and H263-1998
1995      published in other documents subsequent to RFC 1890.  These
1996      additional payload formats are referenced only by dynamic payload
1997      type numbers.
1998
1999   o  The descriptions of the payload formats for G722, G728, GSM, VDVI
2000      were expanded.
2001
2002   o  The payload format for 1016 audio was removed and its static
2003      payload type assignment 1 was marked "reserved" because two
2004      interoperable implementations were not found.
2005
2006   o  Requirements for congestion control were added in Section 2.
2007
2008   o  This profile follows the suggestion in the revised RTP spec that
2009      RTCP bandwidth may be specified separately from the session
2010      bandwidth and separately for active senders and passive receivers.
2011
2012   o  The mapping of a user pass-phrase string into an encryption key
2013      was deleted from Section 2 because two interoperable
2014      implementations were not found.
2015
2016
2017
2018Schulzrinne & Casner        Standards Track                    [Page 36]
2019
2020RFC 3551                    RTP A/V Profile                    July 2003
2021
2022
2023   o  The "quadrophonic" sample ordering convention for four-channel
2024      audio was removed to eliminate an ambiguity as noted in Section
2025      4.1.
2026
2027   Non-functional changes:
2028
2029   o  In Section 4.1, it is now explicitly stated that silence
2030      suppression is allowed for all audio payload formats.  (This has
2031      always been the case and derives from a fundamental aspect of
2032      RTP's design and the motivations for packet audio, but was not
2033      explicit stated before.)  The use of comfort noise is also
2034      explained.
2035
2036   o  In Section 4.1, the requirement level for setting of the marker
2037      bit on the first packet after silence for audio was changed from
2038      "is" to "SHOULD be", and clarified that the marker bit is set only
2039      when packets are intentionally not sent.
2040
2041   o  Similarly, text was added to specify that the marker bit SHOULD be
2042      set to one on the last packet of a video frame, and that video
2043      frames are distinguished by their timestamps.
2044
2045   o  RFC references are added for payload formats published after RFC
2046      1890.
2047
2048   o  The security considerations and full copyright sections were
2049      added.
2050
2051   o  According to Peter Hoddie of Apple, only pre-1994 Macintosh used
2052      the 22254.54 rate and none the 11127.27 rate, so the latter was
2053      dropped from the discussion of suggested sampling frequencies.
2054
2055   o  Table 1 was corrected to move some values from the "ms/packet"
2056      column to the "default ms/packet" column where they belonged.
2057
2058   o  Since the Interactive Multimedia Association ceased operations, an
2059      alternate resource was provided for a referenced IMA document.
2060
2061   o  A note has been added for G722 to clarify a discrepancy between
2062      the actual sampling rate and the RTP timestamp clock rate.
2063
2064   o  Small clarifications of the text have been made in several places,
2065      some in response to questions from readers.  In particular:
2066
2067      -  A definition for "media type" is given in Section 1.1 to allow
2068         the explanation of multiplexing RTP sessions in Section 6 to be
2069         more clear regarding the multiplexing of multiple media.
2070
2071
2072
2073
2074Schulzrinne & Casner        Standards Track                    [Page 37]
2075
2076RFC 3551                    RTP A/V Profile                    July 2003
2077
2078
2079      -  The explanation of how to determine the number of audio frames
2080         in a packet from the length was expanded.
2081
2082      -  More description of the allocation of bandwidth to SDES items
2083         is given.
2084
2085      -  A note was added that the convention for the order of channels
2086         specified in Section 4.1 may be overridden by a particular
2087         encoding or payload format specification.
2088
2089      -  The terms MUST, SHOULD, MAY, etc. are used as defined in RFC
2090         2119.
2091
2092   o  A second author for this document was added.
2093
209410. Security Considerations
2095
2096   Implementations using the profile defined in this specification are
2097   subject to the security considerations discussed in the RTP
2098   specification [1].  This profile does not specify any different
2099   security services.  The primary function of this profile is to list a
2100   set of data compression encodings for audio and video media.
2101
2102   Confidentiality of the media streams is achieved by encryption.
2103   Because the data compression used with the payload formats described
2104   in this profile is applied end-to-end, encryption may be performed
2105   after compression so there is no conflict between the two operations.
2106
2107   A potential denial-of-service threat exists for data encodings using
2108   compression techniques that have non-uniform receiver-end
2109   computational load.  The attacker can inject pathological datagrams
2110   into the stream which are complex to decode and cause the receiver to
2111   be overloaded.
2112
2113   As with any IP-based protocol, in some circumstances a receiver may
2114   be overloaded simply by the receipt of too many packets, either
2115   desired or undesired.  Network-layer authentication MAY be used to
2116   discard packets from undesired sources, but the processing cost of
2117   the authentication itself may be too high.  In a multicast
2118   environment, source pruning is implemented in IGMPv3 (RFC 3376) [24]
2119   and in multicast routing protocols to allow a receiver to select
2120   which sources are allowed to reach it.
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130Schulzrinne & Casner        Standards Track                    [Page 38]
2131
2132RFC 3551                    RTP A/V Profile                    July 2003
2133
2134
213511. IANA Considerations
2136
2137   The RTP specification establishes a registry of profile names for use
2138   by higher-level control protocols, such as the Session Description
2139   Protocol (SDP), RFC 2327 [6], to refer to transport methods.  This
2140   profile registers the name "RTP/AVP".
2141
2142   Section 3 establishes the policy that no additional registration of
2143   static RTP payload types for this profile will be made beyond those
2144   added in this document revision and included in Tables 4 and 5.  IANA
2145   may reference that section in declining to accept any additional
2146   registration requests.  In Tables 4 and 5, note that types 1 and 2
2147   have been marked reserved and the set of "dyn" payload types included
2148   has been updated.  These changes are explained in Sections 6 and 9.
2149
215012.  References
2151
215212.1 Normative References
2153
2154   [1]  Schulzrinne, H., Casner, S., Frederick, R. and V. Jacobson,
2155        "RTP:  A Transport Protocol for Real-Time Applications", RFC
2156        3550, July 2003.
2157
2158   [2]  Bradner, S., "Key Words for Use in RFCs to Indicate Requirement
2159        Levels", BCP 14, RFC 2119, March 1997.
2160
2161   [3]  Apple Computer, "Audio Interchange File Format AIFF-C", August
2162        1991.  (also ftp://ftp.sgi.com/sgi/aiff-c.9.26.91.ps.Z).
2163
216412.2 Informative References
2165
2166   [4]  Braden, R., Clark, D. and S. Shenker, "Integrated Services in
2167        the Internet Architecture: an Overview", RFC 1633, June 1994.
2168
2169   [5]  Blake, S., Black, D., Carlson, M., Davies, E., Wang, Z. and W.
2170        Weiss, "An Architecture for Differentiated Service", RFC 2475,
2171        December 1998.
2172
2173   [6]  Handley, M. and V. Jacobson, "SDP: Session Description
2174        Protocol", RFC 2327, April 1998.
2175
2176   [7]  Casner, S. and P. Hoschka, "MIME Type Registration of RTP
2177        Payload Types", RFC 3555, July 2003.
2178
2179   [8]  Freed, N., Klensin, J. and J. Postel, "Multipurpose Internet
2180        Mail Extensions (MIME) Part Four: Registration Procedures", BCP
2181        13, RFC 2048, November 1996.
2182
2183
2184
2185
2186Schulzrinne & Casner        Standards Track                    [Page 39]
2187
2188RFC 3551                    RTP A/V Profile                    July 2003
2189
2190
2191   [9]  Zopf, R., "Real-time Transport Protocol (RTP) Payload for
2192        Comfort Noise (CN)", RFC 3389, September 2002.
2193
2194   [10] Deleam, D. and J.-P. Petit, "Real-time implementations of the
2195        recent ITU-T low bit rate speech coders on the TI TMS320C54X
2196        DSP: results, methodology, and applications", in Proc. of
2197        International Conference on Signal Processing, Technology, and
2198        Applications (ICSPAT) , (Boston, Massachusetts), pp. 1656--1660,
2199        October 1996.
2200
2201   [11] Mouly, M. and M.-B. Pautet, The GSM system for mobile
2202        communications Lassay-les-Chateaux, France: Europe Media
2203        Duplication, 1993.
2204
2205   [12] Degener, J., "Digital Speech Compression", Dr. Dobb's Journal,
2206        December 1994.
2207
2208   [13] Redl, S., Weber, M. and M. Oliphant, An Introduction to GSM
2209        Boston: Artech House, 1995.
2210
2211   [14] Hoffman, D., Fernando, G., Goyal, V. and M. Civanlar, "RTP
2212        Payload Format for MPEG1/MPEG2 Video", RFC 2250, January 1998.
2213
2214   [15] Jayant, N. and P. Noll, Digital Coding of Waveforms--Principles
2215        and Applications to Speech and Video Englewood Cliffs, New
2216        Jersey: Prentice-Hall, 1984.
2217
2218   [16] McKay, K., "RTP Payload Format for PureVoice(tm) Audio", RFC
2219        2658, August 1999.
2220
2221   [17] Perkins, C., Kouvelas, I., Hodson, O., Hardman, V., Handley, M.,
2222        Bolot, J.-C., Vega-Garcia, A. and S. Fosse-Parisis, "RTP Payload
2223        for Redundant Audio Data", RFC 2198, September 1997.
2224
2225   [18] Speer, M. and D. Hoffman, "RTP Payload Format of Sun's CellB
2226        Video Encoding", RFC 2029, October 1996.
2227
2228   [19] Berc, L., Fenner, W., Frederick, R., McCanne, S. and P. Stewart,
2229        "RTP Payload Format for JPEG-Compressed Video", RFC 2435,
2230        October 1998.
2231
2232   [20] Turletti, T. and C. Huitema, "RTP Payload Format for H.261 Video
2233        Streams", RFC 2032, October 1996.
2234
2235   [21] Zhu, C., "RTP Payload Format for H.263 Video Streams", RFC 2190,
2236        September 1997.
2237
2238
2239
2240
2241
2242Schulzrinne & Casner        Standards Track                    [Page 40]
2243
2244RFC 3551                    RTP A/V Profile                    July 2003
2245
2246
2247   [22] Bormann, C., Cline, L., Deisher, G., Gardos, T., Maciocco, C.,
2248        Newell, D., Ott, J., Sullivan, G., Wenger, S. and C. Zhu, "RTP
2249        Payload Format for the 1998 Version of ITU-T Rec. H.263 Video
2250        (H.263+)", RFC 2429, October 1998.
2251
2252   [23] Schulzrinne, H., Rao, A. and R. Lanphier, "Real Time Streaming
2253        Protocol (RTSP)", RFC 2326, April 1998.
2254
2255   [24] Cain, B., Deering, S., Kouvelas, I., Fenner, B. and A.
2256        Thyagarajan, "Internet Group Management Protocol, Version 3",
2257        RFC 3376, October 2002.
2258
225913. Current Locations of Related Resources
2260
2261   Note:  Several sections below refer to the ITU-T Software Tool
2262   Library (STL).  It is available from the ITU Sales Service, Place des
2263   Nations, CH-1211 Geneve 20, Switzerland (also check
2264   http://www.itu.int).  The ITU-T STL is covered by a license defined
2265   in ITU-T Recommendation G.191, "Software tools for speech and audio
2266   coding standardization".
2267
2268   DVI4
2269
2270   An archived copy of the document IMA Recommended Practices for
2271   Enhancing Digital Audio Compatibility in Multimedia Systems (version
2272   3.0), which describes the IMA ADPCM algorithm, is available at:
2273
2274      http://www.cs.columbia.edu/~hgs/audio/dvi/
2275
2276   An implementation is available from Jack Jansen at
2277
2278      ftp://ftp.cwi.nl/local/pub/audio/adpcm.shar
2279
2280   G722
2281
2282   An implementation of the G.722 algorithm is available as part of the
2283   ITU-T STL, described above.
2284
2285   G723
2286
2287   The reference C code implementation defining the G.723.1 algorithm
2288   and its Annexes A, B, and C are available as an integral part of
2289   Recommendation G.723.1 from the ITU Sales Service, address listed
2290   above.  Both the algorithm and C code are covered by a specific
2291   license.  The ITU-T Secretariat should be contacted to obtain such
2292   licensing information.
2293
2294
2295
2296
2297
2298Schulzrinne & Casner        Standards Track                    [Page 41]
2299
2300RFC 3551                    RTP A/V Profile                    July 2003
2301
2302
2303   G726
2304
2305   G726 is specified in the ITU-T Recommendation G.726, "40, 32, 24, and
2306   16 kb/s Adaptive Differential Pulse Code Modulation (ADPCM)".  An
2307   implementation of the G.726 algorithm is available as part of the
2308   ITU-T STL, described above.
2309
2310   G729
2311
2312   The reference C code implementation defining the G.729 algorithm and
2313   its Annexes A through I are available as an integral part of
2314   Recommendation G.729 from the ITU Sales Service, listed above.  Annex
2315   I contains the integrated C source code for all G.729 operating
2316   modes.  The G.729 algorithm and associated C code are covered by a
2317   specific license.  The contact information for obtaining the license
2318   is available from the ITU-T Secretariat.
2319
2320   GSM
2321
2322   A reference implementation was written by Carsten Bormann and Jutta
2323   Degener (then at TU Berlin, Germany).  It is available at
2324
2325      http://www.dmn.tzi.org/software/gsm/
2326
2327   Although the RPE-LTP algorithm is not an ITU-T standard, there is a C
2328   code implementation of the RPE-LTP algorithm available as part of the
2329   ITU-T STL.  The STL implementation is an adaptation of the TU Berlin
2330   version.
2331
2332   LPC
2333
2334   An implementation is available at
2335
2336      ftp://parcftp.xerox.com/pub/net-research/lpc.tar.Z
2337
2338   PCMU, PCMA
2339
2340   An implementation of these algorithms is available as part of the
2341   ITU-T STL, described above.
2342
234314. Acknowledgments
2344
2345   The comments and careful review of Simao Campos, Richard Cox and AVT
2346   Working Group participants are gratefully acknowledged.  The GSM
2347   description was adopted from the IMTC Voice over IP Forum Service
2348   Interoperability Implementation Agreement (January 1997).  Fred Burg
2349   and Terry Lyons helped with the G.729 description.
2350
2351
2352
2353
2354Schulzrinne & Casner        Standards Track                    [Page 42]
2355
2356RFC 3551                    RTP A/V Profile                    July 2003
2357
2358
235915. Intellectual Property Rights Statement
2360
2361   The IETF takes no position regarding the validity or scope of any
2362   intellectual property or other rights that might be claimed to
2363   pertain to the implementation or use of the technology described in
2364   this document or the extent to which any license under such rights
2365   might or might not be available; neither does it represent that it
2366   has made any effort to identify any such rights.  Information on the
2367   IETF's procedures with respect to rights in standards-track and
2368   standards-related documentation can be found in BCP-11.  Copies of
2369   claims of rights made available for publication and any assurances of
2370   licenses to be made available, or the result of an attempt made to
2371   obtain a general license or permission for the use of such
2372   proprietary rights by implementors or users of this specification can
2373   be obtained from the IETF Secretariat.
2374
2375   The IETF invites any interested party to bring to its attention any
2376   copyrights, patents or patent applications, or other proprietary
2377   rights which may cover technology that may be required to practice
2378   this standard.  Please address the information to the IETF Executive
2379   Director.
2380
238116. Authors' Addresses
2382
2383   Henning Schulzrinne
2384   Department of Computer Science
2385   Columbia University
2386   1214 Amsterdam Avenue
2387   New York, NY 10027
2388   United States
2389
2390   EMail: schulzrinne@cs.columbia.edu
2391
2392
2393   Stephen L. Casner
2394   Packet Design
2395   3400 Hillview Avenue, Building 3
2396   Palo Alto, CA 94304
2397   United States
2398
2399   EMail: casner@acm.org
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410Schulzrinne & Casner        Standards Track                    [Page 43]
2411
2412RFC 3551                    RTP A/V Profile                    July 2003
2413
2414
241517. Full Copyright Statement
2416
2417   Copyright (C) The Internet Society (2003).  All Rights Reserved.
2418
2419   This document and translations of it may be copied and furnished to
2420   others, and derivative works that comment on or otherwise explain it
2421   or assist in its implementation may be prepared, copied, published
2422   and distributed, in whole or in part, without restriction of any
2423   kind, provided that the above copyright notice and this paragraph are
2424   included on all such copies and derivative works.  However, this
2425   document itself may not be modified in any way, such as by removing
2426   the copyright notice or references to the Internet Society or other
2427   Internet organizations, except as needed for the purpose of
2428   developing Internet standards in which case the procedures for
2429   copyrights defined in the Internet Standards process must be
2430   followed, or as required to translate it into languages other than
2431   English.
2432
2433   The limited permissions granted above are perpetual and will not be
2434   revoked by the Internet Society or its successors or assigns.
2435
2436   This document and the information contained herein is provided on an
2437   "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
2438   TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
2439   BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
2440   HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
2441   MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
2442
2443Acknowledgement
2444
2445   Funding for the RFC Editor function is currently provided by the
2446   Internet Society.
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466Schulzrinne & Casner        Standards Track                    [Page 44]
2467
2468