idnits 2.17.1 draft-ietf-payload-vp9-12.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (1 April 2021) is 1119 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) -- Possible downref: Non-RFC (?) normative reference: ref. 'VP9-BITSTREAM' Summary: 0 errors (**), 0 flaws (~~), 1 warning (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 AVTCore Working Group J. Uberti 3 Internet-Draft S. Holmer 4 Intended status: Standards Track M. Flodman 5 Expires: 3 October 2021 D. Hong 6 Google 7 J. Lennox 8 8x8 / Jitsi 9 1 April 2021 11 RTP Payload Format for VP9 Video 12 draft-ietf-payload-vp9-12 14 Abstract 16 This memo describes an RTP payload format for the VP9 video codec. 17 The payload format has wide applicability, as it supports 18 applications from low bit-rate peer-to-peer usage, to high bit-rate 19 video conferences. It includes provisions for temporal and spatial 20 scalability. 22 Status of This Memo 24 This Internet-Draft is submitted in full conformance with the 25 provisions of BCP 78 and BCP 79. 27 Internet-Drafts are working documents of the Internet Engineering 28 Task Force (IETF). Note that other groups may also distribute 29 working documents as Internet-Drafts. The list of current Internet- 30 Drafts is at https://datatracker.ietf.org/drafts/current/. 32 Internet-Drafts are draft documents valid for a maximum of six months 33 and may be updated, replaced, or obsoleted by other documents at any 34 time. It is inappropriate to use Internet-Drafts as reference 35 material or to cite them other than as "work in progress." 37 This Internet-Draft will expire on 3 October 2021. 39 Copyright Notice 41 Copyright (c) 2021 IETF Trust and the persons identified as the 42 document authors. All rights reserved. 44 This document is subject to BCP 78 and the IETF Trust's Legal 45 Provisions Relating to IETF Documents (https://trustee.ietf.org/ 46 license-info) in effect on the date of publication of this document. 47 Please review these documents carefully, as they describe your rights 48 and restrictions with respect to this document. Code Components 49 extracted from this document must include Simplified BSD License text 50 as described in Section 4.e of the Trust Legal Provisions and are 51 provided without warranty as described in the Simplified BSD License. 53 Table of Contents 55 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 56 2. Conventions, Definitions and Acronyms . . . . . . . . . . . . 3 57 3. Media Format Description . . . . . . . . . . . . . . . . . . 3 58 4. Payload Format . . . . . . . . . . . . . . . . . . . . . . . 5 59 4.1. RTP Header Usage . . . . . . . . . . . . . . . . . . . . 5 60 4.2. VP9 Payload Descriptor . . . . . . . . . . . . . . . . . 6 61 4.2.1. Scalability Structure (SS): . . . . . . . . . . . . . 11 62 4.3. Frame Fragmentation . . . . . . . . . . . . . . . . . . . 12 63 4.4. Scalable encoding considerations . . . . . . . . . . . . 13 64 4.5. Examples of VP9 RTP Stream . . . . . . . . . . . . . . . 13 65 4.5.1. Reference picture use for scalable structure . . . . 13 66 5. Feedback Messages and Header Extensions . . . . . . . . . . . 14 67 5.1. Reference Picture Selection Indication (RPSI) . . . . . . 14 68 5.2. Full Intra Request (FIR) . . . . . . . . . . . . . . . . 15 69 5.3. Layer Refresh Request (LRR) . . . . . . . . . . . . . . . 15 70 6. Payload Format Parameters . . . . . . . . . . . . . . . . . . 16 71 6.1. Media Type Definition . . . . . . . . . . . . . . . . . . 16 72 6.2. SDP Parameters . . . . . . . . . . . . . . . . . . . . . 19 73 6.2.1. Mapping of Media Subtype Parameters to SDP . . . . . 19 74 6.2.2. Offer/Answer Considerations . . . . . . . . . . . . . 20 75 7. Security Considerations . . . . . . . . . . . . . . . . . . . 20 76 8. Congestion Control . . . . . . . . . . . . . . . . . . . . . 21 77 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 21 78 10. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 21 79 11. References . . . . . . . . . . . . . . . . . . . . . . . . . 21 80 11.1. Normative References . . . . . . . . . . . . . . . . . . 21 81 11.2. Informative References . . . . . . . . . . . . . . . . . 23 82 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 23 84 1. Introduction 86 This memo describes an RTP payload specification applicable to the 87 transmission of video streams encoded using the VP9 video codec 88 [VP9-BITSTREAM]. The format described in this document can be used 89 both in peer-to-peer and video conferencing applications. 91 The VP9 video codec was developed by Google, and is the successor to 92 its earlier VP8 [RFC6386] codec. Above the compression improvements 93 and other general enhancements above VP8, VP9 is also designed in a 94 way that allows spatially-scalable video encoding. 96 2. Conventions, Definitions and Acronyms 98 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 99 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 100 "OPTIONAL" in this document are to be interpreted as described in BCP 101 14 [RFC2119] [RFC8174] when, and only when, they appear in all 102 capitals, as shown here. 104 3. Media Format Description 106 The VP9 codec can maintain up to eight reference frames, of which up 107 to three can be referenced by any new frame. 109 VP9 also allows a frame to use another frame of a different 110 resolution as a reference frame. (Specifically, a frame may use any 111 references whose width and height are between 1/16th that of the 112 current frame and twice that of the current frame, inclusive.) This 113 allows internal resolution changes without requiring the use of key 114 frames. 116 These features together enable an encoder to implement various forms 117 of coarse-grained scalability, including temporal, spatial and 118 quality scalability modes, as well as combinations of these, without 119 the need for explicit scalable coding tools. 121 Temporal layers define different frame rates of video; spatial and 122 quality layers define different and possibly dependent 123 representations of a single input frame. Spatial layers allow a 124 frame to be encoded at different resolutions, whereas quality layers 125 allow a frame to be encoded at the same resolution but at different 126 qualities (and thus with different amounts of coding error). VP9 127 supports quality layers as spatial layers without any resolution 128 changes; hereinafter, the term "spatial layer" is used to represent 129 both spatial and quality layers. 131 This payload format specification defines how such temporal and 132 spatial scalability layers can be described and communicated. 134 Temporal and spatial scalability layers are associated with non- 135 negative integer IDs. The lowest layer of either type has an ID of 136 0, and is sometimes referred to as the "base" temporal or spatial 137 layer. 139 Layers are designed (and MUST be encoded) such that if any layer, and 140 all higher layers, are removed from the bitstream along either of the 141 two dimensions, the remaining bitstream is still correctly decodable. 143 For terminology, this document uses the term "frame" to refer to a 144 single encoded VP9 frame for a particular resolution/quality, and 145 "picture" to refer to all the representations (frames) at a single 146 instant in time. A picture thus consists of one or more frames, 147 encoding different spatial layers. 149 Within a picture, a frame with spatial layer ID equal to SID, where 150 SID > 0, can depend on a frame of the same picture with a lower 151 spatial layer ID. This "inter-layer" dependency can result in 152 additional coding gain compared to the case where only traditional 153 "inter-picture" dependency is used, where a frame depends on 154 previously coded frame in time. For simplicity, this payload format 155 assumes that, within a picture and if inter-layer dependency is used, 156 a spatial layer SID frame can depend only on the immediately previous 157 spatial layer SID-1 frame, when S > 0. Additionally, if inter- 158 picture dependency is used, a spatial layer SID frame is assumed to 159 only depend on a previously coded spatial layer SID frame. 161 Given above simplifications for inter-layer and inter-picture 162 dependencies, a flag (the D bit described below) is used to indicate 163 whether a spatial layer SID frame depends on the spatial layer SID-1 164 frame. Given the D bit, a receiver only needs to additionally know 165 the inter-picture dependency structure for a given spatial layer 166 frame in order to determine its decodability. Two modes of 167 describing the inter-picture dependency structure are possible: 168 "flexible mode" and "non-flexible mode". An encoder can only switch 169 between the two on the first packet of a key frame with temporal 170 layer ID equal to 0. 172 In flexible mode, each packet can contain up to 3 reference indices, 173 which identify all frames referenced by the frame transmitted in the 174 current packet for inter-picture prediction. This (along with the D 175 bit) enables a receiver to identify if a frame is decodable or not 176 and helps it understand the temporal layer structure. Since this is 177 signaled in each packet it makes it possible to have very flexible 178 temporal layer hierarchies and patterns which are changing 179 dynamically. 181 In non-flexible mode, the inter-picture dependency (the reference 182 indices) of a Picture Group (PG) MUST be pre-specified as part of the 183 scalability structure (SS) data. In this mode, each packet has an 184 index to refer to one of the described pictures in the PG, from which 185 the pictures referenced by the picture transmitted in the current 186 packet for inter-picture prediction can be identified. 188 (Editor's Note: A "Picture Group", as used in this document, is not 189 the same thing as a the term "Group of Pictures" as it is 190 traditionally used in video coding, i.e. to mean an independently- 191 decoadable run of pictures beginning with a keyframe. Suggestions 192 for better terminology are welcome.) 194 The SS data can also be used to specify the resolution of each 195 spatial layer present in the VP9 stream for both flexible and non- 196 flexible modes. 198 4. Payload Format 200 This section describes how the encoded VP9 bitstream is encapsulated 201 in RTP. To handle network losses usage of RTP/AVPF [RFC4585] is 202 RECOMMENDED. All integer fields in the specifications are encoded as 203 unsigned integers in network octet order. 205 4.1. RTP Header Usage 207 The general RTP payload format for VP9 is depicted below. 209 0 1 2 3 210 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 1 211 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 212 |V=2|P|X| CC |M| PT | sequence number | 213 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 214 | timestamp | 215 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 216 | synchronization source (SSRC) identifier | 217 +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+ 218 | contributing source (CSRC) identifiers | 219 | .... | 220 +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+ 221 | VP9 payload descriptor (integer #octets) | 222 : : 223 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 224 | : VP9 pyld hdr | | 225 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | 226 | | 227 + | 228 : Bytes 2..N of VP9 payload : 229 | | 230 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 231 | : OPTIONAL RTP padding | 232 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 233 Figure 1 235 The VP9 payload descriptor will be described in Section 4.2; the VP9 236 payload header is described in [VP9-BITSTREAM]. OPTIONAL RTP padding 237 MUST NOT be included unless the P bit is set. The figure 238 specifically shows the format for the first packet in a frame. 239 Subsequent packets will not contain the VP9 payload header, and will 240 have later octets in the frame payload. 242 Marker bit (M): MUST be set to 1 for the final packet of the highest 243 spatial layer frame (the final packet of the picture), and 0 244 otherwise. Unless spatial scalability is in use for this picture, 245 this will have the same value as the E bit described below. Note 246 this bit MUST be set to 1 for the target spatial layer frame if a 247 stream is being rewritten to remove higher spatial layers. 249 Payload Type (PT): In line with the policy in Section 3 of 250 [RFC3551], applications using the VP9 RTP payload profile MUST 251 assign a dynamic payload type number to be used in each RTP 252 session and provide a mechanism to indicate the mapping. See 253 Section 6.2 for the mechanism to be used with the Session 254 Description Protocol (SDP) [RFC8866]. 256 Timestamp: The RTP timestamp indicates the time when the input frame 257 was sampled, at a clock rate of 90 kHz. If the input picture is 258 encoded with multiple layer frames, all of the frames of the 259 picture MUST have the same timestamp. 261 If a frame has the VP9 show_frame field set to 0 (i.e., it is 262 meant only to populate a reference buffer, without being output) 263 its timestamp MAY alternately be set to be the same as the 264 subsequent frame with show_frame equal to 1. (This will be 265 convenient for playing out pre-encoded content packaged with VP9 266 "superframes", which typically bundle show_frame==0 frames with a 267 subsequent show_frame==1 frame.) Every frame with show_frame==1, 268 however, MUST have a unique timestamp modulo the 2^32 wrap of the 269 field. 271 The remaining RTP Fixed Header Fields (V, P, X, CC, sequence number, 272 SSRC and CSRC identifiers) are used as specified in Section 5.1 of 273 [RFC3550]. 275 4.2. VP9 Payload Descriptor 277 In flexible mode (with the F bit below set to 1), The first octets 278 after the RTP header are the VP9 payload descriptor, with the 279 following structure. 281 0 1 2 3 4 5 6 7 282 +-+-+-+-+-+-+-+-+ 283 |I|P|L|F|B|E|V|Z| (REQUIRED) 284 +-+-+-+-+-+-+-+-+ 285 I: |M| PICTURE ID | (REQUIRED) 286 +-+-+-+-+-+-+-+-+ 287 M: | EXTENDED PID | (RECOMMENDED) 288 +-+-+-+-+-+-+-+-+ 289 L: | TID |U| SID |D| (CONDITIONALLY RECOMMENDED) 290 +-+-+-+-+-+-+-+-+ -\ 291 P,F: | P_DIFF |N| (CONDITIONALLY REQUIRED) - up to 3 times 292 +-+-+-+-+-+-+-+-+ -/ 293 V: | SS | 294 | .. | 295 +-+-+-+-+-+-+-+-+ 297 Figure 2 299 In non-flexible mode (with the F bit below set to 0), The first 300 octets after the RTP header are the VP9 payload descriptor, with the 301 following structure. 303 0 1 2 3 4 5 6 7 304 +-+-+-+-+-+-+-+-+ 305 |I|P|L|F|B|E|V|Z| (REQUIRED) 306 +-+-+-+-+-+-+-+-+ 307 I: |M| PICTURE ID | (RECOMMENDED) 308 +-+-+-+-+-+-+-+-+ 309 M: | EXTENDED PID | (RECOMMENDED) 310 +-+-+-+-+-+-+-+-+ 311 L: | TID |U| SID |D| (CONDITIONALLY RECOMMENDED) 312 +-+-+-+-+-+-+-+-+ 313 | TL0PICIDX | (CONDITIONALLY REQUIRED) 314 +-+-+-+-+-+-+-+-+ 315 V: | SS | 316 | .. | 317 +-+-+-+-+-+-+-+-+ 319 Figure 3 321 I: Picture ID (PID) present. When set to one, the OPTIONAL PID MUST 322 be present after the mandatory first octet and specified as below. 323 Otherwise, PID MUST NOT be present. If the SS field was present 324 in the stream's most recent start of a keyframe (i.e., non- 325 flexible scalability mode is in use), then the PID MUST also be 326 present in every packet. 328 P: Inter-picture predicted frame. When set to zero, the frame does 329 not utilize inter-picture prediction. In this case, up-switching 330 to a current spatial layer's frame is possible from directly lower 331 spatial layer frame. P SHOULD also be set to zero when encoding a 332 layer synchronization frame in response to an LRR 333 [I-D.ietf-avtext-lrr] message (see Section 5.3). When P is set to 334 zero, the TID field (described below) MUST also be set to 0 (if 335 present). Note that the P bit does not forbid intra-picture, 336 inter-layer prediction from earlier frames of the same picture, if 337 any. 339 L: Layer indices present. When set to one, the one or two octets 340 following the mandatory first octet and the PID (if present) is as 341 described by "Layer indices" below. If the F bit (described 342 below) is set to 1 (indicating flexible mode), then only one octet 343 is present for the layer indices. Otherwise if the F bit is set 344 to 0 (indicating non-flexible mode), then two octets are present 345 for the layer indices. 347 F: Flexible mode. F set to one indicates flexible mode and if the P 348 bit is also set to one, then the octets following the mandatory 349 first octet, the PID, and layer indices (if present) are as 350 described by "Reference indices" below. This MUST only be set to 351 1 if the I bit is also set to one; if the I bit is set to zero, 352 then this MUST also be set to zero and ignored by receivers. The 353 value of this F bit MUST only change on the first packet of a key 354 picture. A key picture is a picture whose base spatial layer 355 frame is a key frame, and which thus completely resets the encoder 356 state. This packet will have its P bit equal to zero, SID or D 357 bit (described below) equal to zero, and B bit (described below) 358 equal to 1. 360 B: Start of a frame. MUST be set to 1 if the first payload octet of 361 the RTP packet is the beginning of a new VP9 frame, and MUST NOT 362 be 1 otherwise. Note that this frame might not be the first frame 363 of a picture. 365 E: End of a frame. MUST be set to 1 for the final RTP packet of a 366 VP9 frame, and 0 otherwise. This enables a decoder to finish 367 decoding the frame, where it otherwise may need to wait for the 368 next packet to explicitly know that the frame is complete. Note 369 that, if spatial scalability is in use, more frames from the same 370 picture may follow; see the description of the M bit above. 372 V: Scalability structure (SS) data present. When set to one, the 373 OPTIONAL SS data MUST be present in the payload descriptor. 374 Otherwise, the SS data MUST NOT be present. 376 Z: Not a reference frame for upper spatial layers. If set to 1, 377 indicates that frames with higher spatial layers SID+1 of the 378 current and following pictures do not depend on the current 379 spatial layer SID frame. This enables a decoder which is 380 targeting a higher spatial layer to know that it can safely 381 discard this packet's frame without processing it, without having 382 to wait for the "D" bit in the higher-layer frame (see below). 384 The mandatory first octet is followed by the extension data fields 385 that are enabled: 387 M: The most significant bit of the first octet is an extension flag. 388 The field MUST be present if the I bit is equal to one. If set, 389 the PID field MUST contain 15 bits; otherwise, it MUST contain 7 390 bits. See PID below. 392 Picture ID (PID): Picture ID represented in 7 or 15 bits, depending 393 on the M bit. This is a running index of the pictures. The field 394 MUST be present if the I bit is equal to one. If M is set to 395 zero, 7 bits carry the PID; else if M is set to one, 15 bits carry 396 the PID in network byte order. The sender may choose between a 7- 397 or 15-bit index. The PID SHOULD start on a random number, and 398 MUST wrap after reaching the maximum ID. The receiver MUST NOT 399 assume that the number of bits in PID stay the same through the 400 session. 402 In the non-flexible mode (when the F bit is set to 0), this PID is 403 used as an index to the picture group (PG) specified in the SS 404 data below. In this mode, the PID of the key frame corresponds to 405 the first specified frame in the PG. Then subsequent PIDs are 406 mapped to subsequently specified frames in the PG (modulo N_G, 407 specified in the SS data below), respectively. 409 All frames of the same picture MUST have the same PID value. 411 Frames (and their corresponding pictures) with the VP9 show_frame 412 field equal to 0 MUST have distinct PID values from subsequent 413 pictures with show_frame equal to 1. Thus, a Picture as defined 414 in this specification is different than a VP9 Superframe. 416 All frames of the same picture MUST have the same value for 417 show_frame. 419 Layer indices: This information is optional but recommended whenever 420 encoding with layers. For both flexible and non-flexible modes, 421 one octet is used to specify a layer frame's temporal layer ID 422 (TID) and spatial layer ID (SID) as shown both in Figure 2 and 423 Figure 3. Additionally, a bit (U) is used to indicate that the 424 current frame is a "switching up point" frame. Another bit (D) is 425 used to indicate whether inter-layer prediction is used for the 426 current frame. 428 In the non-flexible mode (when the F bit is set to 0), another 429 octet is used to represent temporal layer 0 index (TL0PICIDX), as 430 depicted in Figure 3. The TL0PICIDX is present so that all 431 minimally required frames - the base temporal layer frames - can 432 be tracked. 434 The TID and SID fields indicate the temporal and spatial layers 435 and can help middleboxes and endpoints quickly identify which 436 layer a packet belongs to. 438 TID: The temporal layer ID of current frame. In the case of non- 439 flexible mode, if PID is mapped to a picture in a specified PG, 440 then the value of TID MUST match the corresponding TID value of 441 the mapped picture in the PG. 443 U: Switching up point. If this bit is set to 1 for the current 444 picture with temporal layer ID equal to TID, then "switch up" 445 to a higher frame rate is possible as subsequent higher 446 temporal layer pictures will not depend on any picture before 447 the current picture (in coding order) with temporal layer ID 448 greater than TID. 450 SID: The spatial layer ID of current frame. Note that frames 451 with spatial layer SID > 0 may be dependent on decoded spatial 452 layer SID-1 frame within the same picture. Different frames of 453 the same picture MUST have distinct spatial layer IDs, and 454 frames' spatial layers MUST appear in increasing order within 455 the frame. 457 D: Inter-layer dependency used. MUST be set to one if current 458 spatial layer SID frame depends on spatial layer SID-1 frame of 459 the same picture. MUST only be set to zero if current spatial 460 layer SID frame does not depend on spatial layer SID-1 frame of 461 the same picture. For the base layer frame (with SID equal to 462 0), this D bit MUST be set to zero. 464 TL0PICIDX: 8 bits temporal layer zero index. TL0PICIDX is only 465 present in the non-flexible mode (F = 0). This is a running 466 index for the temporal base layer pictures, i.e., the pictures 467 with TID set to 0. If TID is larger than 0, TL0PICIDX 468 indicates which temporal base layer picture the current picture 469 depends on. TL0PICIDX MUST be incremented when TID is equal to 470 0. The index SHOULD start on a random number, and MUST restart 471 at 0 after reaching the maximum number 255. 473 Reference indices: When P and F are both set to one, indicating a 474 non-key frame in flexible mode, then at least one reference index 475 has to be specified as below. Additional reference indices (total 476 of up to 3 reference indices are allowed) may be specified using 477 the N bit below. When either P or F is set to zero, then no 478 reference index is specified. 480 P_DIFF: The reference index (in 7 bits) specified as the relative 481 PID from the current picture. For example, when P_DIFF=3 on a 482 packet containing the picture with PID 112 means that the 483 picture refers back to the picture with PID 109. This 484 calculation is done modulo the size of the PID field, i.e., 485 either 7 or 15 bits. 487 N: 1 if there is additional P_DIFF following the current P_DIFF. 489 4.2.1. Scalability Structure (SS): 491 The scalability structure (SS) data describes the resolution of each 492 frame within a picture as well as the inter-picture dependencies for 493 a picture group (PG). If the VP9 payload descriptor's "V" bit is 494 set, the SS data is present in the position indicated in Figure 2 and 495 Figure 3. 497 +-+-+-+-+-+-+-+-+ 498 V: | N_S |Y|G|-|-|-| 499 +-+-+-+-+-+-+-+-+ -\ 500 Y: | WIDTH | (OPTIONAL) . 501 + + . 502 | | (OPTIONAL) . 503 +-+-+-+-+-+-+-+-+ . - N_S + 1 times 504 | HEIGHT | (OPTIONAL) . 505 + + . 506 | | (OPTIONAL) . 507 +-+-+-+-+-+-+-+-+ -/ 508 G: | N_G | (OPTIONAL) 509 +-+-+-+-+-+-+-+-+ -\ 510 N_G: | TID |U| R |-|-| (OPTIONAL) . 511 +-+-+-+-+-+-+-+-+ -\ . - N_G times 512 | P_DIFF | (OPTIONAL) . - R times . 513 +-+-+-+-+-+-+-+-+ -/ -/ 515 Figure 4 517 N_S: N_S + 1 indicates the number of spatial layers present in the 518 VP9 stream. 520 Y: Each spatial layer's frame resolution present. When set to one, 521 the OPTIONAL WIDTH (2 octets) and HEIGHT (2 octets) MUST be 522 present for each layer frame. Otherwise, the resolution MUST NOT 523 be present. 525 G: PG description present flag. 527 -: Bit reserved for future use. MUST be set to zero and MUST be 528 ignored by the receiver. 530 N_G: N_G indicates the number of pictures in a Picture Group (PG). 531 If N_G is greater than 0, then the SS data allows the inter- 532 picture dependency structure of the VP9 stream to be pre-declared, 533 rather than indicating it on the fly with every packet. If N_G is 534 greater than 0, then for N_G pictures in the PG, each picture's 535 temporal layer ID (TID), switch up point (U), and the R reference 536 indices (P_DIFFs) are specified. 538 The first picture specified in the PG MUST have TID set to 0. 540 G set to 0 or N_G set to 0 indicates that either there is only one 541 temporal layer or no fixed inter-picture dependency information is 542 present going forward in the bitstream. 544 Note that for a given picture, all frames follow the same inter- 545 picture dependency structure. However, the frame rate of each 546 spatial layer can be different from each other and this can be 547 controlled with the use of the D bit described above. The 548 specified dependency structure in the SS data MUST be for the 549 highest frame rate layer. 551 In a scalable stream sent with a fixed pattern, the SS data SHOULD be 552 included in the first packet of every key frame. This is a packet 553 with P bit equal to zero, SID or D bit equal to zero, and B bit equal 554 to 1. The SS data MUST only be changed on the picture that 555 corresponds to the first picture specified in the previous SS data's 556 PG (if the previous SS data's N_G was greater than 0). 558 4.3. Frame Fragmentation 560 VP9 frames are fragmented into packets, in RTP sequence number order, 561 beginning with a packet with the B bit set, and ending with a packet 562 with the E bit set. There is no mechanism for finer-grained access 563 to parts of a VP9 frame. 565 4.4. Scalable encoding considerations 567 In addition to the use of reference frames, VP9 has several 568 additional forms of inter-frame dependencies, largely involving 569 probability tables for the entropy and tree encoders. In VP9 syntax, 570 the syntax element "error_resilient_mode" resets this additional 571 inter-frame data, allowing a frame's syntax to be decoded 572 independently. 574 Due to the requirements of scalable streams, a VP9 encoder producing 575 a scalable stream needs to ensure that a frame does not depend on a 576 previous frame (of the same or a previous picture) that can 577 legitimately be removed from the stream. Thus, a frame that follows 578 a removable frame (in full decode order) MUST be encoded with 579 "error_resilient_mode" set to true. 581 For spatially-scalable streams, this means that 582 "error_resilient_mode" needs to be turned on for the base spatial 583 layer; it can however be turned off for higher spatial layers, 584 assuming they are sent with inter-layer dependency (i.e. with the "D" 585 bit set). For streams that are only temporally-scalable without 586 spatial scalability, "error_resilient_mode" can additionally be 587 turned off for any picture that immediately follows a temporal layer 588 0 frame. 590 4.5. Examples of VP9 RTP Stream 592 4.5.1. Reference picture use for scalable structure 594 As discussed in Section 3, the VP9 codec can maintain up to eight 595 reference frames, of which up to three can be referenced or updated 596 by any new frame. This section illustrates one way that a scalable 597 structure (with three spatial layers and three temporal layers) can 598 be constructed using these reference frames. 600 +==========+=========+============+=========+ 601 | Temporal | Spatial | References | Updates | 602 +==========+=========+============+=========+ 603 | 0 | 0 | 0 | 0 | 604 +----------+---------+------------+---------+ 605 | 0 | 1 | 0,1 | 1 | 606 +----------+---------+------------+---------+ 607 | 0 | 2 | 1,2 | 2 | 608 +----------+---------+------------+---------+ 609 | 2 | 0 | 0 | 6 | 610 +----------+---------+------------+---------+ 611 | 2 | 1 | 1,6 | 7 | 612 +----------+---------+------------+---------+ 613 | 2 | 2 | 2,7 | - | 614 +----------+---------+------------+---------+ 615 | 1 | 0 | 0 | 3 | 616 +----------+---------+------------+---------+ 617 | 1 | 1 | 1,3 | 4 | 618 +----------+---------+------------+---------+ 619 | 1 | 2 | 2,4 | 5 | 620 +----------+---------+------------+---------+ 621 | 2 | 0 | 3 | 6 | 622 +----------+---------+------------+---------+ 623 | 2 | 1 | 4,6 | 7 | 624 +----------+---------+------------+---------+ 625 | 2 | 2 | 5,7 | - | 626 +----------+---------+------------+---------+ 628 Table 1: Example scalability structure 630 This structure is constructed such that the "U" bit can always be 631 set. 633 5. Feedback Messages and Header Extensions 635 5.1. Reference Picture Selection Indication (RPSI) 637 The reference picture selection index is a payload-specific feedback 638 message defined within the RTCP-based feedback format. The RPSI 639 message is generated by a receiver and can be used in two ways. 640 Either it can signal a preferred reference picture when a loss has 641 been detected by the decoder -- preferably then a reference that the 642 decoder knows is perfect -- or, it can be used as positive feedback 643 information to acknowledge correct decoding of certain reference 644 pictures. The positive feedback method is useful for VP9 used for 645 point to point (unicast) communication. The use of RPSI for VP9 is 646 preferably combined with a special update pattern of the codec's two 647 special reference frames -- the golden frame and the altref frame -- 648 in which they are updated in an alternating leapfrog fashion. When a 649 receiver has received and correctly decoded a golden or altref frame, 650 and that frame had a PictureID in the payload descriptor, the 651 receiver can acknowledge this simply by sending an RPSI message back 652 to the sender. The message body (i.e., the "native RPSI bit string" 653 in [RFC4585]) is simply the PictureID of the received frame. 655 Note: because all frames of the same picture must have the same 656 inter-picture reference structure, there is no need for a message to 657 specify which frame is being selected. 659 5.2. Full Intra Request (FIR) 661 The Full Intra Request (FIR) [RFC5104] RTCP feedback message allows a 662 receiver to request a full state refresh of an encoded stream. 664 Upon receipt of an FIR request, a VP9 sender MUST send a picture with 665 a keyframe for its spatial layer 0 layer frame, and then send frames 666 without inter-picture prediction (P=0) for any higher layer frames. 668 5.3. Layer Refresh Request (LRR) 670 The Layer Refresh Request [I-D.ietf-avtext-lrr] allows a receiver to 671 request a single layer of a spatially or temporally encoded stream to 672 be refreshed, without necessarily affecting the stream's other 673 layers. 675 +---------------+---------------+ 676 |0|1|2|3|4|5|6|7|0|1|2|3|4|5|6|7| 677 +---------------+---------+-----+ 678 | RES | TID | RES | SID | 679 +---------------+---------+-----+ 681 Figure 5 683 Figure 5 shows the format of LRR's layer index fields for VP9 684 streams. The two "RES" fields MUST be set to 0 on transmission and 685 ingnored on reception. See Section 4.2 for details on the TID and 686 SID fields. 688 Identification of a layer refresh frame can be derived from the 689 reference IDs of each frame by backtracking the dependency chain 690 until reaching a point where only decodable frames are being 691 referenced. Therefore it's recommended for both the flexible and the 692 non-flexible mode that, when upgrade frames are being encoded in 693 response to a LRR, those packets should contain layer indices and the 694 reference fields so that the decoder or an MCU can make this 695 derivation. 697 Example: 699 LRR {1,0}, {2,1} is sent by an MCU when it is currently relaying 700 {1,0} to a receiver and which wants to upgrade to {2,1}. In response 701 the encoder should encode the next frames in layers {1,1} and {2,1} 702 by only referring to frames in {1,0}, or {0,0}. 704 In the non-flexible mode, periodic upgrade frames can be defined by 705 the layer structure of the SS, thus periodic upgrade frames can be 706 automatically identified by the picture ID. 708 6. Payload Format Parameters 710 This payload format has three optional parameters. 712 6.1. Media Type Definition 714 This registration is done using the template defined in [RFC6838] and 715 following [RFC4855]. 717 Type name: 718 video 720 Subtype name: 721 VP9 723 Required parameters: 724 None. 726 Optional parameters: 727 The max-fr and max-fs parameters are used to signal the 728 capabilities of a receiver implementation. If the implementation 729 is willing to receive media, both parameters MUST be provided. 730 These parameters MUST NOT be used for any other purpose. A media 731 sender SHOULD NOT send media with a frame rate or frame size 732 exceeding the max-fr and max-fs values signaled. (There may be 733 scenarios, such as pre-encoded media or selective forwarding 734 middleboxes [RFC7667], where a media sender does not have media 735 available that fits within a receivers max-fs and max-fr value; in 736 such scenarios, a sender MAY exceed the signaled values.) 738 max-fr: The value of max-fr is an integer indicating the maximum 739 frame rate in units of frames per second that the decoder is 740 capable of decoding. 742 max-fs: The value of max-fs is an integer indicating the maximum 743 frame size in units of macroblocks that the decoder is capable 744 of decoding. 746 The decoder is capable of decoding this frame size as long as 747 the width and height of the frame in macroblocks are less than 748 int(sqrt(max-fs * 8)) - for instance, a max-fs of 1200 (capable 749 of supporting 640x480 resolution) will support widths and 750 heights up to 1552 pixels (97 macroblocks). 752 profile-id: The value of profile-id is an integer indicating the 753 default coding profile, the subset of coding tools that may 754 have been used to generate the stream or that the receiver 755 supports). Table 2 lists all of the profiles defined in 756 section 7.2 of [VP9-BITSTREAM] and the corresponding integer 757 values to be used. 759 If no profile-id is present, Profile 0 MUST be inferred. (The 760 profile-id parameter was added relatively late in the 761 development of this specification, so some existing 762 implementations may not send it.) 764 Informative note: See Table 3 for capabilities of coding 765 profiles defined in section 7.2 of [VP9-BITSTREAM]. 767 Encoding considerations: 768 This media type is framed in RTP and contains binary data; see 769 Section 4.8 of [RFC6838]. 771 Security considerations: 772 See Section 7 of RFC xxxx. 774 [RFC Editor: Upon publication as an RFC, please replace "XXXX" 775 with the number assigned to this document and remove this note.] 777 Interoperability considerations: 778 None. 780 Published specification: 781 VP9 bitstream format [VP9-BITSTREAM] and RFC XXXX. 783 [RFC Editor: Upon publication as an RFC, please replace "XXXX" 784 with the number assigned to this document and remove this note.] 786 Applications which use this media type: 787 For example: Video over IP, video conferencing. 789 Fragment identifier considerations: 790 N/A. 792 Additional information: 793 None. 795 Person & email address to contact for further information: 796 Jonathan Lennox 798 Intended usage: 799 COMMON 801 Restrictions on usage: 802 This media type depends on RTP framing, and hence is only defined 803 for transfer via RTP [RFC3550]. 805 Author: 806 Jonathan Lennox 808 Change controller: 809 IETF AVTCore Working Group delegated from the IESG. 811 +=========+============+ 812 | Profile | profile-id | 813 +=========+============+ 814 | 0 | 0 | 815 +---------+------------+ 816 | 1 | 1 | 817 +---------+------------+ 818 | 2 | 2 | 819 +---------+------------+ 820 | 3 | 3 | 821 +---------+------------+ 823 Table 2: Table 1. 824 Table of profile-id 825 integer values 826 representing the VP9 827 profile corresponding 828 to the set of coding 829 tools supported. 831 +=========+===========+=================+==========================+ 832 | Profile | Bit Depth | SRGB Colorspace | Chroma Subsampling | 833 +=========+===========+=================+==========================+ 834 | 0 | 8 | No | YUV 4:2:0 | 835 +---------+-----------+-----------------+--------------------------+ 836 | 1 | 8 | Yes | YUV 4:2:0,4:4:0 or 4:4:4 | 837 +---------+-----------+-----------------+--------------------------+ 838 | 2 | 10 or 12 | No | YUV 4:2:0 | 839 +---------+-----------+-----------------+--------------------------+ 840 | 3 | 10 or 12 | Yes | YUV 4:2:0,4:4:0 or 4:4:4 | 841 +---------+-----------+-----------------+--------------------------+ 843 Table 3: Table 2. Table of profile capabilities. 845 6.2. SDP Parameters 847 The receiver MUST ignore any fmtp parameter unspecified in this memo. 849 6.2.1. Mapping of Media Subtype Parameters to SDP 851 The media type video/VP9 string is mapped to fields in the Session 852 Description Protocol (SDP) [RFC8866] as follows: 854 * The media name in the "m=" line of SDP MUST be video. 856 * The encoding name in the "a=rtpmap" line of SDP MUST be VP9 (the 857 media subtype). 859 * The clock rate in the "a=rtpmap" line MUST be 90000. 861 * The parameters "max-fr" and "max-fs" MUST be included in the 862 "a=fmtp" line of SDP if the receiver wishes to declare its 863 receiver capabilities. These parameters are expressed as a media 864 subtype string, in the form of a semicolon separated list of 865 parameter=value pairs. 867 * The OPTIONAL parameter profile-id, when present, SHOULD be 868 included in the "a=fmtp" line of SDP. This parameter is expressed 869 as a media subtype string, in the form of a parameter=value pair. 870 When the parameter is not present, a value of 0 MUST be inferred 871 for profile-id. 873 6.2.1.1. Example 875 An example of media representation in SDP is as follows: 877 m=video 49170 RTP/AVPF 98 878 a=rtpmap:98 VP9/90000 879 a=fmtp:98 max-fr=30;max-fs=3600;profile-id=0 881 6.2.2. Offer/Answer Considerations 883 When VP9 is offered over RTP using SDP in an Offer/Answer model 884 [RFC3264] for negotiation for unicast usage, the following 885 limitations and rules apply: 887 * The parameter identifying a media format configuration for VP9 is 888 profile-id. This media format configuration parameter MUST be 889 used symmetrically; that is, the answerer MUST either maintain 890 this configuration parameter or remove the media format (payload 891 type) completely if it is not supported. 893 * The max-fr and max-fs parameters are used declaratively to 894 describe receiver capabilities, even in the Offer/Answer model. 895 The values in an answer are used to describe the answerer's 896 capabilities, and thus their values are set independently of the 897 values in the offer. 899 * To simplify the handling and matching of these configurations, the 900 same RTP payload type number used in the offer SHOULD also be used 901 in the answer and in a subsequent offer, as specified in 902 [RFC3264]. An answer or subsequent offer MUST NOT contain the 903 payload type number used in the offer unless the profile-id value 904 is exactly the same as in the original offer. However, max-fr and 905 max-fs parameters MAY be changed in subsequent offers and answers, 906 with the same payload type number, if an endpoint wishes to change 907 its declared receiver capabilties. 909 7. Security Considerations 911 RTP packets using the payload format defined in this specification 912 are subject to the security considerations discussed in the RTP 913 specification [RFC3550], and in any applicable RTP profile such as 914 RTP/AVP [RFC3551], RTP/AVPF [RFC4585], RTP/SAVP [RFC3711], or RTP/ 915 SAVPF [RFC5124]. SAVPF [RFC5124]. However, as "Securing the RTP 916 Protocol Framework: Why RTP Does Not Mandate a Single Media Security 917 Solution" [RFC7202] discusses, it is not an RTP payload format's 918 responsibility to discuss or mandate what solutions are used to meet 919 the basic security goals like confidentiality, integrity and source 920 authenticity for RTP in general. This responsibility lays on anyone 921 using RTP in an application. They can find guidance on available 922 security mechanisms in Options for Securing RTP Sessions [RFC7201]. 923 Applications SHOULD use one or more appropriate strong security 924 mechanisms. The rest of this security consideration section 925 discusses the security impacting properties of the payload format 926 itself. 928 This RTP payload format and its media decoder do not exhibit any 929 significant non-uniformity in the receiver-side computational 930 complexity for packet processing, and thus are unlikely to pose a 931 denial-of-service threat due to the receipt of pathological data. 932 Nor does the RTP payload format contain any active content. 934 8. Congestion Control 936 Congestion control for RTP SHALL be used in accordance with RFC 3550 937 [RFC3550], and with any applicable RTP profile; e.g., RFC 3551 938 [RFC3551]. The congestion control mechanism can, in a real-time 939 encoding scenario, adapt the transmission rate by instructing the 940 encoder to encode at a certain target rate. Media aware network 941 elements MAY use the information in the VP9 payload descriptor in 942 Section 4.2 to identify non-reference frames and discard them in 943 order to reduce network congestion. Note that discarding of non- 944 reference frames cannot be done if the stream is encrypted (because 945 the non-reference marker is encrypted). 947 9. IANA Considerations 949 The IANA is requested to register the media type registration "video/ 950 vp9" as specified in Section 6.1. The media type is also requested 951 to be added to the IANA registry for "RTP Payload Format MIME types" 952 . 954 10. Acknowledgments 956 Alex Eleftheriadis, Yuki Ito, Won Kap Jang, Sergio Garcia Murillo, 957 Roi Sasson, Timothy Terriberry, Emircan Uysaler, and Thomas Volkert 958 commented on the development of this document and provided helpful 959 comments and feedback. 961 11. References 963 11.1. Normative References 965 [I-D.ietf-avtext-lrr] 966 Lennox, J., Hong, D., Uberti, J., Holmer, S., and M. 967 Flodman, "The Layer Refresh Request (LRR) RTCP Feedback 968 Message", Work in Progress, Internet-Draft, draft-ietf- 969 avtext-lrr-07, 2 July 2017, . 972 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 973 Requirement Levels", BCP 14, RFC 2119, 974 DOI 10.17487/RFC2119, March 1997, 975 . 977 [RFC3264] Rosenberg, J. and H. Schulzrinne, "An Offer/Answer Model 978 with Session Description Protocol (SDP)", RFC 3264, 979 DOI 10.17487/RFC3264, June 2002, 980 . 982 [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. 983 Jacobson, "RTP: A Transport Protocol for Real-Time 984 Applications", STD 64, RFC 3550, DOI 10.17487/RFC3550, 985 July 2003, . 987 [RFC4585] Ott, J., Wenger, S., Sato, N., Burmeister, C., and J. Rey, 988 "Extended RTP Profile for Real-time Transport Control 989 Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585, 990 DOI 10.17487/RFC4585, July 2006, 991 . 993 [RFC4855] Casner, S., "Media Type Registration of RTP Payload 994 Formats", RFC 4855, DOI 10.17487/RFC4855, February 2007, 995 . 997 [RFC5104] Wenger, S., Chandra, U., Westerlund, M., and B. Burman, 998 "Codec Control Messages in the RTP Audio-Visual Profile 999 with Feedback (AVPF)", RFC 5104, DOI 10.17487/RFC5104, 1000 February 2008, . 1002 [RFC6838] Freed, N., Klensin, J., and T. Hansen, "Media Type 1003 Specifications and Registration Procedures", BCP 13, 1004 RFC 6838, DOI 10.17487/RFC6838, January 2013, 1005 . 1007 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 1008 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 1009 May 2017, . 1011 [RFC8866] Begen, A., Kyzivat, P., Perkins, C., and M. Handley, "SDP: 1012 Session Description Protocol", RFC 8866, 1013 DOI 10.17487/RFC8866, January 2021, 1014 . 1016 [VP9-BITSTREAM] 1017 Grange, A., de Rivaz, P., and J. Hunt, "VP9 Bitstream & 1018 Decoding Process Specification", Version 0.6, 31 March 1019 2016, 1020 . 1024 11.2. Informative References 1026 [RFC3551] Schulzrinne, H. and S. Casner, "RTP Profile for Audio and 1027 Video Conferences with Minimal Control", STD 65, RFC 3551, 1028 DOI 10.17487/RFC3551, July 2003, 1029 . 1031 [RFC3711] Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K. 1032 Norrman, "The Secure Real-time Transport Protocol (SRTP)", 1033 RFC 3711, DOI 10.17487/RFC3711, March 2004, 1034 . 1036 [RFC5124] Ott, J. and E. Carrara, "Extended Secure RTP Profile for 1037 Real-time Transport Control Protocol (RTCP)-Based Feedback 1038 (RTP/SAVPF)", RFC 5124, DOI 10.17487/RFC5124, February 1039 2008, . 1041 [RFC6386] Bankoski, J., Koleszar, J., Quillio, L., Salonen, J., 1042 Wilkins, P., and Y. Xu, "VP8 Data Format and Decoding 1043 Guide", RFC 6386, DOI 10.17487/RFC6386, November 2011, 1044 . 1046 [RFC7201] Westerlund, M. and C. Perkins, "Options for Securing RTP 1047 Sessions", RFC 7201, DOI 10.17487/RFC7201, April 2014, 1048 . 1050 [RFC7202] Perkins, C. and M. Westerlund, "Securing the RTP 1051 Framework: Why RTP Does Not Mandate a Single Media 1052 Security Solution", RFC 7202, DOI 10.17487/RFC7202, April 1053 2014, . 1055 [RFC7667] Westerlund, M. and S. Wenger, "RTP Topologies", RFC 7667, 1056 DOI 10.17487/RFC7667, November 2015, 1057 . 1059 Authors' Addresses 1061 Justin Uberti 1062 Google, Inc. 1063 747 6th Street South 1064 Kirkland, WA 98033 1065 United States of America 1067 Email: justin@uberti.name 1068 Stefan Holmer 1069 Google, Inc. 1070 Kungsbron 2 1071 SE-111 22 Stockholm 1072 Sweden 1074 Email: holmer@google.com 1076 Magnus Flodman 1077 Google, Inc. 1078 Kungsbron 2 1079 SE-111 22 Stockholm 1080 Sweden 1082 Email: mflodman@google.com 1084 Danny Hong 1085 Google, Inc. 1086 1585 Charleston Road 1087 Mountain View, CA 94043 1088 United States of America 1090 Email: dannyhong@google.com 1092 Jonathan Lennox 1093 8x8, Inc. / Jitsi 1094 1350 Broadway 1095 New York, NY 10018 1096 United States of America 1098 Email: jonathan.lennox@8x8.com