1、INTERNATIONAL TELECOMMUN CATION UN ION ITU-T TELECOMMUNICATION STAN DA RD I ZATI ON S ECTO R OF ITU J.57 (ex CMTT.724) (06/90) TELEVISION AND SOUND TRANSMISSION TRANSMISSION OF DIGITAL STUDIO QUALITY SOUND SIGNALS OVER HI CHANNELS ITU-T Recommendation J.57 (Formerly Recommendation ITU-R CMTT.724) FO
2、REWORD The ITU Telecommunication Standardization Sector (ITU-T) is a permanent organ of the International Telecom- munication Union. The ITU-T is responsible for studying technical, operating and tariff questions and issuing Recommendations on them with a view to standardizing telecommunications on
3、a worldwide basis. The World Telecommunication Standardization Conference (WTSC), which meets every four years, established the topics for study by the ITU-T Study Groups which, in their turn, produce Recommendations on these topics. ITU-T Recommendation 5.57 (formerly Recommendation ITU-R CMTT.724)
4、 was elaborated by the former ITU-R Study Group CMTT. See Note 1 below. NOTES 1 As a consequence of a reform process within the International Telecommunication Union (ITU), the CCITT ceased to exist as of 28 February 1993. In its place, the ITU Telecommunication Standardization Sector (ITU-T) was cr
5、eated as of 1 March 1993. Similarly, in this reform process, the CCIR and the IFRB have been replaced by the Radiocommunication Sector (ITU-R). Conforming to a joint decision by the World Telecommunication Standardization Conference (Helsinki, March 1993) and the Radiocommunication Assembly (Geneva,
6、 November 1993), the ITU-R Study Group CMTT was transferred to ITU-T as Study Group 9, except for the satellite news gathering (SNG) study area which was transferred to ITU-R Study Group 4. 2 In this Recommendation, the expression “Administration” is used for conciseness to indicate both a telecommu
7、nication administration and a recognized operating agency. O ITU 1990 All rights reserved. No part of this publication may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying and microfilm, without permission in writing from the ITU. Recommendation
8、 5.57l) TRANSMISSION OF DIGITAL STUDIO QUALITY SOUND SIGNALS OVER H1 CHANNELS*) (1990) The CCIR, CONSIDERING that the source encoding for digital sound signals in broadcasting studios is given in Recommendation 646; that the 2-channel digital audio interface is specified in Recommendation 647; that
9、a format for digital studio-quality sound signal connections should be based on these Recommendations; that the sound quality and the auxiliary information carried over the Recommendation 647 interface should be that the hierarchical bit rates for digital networks are given in CCITT Recommendation G
10、.702; that a digital hierarchy for interworking between networks using different transmission hierarchies is given in (a) (6) (e) (4 maintained as far as possible; (e) (J3 CCITT Recommandation G.802; (d (h) that the access levels for the H1 channels in the ISDN are given in CCITT Recommendation 1.41
11、2; that simple interworking between the different hierarchies is necessary; that account must be taken of network impairments such as signal bit errors, error bursts and controlled slips; that, for some applications, the introduction of excessive delay can cause operational problems, (k) RECOMMENDS
12、that for transmission of digital studio-quality sound signals the coding and multiplexing format given in Annex I should be used. ANNEX I CODING AND MULTIPLEXING FORMAT 1. Introduction which should be read in conjunction with this text. contribution format: - Preamble (bit 0-3): not transmitted - Bi
13、t4-7: not transmitted - Audio sample word (bit 8-27): companded - Validity flag (bit 28): not transmitted - User data (bit 29): transmitted transparently in H12 only - Channel status (bit 30): subjected to data compression - Parity bit (bit 31): The transmission format is based on the 2-channel digi
14、tal audio interface described in Recommendation 647, The 32 bits in each sub-frame of the Recommendation 647 interface are treated in the following way in the not transmitted. Replaced by a parity bit for the audio sample word. Formerly Recommendation lTU-R CMTT.724. *) The United States of America
15、withholds its approval of this Recommendation. Recommendation 5.57 (06/90) 1 2. Compatibility between H11 and H12-based systems The H12 level provides a total of 20 bits per sample, and H11 provides a total of 16 bits per sample. To simplify interworking between H11 and H12 channels, the companding
16、of the audio signal is such that the samples are compressed for transmission in the H11 channel. In the H12 channel, extra bits may be conveyed, to improve resolution of the audio coding and provide a user data channel. The essential data occupies the entire available capacity of the H11 channel, an
17、d the first 24 available octets of each frame of the H12 channel. 3. Coding law Near instantaneous companding from 20 to 15 bitshample is applied. A 1 ms companding block is used with 8 coding ranges. The coding table is shown in Fig. la and the transmitted bits in Fig. lb. Unused bits are set to 1
18、(one). The 1 ms companding block introduces a fundamental delay of 2 ms per codec. In practice, the total delay will be slightly longer. 4. Sample error detection One parity bit is applied to the 7 most significant bits of each transmitted sound sample such that the parity group is odd. 5. Sample in
19、terleaving The companding block contains 96 sound samples (48 from each sound signal). The sound samples within the companding block are organized in 8 successive frames complying with CCITT Recommendation G.704. Each frame contains 6 samples from each sound signal, with the associated parity bits.
20、Adjacent sound samples within the companding block are separated by four Recommendation G.704 frames. This is shown diagrammatically in Fig. 2. The first four frames carry all the odd-numbered samples from both sound signals; the second four frames carry all the even numbered samples. In the event o
21、f an error burst corrupting consecutive bits for a period equivalent to up to four frames, the erroneous samples should be concealed by interpolation between adjacent samples (from part of the block unaffected by the error burst) . 6. Signalling in parity Chambers, 19851 There are 96 parity bits per
22、 1 ms companding block. Some extra data bits are carried by modifying the parity bits as follows: 6.1 Scale factor transmission Each bit of each 3-bit scale factor word is carried in the parity of 8-sound samples defined in Q 6.4, according to the following rule: a scale factor bit which is ?O? caus
23、es the parity of the 8 samples to be unchanged; a scale factor bit which is ?1? modifies the parity. At the decoder, a majority decision process is used to determine the scale factor bits and restore the original parity bit. The samples are then checked for errors in the normal way. 6.2 Channel stat
24、us Compressed channel status (see Q 9) is carried in exactly the same manner as the scale factor bits. 6.3 Framing signals Multiframe alignment signals (MFA) (see Q 7) and frame slip detection signals (FSD) (see Q 8) are carried by modifying the parity of single samples. These signals do not have th
25、e benefit of majority decision decoding, but are inherently predictable and may be decoded reliably. 6.4 Signalling within the Companding block The sample-interleaved companding block is represented in Fig. 3 by the rectangle at the bottom of the diagram. Each row within the rectangle represents a R
26、ecommendation G.704 frame, and is subdivided into 12 squares representing the 12 samples. The diagram further shows the modification of the parity bits associated with the 12 samples in each frame by the signalling in parity mentioned above. 2 Recommendation 5.57 (06/90) Sign bit LSB I 1 o1 xxxxxxxx
27、xxxxx O01 xxxxxxxxxxxx 0001xxxxxxxxxxx 00001 xxxxxxxxxxxx 000001xxxxxxxxxxx 0000001xxxxxxxxxxxxx 00000001xxxxxxxxxxxx ooooooooxxxxxxxxxxxx 11111111xxxxxxxxxxxx 11111110xxxxxxxxxxxx 1111110xxxxxxxxxxxx 11111oxxxxxxxxxxx 1111oxxxxxxxxxxxx 111oxxxxxxxxxxx 110xxxxxxxxxxxx 1oxxxxxxxxxxxxx -I- Bits trunca
28、ted in H11 channels Bits truncated in HI 1 and H1 FIGURE la - Coding table Sign bit LSB I I o1 xxxxxxxxxxxxx 01xxxxxxxxxxxxx 01xxxxxxxxxxxxx 01xxxxxxxxxxxxx 01xxxxxxxxxxxxx o1 xxxxxxxxxxxxx o1 xxxxxxxxxxxx1 oxxxxxxxxxxxx11 1 xxxxxxxxxxxx11 10xxxxxxxxxxxx1 1oxxxxxxxxxxxxx 1oxxxxxxxxxxxxx 1oxxxxxxxxxx
29、xxx 10xxxxxxxxxxxxx 10xxxxxxxxxxxxx 1oxxxxxxxxxxxxx - b* b,4 47 + Bits transmitted in H11 channels -+ 6 Bits transmitted in H12 channels _+ Scale facto1 s, SI so 000 001 010 o1 1 1 O0 1 o1 110 111 111 110 1 o1 1 O0 o1 1 010 001 000 - X=l or O dol-sc Scale factor s, SI so 000 001 010 011 1 O0 1 o1 11
30、0 111 111 110 1 o1 1 O0 011 010 001 000 - X=l or O FIGURE 1 b - Transmitted bits d02-sc Recommendation 5.57 (06/90) 3 O ms Odd-numbered samples Even-numbered samples FIGURE 2 - Sample interleaving within the companding block 1 ms = 1 companding block d03-sc Oms I O mc Order of transmission _) of sam
31、ples in frames Signalling in parity data 0 Example data ,m MFA: mutiframe alignment signai SF, : scale factor (channel A) CA D FSD : frame siip detection signal : compressed channel status data : additional data (not defined) Modified parity matrix Order of transmission of frames in companding block
32、 1 1 ms FIGURE 3 - Signalling in panty matrix and data d04-sc 4 Recommendation 5.57 (06/90) 7. Synchronization and frame alignment The sampling frequency of the sound signal must be synchronous with the bit clock of the transmission system. Frame alignment of the companding block and multiframe (see
33、 Q 9) is accomplished by application of a 1536-bit chain code, MFA. The generator is shown in Fig. 4a and the corresponding circuit for locking in Fig. 4b Chambers, 19851. The start word of the MFA generator is 10000110100 and it corresponds to the first frame of the multiframe. The generator will p
34、roduce a sync pulse after 1536 bits (192 ms) and automatically reset to the start word. The start of the multiframe may be locked to the interface Z-preamble as shown in Fig. 6. The 1 ms companding block sync is generated by decoding the corresponding register contents. The multiframe alignment sign
35、al (MFA) is signalled in parity (see Q 6). I 11 bit shift register F 1 MFA sync pulse FIGURE 4a - Chain code generator U Unlocked L Locked I chain Incoming+ code - -! 1 11 bit shift register AIBICIDIEIFIGIHIJ IKIL XOR Chaincode =E -F 0 MFA sync pulse FIGURE 4b - Chain code synchronization circuit d0
36、6-sc Recommendation 5.57 (06/90) 5 8. Frame slip detection and management A controlled slip is defined as suppression or repetition of a Recommendation G.704 frame. A fixed of bits, FSD, ( . . . 110011001100 . . . ) is signalled in parity to assist decoders to detect controlled slips during transmis
37、sion. The FSD is transmitted as shown in Fig. 3 (i.e. the FSD is frame-aligned with the companding block). With a companding block of 1 ms and traditional framing methods, it would normally take a number of such frames (7-8 ms) to detect a slip and re-align the companding frame. With the FSD, it wil
38、l only take a few Recommendation G.704 frames to detect that a slip has occurred as the phase of the sequence will shift + or -90” (depending on whether a frame has been suppressed or repeated). Through a modulo-2 operation on the received and the expected FSD it is possible to detect, within two fr
39、ames or more precisely, in which “pair” (1 and 2, 3 and 4,5 and 6, or 7 and 8) of frames within the companding block the slip has occurred. A suggested strategy for interpolation after a detected slip is shown in Fig. 5a and 5b, where the decoder produces a decoded block of the same length as the re
40、ceived block. Only channel A is shown; channel B is handled identically (Fig. 2 shows the transmitted sample sequence). Positive slip - one frame is repeated u- x x x x )( x -Frame1012 25A 27A 29A 31A 33A 35A Isdeleted 37A . . . . 45A 47A 2A 4A 6A 8A 10A 12A 14A 16A . _- 7-frame dccoded companding b
41、lock - . . . 46A 48A 1 Unrecoverable samples 2A X 4A 6A X 8A 10A X 12A 14A X 16A 18A X 20A 22A X 24A 25A 26A 27A 28A 29A 30A 31A 3% . . . # 46A 47A 48A 25A 27A . . . . . . 45A 47A - 14A 16A . . . . 46A 48A Received companding block 6 extra samples produced by interpolation I 1 - Frame 1 or 2 is repe
42、ated - 13A 14A 15A 16A 17A 18A 19A 20A . . . 9-frame decoded companding block I . . . 46A 47A 48A Reconstructed companding block FIGURE 5 a - Strategy for interpolation after a detected positive slip d07-sc FIGURE 5 b - Strategy for interpolation er a detected negative slip Note that it is not possi
43、ble to tell which frame has been deleted. Consequently, of the 12 transmitted samples, 6 are not received and 6 cannot be identified. In the reconstructed companding block, 6 samples are simply omitted (to adjust the length of the reconstructed block) and 6 are replaced by interpolation. d08-sc 6 Re
44、commendation 5.57 (06/90) 9. Channel status Channel status data in the digital audio interface consists of a 192 bit (24 byte) cycle, which is repeated in 4 ms (one block of the interface). The channel status is signalled in parity according to the description in Q 6.2 and 6.4. This method of signal
45、ling provides one channel status bit for each audio signal per 1 ms companding block, enabling the system to carry one channel status block every 192 ms. This is shown in Fig. 6. Because only one data block out of 48 is transmitted, the two counters (local sample and time of day address code) must b
46、e incremented in the decoder by the appropriate amount. The start of the multiframe is signalled by the multiframe alignment signal defined in Q 7. The time codes carried in the compressed channel status refer to the timing of the first sample in the multiframe. Input interface signai 4- 48 blocks =
47、 192 ms _) 2 X 192 decoded C bits n Output interface signal Z preamble 7 Regenerated C bits 3 FIGURE 6 - Transmission of channel status data d09-sc 10. User data The user data bit in each sub-frame of the interface is transmitted transparently in the H12 channel only. 11. Frame structure and bit int
48、erleaving The Recommendation G.704 frame for the H11 channel contains 192 usable bits, and that for the H12 channel contains 240 usable bits. Twelve 15-bit sound samples, each accompanied by its parity bit, may be carried in either type of frame. Additionally, the H12 frame may carry sufficient extr
49、a audio bits to increase the length of the compressed audio samples to 18 bits, and one user data bit per sample. Recommendation 5.57 (06/90) 7 The organization of the 24 octets of data which are common to both H11 and H12 channels are identical in both types of frame, to facilitate remultiplexing at the interface between H11 and H12 channels. The common data occupies the entire available capacity of the H11 frame, and the first 24 available octets in the H12 frame, as shown in Fig. 7. The remaining 6 octets in the H12
