ANSI INCITS ISO IEC 11172-3-1993 Information technology - Coding of moving pictures and associated audio for digital storage media at up to about 1 5 Mbit s - Part 3 Audio.pdf

1、INTERNATIONAL STANDARD ISO/IEC 11172-3 First edition 1993-08-01 Information technology - Coding of moving pictures and associated audio for digital storage media at up to about I,5 Mbit/s - Part 3: Audio Technologies de /information - Codage de /image animee et du son associ6 pour /es supports de st

2、ockage num lbig_values*2+countl*4; 1+=4) ( hcodlvlIwlllxlIlyll if (v!=O) signv if (w!=o) signw if (x!=O) signx if (y!=O) signy isl = v isl+l = w isl+2 = x isl+3 = y I for (; 1576; I+) isl = 0 019 113 1 113 1 16 bslbf 1 bslbf 1 bslbf 1 bslbf 1 bslbf bslbf uimsbf bslbf uimsbf bslbf 2.4.1.8 Ancillary d

3、ata sy11tax ancihyJata() ( if (layer = 1) II (layer = 2) for (b=O; bno-of-ancillary-bits; b+) ancillary-bit 1 No. of bits Mnemonic 1 bslbf 19 ISO/IEC 11172-3: 1993 (E) 0 ISO/IEC 2.4.2 Semantics for the audio bitstream syntax 2.4.2.1 Audio sequence general frame - Layer I and Layer II: Part of the bi

4、tstream that is decodable by itself. In Layer I it contains information for 384 samples and in Layer II for 1 152 samples. It starts with a syncword, and ends just before the next syncword. It consists of an integer number of slots (four bytes in Layer I, one byte in Layer II). - Layer III: Part of

5、the bitstream that is decodable with the use of previously acquired mJn information. In Layer III it contains information for 1 152 samples. Although the distance between the start of consecutive syncwords is an integer number of slots (one byte in Layer III), the audio information belonging to one

6、frame is generally not contained between two successive syncwords. 2.4.2.2 Audio frame header - Part of the bitstream containing synchronization and state information. error-check - Part of the bitstream containing information for error detection. audio-data - Part of the bitstream containing inform

7、ation on the audio samples. ancillary-data - P,art of the bitstream that may be used for ancillary data. 2.4.2.3 Header The Fist 32 bits (four bytes) are header information which is common to all layers. syncword - The bit string 1111 1111 1111. ID - One bit to indicate the ID of the algorithm. Equa

8、ls 1 for ISO/IEC 11172-3 audio, 0 is reserved. Layer - 2 bits to indicate which layer is used, according to the following. To ch,ange the layer, a reset of the audio decoder may bc required. protection-bit - One bit to indicate whether redundancy has been added in the audio bitstreCam to facilitate

9、error detection and concealment. Equals 1 if no redundancy has been added, 0 if redundancy has been added bitrate-index - Indicates the bitrate. The all zero value indicates the free format condition, in which a fixed bitrate which does not need to be in the list can be used. Fixed means that a fram

10、e contains either N or N+l slots, depending on the value of the padding bit. The bitrate-index is an index to a table, which is different for the different layers. The bitrate-index indicates the total bitrate irrespective of the mode (stereo, joint-stereo, dual-chzumel, single-channel). 20 0 ISOAEC

11、 ISOAEC 11172-3: 1993 (E) bitrate-index 0000 0001 0010 0011 0100 0101 0110 0111 1000 1001 1010 1011 1100 I 101 1110 1111 bitrate specified (kbits/s) Layer I !itx tie tk!e 32 32 32 64 48 40 96 56 48 128 64 56 160 80 64 192 96 80 224 112 96 256 128 112 288 160 128 320 192 160 352 224 192 384 256 224 4

12、16 320 256 448 384 320 fofiidden forbidden forbidden Layer II Layer III L In order to provide the smallest possible delay and complexity, the decoder is not required to support a continuously variable bitrate when in Layer I or II. Layer III supports variable bitrate by switching the bitrate-index.

13、The switching of the b padding = 110; for each subsequent audio frame: if (Layer = 1) dif = (12 * bitrate) % sampling-frequency; else dif = (144 * bitmte) % sampling-frequency; rest = rest - dif; if (rest c 0) ( padding = yes; rest = rest + samplingfrequency; I else padding = no; private-bit - Bit f

14、or private use. This bit will not be used in the future by ISO/IEC. mode - Indicates the mode according to the following table. In Layer I and II the joint-stereo mode is intensity-stereo, in Layer III it is intensity-stereo and/or ms-stereo. -1 jomt-stereo (uitensity-stereo and/or ms-stereo) In Lay

15、er I, in all modes except joint stereo, the value of bound equals 32. In layer II, in all modes except joint-stereo, the value of bound equals sblimit. In joint-stereo mode the bound is determined by the mode_extension. mode-extension - These bits are used in joint-stereo mode. In Layer I and II the

16、y indicate which subbands are in intensity-stereo. All other subbands are coded in stereo. mode-extension 00 subbands 4-3 1 in intensity-stereo, bound=4 01 subbands 8-3 1 in intensity-stereo, bound=8 10 subb,ands 12-31 in intensity-stereo, bound=12 11 subbands 16-3 1 in intensity-stereo, bound=16 In

17、 Layer III they indicate which type of joint stereo coding method is applied. The frequency ranges over which the intensity-stereo and ms-stereo modes are applied are implicit in the algorithm. For more infonnation see 2.4.3.4. mode-extension intensity-stereo 00 Off 01 011 10 Off ms-stereo Off Off 0

18、11 I 11 I 011 I 011 I 22 0 ISOAEC ISOAEC 11172-3: 1993 (E) Note that the mode “stereo” is used if the mode bits specify stereo or equivalently if the mode bits specify joint stereo and the mode-extension specifies intensity-stereo “off” and ms-stereo “off”. copyright - If this bit equals O, there is

19、 no copyright on the ISO/IEC 11172-3 bitstream, 1 means copyright protected. original/copy - This bit equals 0 if the bitstream is a copy, 1 if it is an original. emphasis - Indicates the type of de-emphasis that shall be used. di 2.4.2.4 Error check crc-check - A 16 bit parity-check word is used fo

20、r optional error detection within the encoded bitstream. 2.4.2.5 Audio data, Layer I allocationchsb - Indicates the number of bits used to code the samples in subband sb of channel ch. For subbands in intensity-stereo mode the bitstream comils only one allocation data element per subband. allocation

21、chsb bits per sample 0 0 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 forbidden Note: For code 0000 no samples are transferred. scalefactorchsb - Indicates the factor of subband sb of channel ch by which the requantized samples of subband sb in channel ch shall be multiplied

22、. The six bits constitute an unsigned integer, index to table B.l “Layer I, II scalefactors”. samplechsbs - Coded representation of the s-th sample in subband sb of channel ch. For subbands in intensity-stereo mode the coded representation of the sample is valid for both channels. 2.4.2.6 Audio data

23、, Layer II allocationchsb - Corltails infonnation concerning the quantizers used for the samples in subband sb in ch,annel ch, whether the infonnation on three consecutive samples has been grouped to one code, and on the number of bits used to code the samples. The meaning and length of this field d

24、epends on the number of the subband, the bitrate, and the sampling frequency. The bits in this field form an unsigned integer used as an index to the relevant table in table B.2 “Layer II bit allocation tables”, which gives the number of levels used for quantization. For subbands in intensity-stereo

25、 mode the bitstream contains only one allocation data element per subband. 23 ISOIIEC 11172-3: 1993 (E) 0 ISO/IEC scfsichsb - Scalefactor selection infonnation. This gives infonnation on the number of scalefactors transferred for subband sb in channel ch and for which ptarts of the signal in this fr

26、ame they are valid. The frame is divided into three equal parts of 12 subband samples each per subband. scfsisb 00 three scalefactors transmitted, for parts 0,1,2 respectively. 01 two scalefactors transmitted, first one valid for parts 0 and 1, second one for part 2. 10 one scalefactor transmitted,

27、valid for all three parts. 11 two scalefactors transmitted, fust one valid for part 0, the second one for parts 1 and 2. scalefactorchsbp - Indicates the factor by which the requantized samples of subband sb in channel ch and of part p of the frame should be multiplied. The six bits constitute an un

28、signed integer, index to table B .l “Layer I, II scalefactors. groupingchsb - Is a function that determines, whether grouping is in effect for coding of s,amples in subband sb of channel ch. Grouping means, that three consecutive samples (a triplet) of the current subband sb in ch,annel ch in the cu

29、rrent granule gr are coded and transmitted using one common codeword and not using three separate codewords. Groupingchsb is true, if in the Bit Allocation table currently in use (see B.2) the value found under the sb (row) and the allocationsb (column) is either 3, 5, or 9. Otherwise it is false. F

30、or subbands in intensity-stereo mode the grouping is valid for both channels. samplecodechsbgr - Coded representation of the three consecutive sCampIes in the granule gr in subband sb of channel ch. For subbands in intensity-stereo mode the coded representation of the samplecode is valid for both ch

31、annels. samplechlsbs - Coded representation of the s-tb sCampIe in subband sb of channel ch. For subbands in intensity-stereo mode the coded representation of the sample is valid for both channels. 2.4.2.7 Audio data, Layer Ill main-data-begin - The value of main-data-begin is used to determine the

32、location of the first bit of main data of a frame. The mahi-data-begin value specifies the location as a negative offset in bytes from the first byte of the audio sync word. The number of bytes belonging to the header and side infonnation is not taken into account. For example, if main-data-begin =

33、0, then main data starts after the side infonnation. Examples are given in figure A.7.a and tigure A.7.b. private-bits - Bits for private use. These bits will not be used in the future by ISO/IEC. The number of private-bits depends on the number of channels. The number of bits allocated for private-

34、bits is detennined to equalize the total number of bits used for side-infonnation. scfsichscfsi-band - In Layer III, the scalefactor selection infonnation works simiharly to audio Layer II. The main difference is the use of the variable scfsi-band to apply scfsi to groups of scalefactors instead of single scalefactors. The application of scalefactors to granules is controlled by scfsi. 24