CAN CSA-ISO IEC 14496-3-2010 Information technology - Coding of audio-visual objects - Part 3 Audio (Technical Corrigendum 1 December 2010 Technical Corrigendum 2 January 2013).pdf

《CAN CSA-ISO IEC 14496-3-2010 Information technology - Coding of audio-visual objects - Part 3 Audio (Technical Corrigendum 1 December 2010 Technical Corrigendum 2 January 2013).pdf》由会员分享，可在线阅读，更多相关《CAN CSA-ISO IEC 14496-3-2010 Information technology - Coding of audio-visual objects - Part 3 Audio (Technical Corrigendum 1 December 2010 Technical Corrigendum 2 January 2013).pdf（1442页珍藏版）》请在麦多课文档分享上搜索。

1、Information technology Coding of audio-visual objects Part 3: AudioCAN/CSA-ISO/IEC 14496-3-10(ISO/IEC 14496-3:2009, IDT)National Standard of CanadaNOT FOR RESALE.PUBLICATION NON DESTINE LA REVENTE.Legal Notice for StandardsCanadian Standards Association (CSA) standards are developed through a consen

2、sus standards development process approved by the Standards Council of Canada. This process brings together volunteers representing varied viewpoints and interests to achieve consensus and develop a standard. Although CSA administers the process and establishes rules to promote fairness in achieving

3、 consensus, it does not independently test, evaluate, or verify the content of standards.Disclaimer and exclusion of liabilityThis document is provided without any representations, warranties, or conditions of any kind, express or implied, including, without limitation, implied warranties or conditi

4、ons concerning this documents fitness for a particular purpose or use, its merchantability, or its non-infringement of any third partys intellectual property rights. CSA does not warrant the accuracy, completeness, or currency of any of the information published in this document. CSA makes no repres

5、entations or warranties regarding this documents compliance with any applicable statute, rule, or regulation. IN NO EVENT SHALL CSA, ITS VOLUNTEERS, MEMBERS, SUBSIDIARIES, OR AFFILIATED COMPANIES, OR THEIR EMPLOYEES, DIRECTORS, OR OFFICERS, BE LIABLE FOR ANY DIRECT , INDIRECT, OR INCIDENTAL DAMAGES,

6、 INJURY, LOSS, COSTS, OR EXPENSES, HOWSOEVER CAUSED, INCLUDING BUT NOT LIMITED TO SPECIAL OR CONSEQUENTIAL DAMAGES, LOST REVENUE, BUSINESS INTERRUPTION, LOST OR DAMAGED DATA, OR ANY OTHER COMMERCIAL OR ECONOMIC LOSS, WHETHER BASED IN CONTRACT, TORT (INCLUDING NEGLIGENCE), OR ANY OTHER THEORY OF LIAB

7、ILITY , ARISING OUT OF OR RESUL TING FROM ACCESS TO OR POSSESSION OR USE OF THIS DOCUMENT, EVEN IF CSA HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES, INJURY, LOSS, COSTS, OR EXPENSES.In publishing and making this document available, CSA is not undertaking to render professional or other servic

8、es for or on behalf of any person or entity or to perform any duty owed by any person or entity to another person or entity. The information in this document is directed to those who have the appropriate degree of experience to use and apply its contents, and CSA accepts no responsibility whatsoever

9、 arising in any way from any and all use of or reliance on the information contained in this document. CSA is a private not-for-profit company that publishes voluntary standards and related documents. CSA h as no power, nor does it undertake, to enforce compliance with the contents of the standards

10、or other documents it publishes. Intellectual property rights and ownershipAs between CSA and the users of this doc ument (whether it be i n printed or electronic form), CSA is the o wner, or the authorized licensee, of all works contained herein that are protected by copyright, all trade-marks (exc

11、ept as otherwise noted to the contrary), and all inventions and trade secrets that may be contained in this document, whether or not such inventions and trade secrets are protected by patents and applications for patents. Without limitation, the unauthorized use, modification, copying, or disclosure

12、 of this document may violate laws that protect CSAs and/or others intellectual property and may give rise to a right in CSA and/or others to seek legal redress for such use, modification, copying, or disclosure. To the extent permitted by licence or by law, CSA reserves all intellectual property ri

13、ghts in this document.Patent rightsAttention is drawn to the possibility that some of the elements of this standard may be the subject of patent rights. CSA shall not be held responsible for identifying any or all such patent rights. Users of this standard are expressly advised that determination of

14、 the validity of any such patent rights is entirely their own responsibility.Authorized use of this documentThis document is being provided by CSA for informational and non-commercial use only. The user of this document is authorized to do only the following:If this document is in electronic form:.l

15、oad this document onto a computer for the sole purpose of reviewing it;.search and browse this document; and.print this document if it is in PDF format. Limited copies of this doc ument in print or paper form may be distributed only to persons who are authorized by CSA to h ave such copies, and only

16、 if this Legal Notice appears on each such copy.In addition, users may not and may not permit others to.alter this document in any way or remove this Legal Notice from the attached standard;.sell this document without authorization from CSA; or.make an electronic copy of this document.If you do not

17、agree with any of the terms and conditions contained in this Legal Notice, you may not load or use this document or make any copies of the co ntents hereof, and if you do make such copies, you are required to destroy them immediately. Use of this document constitutes your acceptance of the terms and

18、 conditions of this Legal Notice.Information technology Coding of audio-visual objects Part 3: AudioTechnical Corrigendum 2:2013 (IDT) toNational Standard of CanadaCAN/CSA-ISO/IEC 14496-3-10(ISO/IEC 14496-3:2009, IDT)NOT FOR RESALE. / PUBLICATION NON DESTINE LA REVENTE.Standards Update ServiceTechni

19、cal Corrigendum 2:2013 toCAN/CSA-ISO/IEC 14496-3-10January 2013Title: Information technology Coding of audio-visual objects Part 3: AudioPagination: 6 pagesTo register for e-mail notification about any updates to this publicationgo to shop.csa.caclick on CSA Update ServiceThe List ID that you will n

20、eed to register for updates to this publication is 2420892.If you require assistance, please e-mail techsupportcsagroup.org or call 416-747-2233.Visit CSA Groups policy on privacy at csagroup.org/legal to find out how we protect your personal information.INTERNATIONAL STANDARD ISO/IEC 14496-3:2009TE

21、CHNICAL CORRIGENDUM 2 Published 2011-10-15 INTERNATIONAL ORGANIZATION FOR STANDARDIZATION ORGANISATION INTERNATIONALE DE NORMALISATIONINTERNATIONAL ELECTROTECHNICAL COMMISSION COMMISSION LECTROTECHNIQUE INTERNATIONALEInformation technology Coding of audio-visual objects Part 3: Audio TECHNICAL CORRI

22、GENDUM 2 Technologies de linformation Codage des objets audio-visuels Partie 3: Codage audio RECTIFICATIF TECHNIQUE 2 Technical Corrigendum 2 to ISO/IEC 14496-3:2009 was prepared by Joint Technical Committee ISO/IEC JTC 1, Information technology, Subcommittee SC 29, Coding of audio, picture, multime

23、dia and hypermedia information. ICS 35.040 Ref. No. ISO/IEC 14496-3:2009/Cor.2:2011(E) ISO/IEC 2011 All rights reserved Technical Corrigendum 2:2013 to CAN/CSA-ISO/IEC 14496-3-10ISO/IEC 14496-3:2009/Cor.2:2011(E) 2 ISO/IEC 2011 All rights reservedReplace Table 11.3 with: Table 11.3 Syntax of block_d

24、ata Syntax No. of bits Mnemonic block_data() block_type; 1 uimsbf if (block_type = 0) const_block; uimsbf js_block; (reserved) 5 if (const_block = 1) if (resolution = 0) / 8 bits const_val; 8 simsbf else if (resolution = 1) / 16 bits const_val; 16 simsbf else if (resolution = 2 | floating = 1) / 24

25、bits const_val; 24 simsbf else / 32 bits const_val; 32 simsbf else js_block; 1 uimsbf if (bgmc_mode = 0) else if (bgmc_mode = 1) 2 uimsbf sub_blocks = 1 0) smp_val0; varies Rice code if (start 1) res1; varies Rice code if (start 2) res2; varies Rice code if (bgmc_mode) for (n = start; n sb_length; n

26、+) msbn; varies BGMC for (k=1; k sub_blocks; k+) for (n = k * sb_length; n (k+1) * sb_length; n+) msbn; varies BGMC for (n = start; n sb_length; n+) if (msbn != tail_code) lsbn; varies uimsbf else tailn; varies Rice code for (k=1; k sub_blocks; k+) for (n = k * sb_length; n (k+1) * sb_length; n+) if

27、 (msbn != tail_code) lsbn; varies uimsbf Technical Corrigendum 2:2013 to CAN/CSA-ISO/IEC 14496-3-10ISO/IEC 14496-3:2009/Cor.2:2011(E) 4 ISO/IEC 2011 All rights reservedSyntax No. of bits Mnemonic else tailn; varies Rice code Else for (n = start; n block_length; n+) resn; varies Rice code if (RLSLMS)

28、 RLSLMS_extension_data() if (!mc_coding | js_switch) byte_align; 07 bslbf Note: random_access_block is true if the current block belongs to a random access frame (frame_id % random_access = 0) and is the first (or only) block of a channel in this frame. If non-adaptive prediction order is used (adap

29、t_order = 0), then in random access frames the block length switching must be constrained so that no blocks in the frame need samples from the previous frame for the prediction process. The condition start = sb_length must be true in all frames. If mc_coding is used, prohibit the use of zero block a

30、nd const block (block_type = 0) as a slave channel, but permit it as a master channel. RLSLMS shall not be used together with block_switching and mc_coding. In 11.4.3 Payloads for Floating-Point Data, replace Note after Table 11.6 Syntax of diff_float_data (changes highlighted): Note: “byte_align” s

31、tands for padding of bits to the next byte boundary. “FlushDict()“ is the function that clears and initializes the dictionary and variables of the Masked-LZ decompression module (see subclause 11.6.9). with: Note: “random_access_block” is defined as (random_access != 0 dec_chars nchars; ) string_cod

32、e; 915 uimsbf Note: “nchars” is the number of characters to be decoded (see 11.6.9). Technical Corrigendum 2:2013 to CAN/CSA-ISO/IEC 14496-3-10ISO/IEC 14496-3:2009/Cor.2:2011(E) ISO/IEC 2011 All rights reserved 5Replace Table 11.14 with: Table 11.14 Elements of diff_float_data Field #Bits Descriptio

33、n / Values use_acf 1 1: acf_flagc is present 0: acf_flagc is not present acf_flagc 1 1: acf_mantissac is present 0: acf_mantissac is not present acf_mantissac 23 Full mantissa data of common multiplier highest_bytec 2 Highest nonzero bytes of mantissa in a frame partA_flagc 1 1: Samples exist in Par

34、t-A 0: No sample exists or all zero in Part-A shift_ampc 1 1: shift_valuec is present 0: shift_valuec is not present shift_valuec 8 Shift value: The shift value is biased by 127. The value (shift_valuec-127) is added to the exponent of all floating-point values of channel c after conversion of decod

35、ed integer to floating-point values, and before addition of integer and the difference data. In 11.6.9.1 Encoder for Floating-Point data, replace following sentences (changes highlighted): If the input signal is 32-bit floating-point, input values are decomposed as shown in 11.14 into three parts: A

36、n estimated common multiplier A, a truncated integer multiplicand sequence Y, and a difference signal Z. The same compression scheme as for normal integer input is applied for the truncated and normalized integer multiplicand sequence. with: If the input signal is 32-bit floating-point, input values

37、 are decomposed as shown in 11.14 into three parts: An estimated common multiplier A, a 24-bit truncated integer multiplicand sequence Y, and a difference signal Z. The same compression scheme as for normal 24-bit integer input is applied for the truncated and normalized integer multiplicand sequenc

38、e. In 11.6.9.3.2.2 Normalization parameters, replace the first sentence of the second paragraph (changes highlighted): First, use_acd is decoded. with: First, use_acf is decoded. In 11.6.9.3.2.5 Masked-LZ decompression, replace the first sentence of the third paragraph (changes highlighted): The ran

39、ge of the code_bits is varied from 9 to 14 bits, since the index of the dictionary is coded as 9 to 15 bits depending on the number of the entries stored in the dictionary. with: The range of code_bits is varied from 9 to 15 bits, since the index of the dictionary is coded as 9 to 15 bits depending

40、on the number of the entries stored in the dictionary. Technical Corrigendum 2:2013 to CAN/CSA-ISO/IEC 14496-3-10ISO/IEC 14496-3:2009/Cor.2:2011(E) 6 ISO/IEC 2011 All rights reservedand the first sentence of the fourth paragraph: The deocder reads (code_bits) bits from bit stream, and gets string_co

41、de. with: The decoder reads (code_bits) bits from bit stream, and gets string_code. Replace 11.6.9.3.3.3 Multiplication of the common multiplier with: After conversion, the common multiplier A is reconstructed from acf_mantissac and multiplied to Fcn, and the result is set to Fcn. The computing proc

42、edure of multiplication is as follows. Step 1: Sign bit setting: The sign of the result is the same as that of Fcn. Step 2: Multiplication of mantissa: (acf_mantissac | 0x0800000) is multiplied to (mantissa bits of Fcn | 0x0800000) in a 64-bit integer resister. Step 3: Normalization: The result of 6

43、4-bit integer multiplication is normalized to 24-bit precision, and represented with 23 bits after discarding the top bit. Since 1.0 = (acf_mantissac | 0x0800000)*2-23, (mantissa part of Fcn | 0x0800000) *2-23 2.0, the result of the multiplication is in the range 1, 4). Consequently, it might be nec

44、essary to normalize by repeatedly shifting one bit to the right and incrementing the exponent. Step 4: Rounding; The rounding mode “round to nearest, to even when tie“ is applied to round off the normalized mantissa of the result. Normalization process might be needed after the rounding. Replace 11.

45、6.9.3.3.4 Addition of difference value of mantissa with: After the multiplication, the reconstructed difference value of the mantissa Dcn is added to the floating-point data Fcn, and the result is set to Fcn. The computing procedure of addition is as follows. Step 1: Addition of mantissa: (Dcn) is a

46、dded to (mantissa bits of Fcn | 0x0800000) in a 32-bit integer resister. Step 2: Normalization: The result of 32-bit integer addition is normalized to 24-bit precision, and represented with 23 bits after discarding the top bit. Since (Dcn)*2-23 1.0, and 1.0 = (mantissa part of Fcn | 0x0800000) *2-23

47、 2.0, the result of the multiplication is in the range 1, 3). Consequently, it might be necessary to normalize by repeatedly shifting one bit to the right and incrementing the exponent. Step 4: Truncation: The rounding is not needed for this addition because whenever shifting occurs, the LSB of the

48、resulting mantissa equals 0. Technical Corrigendum 2:2013 to CAN/CSA-ISO/IEC 14496-3-10Copyright NoticeThis Amendment to the International Standard contains information copyright protected by the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC). Except as permitted under the laws of Canada, no extract of this Amendment to the International Standard may be reproduced,