Attachment J6 - Specification 1208.pdf

LIBRARY OF CONGRESS

National Library Service for the Blind and Physically Handicapped

Specification 1208:2014

Preparation of audio files from Commercial Audiobooks

Version 1.0, January 2014

NLS/BPH Specification 1208:2014 January 2014

Issue Date Change Description

Revision History Rev. 1.0 Jan/30/2014 First Release

Document No.

1208:2014

Release Approval:

Head, Quality Assurance

(Robert Norton)

Head, Engineering Section

(John Brown)

Chief, Materials Development Division

(Michael Katzmann)

Director

(Karen Keninger)

Table of Contents

1. Scope

2. Reference Documents

2.1 Specifications

2.2 Standards

2.3 Definition of Terms

3. Requirements

3.1 General Audio Characteristics

3.1.1 Signal Level for Spoken Text

3.1.2 Background Noise Level in the Commercial Master

3.1.3 Distribution Announcements

3.2 Primary Audio WAV Files

3.2.1 Primary Audio File Configuration

3.2.2 Primary Audio File Limit and Duration

3.2.3 Primary Audio File Content

3.2.3.1 Content of DTB Announcement File (Except for Children’s Books)

3.2.3.2 Ending Announcement (Except for Children’s Books)

3.2.3.3 Content of DTB Announcement File (Children’s Books)

3.2.3.4 Ending Announcement (Children’s Books)

3.3 Headings WAV File

3.3.1 Headings File Configuration

3.3.2 Headings File Content

4. Quality Assurance

4.1 Contractor Inspection

4.1.1 General Inspection of Commercial Masters

4.1.1.1 Unacceptable Audible Characteristics of the Recording

4.1.1.2 Unacceptable Acoustical Quality and Speech Intelligibility

4.1.2 Test Equipment and Inspection Facilities

4.1.3 Responsibility for Compliance

4.1.4 Reporting of Inspection Results

4.1.5 Quality Assurance Procedures (QAP)

4.2 Library of Congress (NLS/BPH) Inspection

4.2.1 WAV file Inspection

4.2.2 Facility Inspection

4.3 Acceptance Inspection

4.3.1 Contractor’s Acceptance Inspection

4.3.2 NLS Acceptance Inspection

4.4 Methods of Inspection

4.4.1 Primary Audio File Configuration

4.4.2 Primary Audio File Duration

4.4.3 Signal Levels

4.4.3.1 Reference Level

4.4.3.2 Spoken Text

4.4.3.3 Background Noise

4.4.3.4 Headings (hdgs) Clip File

4.5 Warranty

1. Scope This specification defines requirements for the preparation audio files used in creation of digital talking books (DTBs) produced for the National Library Service for the Blind and Physically Handicapped (NLS), Library of Congress, when the source of the audio files is a commercial audiobook. The digital audio produced to satisfy the requirements of this specification is encoded by linear pulse-code modulation and the bitstream is contained within a file conforming to the RIFF WAV (WAV) specification regardless of the format of the source audio files.

The audio content of a DTB consists of the primary audio files and the headings audio file. The primary audio files contain the complete content of the book, usually in chronological play order. The headings file is derived from the primary audio files and contains clips referenced in the DTB’s Navigation Control File (NCX). These clips are the book’s title, author, and all the audio labels of navigation points and list items referenced by the

NCX.

NLS Specifications 1203, 1205, and 1206 define requirements for the set of files, protection, and uploading of a DTB. In some areas, these requirements augment those contained in this specification. It is therefore essential that producers of DTBs reference both this specification and NLS Specifications 1203, 1205, and 1206 to ensure full compliance with all NLS requirements.

The requirements of this specification, subject to objective and subjective evaluations, must be fulfilled. NLS may reject any master recording that fails to meet the requirements specified herein. Specific artistic points not covered by these specifications are subject to judgment by NLS staff and also can be cause for rejection of a particular title.

2. Reference Documents The following documents and publications form a part of this specification. In the event of conflict between the documents and publications referenced herein and the content of this specification, the content of this specification shall be considered a superseding requirement.

2.1 Specifications

American National Standards Institute (ANSI)

ANSI/NISO Z39.86-2002

Specifications for the Digital Talking Book

The document cited above is available from:

American National Standards Institute, Inc.

25 West 43rd Street, 4th Floor New York, NY 10036 Tel: (212) 642-4900 Fax: (212) 398-0023 www.niso.org/workrooms/daisy/Z39-86-2002.html

National Library Service for the Blind and Physically Handicapped

NLS Specification 1203 (current revision) Construction of Digital Talking Books

NLS Specification 1205 (current revision) Protected Digital Talking Book

NLS Specification 1206 (current revision) Uploading Digital Talking Books and Magazines

The documents cited above are available from:

National Library Service for the Blind and Physically Handicapped Library of Congress 1291 Taylor Street NW Washington, DC 20542-0002

Tel: (202) 707-5100 Fax: (202) 707-0712 and at www.loc.gov/nls/specs/

2.2 Standards

American National Standards Institute (ANSI)

ANSI/ASQ Z1.4-2003

Sampling Procedures and Tables for Inspection by Attributes

The document cited above is available from:

American Society for Quality Quality Press 600 N. Plankinton Ave. PO Box 3005 Milwaukee, WI 53201-2914 and at www.asq.org/quality-press/index.html

Audio Engineering Society

AES17

AES standard method for digital audio engineering Measurement of digital audio equipment

The documents cited above are available from:

Audio Engineering Society 60 East 42nd Street, Room 2520 New York, NY 10165-2520 and at www.aes.org/publications/standards/

European Broadcasting Union

Specification of the Broadcast WAVE Format available at http://tech.ebu.ch/docs/tech/tech3285.pdf

2.3 Definition of Terms

Commercial master – The digital audio files (of any file format) containing the recorded narration provided by a commercial agency, the source of the primary audio files.

WAV file – Digital audio encoded by linear pulse-code modulation contained within a file conforming to the RIFF WAV (WAV) specification.

PAR – Production Authorization Record.

Primary audio file – Digital audio (WAV) file needed to make a complete digital talking book (DTB) containing the audio of the commercial master.

Headings audio file – Digital audio (WAV) file containing audio clips from the primary audio files of the title, author, and all audio labels for navigation points and navigation lists.

DTB – The complete set of files that make up a digital talking book as defined by NLS Specification 1203:2013, 1205, and ANSI/NISO Z39.86-2002.

3. Requirements The requirements below define the characteristics and configuration of all primary files and the headings file derived from commercial masters.

3.1 General Audio Characteristics

All primary files and the headings file shall meet the requirements listed below. The contractor shall edit the commercial master to satisfy these requirements. Failure to do so will result in a rejection by NLS Quality Assurance Section (QAS). Commercial masters may be edited to meet the requirements in section 3.1.1, 3.1.2, and 3.1.3 without approval from NLS.

3.1.1 Signal Level for Spoken Text

a. The average RMS signal level for the spoken text shall be not less than -24 dB FS, nor greater than -16 dB FS

b. No clipping or distortion shall occur

c. The recording shall have no abrupt level changes as a result of file (side) transitions

3.1.2 Background Noise Level in the Commercial Master

The RMS background noise level shall not exceed -72 dB FS, A-weighted. Background noise includes both acoustical and electrical noise.

3.1.3 Distribution Announcements

Any announcement pertaining to commercial master distribution on a hard medium, such as cassette or CD, must be removed by the contractor.

3.2 Primary Audio WAV Files

3.2.1 Primary Audio File Configuration

Each primary audio file shall be written as linear PCM data in WAV-file format at a sample rate of 44.1 kHz, 16 bits per sample, with dither, block alignment = 2.

Primary audio files shall be named according to one of the methods described below. All audio files within a DTB must be named consistently according to one of these methods. For each method, all alphabetic characters shall be in lowercase.

a. A primary audio file shall be named “nnnnnsmm.wav,” where “nnnnn” is the book number and “mm” is the two-digit sequence indicator (e.g., 56123s07.wav). Leading zeros shall be added to assure a five-digit book number minimum. The first file number shall start with 01 and continue consecutively until the book recording is complete.

b. A primary audio file shall be named “nnnnn-mmmm.wav,” where “nnnnn” is the book number and “-mmmm” is the four-digit sequence indicator (e.g., 56123-0007.wav). Leading zeros shall be added to assure a five-digit book number minimum. The first file number shall start with -0001 and continue consecutively until the book recording is complete.

3.2.2 Primary Audio File Limit and Duration

A DTB shall contain no more than 200 primary audio files. Refer to section 3.2.1.3 of NLS Specification 1203:2013 for the total file limit of a DTB.

Primary audio files shall not exceed 90 minutes.

3.2.3 Primary Audio File Content

The first primary audio file shall begin with the DTB announcement. The content of the commercial masters shall follow the DTB announcement in the first primary audio file and all of the subsequent primary audio files.

3.2.3.1 Content of DTB Announcement File (Except for Children’s Books)

The contractor shall record the following opening announcements for all titles except children’s books and where noted:

a. Title

b. “By (author name)”

c. The prefix “DB” plus the book number, e.g., “DB56123”

d. “[Audio Publisher] originally produced this unabridged recording for the commercial market. It has been adapted for the NLS talking-book program.

Elements from the original recording—including music, sound effects, announcements, or other additional material—may be present. Traditional NLS talking-book announcements have been added.

“Please note that, although this recording is unabridged, some information—such as photo captions, end notes, bibliographies, appendices, or other supplementary material—may have been omitted by the audio publisher.

“This recording is used courtesy of [Audio Publisher].”

e. “Read by __________________________.”(The name of the narrator of the commercial master. If there are several commercial master narrators give all names in order of appearance. Use “various readers” when there are more than four narrators.)

f. “Approximate reading time: ____ hours, ____ minutes.” (Rounded to the nearest five minutes, accurate to within +/- fifteen minutes of the value of “dtb:totalTime.”

See NLS Specification 1203 OPF metadata.)

g. For DTBs that have all book items at level one: “This book contains markers allowing direct access to the [book item(s)].”

For DTBs that have more than one level of hierarchical navigation: “This book contains markers allowing direct access to: at level 1 the [book item(s)], at level 2 the [book item(s)], . . ., and at level (lowest hierarchical level for the recording) the [book item(s)] and [page numbers (if marked in the recording)].”

This announcement shall include all hierarchical navPoints. Nonhierarchical page numbers, if marked, in the recording shall also be referenced in this announcement. The term for the “book item” shall be appropriate to the commercial master.

Producers are responsible for determining the appropriate navPoints for each DTB and for ensuring that all DTB navPoints are reflected in the Production Authorization Record (PAR) If NLS has suggested navPoints in the PAR, and the producer determines that changes to the suggestions or additional navPoints are necessary, the producer must either request that a PAR revision be made by their contract monitor or revise the PAR directly if given that ability by NLS. A revised PAR shall be submitted for QAS review.

h. “Library of Congress annotation: . . .” (To be supplied by NLS.)

3.2.3.2 Ending Announcement (Except for Children’s Books)

The following announcements shall be recorded on the last primary audio file after the end of the commercial recording, in the order stated:

a. “End of (Title) by [Author(s) and spelling of author’s name].” (e.g., “End of Troubling a Star by Madeleine L’Engle, M-A-D-E-L-E-I-N-E, L-’-E-N-G-L- E.” In the case of multiple authors all authors’ names shall be pronounced, then spelled.)

b. “Further reproduction or distribution in other than a specialized format is prohibited.”

c. “If you found this book to be defective, please contact your cooperating network library.”

3.2.3.3 Content of DTB Announcement File (Children’s Books)

The following specifications apply only to commercial masters that have been identified as “For preschool–grade 2” in Library of Congress annotation of the book

PAR.

For children’s books, the contractor shall record the following opening announcements for use with the DTB:

a. Title

b. “By (author name).”

c. The prefix “DB” plus the book number, e.g., “DB68886”

d. “Reading time, _____ minutes.” (Rounded to the nearest minute, accurate to within +/- five minutes of the value of “dtb:totalTime.” See NLS Specification 1203 OPF metadata.)

3.2.3.4 Ending Announcement (Children’s Books)

The following announcements shall be recorded on the last primary audio file after the end of the commercial recording, in the order stated:

a. “End of (Title) by (Author and spelling of author’s name).” (e.g. “End of Troubling a Star by Madeleine L’Engle, M-A-D-E-L-E-I-N-E, L-’ -E-N-G-L-E.”

In the case of multiple authors all authors’ names shall be pronounced then spelled.) Include illustration credit, if any, as read in the opening announcements.

b. “Further reproduction or distribution in other than a specialized format is prohibited.”

c. “If you found this book to be defective, please contact your cooperating network library.”

3.3 Headings WAV File

3.3.1 Headings File Configuration

The headings file shall be written as linear PCM data in WAV-file format at a sample rate of 44.1 kHz, 16 bits per sample, with dither, block alignment = 2.

The headings file shall be named “nnnnnhdgs.wav” where “nnnnn” is the book number (e.g., 68886hdgs.wav). Leading zeros shall be added to ensure a five-digit book number minimum. All alphabetic characters shall be in lowercase.

3.3.2 Headings File Content

The headings file shall contain the audio clips of the navigation control file (NCX) docTitle, docAuthor, and audio navLabels for navPoints and navTargets.

4. Quality Assurance

4.1 Contractor Inspection

The contractor is responsible for the performance of all inspection requirements specified herein.

4.1.1 General Inspection of Commercial Masters

The contractor shall immediately notify NLS Quality Assurance Section (QAS) if any of the following items are found in the commercial master during the course of production. The contractor shall make a recommendation as to whether the item can be corrected and if so shall propose a solution. NLS QAS shall review the proposed solution and will provide written authorization to implement the solution. The contractor shall not edit or process commercial masters to correct these items without approval from NLS QAS. Final authority for judging the acceptability of a commercial master (including any corrections authorized in advance by NLS) shall rest with the NLS Quality Assurance Section.

4.1.1.1 Unacceptable Audible Characteristics of the Recording

a. Hum

b. Egregious background noise in the narration environment

c. A difference between a section of sound track that has a correction and any other section of the same sound track

d. Gaps, seams, or transients in edits

e. Egregious effects caused by sound-processing devices (such as automatic gain controls, compressors/limiters, noise gates, equalizers, etc.)

f. Radio frequency interference (RFI) or electromagnetic interference (EMI)

4.1.1.2 Unacceptable Acoustical Quality and Speech Intelligibility

a. The recording is judged to have an overly muffled, bassy, or tinny quality

b. The recording exhibits an objectionable degree of sibilance

c. The recording has other undesirable acoustical characteristics judged to be audible to an objectionable degree

d. The spoken text is of poor clarity

4.1.2 Test Equipment and Inspection Facilities

The contractor shall ensure that test and inspection facilities of sufficient accuracy, quality, and quantity are established and maintained to permit performance of required inspections. Test equipment shall be calibrated annually. NLS reserves the right to request copies, for review, of any documentation related to this specification. These shall be provided at no cost to NLS.

4.1.3 Responsibility for Compliance

The inspections set forth in this specification shall become a part of the contractor’s overall inspection system or quality program. The absence of any inspection requirements in the specification shall not relieve the contractor of the responsibility of assuring that all products submitted to NLS for acceptance comply with all requirements of the contract. Sampling in quality conformance does not authorize the submission of known defective material, either indicated or actual, nor does it commit NLS to the acceptance of defective material. Should NLS determine that a significant fault or faults are found in production materials, within the warranty period, then correction of the fault or faults and inspections or controls for prevention shall be instituted without additional charge to NLS.

4.1.4 Reporting of Inspection Results

The contractor shall maintain complete records (in English) of all inspection results for the duration of the contract, including the warranty period. NLS reserves the right to request copies, for review, of any documentation related to this specification. These shall be provided at no cost to

NLS.

4.1.5 Quality Assurance Procedures (QAP)

The contractor shall prepare a written document that describes the quality-assurance procedures used to achieve the requirements of the specification. The contractor shall submit this document to the NLS Quality Assurance Section at the beginning of each new contract or at any time required by NLS. NLS reserves the right to require the contractor to improve quality-assurance procedures.

4.2 Library of Congress (NLS/BPH) Inspection

4.2.1 WAV file Inspection

NLS reserves the right to reject any WAV file that fails to meet the requirements of this specification. Each WAV file (see NLS Specification 1206) shall be uploaded to NLS for inspection of specification compliance and/or narration approval.

4.2.2 Facility Inspection

NLS reserves the right to inspect plant facilities or recording processes at any reasonable time. An NLS representative shall have the authority, without advance notice, to select at random a sample of the books at any time during the course of the contract for testing to the specified requirements.

4.3 Acceptance Inspection

4.3.1 Contractor’s Acceptance Inspection

Acceptance inspection shall be performed by the contractor on every commercial master and shall consist of the inspections listed in the table below.

4.3.2 NLS Acceptance Inspection

Acceptance inspection will be performed by NLS on every commercial master and will consist of the inspections listed in the table below. The complete master recording (WAV files, unprotected DTB, and protected DTB) shall be uploaded into an automated validator suite (autotest) to assure compliance to NLS specifications and ANSI/NISO Z39.86-2002. Master recordings that pass autotest will be queued in the NLS QA section for full review.

Requirement Section Test Method

All applicable requirements of this specification 3.1 Audio Review

Primary Audio file configuration 3.2.1 4.4.1

Primary Audio file duration 3.2.2 4.4.2

Signal level – text 3.1.1 4.4.3.2

Background noise level 3.1.2 4.4.3.3

File of headings clips 3.3 4.4.3.4

Acceptance Inspection

4.4 Methods of Inspection

4.4.1 Primary Audio File Configuration

Examine with an audio editor for conformance with the requirements of Section 3.2.1

3.2.2

4.4.2 Primary Audio File Duration

Examine with an audio editor for conformance with the requirements of Section .

4.4.3 Signal Levels

4.4.3.1 Reference Level

Full-scale signal level is equal to 0 dB FS and is defined in AES17. It is the amplitude of a 997 Hz sine wave whose positive peak value reaches the positive full scale, leaving the negative maximum code unused. (The reference is based on a sine wave, not a square wave.)

RMS signal level in dB FS is calculated using the formula:

L N x n N ref n 20 1 20 1

2log ( ) log ( )

Where x is a set of N samples of the signal to be measured and ref is a set of N samples of the 0 dB FS sine wave. Average RMS signal level shall be measured using a section of the signal that is 1 minute in length.

4.4.3.2 Spoken Text

Select a section of the primary audio file that is 1 minute in length and measure the un-weighted RMS level in db FS for conformance to the requirement of Section 3.1.1

3.1.2

3.3

4.4.3.3 Background Noise

Select a quiet section of the primary audio file at least 1 second long between words.

Apply an A-weighting filter to the selection and measure the RMS noise for conformance to the requirement of Section .

4.4.3.4 Headings (hdgs) Clip File

Examine the heading clip file for conformance with the requirements of Section

4.5 Warranty

The contractor shall agree to unconditionally warrant that the commercial master conforms to the requirements set forth in this specification for a period of two (2) years from the date the book is approved by the NLS Quality Assurance Section.