Ultratec Exhibit 1008
`Ultratec v Sorenson IP Holdings Page 1 of 21
`
`

`

`National Association of Broadcasters
`
`ENGINEERING HANDBOOK
`10th Edition
`
`Editor-in-Chief
`
`EDMUND A. WILLIAMS
`
`Associate Editors
`
`GRAHAM A. JONES
`DAVID H. LAYER
`THOMAS G. OSENKOWSKY
`
`National Association of N;tB·
`
`BROADCASTERS
`
`AMSTERDAM • BOSTON • HEIDLEBERG • LONDON
`NEW YORK • OXFORD • PARIS • SAN DIEGO
`SAN FRANOSCO • SINGAPORE • SYDNEY • TOKYO
`Focal Pres.5 is an imprint of Elsevier
`
`u
`
`Ultratec Exhibit 1008
`Ultratec v Sorenson IP Holdings Page 2 of 21
`
`

`

`Senior Acquisitions Editor: Angelina Ward
`Publishing Services Manager: George Morrison
`Senior Project Manager: Paul Gottehrer
`Assistant Editor: Doug Shults
`Marketing Manager: Christine Degon Veroulis
`Compositor: MuJtiscience Press, lnc.
`
`Focal Press is an imprint of Elsevier
`30 Corporate Drive, Suite 400, Burlington, MA 01803, USA
`Linacre H ouse, Jordan Hill, Oxford OX2 80P, UK
`
`Copyright © 2007, Elsevier. All rights reserved.
`
`Copyright 1935, 1938, 1946, 1960, 1975, 1985, 1992 Nalional Association of Broadcasters. Ali rights reserved.
`Printed in the United States of America. No part of this book may be reproduced in any form or by any means without
`permission in writing from the National Association of Broadcasters, 1771 N Street, NW Washington, DC 20036-2891.
`
`This publication is not intended to be a substitute for the professional advice of an engineer or attorney. While the Editors have
`endeavored to include information current at the time of publication, statutes, rules, regulations and engineering knowledge
`will undoubtedly change following publication. It is the Reader's responsibility to obtain the most up-to-date information on
`any topic.
`
`NAB and the contributors to this publication make no representations and disclaims all warranties, express or implied, (except
`as otherwise provided by law) regarding the information contained herein, including without limitation the implied warranty of
`fitness for a particular purpose and the implied warranty of merchantability. To the extent permitted by law, neither NAB nor
`any contributor to this publication shall be liable to any other person or entity for any damages, including without limitation any
`direct, indirect, special, consequential, exemplary or punitive damages under any theory of law.
`
`For information on other publications available through the National Association of Broadcasters, call 202-429-7424,
`800-368-5644 or visit the NAB Store website at http:/ / www.NABstore.com.
`
`No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means,
`electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher.
`
`Permissions may be sought directly from Elsevier's Science & Technology Rights
`Department in Oxford, UK: phone: (+44) 1865 843830, fax: (+44) 1865 853333,
`E-mail: permissions@elsevier.com. You may also complete your request online
`via the Elsevier homepage (http:/ / elsevier.com), by selecting "Support & Contact"
`then "Copyright and Permission" and then "Obtaining Permissions."
`
`@ Recognizing the importance of preserving what has been written,
`·'-/ Elsevier prints its books on acid-free paper whenever possible.
`
`Library of Congress Cataloging-in-Publication Data
`Application submitted
`
`British Library Cataloguing-in-Publication Data
`A catalogue record for this book is available from the British Library.
`
`I
`
`ISBN 13: 978-0-24-0-80751-5
`ISBN 10: 0-240-80751-0
`ISBN 13: 978-0-240-80982-3 (CD-ROM)
`ISBN 10: 0-240-80982-3 (CO-ROM)
`
`For information on all Focal Press publications
`visit our website a t www.books.elsevier.com
`
`0708091011 54321
`
`Printed in the United States of America
`
`Working together to grow
`libraries in developing countries
`www.dsevier.com I www.bookaid.org I www.sabre.org
`
`Ultratec Exhibit 1008
`Ultratec v Sorenson IP Holdings Page 3 of 21
`
`

`

`CHAPTER
`
`5.24
`
`Closed Captioning Systems
`
`JEFF HUTCHINS*
`Accessible Media Industry Coalition
`Pittsburgh, Pennsylvania
`
`ALAN LAMBSHEAD**
`Evertz Microsystems
`Burlington, Ontario, Canada
`
`1
`
`INTRODUCTION
`Captioning, according to the Accessible Media Indus(cid:173)
`try Coalition, is "the process of converting the audio
`content of a television broadcast, webcast, film, video,
`CD-ROM, DVD, live event, or other productions into
`text, which is displayed on a screen or monitor." Cap(cid:173)
`tions not only display words as the text equivalent of
`spoken dialogue or narration, but also include speaker
`identification and sound effects. It is important that
`the captions be
`; ' Synchronized and appear at approximately the
`f sam~ time as the audio is available;
`' ~qmvalent and equal in content to that of the audio,
`mcluding speaker identification and sound effects;
`' Accessible and readily available to those who need
`them.
`. Ca~tions may be displayed on or adjacent to the
`1
`1deo unage. Open captions are a permanent part of the
`.
`~age '.fild cannot be turned off. Closed captions are
`deansm1tted with the audio and video, but require a
`thcoder to ~etect, decipher, and display the captions so
`10atonlyv1ewers who wish to see the captions will do
`~i ~~data for closed captions is transmitted with the
`bl evi~ion_ program, either on line 21 in the vertical
`~g interval (VBI) of an analog program, or in a
`lf&,p
`
`separate data packet accompanying the audio and
`video of a digital program.
`Captioning is distinct from subtitling in that cap(cid:173)
`tioning includes aural information other than just dia(cid:173)
`logue, such as sound effects, music effects, and
`indications of who is speaking, all intended to aid the
`viewer who is unable to hear the soundtrack. The orig(cid:173)
`inal target audience for captioning was primarily peo(cid:173)
`ple who are deaf or hearing impaired (about 20
`million in 1980). The market has since expanded to
`include people learning English as a second language;
`those learning to read, especially students with read(cid:173)
`ing disabilities; people in noisy places (like bars and
`airports) or quiet places (like hospitals and spas). Cap(cid:173)
`tions generally appear in the lower portion of the tele(cid:173)
`vision screen and vary in size in proportion to the size
`of the television screen. The caption characters are
`sized to be easily visible, typically white letters against
`a black background (for analog television), and usu(cid:173)
`ally do not obstruct essential parts of the picture. Cap(cid:173)
`tions for digital television (DTV) can have a wider
`range of font styles, sizes, and colors than those for
`NTSC transmissions (see Figure 5.24-1).
`Closed captioning may be added in real time to a
`live program or may be added before transmission as
`part of postproduction for a prerecorded program.
`
`.. )/f:.
`~
`,J~~~Utchins updated the analog captioning portion of this chapter drawing on material first published in the 9th edition of the NAB Engi(cid:173)
`''A!anandbook, Chapter 5.13, "Closed Captioning and Extended Services," by Amnon Salomon and Gerald Freda.
`~Pro ~~bshead adapted the DTV captioning portion of this chapter from material first published in the NAB Broadcast .Engineering Confer(cid:173)
`cee mgs 2004, "Implementing Closed Captioning for DTV," by Graham Jones.
`
`NAB ENGINEERING HANDBOOK
`Copyright© 2007 Focal Press.
`All rights of reproduction in any form reserved.
`
`1435
`
`Ultratec Exhibit 1008
`Ultratec v Sorenson IP Holdings Page 4 of 21
`
`

`

`SECTION 5: VIDEO PRODUCTION AND STUDIO TECHNOLOGY
`
`This act required that, effective July 1, 1993
`sion receivers with picture screens 13 inche ' a]] televj.
`must be equipped to display closed captios or greater
`sion transmission, and it required the FC~ed televj.
`rules to implement this requirement.
`to enact
`FCC Report and Order 6 of General Docket
`adopted April 12, 1991, and released to the 91bl.was
`April 15, 1991. The Order became effectiv~u Uc on
`1993. The Order amends Part 15 of the FCC'
`July 1,
`adding a new Section 15.119 to set out the F~tles by
`for captions and caption decoders. The hi hli Rules
`this section as it relates to NTSC analog tele~i .ghts of
`tioning are as follows:
`sion cap-
`
`• Effective July 1, 1993, TV receivers 13 inches
`larger must have caption decoders;
`or
`• Closed caption information to be carried in lin 2
`of field l;
`e 1
`• Decoders to have user-selectable Caption displ
`modes. Text display mode is optional;
`ay
`• Caption and Text modes may contain data in eith
`of two operating channels, usually referred to as er ,.,.
`Cl, C2, Tl, and T2;
`• Receivers must decode at least Cl and C2 captions;
`• Captions to be displayed in "boxes" on the screen
`within the Safe Title Area defined by SMPTE 27.3
`(now replaced by SMPTE RP218-2002);
`• Caption decoding circuitry must be tolerant of ~
`cable security systems that alter line numbering,
`etc.
`The complete text of the FCC Report and Order
`(R&O) may be obtained from the U.S. Government
`Printing Office. The text was also published in the Fed·
`eral Register, Vol. 56, No. 114, p. 27200. The FCC Rules
`and Regulations are contained in the Code of Federal
`Regulations, Part 47, Telecommunications. See Chapter
`1.2 of this handbook, "Broadcast-Related Organiza·
`tions and Information," for information on contacting
`tlfe U.S. Government Printing Office and other organi·
`h~tions mentioned in this chapter.1
`. .
`.
`.
`
`Update Since Passage of Decoder Circuitry Act
`Since the passage of the Decoder Circuitry. ~ct, th~
`FCC has ruled several additions. These _ad~1tio~~n·
`documented with other technical details m a
`608
`sumer Electronics Association document CEA- th'
`titled Line 21 Data Services (formerly ~~-608)iJ~oc~
`time of writing (2006), the current revision oft 5
`ument is CEA-608-D.
`.
`. has
`The regulation and use of line 21 for cap~?nmf fea(cid:173)
`been expanded to include both fields. Ad~i~o~:i Ian·
`tures have been added to deal with a~ditio deliver
`guages. Also, within field 2 is a capacity to. addi(cid:173)
`4! ~!udes,
`
`additional captioning services (C3, C4, T3
`tion to extended data services (XDS). X
`
`T5
`0,
`
`U1
`
`16 2006, rna~
`uary fc~.gov ;oet1
`1The entire Part 15 regulations, as updated Feb
`also be downloaded as a PDF file from http:/ WWW·
`info/ rules/partl5/ partlS-2-16-06.pdf.
`
`1r
`
`FIGURE 5.24-1 DTV caption data processed by a
`decoder and displayed as text across the lower portion
`of the picture.
`
`A BRIEF HISTORY OF CLOSED
`CAPTIONING
`The Public Broadcasting Service (PBS) developed
`closed captioned technology for NTSC analog televi(cid:173)
`sion during the period 1973-1979 with funding sup(cid:173)
`port from the federal government (Department of
`Health, Education, and Welfare). Field test transmis(cid:173)
`sions were conducted on all aspects of caption genera(cid:173)
`tion, encoding, decoding, and display features of the
`service.
`During those years, a small number of programs
`were open-captioned by The Caption Center at PBS
`station WGBH and carried on PBS. The newly cre(cid:173)
`ated National Captioning Institute (NCI), in cooper(cid:173)
`ation with the ABC, CBS, NBC, and PBS networks,
`launched the closed captioning service in March
`1980 with approximately 16 hours per week of cap(cid:173)
`tioned programming. The first consumer product
`containing the decoding feature, called TeleCaption®,
`was sold by Sears, Roebuck and Co. and was a sepa(cid:173)
`rate box that worked with a standard television.
`
`FCC Rules and CEA Standards
`The line 21 captioning data signal of NTSC analog
`television signals is protected from interference from
`any other VBI service, test signal, or spillover from
`active video under FCC Rules and Regulations, Sec(cid:173)
`tion 73.682(a)(22), adopted in 1976. These rules also
`established the transmission standards for captioning
`and list the uses of the data channel.
`Following are highlights from rules associated with
`the Television Decoder Circuitry Act of 1990 that
`amended Part 15 of the FCC rules (Radio Frequency
`Devices, which relate to television receivers).
`
`Television Decoder Circuitry Act of 1990
`The U.S. Congress passed the Television Decoder Cir(cid:173)
`cuitry Act of 1990 (Pub. L. 101-431, 104 Stat. 960 (1990)).
`
`1436
`
`Ultratec Exhibit 1008
`Ultratec v Sorenson IP Holdings Page 5 of 21
`
`

`

`CHAPTER 5.24: CLOSED CAPTIONING SYSTEMS
`
`t limited to, content advisory (V-chip ), Time
`vut is 110 d source of transmission.
`01DaY:rsc television, the Accessible Media Industry
`recommends using Cl (also called CC-1) for
`f~r.
`coalitl.011 ary captioning service, and C3 (or CC-3) on
`the pr~or any secondary captioning s.ei:rice. ~and­
`field 2 !iJllitations make it extremely difficult, if not
`11idth 'ble to use Cl and C2 simultaneously for two
`. poss1
`'
`.
`.
`1 captioning services.
`an
`;epara e
`
`The roll-up style is most often used for captioning of
`live programs. From one to four rows of text display
`the captions in a scrolling manner.
`The paint-on style is sometimes used in analog sys(cid:173)
`tems because it allows characters to be displayed
`immediately, avoiding the transmission lag of pop-on
`style that occurs thanks to the 60 characters per second
`data rate of line 21. Paint-on can also be used to create
`special effects, such as a simulated "reveal" of text
`added to a displayed caption .
`
`. 'tal Television Captioning
`01g1
`1997, the next generation of closed captioning
`~!~\al television was adopted. The new Consumer
`1or trginics Association caption standard, CEA-708,
`8~d 0vigital Television (DTV) Closed Captioning, was
`u ,:elo ed by an industry group composed of caption
`de . epproviders, receiver manufacturers, broadcast-
`:ervIC
`·
`· and encoder manufactu-:ers. It en~b~es impr~"'.ed
`:~ed captioning to ~e provided for d1g1t~l ~e.lev1s1on
`ansmissions, including both standard defrmtion (SD)
`. ~id high definition (HD) ~ele".ision, while cont~nuing
`i to provide~ method of delivering CEA-608 captions to
`allalog receivers.
`DIV closed captions (known as DTVCC) have
`11eatly enhanced formatting and display capabilities
`~ompared to line 21 captions, with up to 63 services
`f€I program; eight independently controlled display
`~indows; and an extended range of characters and
`multiple fonts, sizes, and background and character
`colors and edges. The standard specifies how caption
`information is to be coded and processed, minimum
`implementation recommendations for DTV closed
`caption decoders, and recommended practices for cap(cid:173)
`tion encoder and decoder manufacturers. It also
`. requires that decoders give users control over caption
`font, color, size, and location that may override the
`rarameters as transmitted.
`During the transition from analog to digital broad(cid:173)
`casting, analog NTSC and DTV transmissions will
`coexist, as will analog and digital cable distribution
`~st~ms. In addition, analog and digital television pro(cid:173)
`~u:~~n and distribution systems may coexist in many
`iacili~es and networks, frequently with both SD and
`HD video formats. This situation creates added com(cid:173)
`;lexity for closed captioning as well as other aspects
`~system implementation. A discussion of how to
`mplement DTVCC, together with a summary of the
`~levant FCC rules, is in a later section of this chapter.
`.,,;, ,.
`
`·~···
`
`~rhr
`DISPLAY FORMAT
`. ~alog CEA-608 captions can operate in any of three
`~rent styles: pop-on, roll-up, and paint-on. DTV
`Jir ·'.08 captions also allow text to scroll in multiple
`ections.
`:ec~ Pap-on style is used when the captions are pre(cid:173)
`~o ed: Captions are transmitted to a nondisplayed
`:apti ry m advance of the time they are to appear. The
`~tr on d~ta. stored in this memory is then displayed
`· ansmiss1on of a single display code.
`
`CEA-608 Captions
`CEA-608 captions (as well as material displayed in
`line 21 Text Services) may have any of several
`attributes in addition to block monochrome charac(cid:173)
`ters. Attributes include upper- and lowercase, six dif(cid:173)
`ferent colors, italics (or slanted text), underline, and
`flash. The caption attributes are determined during
`caption or text authoring and are communicated to the
`decoder using special control codes in the data stream .
`The display of CEA-608 captions consists of 15 rows
`with up to 32 monospace characters per row. Decoder
`controls permit the viewer to select Captions or Text.
`In the Captions mode, a maximum of four rows is
`used onto which each caption pops on when prere(cid:173)
`corded captions are received and rolls up when the
`captions are live. The four rows can appear anywhere
`on the 15 displayable rows which occupy most of the
`screen area. The first row starts at line 43 in each field.
`Each row occupies 13 lines of a field scan or 26 lines of
`the 525 lines. Each row of characters is displayed
`within a black surround box to enhance the readability
`against the normal video background. The box
`extends one character position to the left of the initial
`character in each row and one character to the right of
`. each row.
`
`CEA-708 Captions
`The display of one CEA-708 caption service consists of
`up to eight windows of variable dimensions and posi(cid:173)
`tions that contain the text. Caption text may be pro(cid:173)
`portional or monospace, with or without serifs, in up
`to three character sizes. Up to 64 colors are allowed, as
`well as a range of transparency levels, for characters,
`character edges, and window backgrounds. Support
`for these features among decoders varies, however,
`because FCC rules require decoders only to support a
`subset of the features defined in CEA-708.
`
`CAPTION DATA ENCODERS
`Encoding is the process of inserting the caption data
`into the VBI of an analog television signal or into the
`appropriate data packet of a DTV signal. The data
`contains the caption text in addition to positional
`instructions and display attributes (for instance,
`color and italics). The encoder is placed in the video
`path of the program to be captioned. There are two
`
`1437
`
`Ultratec Exhibit 1008
`Ultratec v Sorenson IP Holdings Page 6 of 21
`
`

`

`SECTION 5: VIDEO PRODUCTION AND STUDIO TECHNOLOGY
`
`types of the line 21 VBI encoders: smart encoder and
`simple encoder.
`
`Simple Encoder
`The simple encoder was the first type of line 21
`encoder. It unconditionally strips any data already
`encoded on line 21. It generates the line 21 data signal,
`to be inserted on a video signal, from caption data
`received at an RS-232 serial input. This encoder is used
`exclusively in postproduction applications for insert(cid:173)
`ing pre-authored captions on program material where
`there is no previously recorded caption data. The sim(cid:173)
`ple encoder cannot add captions to a video signal
`already containing line 21 data.
`
`Smart Encoder
`The smart encoder is used to insert caption data into
`line 21 of fields 1 or 2 of the VBI. Under user controls,
`it can strip all previously encoded data contained in
`the video, or intelligently leave one or more data chan(cid:173)
`nels intact while inserting new data. (Note that all pre(cid:173)
`viously encoded data that is to be passed through
`intact will actually be stripped and reinserted typi(cid:173)
`cally two video frames later.)
`The smart encoder typically receives new data for
`encoding through a serial data port or via telephone
`line. Using a smart encoder, locally produced Caption
`and Text Mode Services may be added to an already
`closed captioned video program or if noncaption text
`is contained on line 21. If captions are present on the
`incoming program, noncaption data may be inter(cid:173)
`leaved into the gaps between the captions by the
`smart encoder.
`· Smart encoders are used in the process of inserting
`real-time captions, live-display captions, and captions
`derived from scripts generated by newsroom comput(cid:173)
`ers (also used to feed teleprompters), and can also be
`used by off-line captioning systems that control the
`timing and preparation of data (including the genera(cid:173)
`tion of all control codes needed to manage the caption
`display).
`
`Digital Encoders
`Since the advent of DTV captions, encoder manufac(cid:173)
`turers have produced a new line of encoders to deal
`with analog and digital broadcasts. See the section on
`implementing DTVCC for more information on vari(cid:173)
`ous types of digital encoders.
`
`AUTHORING CAPTION DATA
`Several techniques are employed to create caption
`data based on whether the broadcast is live or prere(cid:173)
`corded. There are also variations within the live and
`prerecorded captioning technologies.
`
`Prerecorded Captioning
`Prerecorded captioning (off-line or nonlive
`.
`involves the preparation of closed captio captioning)
`grams that have been recorded prior to th n~ for pr0-
`The captions are created and then mer e eir t~lecast.
`video signal and a closed captioned vide~t d With the
`of the program is made. A typical block dia ape Version
`prerecorded captioning. process is showfr~m ~f the
`5.24-2. Th.e ~ain step_s in theyrocess are as foll Figure
`A captiorung facility receives a time-cod d ows.
`a program master either on videotape or a e c~py of
`video file. A caption author reviews each ss a dig1tal(cid:173)
`scribes the soundtrack, and formats the trancsen~, ~an-
`·
`·
`d.
`cnpt int
`iscrete captions usmg a personal computer S
`0
`quality-control checks are performed, induct: everal
`ing, grammar, syntax, timing, and screen po~% sp~U­
`of the captions. The captions are saved as a fil 0ni:ig
`their corresponding display time codes. The ~~1th
`fil 15
`then transferred by any of the usual means f
`transfer (LAN, WAN, modem, or by e-mail ctf~ e
`flash memory) to the location where the progra~ .' ~r
`15 0
`have the captions added or "encoded."
`The recorded program and caption data are th
`merged, using the time c~des to trigger the captio~~
`to produce a closed captioned master video signal
`which is typically recorded but may be transmitted
`live to air. Captions that are produced in advance are
`displayed on a caption-equipped receiver as pop-on
`captions, each one complete and timed to coincide
`with the spoken dialogue on the screen.
`
`Live Captioning
`Live captioning involves the addition of caption data to
`the television signal at the time of a live transmission
`or broadcast. Examples of live captioned program·
`ming include news programs, sporting events, lire
`events (such as meetings, news conferences, and
`award shows), and special bulletins or reports. There
`ff· are three methods of captioning live programs: r~al·
`time, live-display, and newsroom computer. A typKal
`block diagram of the live captioning process is shown
`in Figure 5.24-3.
`
`/
`
`Real-Time Captions
`Real-time captions are created and transmitted silll:ul·
`taneously. There are several methods of gene(a~ng
`real-time captions: (1) by a court reporter -~~
`called a "stenocaptioner") using a stenotype ors~ a
`machine that deciphers phonetic short~and !(Feho0~.
`special keyboard; (2) by a "voicecaptioner w eats
`tens to the soundtrack and simultaneously ~e~ has
`each word into a speech-recognition syste~ to~ ut;
`been modified to generate a formatted caption tin~ to
`or (3) by a speech-recognition syste~ attemp
`transcribe directly from program audio.
`
`0
`
`an
`Steno captioning
`.
`tioner c
`With a stenotype machine, the real-time rp etic codes
`key in up to 260 words per minute. The P on
`
`1438
`
`Ultratec Exhibit 1008
`Ultratec v Sorenson IP Holdings Page 7 of 21
`
`

`

`CHAPTER 5.24: CLOSED CAPTIONING SYSTEMS
`
`Videotape
`Recorder
`
`Video
`
`Video
`Monitor
`
`Time-code
`Reader
`
`Personal
`Computer
`
`Personal
`Computer
`
`VTRVideo
`Master
`
`Caption
`Data
`
`Character
`Generator
`
`Encoder
`with Time-code
`Reader
`
`Caption
`Data
`
`Time-code
`
`-
`
`Time-code
`Corrected Video
`Prerecorded captioning system. Top: caption authoring; bottom: encoding.
`
`'~
`
`·.~
`
`Encoding
`
`Authoring
`and Local
`Encoding
`'.&:'
`'1:'~
`·.~ Modem
`~ ·······--------
`Modem
`
`I.t·/
`...,,-.~
`
`._,
`
`Off-site
`Encoding
`
`',;!£ j
`'
`~ .. l :
`~'
`i~~
`l1'
`
`Authoring
`""*r
`,~1li
`·~~~
`JW. I'
`£gA
`········---------- --------------------------·-------------------------------------------------------
`;::, · Diskette
`~::·~;Modem
`i~ _i:f!ft
`
`Video with
`Caption Data
`
`VTRVideo
`Sub master
`
`Video with
`Caption Data
`
`Monitor with
`Captions
`Displayed
`
`Live
`Video
`
`I
`
`I
`
`Stenotype
`Machine
`
`Personal
`Computer
`
`)I
`
`Audio
`Monitor
`
`Caption
`Data
`
`)I
`
`Video
`Monitor
`
`Video with
`Caption Data
`
`~
`
`Smart
`Encoder
`
`-------- ----------------------------------------------------------------------------------
`
`Caption
`Data
`
`Smart
`Encoder
`
`..
`
`Transmitler
`
`Video with
`Caption Data
`
`Monitor with
`Captions
`Displayed
`
`Master
`Video
`
`I
`
`I
`FIGURE 5.24-3 Live captioning system. Top: authoring and local encoding; bottom: off-site encoding.
`
`1439
`
`Ultratec Exhibit 1008
`Ultratec v Sorenson IP Holdings Page 8 of 21
`
`

`

`SECTION 5: VIDEO PRODUCTION AND STUDIO TECHNOLOGY
`
`are translated into English words by a computer that
`has been programmed with the phonetic codes and a
`dictionary customized to the individual court reporter.
`From the computer, the words and caption control
`codes are sent via a data circuit to a caption encoder
`where they are encoded into the television signal as
`real-time, or roll-up, captions.
`A typical lag time between the time a word is spo(cid:173)
`ken and the same word appears in a caption on a
`viewer's screen is two to three seconds.
`Errors in real-time stenotype-generated captions
`are generally the result of one of four factors: (1) the
`captioner may hit the wrong keys; (2) the captioner
`may be unable to keep pace with the speakers; (3) the
`captioner may not hear the spoken word correctly; or
`(4) the computer may not have a particular word in its
`"dictionary," which is often the case when unfamiliar
`proper nouns (such as the names of people or places)
`are spoken.
`
`Voicecapti oning
`Captioning by a voicecaptioner is similar to stenocap(cid:173)
`tioning except that input is by voice rather than key(cid:173)
`board. The voicecaptioner trains a customized speech(cid:173)
`recognition system to understand his/her voice accu(cid:173)
`rately and quickly. Ideally, the speaker uses _a high(cid:173)
`quality microphone in a controlled, sound-isolated
`environment. The speech recognition software must
`·be "trained" to the voice of the voicecaptioner. A sys-
`tem that must resolve only one voice has many advan(cid:173)
`tages over a system that must recognize multiJ?le
`voices, especially if those voices are recorded or deliv(cid:173)
`ered via low-fidelity methods.
`Errors in real-time voice-generated captions are
`generally the result of one of three factors: (1) the cap(cid:173)
`tioner may say the wrong word; (2) the captioner may
`say the wrong form of a word (failing to distinguish
`between conflicts such as "their" and "they're" and
`"there"); or (3) the computer may not have a particular
`word in its "dictionary," which is often the case when
`unfamiliar proper nouns (such as the names of people
`or places) are spoken.
`
`Speech-Recognition Captioning
`At the time of this writing, speech-recognition systems
`that attempt to transcribe program audio directly with
`no human intervention are beginning to appear.
`Although it is anticipated that such speech-to-text sys-
`
`terns will improve greatly in coming year
`these systems are generally far less ace~' at this tiin
`~arther be~d pr?gr~m audio than the s~:te anct la;
`mg and v01cecaphomng systems.
`nocaption.
`
`Live-Display Captions
`Live-display captioning is used when an ace
`is available in advance 0£ a televised live ev:te Script
`a speech or newscast. The scripted words t such as
`verted to captions by an editor and stored are con.
`puter disk. When the live event is televised ~ha c~rn­
`~anual~y trigg~rs e~ch c_aption for inserti~n .e editor
`video signal. With hve-d1splay captioning th IIlto the
`are timed to appear on the television receivere Words
`are being spoken. If last-minute changes are :sdth~y
`.
`.
`b
`uia e lJl
`th
`e ~cnpt, it may e necessary to switch to real-tirn
`caphomng.
`e
`
`Newsroom Computer Captions
`
`The third method of live captioning is the generati
`of captions through the use of newsroom compu~n
`equipment. Many television newsrooms convert nei ~
`stories from their word processors into data for use~·~
`a prompting machine for the news reporter to read
`while on the air. At the same time the reporter reads
`the script from the prompter, the computer controlling
`the teleprompter passes the data through a serial inter(cid:173)
`face to the caption encoder that inserts the captions
`into the video signal (see Figure 5.24-4).
`There are some dangers in this process. It is not as
`automatic as might be assumed. Some newsrooms add
`cues to the prompting text designed for the on-air per(cid:173)
`sonnel, some of which might be inappropriate to
`appear as captions. Program breaks or changes of
`scene may require changes in the prompting operatio~
`that could interrupt the captioning process. Finally, 1t
`may be necessary for an operator to sign~l ~e caption(cid:173)
`ing encoder to add captions at the begmrnng of the
`)ocal program _and to revert to the pass-thro~gh o~
`. , upstream captions at the end so that prev10usly
`f encoded captions (for example, on a ne~or~ newscas~
`that follows local news) will be transrmtted mtact. It b
`desirable for the operators and program pro?ucers to
`have an off-air monitor to ensure the integnty of the
`captioning process.
`d
`Note that FCC rules prohibit ABC, CBS'. ~ox, an
`NBC and their affiliates in the top 25 telev1s10n mar-
`
`I
`
`I
`
`1440
`
`Video
`Source
`
`I
`
`)I
`
`Monitor with
`Captions
`Displayed
`
`Prompter
`
`Caption
`Data
`
`)I
`
`Smart
`Encoder
`
`I
`FIGURE 5.24-4 Newsroom captioning with teleprompter.
`
`Transmitter
`
`Video with
`Ca lion Data
`p
`
`ti!.
`
`Ultratec Exhibit 1008
`Ultratec v Sorenson IP Holdings Page 9 of 21
`
`

`

`~-
`
`BLANKING
`LEVEL
`
`PROGRAM
`COLOR
`BURST
`
`50 -
`
`-
`
`25 -
`20 -
`
`o-
`
`CHAPTER 5.24: CLOSED CAPTIONING SYSTEMS
`
`7 CYCLES OF 0.503 MHz
`(CLOCK RUN-IN)
`
`TWO- 7 BIT+ PARITY ASCII
`CHARACTERS
`!DATA)
`
`..., __ 12.910uS __ _..
`!0.20H)
`
`i-------- 33.764us - - - - - - " " " '
`{0.53H)
`
`27.452uS
`!0.43H)
`I"'
`10.074uS ll""t------------- 51.26BuS
`!0.16H)
`
`, f
`
`3.972uS
`l0.06Hl
`
`FIGURE 5.24-5 Line 21 data signal waveform specification.
`
`61.342uS
`{O 965H)
`
`'a
`
`~11,
`~'
`
`I. ·.·. ,;. ~:~.l.· .. ·
`
`\~ts from counting news programming using elec(cid:173)
`i ;onic newsroom techniques toward their captioning
`:equirements. Rather, these networks and stations
`· ~ust provide real-time captioning using a stenogra(cid:173)
`. ;tier to convert the entire audio portion of the live
`~ugram to captions. Others may use electronic news-
`~ :OOm techniques to meet captioning mandates.
`
`rJ! .~.
`"/" LINE 21 DATA FORMAT
`; "'1'f1· .~:
`~Captions associated with an analog NTSC television
`yrngram are transmitted as an encoded composite
`~ta signal during line 21 on field 1 of the video sig(cid:173)
`JJ!, as shown in Figure 5.24-5. The signal consists of a
`'.:lockrun-in signal, a start bit, and 16 bits of data corre(cid:173)
`;~nding to 2 bytes of 8 bits each (7-bit character code
`•¢us 1 parity bit). Therefore, transmission of actual
`~tais2bytes every 1/30 of a second, equivalent to 60
`i~tes (characters) per second. The data stream also
`
`contains control codes that provide the instructions for
`the timing, screen position, and attributes of the data.
`The clock run-in consists of a seven-cycle sinusoidal
`burst that is frequency and phase locked to the caption
`data clock. The frequency of 32 fH (32 x 15,734.26 Hz=
`0.503496 MHz) is twice that of the data clock and pro(cid:173)
`vides synchronization for the decoder clock. The clock
`run-in signal is followed by the equivalent of 2 data
`bits at logical zero level and then a logical one start bit.
`The last two cycles of the clock run-in, the 2 logical
`zero bits, and the logical one start bit constitute an 8-
`bit frame code signifying the start of data, as shown in
`Figure 5.24-6.
`The 7-bit character data is coded in a nonreturn-to(cid:173)
`zero (NRZ) format. An 8th bit is added to each charac(cid:173)
`ter to provide odd parity for error detection.
`The sequence of identification, control, and charac(cid:173)
`ter code transmission is shown in Figure 5.24~7 and
`Figure 5.24-8. Each caption transmission is preceded
`
`;.le lllllP...
`~
`
`
`
`
`- -__ __ __
`2
`1
`3
`4
`5
`
`CL~CK ~UN-:
`
`,,
`
`9
`
`10
`
`12
`
`13>i
`
`START
`
`I< BIT -i<
`
`DATA....__.
`
`I 0
`JE
`
`