Obsolete Product(s) - Obsolete Product(s)

Transcription

1 AN1841 APPLICATION NOTE Closed Caption and XDS Software 1 INTRODUCTION This document describes the Closed Caption and extended Data Service (XDS) extraction software module available for the ST92x196 family. 2 COMPOSITE VIDEO SIGNAL 2.1 Introduction The composite video signal waveform is the amplitude-versus-time representation of the signal produced jointly by the television camera, the camera-control unit, the synchro generator and, for color operation, the color encoder. Signals considered in this chapter apply to the 525- line NTSC color television system used in the USA. 2.2 Characteristics Picture scan, is from left to right and top to bottom, consisting of 525 horizontal lines per frame and 30 frames per second. Each frame is divided into two alternating fields. The polarity of modulation is negative. 70% of the amplitude range is selected to transmit the brightness information. The 30% left are available for the synchronizing information, needed to lock the picture beam to the camera or other source. In color television transmissions, two synchronization signals are used to control deflection, a third one provides the color-coding information. AN1841/0704 Rev. 2 1/17

2 2.2.1 Horizontal synchro The horizontal synchro is used to blank the source output during the horizontal retrace interval time. Line frequency = Hz/2 x 525 lines = Hz Figure 1. Horizontal synchro HORIZONTAL BLANKING BLANKING LEVEL PROGRAM COLOR BURST HORIZONTAL SYNCHRO 2/17

3 2.2.2 Vertical synchro The vertical synchro is used to blank the source output during the vertical retrace interval time. Also, in order to respect a good interlace, signals are different on two successive fields. In one case the last horizontal pulse is one half-line period preceding the vertical blanking interval, and in the next field, the last horizontal pulse will followed by a full line period. Vertical synchronizing is guaranteed by doubling the horizontal pulse frequency during the vertical retrace. Figure 2. Vertical synchro FIELD 1 VERTICAL BLANKING INTERVAL 3 LINES 3 LINES 3 LINES PRE-EQUALIZING VERTICAL SYNCHRO POST-EQUALIZING LINE 21 FIELD 2 VERTICAL BLANKING INTERVAL 3 LINES 3 LINES 3 LINES PRE-EQUALIZING VERTICAL SYNCHRO POST-EQUALIZING LINE Color synchro The third synchro signal is the color-synchronizing burst which gives the reference phase for the chrominance subcarrier. The synchronizing information is contained in a short burst of the subcarrier frequency which follows the horizontal pulse and is completed before the start of the picture modulation. 3/17

4 3 CLOSED CAPTION SIGNAL 3.1 Introduction As described above, TV lines during the vertical retrace do not contain active video portion (invisible lines). So, these lines are free to carry composite data signal (Teletext, Closed Caption, VPS...). This chapter presents the video signal specification and encoding format for the Closed Captioning signal. It is intended to summarize the information given in EIA608 & EIA 744 specifications of the Federal Communications Commissions Rules and Regulations regarding the use of the line 21 of the television vertical blanking interval for closed captioning. 3.2 General Closed captions of program related material of a broadcast nature will be transmitted in the form of an encoded composite data signal during the unblanked portion of the line 21 of the standard NTSC video signal. The composite data signal consists of a run-in clock signal, a start bit and 16 bits of data. Suitable decoders with the aid of the run-in clock signal and the start bit separate the data signal from the television signal and provide a data stream at a rate of 480 bits/s (16 bits/field1 or field2 with 30 field1 or field2 per seconds). This data stream contains encoded information which provide the instructions for display formatting, language selection and alpha-numeric characters which provide the captions of full page text to be added on the television receiver screen. 3.3 Video signal waveform The composite data signal contained within the active video portion of line period carries a runin clock (data synchronizing signal), a start bit, followed by 16 data bits. The instantaneous data rate is thirty-two times horizontal line scanning frequency (Fh=15734Hz) for a nominal value of bits per seconds. The run-in clock consists of a 7-cycle sinusoidal "burst" which is frequency and phase locked to the Caption data signal. It is intended to provide synchronizing information for the decoder data-clock. The run-in clock signal is followed by an interval of 2 bits duration during which the data signal is at blanking level. At the end of this 2 bits interval signal, a start bit is transmitted. Figure 3. video signal waveform 50 IRE PROGRAM COLOR BURST 7 CYCLES OF 503 KHz (RUN-IN CLOCK) TWO 7 BIT + PARITY ASCII CHARACTERS (DATA) 25 IRE 20 IRE 0 IRE -20 IRE -40 IRE BLANKING LEVEL START BIT te: Legend: IRE: A unit of measurement used to describe the amplitude characteristics of a video signal. Pure white is defined to be 100 IRE and the blanking level is defined to be 0. 4/17

5 3.4 Transmission format The data transmission uses a basic non-return-to-zero (NRZ) code. All control and alpha-numeric characters use the 7-bit code of the USA Standard Code for information interchange (USASCII). An eighth bit is added to each character to provide odd parity for error detection. Bits are transmitted in numerical order from bit 1 to 8. Figure 4. Transmission format FIELD 1 FRAME As described in the video Signal waveform section, each line 21 carrying the composite data signal contains sixteen data bits. Thus, each transmitted television data line contains two complete USASCII characters or control codes (non-printing characters). Control codes are transmitted twice to insure correct reception of control information (An error in control information can affect an entire caption, while individual character errors are less significant). Figure 5. COMPOSITE VIDEO SIGNAL S b1 b2 b3 b4 b5 b6 b7 P b1 b2 b3 b4 b5 b6 b7 P LINE 21: START BIT + CHARACTER + PARITY + CHARACTER + PARITY Line P P first character S second character T During closed caption, test or text transmission, a NUL character will be transmitted when no printing characters or instructions are being sent. The combination of run-in clock burst and the transmitted NUL code or other ASCII character may be used to determine the presence or absence of a captioned, test or text program and to provide automatic decoder display blanking if desired. 5/17

6 4 DATA SLICER CELL 4.1 Introduction Data Slicer has been originally developed for the USA market in order to extract Closed Caption data from a composite video signal. When used in conjunction with the On Screen Display, the Data Slicer allows Closed Caption or subtitling to be easily programmed. 4.2 Data Slicer operation Data extraction can be programmed for a selected line (LN[5:0], in Control Register 1) in field 1 or field 2, from a 2V amplitude video signal (+/- 3dB). When the line counter value matches the value programmed by the user, the selected TV line is fed into the Data Slicer for data extraction. An interrupt is generated at the end or at the beginning of the specified line or at the end or at the beginning of the corresponding one of the opposite field, even if no data has been recovered (the beginning / end of the line interrupt generation is selected by the IRQ_INV bit of the Control register 1). The Monitor Register gives information about the received data, and needs to be checked in the IT subroutine. 4.3 Phase compensation In 2H (double scan) mode, the phase difference between the 2H deflection pulses and the horizontal sync pulses extracted from video can be large. In order to prevent the Data Slicer logic not to detect vertical retrace pulses (see vertical retrace detection), a delay must be added (PHD[2-0], in Control Register 2) to the extracted synchro pulses to compensate a substantial phase difference. 5 DISPLAY DRIVER 5.1 OSD The On Screen Display (OSD) is used to display rows of characters on a Video Display Screen (menu systems, Closed Caption...). To give more flexibility, a 2-row buffer OSDRAM architecture is used. During the time one row buffer is displayed on the screen, the CPU can prepare the content of the other row buffer to be displayed. The vertical placement of the next row of characters is given as a vertical address by programming the Event Line register. At the time the Scan Line counter matches the Event Line value, the OSD switches from one row buffer to the other. An interrupt is generated each time the swap occurs (see On Screen Display) so that you have time to fill the next buffer to be displayed with the correct bytes. 6/17

7 5.2 Display Buffer The Closed Caption data are recovered by the Slicer Cell and computed in the Closed Caption IT subroutine. Information to be displayed is stored in the display_buffer[ ] array. Its size depends on the user choice as far as the number of rows displayed in TEXT mode (from 7 to 15) is concerned. As a result, display_buffer[ ] contains from 8 to 16 lines of 36 characters each. Figure 6. Display Buffer Display Buffer Displayed row The OSD interrupt subroutine is responsible for transferring the content of each display_buffer[ ] row to the OSDRAM row buffers. On top of that, 3 bytes are linked to each row of display_buffer[ ], one for the vertical placement on the TV screen (vertical address in display_address[ ]), one to save the location of the next row to display (display_position[ ], see next section) and at last one to save all display_buffer[ ] related flags (display_flags[ ]). Figure OSDRAM Row Buffer display_control display_address display_position display_flags display_buffer /17

8 5.3 Display Buffer set-up In normal display mode such as local menu systems, OSD rows are linked to their positions in the Display Buffer and to their vertical placements (vertical address) on the screen (row 1 will be the first row to be displayed, row two will be the second row to be displayed and so on). For the Closed Caption, OSD rows are linked only to their vertical placement on the screen, regardless of their positions in the Display Buffer (row four can be the first row to be displayed, row six can be the second row to be displayed and so on). So, with Closed Caption processing, OSD interrupts could be very long because within each interrupt we should check the entire Display Address Buffer in order to find the next OSD row to be displayed. In order to keep the OSD interrupt function as short as possible (in case of Closed Caption processing), the Display Buffer will by set-up once every field, as soon as all of the closed caption rows have been displayed. As a result, each displayable row will be linked to the position of the next row to be displayed (display_position[ ]). Figure 8. display_control display_address display_position display_flags C9 D6 VBI OFF OFF OFF ST92196 ON Closed Caption ON Software ON Demo OFF 1 display_buffer /17

9 As a result, with the example above, the TV will display the following: Figure 9. TV line 47 ST92196 TV line C9 TV line D6 Closed Caption Software 5.4 OSD interrupt subroutine Figure 10. OSD interrupt subroutine OSD_IT TV Menu Display Next Menu Row Update OSD registers IRET Display Interrupt subroutine Display Next C. C. Row Last C.C Row Set up the Display Buffer 9/17

10 6 CLOSED CAPTION PROCESSOR 6.1 Introduction The Closed Caption processor can extract both Closed Caption or Gemstar data. Two bits are used to determine the decoded data type so that user can process them with the appropriate algorithm (CCMODE bit and GSMODE bit in the MR register). An interrupt is generated each time the line number programmed in LN[0:5] (CR1 register) matches the TV line counter value, and data extraction is automatically done by the microcontroller. Data are stored in the data registers (DR1 to DR4). 6.2 Software algorithm Data extraction The software will extract only Closed Caption data according to the user choice: CC1, CC2, CC3, CC4, Text 1, Text 2, Text 3 and Text 4 can be sliced. In addition to that, V-Chip, Time of the Day and Local Time Zone (used for Autoclock setting) data are continuously extracted, as soon as they are present in the XDS packet. XDS data values are stored in 3 arrays: prg_rating[ ], time_of_the_day[ ] and local_time_zone[ ]. The XDS extraction subroutine can be easily modified to decode any other data included in other classes (for example, the Network Name, contained in the Channel Information Class,...) 10/17

11 6.2.2 CC interrupt and Monitor CC status subroutines Figure 11. CC interrupt and Monitor CC status subroutines CC_IT Closed Caption Interrupt subroutine Line 21 in Field 1 or Line 21 in Field 2 Update Scrolling Monitor CC Status Figure 12. Monitor_cc_status Current Field = Field1 but user choose Field 2 Configure CC register to slice line 21, next Field Current Field = Field2 but user choose Field 1 New Data Decoded Update CC Register IRET New Data Decoded Process New Data, CC or XDS Process Data Process XDS Data Refresh Screen RET 11/17

12 6.2.3 Rabbit ears In some cases, synchro pulses can be modified by video tape copyguard system (so called rabbit ears) and then alter the parity bit of the second character. So as to detect parity corruption, a function will check the parity differential between the first character and the second character. Figure 13. Check Parity Bits RET Parity error > 1 Parity error - 2 Parity error >= 8 Reset Parity bit 2 Parity of character 1 = 0 Parity of character 2 = 0 Parity error < 16 Parity error + 1 If parity bit 1 is set, then it is not a parity bit corruption Ifparitybit2isset,itcould be a parity bit corruption If the parity differential is very high, it means that something is "playing" with the parity bit 2. We assume that the parity bit 2 has been set by a video tape CopyGuard, and then, we reset the parity bit 2. In addition to this, 1 bit is available (SCG_EN in Control Register 2) to avoid that CopyGuard pulses reach the CC decoding part Automatic turn-on A parity error counter will be incremented or decremented in the process_new_data & process_no_data subroutines to provide the capability to turn-on or turn-off automatically the Closed Caption display, if the video signal quality is too bad to allow a correct and continuous data extraction. 12/17

13 7 SOFTWARE MODULE INTEGRATION 7.1 Resources The Closed Caption module includes 6 files: caption.c, which contains all the subroutines dedicated for the Closed Caption software, osd.c, which contains the OSD interrupt subroutine and the standard TIMER subroutine. They are given as example. You can re-use them as they are, or modify yours so that the closed captioning is displayed. caption.h, which contains all the Closed Caption constants Initialisation (Including the number of row to be displayed in Text mode : This value is user selectable, and can be between 7 and 15), proto_cc.h, which contains all the prototypes definition for the caption.c file, appli196.h, which contains other constants, cc_color.c, which contains the color palettes definition for Closed Caption (defined as an array of constants). Of course, you can modify the constant definition in the appli196.h file so that the 2 bits used are merged with flags you already have defined. The required memory spaces for the Closed Caption module are the following: ROM size: about 5000 bytes RAM size: 671 bytes if 15 rows are displayed in TEXT mode, 351 if only 7 rows are displayed. Please note that you can share the same RAM area for the Closed Caption data and the OSD display buffer you can have for your menus, because both area will never be used at the same time. This will reduce the amount of RAM used for your application. Nevertheless, you have to take care of the transition between the 2 working modes (Closed Caption and menu): be sure that the shared RAM area is not any mode used and accessed by one working mode before switching to the other one. 7.2 integration guide General Initialization Interrupts When the closed caption data slicer is enabled, the OSD interrupt priority needs to be higher than the Closed Caption interrupt. That is why you need to set the 4 MSB of the EIPLR register (R245 page 0) value to You need also to reset (if needed) the corresponding pending bits (bit 4 and bit 6) in the EIPR register (R243 page 0) and then to enable the corresponding interrupts (bit 4 and bit 6) in the EIMR register (R244 page 0) Slicer Setup To start the Closed Caption data extraction (and / or XDS data extraction) you need to call the init_closed_caption(user_choice) subroutine. The user_choice value will correspond to the data extraction you which: 1= CC1; 2 = CC2; 3 = CC3; 4 = CC4; 5 = TEXT1; 6 = TEXT2; 7 = TEXT3; 8 = TEXT4 13/17

14 You can choose the number of rows displayed in TEXT mode. For this, use the LAST_ROW constant defined in the caption.h file. Warning, the value for it must be between 7 and 15. You can select a time-out value after which the TEXT box is removed from the TV screen if no valid data for the selected TEXT channel has been received. This value is based on a TIMER1 counter incremented by the Standard Timer IT every 0.64s. The TEXT_TIMOUT value is initialized in the caption.h file. You can of course use another time base, but you need to correct the TEXT_TIMOUT value plus the counter name (here TIMER1, used in CC_IT() subroutine) according to you application. Then you can set or reset the TX_CC_DATA bit in the CC_FLAG byte if you want the Closed Caption to be decoded and automatically displayed plus the XDS data to be decoded (bit set), or if you only want the XDS data to be decoded (bit reset, plus initialization done with slicing on field 2, i.e. user_choice = 5 to 8) Flash Function The CC_FLAG contains a bit that is set or reset by the Standard Timer IT, and which is dedicated to the flash feature: OSD_FLASH_ON bit. So you just need to add the update of this bit on the time base you which to have the flash effect on the TV screen. This bit reflects the logic value of the FLA bit in the OSD row attribute (set = flashing characters displayed as spaces, reset = flashing characters displayed as normal characters). Refer to the STDT_IT() subroutine in the osd.c module for example Color Setup When initializing Closed Caption, the basic color palettes (foreground and background) will be set up with the correct color set (init_cc_color_palettes() subroutine in caption.c file) Software Compiling You will need to add the following files in your makefile: osd.c caption.c cc_color.c In addition to this, the script file must be modified: the content of the cc_color.asm file is defined as data and must be linked in ROM, after the text section. For this, add cc_color.o(.data) just after the *(.text) instruction of the text section in your script file. This software has been compiled with compiler version 4.3, with the following options for C files: -tr9 -mlink $(INC) -D$(Option) -c -g -mparmusp -fomit-frame-pointer -O -Wall -Wa,-ahld INC is the include directories, and Option = none Starting Data decoding Closed Caption data decoding As soon as Closed Caption slicer has been initialized, and the Closed Caption interrupt has been enabled, the Closed Caption data are automatically extracted and displayed if the TX_CC_DATA bit of CC_FLAG byte has been set. 14/17

15 XDS data decoding As soon as Closed Caption cell has been initialized, and the Closed Caption interrupt has been enabled, the XDS data are automatically extracted. The main subroutine for slicing XDS data is set_xds_flags(). It tests the value of the received bytes, and sets bits in the XDS_FLAG and XDS_FLAG1 bytes to indicate the XDS data reception status (refer to the caption.h file for the bit description). This subroutine will extract XDS data for: Program Rating : data are stored in the prg_rating[] array. When the program rating data (2 bytes) are all received twice, the PR_VALID bit is set in XDS_FLAG. Then you can use the test_xds_data() function to test if the program rating data are coherent with the EIA specification (test bit 6 of first byte, and if the second data byte is used, test its 6th bit too). If you decided to slice only XDS data (Closed Caption not displayed), the slice_xds_data() subroutine is called automatically. Then, another subroutine is called to display these data on the screen as soon as they are extracted : display_vchip_data(). You can remove the calls to these subroutines if you don t need it. If you decide to keep it, you will need to rewrite it according to your application. The one given is only for example. Time Of the Day : data are stored in the time_of_the_day[] array. When the time of the day data (6 bytes) are all received, the TOD_VALID bit is set in XDS_FLAG. Local Time Zone : data are stored in the local_time_zone[] array. When the local time zone data (2 bytes) are all received twice, the LTZ_VALID bit is set in XDS_FLAG1. If you need to extract more XDS data, you can easily modify the set_xds_flags() subroutine using the same structure as the one already existing. 15/17

16 8 REVISION HISTORY Table 1. Revision History Date Revision Description of Changes March First Issue 7-July Stylesheet update. content change. 16/17

17 The present note which is for guidance only, aims at providing customers with information regarding their products in order for them to save time. As a result, STMicroelectronics shall not be held liable for any direct, indirect or consequential damages with respect to any claims arising from the content of such a note and/or the use made by customers of the information contained herein in connection with their products. Information furnished is believed to be accurate and reliable. However, STMicroelectronics assumes no responsibility for the consequences of use of such information nor for any infringement of patents or other rights of third parties which may result from its use. license is granted by implication or otherwise under any patent or patent rights of STMicroelectronics. Specifications mentioned in this publication are subject to change without notice. This publication supersedes and replaces all information previously supplied. STMicroelectronics products are not authorized for use as critical components in life support devices or systems without express written approval of STMicroelectronics. The ST logo is a registered trademark of STMicroelectronics. All other names are the property of their respective owners 2004 STMicroelectronics - All rights reserved STMicroelectronics group of companies Australia - Belgium - Brazil - Canada - China - Czech Republic - Finland - France - Germany - Hong Kong - India - Israel - Italy - Japan - Malaysia - Malta - Morocco - Singapore - Spain - Sweden - Switzerland - United Kingdom - United States of America 17/17