FLD2
FLD1
FLD0
CVBS9
CVBS8
CVBS7
CVBS6
CVBS5
CVBS4
CVBS3
CVBS2
CVBS1
CVBS0
RESET
PXCK
V
D
V
R
AGND
DD
GND
REF
REFCOMP
Pin Definitions
Pin NamePin NumberValueDescription
CLOCK, SYNC, & CONTROL INPUTS (6 pins)
DCVEN57TTLDigital CVBS Output Enable. When DCVEN is LOW, the
Comp2 output prior to the D/A is routed to D7-0, FLD2-1
providing a digital composite output. When DCVEN is HIGH,
D7-0 and FLD2-1 operate in their normal mode.
HSIN56TTLHorizontal Sync Input. When operating in slave, Genlock, or
DRS-Lock the TMC2192 will start a new horizontal line with
each falling edge of HSIN.
KEY20TTLHard Key selection. When the control register bit HKEN is set
HIGH and the hardware KEY pin is high, the video data
considered to be the foreground. is routed to the COMP2
output. This control signal is data aligned so that the pixel that is
present on the PD port when KEY signal is latched is at the
midpoint of the key transition. When HKEN is LOW, Key is
ignored.
4REV. 1.0.0 8/13/03
PRODUCT SPECIFICATIONTMC2192
Pin Definitions (continued)
Pin NamePin NumberValueDescription
PXCK95TTLPixel Clock Input. PXCK is a clock signal that period is twice
the sample rate of the pixel data. The operating range is 20 to
30 MHz. The clock is internally divided by 2 to generate the
internal pixel clock, PCK. PXCK drives the entire TMC2192
except the asynchronous microprocessor interface.
RESET
VSIN
SYNC & CONTROL OUTPUTS (11 pins)
FLD[2:0]81–83TTLField Identifier. Field Identifier outputs the current field number.
HSOUT74TTLHorizontal Sync Output. The alignment of HSOUT to the pixel
LINE[4:0]76–80TTLVertical Blanking Interval Line Identifier. LINE identifies the
PDC73TTLPixel Data Control.
94TTLMaster Chip Reset. When LOW, All outputs are tri-stated and
the internal state machines and control registers are reset. At
rising edge of RESET, all outputs are active, the preset values
will be loaded into the control registers and the internal states
machines start to operate.
55TTLVertical Sync Input. When operating in slave, Genlock, or
DRS-Lock the TMC2192 will start a new vertical field with each
falling edge of VSIN that is coincident with HSIN.
For all video standards the field identifier will cycle through the
eight counts.
data port or DCVBS port is controlled by control register
TSOUT.
current line number for the first 31 lines. If the line count is
greater than 31 then LINE is 11111b. The first line with a vertical
serration is considered to be line 0.
When PDCDIR = LOW: At a rising edge, The next pixel starts a
controlled ramp of the PD data. At a falling edge, the pixel prior
is the last PD used in the ramp. The rising edge is determined
by the PDCCNT control register, the falling edge of PDC is
determined by the horizontal timing registers.
When PDCDIR = HIGH: PDCIN is used to override the internal
PDC. When HIGH, the internal PDC controls the blank and
unblank window. When LOW, the video remains blanked
regardless of the internal PDC. All edges have the same ramp
control as the internal PDC.
VSOUT75TTLVertical Sync Output. The alignment of VSOUT to the pixel
data port or DCVBS port is controlled by control register
TSOUT.
DATA INPUTS (39 pins)
CVBS[9:0]84–93TTLComposite Data Input
OL[4:0]21–25TTLOverlay Control
PD[23:0]27–38, 41–52TTLComponent Data Input
ANALOG INTERFACE – Video Out (5 pins)
LUMA101.35Vp-pLuma
CHROMA51.35Vp-pChroma
COMP21.35Vp-pComposite D/A with optional keying
REV. 1.0.0 8/13/035
TMC2192PRODUCT SPECIFICATION
Pin Definitions (continued)
Pin NamePin NumberValueDescription
ANALOG INTERFACE – Support (9 pins)
C
BYPLUMA
C
BYPCHROM
C
BYPCOMP
R
REFLUMA
R
REFCHROM
R
REFCOMP
V
REF
MPU INTERFACE (13 pins)
A[1:0]/SA[1:0]61, 62TTLWhen SER (HIGH), OLUT/control/pointer address.
CS/SCL59TTL/R-BUS When SER (HIGH), microprocessor port clock.
D[7:0]63–70TTLBi-directional Data Bus.
RW/SDA60TTL/R-BUS When SER (HIGH), read/write control.
SER58TTLMicroprocessor Select. When LOW, the serial interface is
POWER & GROUND (17 pins)
A
GND
D
V
V
GND
DD
DDA
26, 40, 53, 71, 970.0VDigital ground
110.1 µFReference Bypass Capacitor for LUMA DAC. Connection
point for 0.1 µF Capacitor.
60.1 µFReference Bypass Capacitor for CHROMA DAC. Connection
point for 0.1 µF Capacitor.
30.1 µFReference Bypass Capacitor for COMPOSITE DAC.
Connection point for 0.1 µF Capacitor.
131210 OhmCurrent Setting Resistor. Connection point for external current
setting resistor for LUMA DAC. The resistor is connected
between R
inversely proportional to the value of R
81210 OhmCurrent Setting Resistor. Connection point for external current
setting resistor for CHROMA DAC. The resistor is connected
between R
inversely proportional to the value of R
991210 OhmCurrent Setting Resistor. Connection point for external current
setting resistor for COMPOSITE DAC. The resistor is connected
between R
inversely proportional to the value of R
981.235 VVoltage Reference Input. External voltage reference input,
internal voltage reference output, nominally 1.235V.
When SER (LOW), SA[1:0] of serial chip address SA[6:0].
When SER (LOW), serial bus clock.
When SER (LOW), serial bus bi-directional data.
enabled. When HIGH, the parallel interface is enabled.
4, 9, 14, 15, 18,
19, 100
39, 54, 72, 96+5.0VDigital positive power supply
1, 7, 12, 16, 17 +5.0VAnalog positive power supply
0.0VAnalog ground
REFLUMA
REFCHROM
REFCOMP
and GND. Output video levels are
and GND. Output video levels are
and GND. Output video levels are
.
REF2
REFCHROM
REFCOMP
.
.
6REV. 1.0.0 8/13/03
PRODUCT SPECIFICATIONTMC2192
Functional Description
Input Formats
Control Registers for this section
AddressBit(s)Name
0x057D1OFF
0x056-4INMODE
0x060TSOUT
The TMC2192 supports YCBCR component sources on the
pixel data port. YCBCR input sources are supported in 10 bit
4:2:2, 20 bit 4:2:2, 20 bit 4:4:4, and 24 bit 4:4:4. In the 4:2:2
cases the color difference components are linearly interpolated to 4:4:4 internally.
INMODE
23
777
00
01
9
1x
9
C
B
YC C
BR
YC C
BR
15897
16
0
Figure 1. Input Formats
Demuxing of multiplexed data streams depends on which
synchronization mode the encoder is operating in. For slave
and genlock modes the falling edge of HSIN must be LOW
prior to the CB data in order to demux the data correctly. For
master mode synchronization the falling edge of HSOUT
must be LOW prior to the Y data in order to demux the data
correctly. Finally, in 656 mode the demuxing of the data
stream is determined by the TRS codes, the first sample after
the TRS is considered a C
sample of the CB Y CR YI
B
packet.
The control register D1OFF controls the formatting of the
incoming luminance data at the pixel data port. When
D1OFF is HIGH a blanking level of 64
is subtracted from
10
the luminance and when D1OFF is LOW the incoming the
pixel data is passed through. The inversion of the MSB’s on
the CB and CR components is controlled by the INMODE
control register.
2. INMODE = 01, PD[23:14] = YCBCR running at 27MHz.
The PD port is clocked at twice the pixel rate, with the data
organized as CB Y CR Y, with the cosited Y's following the
CB's. In its CCIR-656 time base mode, the demuxed CB, Y,
and CR data is synchronized to the SAV preamble. The first
R
t
DO
x = (SY+BR+BU+CBP)*2
t
S
Y
C
B
C
R
t
H
Y
x
x
x
x+1
C
x+1
B
C
x+1
R
Y
x+2
C
x+2
B
C
x+2
R
1280n = (SY+BR+BU+CBP+AV)*2
data value, after the SAV preamble, is treated as a CB data
point in the multiplexed CB, Y, CR Y , D1 data stream.
Note: Figure 3, pixel numbering, reflects the SMPTE-125M
pixel numbering.
2192003A
REV. 1.0.0 8/13/037
TMC2192PRODUCT SPECIFICATION
PXCK
PD[23:14]
Y
C
718
B718
Y
C
719
R718
00
FF
FV
00
EAV
HSOUT
(TSOUT = 1)
n = (SY+BR+BU+CBP+AV)*2
PXCK
PD[23:14]
Y
C
Bn
Y
C
n
n+1
Rn
HSIN
HSOUT
(TSOUT = 1)
3. INMODE = 11, PD[9:0] = Y, PD[23:14] = CB/CR
n = (SY+BR+BU+CBP+AV)*2
PXCK
PD[9:0]
Y
Y
n
n+1
0
C
Y
B736
1
736
t
DO
128
t
t
HS
DO
Figure 3. CCIR656 Input Format
0
C
Y
B0
0
t
SP
t
DO
t
HP
Figure 4. 10 bit Input Format
0
Y
0
Y
1
(SY+BR+BU+CBP)*2
t
t
S
H
C
C
FF
00
00
FV
0
Y
B0
0
Y
R0
1
Y
C
2
B2
SAV
65-6294-04
128
t
DO
128
x = (SY+BR+BU+CBP)*2
t
S
C
Bx
x = (SY+BR+BU+CBP)*2
t
S
Y
t
H
Y
C
Y
x
Rx
x+1
t
H
Y
x
x+1
C
Bx+2
65-6294-05
Y
x+2
Y
x+2
PD[23:14]
C
C
Bn
Rn
C
t
SP
C
B0
R0
HSIN
t
HSOUT
(TSOUT = 1)
DO
t
HS
Figure 5. 20 bit 4:2:2 Input Format
4. INMODE = 10, PD[9:0] = Y at PCK, PD[23:14] = CB-CR at PXCK
n = (SY+BR+BU+CBP+AV)*2
PXCK
PD[9:0]
PD[23:14]
C
B
n
HSIN
HSOUT
(TSOUT = 1)
Y
Y
n
n+1
C
C
C
n
n+1
R
B
Rn+1
0
Y
0
C
C
B
0
0
R
t
SP
t
DO
Figure 6. 20 bit 4:4:4 Input Format
C
t
DO
128
x = (SY+BR+BU+CBP)*2
t
S
Y
t
S
C
B
x
t
DO
C
Bx
x
Rx
t
H
t
H
C
x
R
C
Bx+2
65-6294-06
65-6294-07
8REV. 1.0.0 8/13/03
PRODUCT SPECIFICATIONTMC2192
Color Space Matrix
Control Registers for this section
AddressBit(s)Name
0x307-0MCF1L
0x337-0MCF2L
0x357-0MCF3L
0x3A7-4MCF1M
0x3B2-0MCF2M
0x3C2-0MCF3M
Table 1. CSM Coefficient Range
CoefficientGain RangeComment
MCF10 to 2
MCF20 to 111 bit coefficient.
MCF30 to 111 bit coefficient.
Matrix configuration:
Y
composite
U = MCF2 * C
V = MCF3 * C
The color space matrix consists of 3 multipliers with independently adjustable coefficients, and a resolution of
0.00049 (1/2048). The amount of gain varies among coefficients, Table 1 summarizes the gain for each coefficient.
= MCF1 * Y
in
B
R
To aid in the programming of the color space matrix Table 2
provides a set of default input and output values for 100%
color bars. The component values given will be after the pre-
Table 2. Expected Output Values for the CSM with YCBCR Inputs
Inputs5:2 Outputs7:3 Outputs
Color
White876005360056800
Yellow776-44873475-23554503-24957
Cyan61415144837679-33240784-351
Green514-297-375315-156-278340-165-294
Magenta362297375222156278240165294
Red262-151448160-79332173-84351
Blue100448-7361235-5466249-57
Black000000000
Blank64240256
Pedestal440
Sync812
YC
B
C
R
YUVYUV
processing block and prior to the sync and pedestal insertion.
The blank, pedestal, and sync values are given as a reference.
Table 4 gives the default coefficients values for the CSM.
REV. 1.0.0 8/13/039
TMC2192PRODUCT SPECIFICATION
Synchronization Modes
Control Registers for this section
AddressBit(s)Name
0x065-3MODE
0x061TOUT
0x060TSOUT
The TMC2192 offers a variety of synchronization modes;
these are master, slave, genlock, 656 mode, and DRS-Lock.
In master mode, the TMC2192 generates its own timing and
the synchronization is supplied externally by HSOUT and
VSOUT
derives its timing from the input pins HSIN, VSIN. In 656
mode the timing is driven by the synchronization codes
embedded into the data stream.
Master
The TMC2192 drives the output pins HSOUT and VSOUT
to synchronize the incoming video. A new color frame starts
at the rising edge of RESET. The encoder always starts at the
1st vertical serration in field 8 and will freerun the field and
line sequence. The control register bit SRESET can be used
to synchronize the start of the field and line sequence in master mode by resetting the FVHGEN state machine. Output
synchronization signal VSOUT can operate in a traditional
sync mode or in a MPEG style field toggle mode.
Slave
The TMC2192 is driven by the input synchronization pins
HSIN and VSIN. When the falling edge of HSIN and VSIN
occurs at the same rising edge of PXCK the TMC2192 will
start a new field.VSIN can be either a traditional pulse or the
MPEG style field toggle. In both cases the TMC2192 will
flywheel through fields 2, 4, 6, and 8 synchronizing only to
fields 1, 3, 5, and 7.
signals. In slave and genlock modes the TMC2192
CCIR656
The TMC2192 derives all synchronization from the embedded TRS (timing reference signals) information. Blanking of
selected lines is determined by the v bit of the TRS. However
the control registers VBIENx can override and blank the
active video portion of VBI lines regardless of the state of the
v-bit.
Genlock
The TMC2192 is driven by the input synchronization pins
HSIN
and VSIN. When the falling edge of HSIN and VSIN
occurs at the same rising edge of PXCK the TMC2192 will
start a new field.VSIN
can be either a traditional pulse or the
MPEG style field toggle. In both cases the TMC2192 will
flywheel through fields 2, 4, 6, and 8 synchronizing only to
fields 1, 3, 5, and 7. The TMC2192 collects GRS data and
resets its subcarrier phase and frequency to the data embedded in the GRS stream. The GRS detection occurs only on
the CBVS port.
DRS
The TMC2192 is driven by the input synchronization pins
HSIN and VSIN. When the falling edge of HSIN and VSIN
occurs at the same rising edge of PXCK the TMC2192 will
start a new field.VSIN can be either a traditional pulse or the
MPEG style field toggle. In both cases the TMC2192 will
flywheel through fields 2, 4, 6, and 8 synchronizing only to
fields 1, 3, 5, and 7. Subcarrier phase adjustment is determined by the DRS data. The DRS detection can occur on
either the CBVS port or the pixel data port.
Propagation Delay
The propagation delay from the pixel data (PD) input to the
D/A output is 64 PXCK’s. Figure 8 shows the propagation
delay for both master and slave synchronization modes. For
CCIR656 data streams, pixel 736 (pixel 0 in Figure 8) is the
midpoint of sync and is 32 PXCK’s (24 PXCK’s in PAL)
after the EAV TRS.
n = (SY+BR+BU+CBP+AV)*2
PXCK
C
PD[23:14]
HSIN
HSOUT
(TSOUT = 1)
DACx
(ANALOG)
DCVBS
(D[7:0],FLD[2:1])
Bn
Y
C
n
Y
n+1
Rn
Midpoint of the
Falling Edge of Sync
0
C
Y
B0
0
6365
t
DO
COMP
0
COMP
128
t
DO
1
65-6294-09
Figure 7. Propagation Delay through the Encoder
10REV. 1.0.0 8/13/03
PRODUCT SPECIFICATIONTMC2192
Blanking Control
Control Registers for this section
AddressBit(s)Name
0x041-0PDRM
0x062PDCDIR
0x184-0VBIENF1
0x194-0VBIENF2
0x1F7-0PDCCNT
The content of VBIENFx[4:0] selects the first line to contain
an active video region in each field, all subsequent lines for
the remainder of the field are active. To blank an entire field,
the user zeroes the VBIENFx[4:0] control register. In
CCIR656 slave mode, the user can selectively blank any
enabled line by setting its TRS V bit HIGH. For 525-line
systems, NTSC line numbering is employed, with the first
vertical serration starting on line 4. PAL line numbering is
used with 625-line systems, with each field's line 1 being the
start of the first vertical serration.
Any line(s) enabled by the closed caption control are automatically unblanked for the closed caption waveform, irrespective of the corresponding values of VBIENF.
Pixel Data Control
The pixel data control has two modes of operation, as an
input or as an output. The mode of operation is determined
by the PDCDIR control register. When PDC is an input the
internally generated PDC is ANDed with the PDC pin. This
allows the user to blank any active video regions. When PDC
is an output, the internally generated PDC is the output for
the PDC pin.
The internal PDC control will toggle to a logic HIGH at the
pixel specified by PDCNT and toggle to a logic LOW four
pixels prior to the end of the active video region. The starting
point and ending point of the active video region (VA) are
determined by the control registers 10h to 1Fh. When PDC is
used as an input, the sloped edge of the active video region
will occur on the next four pixels following the toggle point.
Edge Shaping
The TMC2192 has three modes of sloped edges on the active
video region and are controlled by PDRM control register.
Table 3. PDC Edge Control
PDRM[1:0]Slope type at PDC (HIGH)Slope type at PDC (LOW)
00The following four pixels have the weighting of
1/8, 1/2, 7/8 and 1 for NTSC and 1/8, 3/8, 5/8,
and 7/8 for PAL.
01The fifth pixel is sampled and scaled 1/8, 1/2,
7/8 and 1 over the next four pixels for NTSC
and 1/8, 3/8, 5/8, and 7/8 over the next four
pixels for PAL.
1xSlope is off, edge control is dictated by the PD
stream from active video start
The following four pixels have the weighting of
1, 7/8, 1/2, and 1/8 for NTSC and 7/8, 5/8, 3/8,
and 1/8 for PAL.
The fifth pixel s sampled and scaled 1, 7/8, 1/2
and 1/8 over the next four pixels for NTSC and
7/8, 5/8, 3/8, and 1/8 over the next four pixels
for PAL.
Slope is off, edge control is dictated by the PD
stream to active video end
REV. 1.0.0 8/13/0311
TMC2192PRODUCT SPECIFICATION
Horizontal Programming
Control registers for this section
AddressBit(s)Name
0x067-6FORMAT
0x197SHORT
0x196T512
0x195HALFEN
0x207-0SY
0x217-0BR
0x227-0BU
0x237-0CBP
0x247-0XBP
0x257-0VA
0x267-0VC
0x277-0VB
0x287-0EL
0x297-0EH
0x2A7-0SL
0x2B7-0SH
0x2C7-0FP
0x2D7-6XBP (MSB’s)
0x2D5-4VA (MSB’s)
0x2D3-2VB (MSB’s)
0x2D1-0VC (MSB’s)
Horizontal interval timing is fully programmable and is
established by loading the timing registers with the duration
of each horizontal element. The duration is expressed in
PCK clock cycles. In this way, any pixel clock rate between
10 MHz and 15 MHz can be accommodated, and any desired
standard or non-standard horizontal video timing may be
produced.
Horizontal timing parameters can be calculated as follows:
t = N x ( PCK period )
= N x ( 2 x PXCK period )
where N is the value loaded into the appropriate timing
register, and PCK is the pixel clock period.
When programming horizontal timing, subtract 5 PCK
periods from the calculated values of CBP and add 5 PCK
periods to the calculated value for VA. The control register
HALFEN
NTSC, PAL-M and line 23 for all other PAL standards when
it is LOW.
enables the 1st half line (UBV) on line 283 for
Table 4. Horizontal Line Equations
Line TypeLine IDLine Length Equals
EE00EL + EH + EL + EH
SE02SL + SH + EL + EH
SS03SL + SH + SL + SH
ES01EL + EH + SL + SH
EB10EL + EH + EL + EH
UBB, -BB0D, 05SY + BR + BU + CBP + VA + FP
UVV, -VV0F, 07SY + BR + BU + CBP + VA + FP
UVE, -VE0C, 04SY + BR + BU + CBP + VC + FP + EL + EH
UBV0ESY + BR + BU + XBP + VB + FP
12REV. 1.0.0 8/13/03
PRODUCT SPECIFICATIONTMC2192
SY
BRBUCBPVAFP
Figure 8. Horizontal Timing
Table 5. Horizontal Timing Specifications
Parameter
NTSC-M
(µs)
PAL-I
(µs)
PAL-M
(µs)
FP1.51.651.9
SY4.74.74.95
BR0.60.90.9
BU2.52.252.25
CBP1.62.551.8
VA52.655651.9551.692
H63.555664.063.492
Vertical interval timing is also fully programmable, and is
established by loading the timing registers with the duration’s of each vertical timing element, the duration expressed
in PCK clock cycles. In this way as with horizontal program-
65-6294-10
ming, any pixel rate between 10 and 15 Mpps can be accommodated, and any desired standard or non-standard vertical
video timing may be produced.
Like horizontal timing parameters, vertical timing parameters are calculated as follows:
t = N x ( PCK period )
= N x ( 2 x PXCK period )
where N is the value loaded into the appropriate timing register, and PCK is the pixel clock period.
The vertical interval comprises several different line types
based upon H, the Horizontal line time.
H = (2 x SL) + (2 x SH) [Vertical sync pulses]
= (2 x EL) + (2 x EH) [Equalization pulses]
H
H/2
EL
EHSLSH
65-6294-11
Figure 9. Horizontal Timing – Vertical Blanking
The VB and VC control registers are added to produce the
half-lines needed in the vertical interval at the beginning and
end of some fields. These must properly mate with components of the normal lines.
1. XBP, VA, VC, and VB are 10 bit values. The 2 MSBs for these four variables are in Timing Register 2D.
2. EH and SL are 9 bit values. A most significant "1" is forced by the TMC2192 since EH and SL must range from 256 to 511.
EH and SL may be extended to 767. Only the eight LSBs are stored in Timing Registers 29 and 2A.
3. Every calculated timing parameter has a minimum value of 5 except EH and SL which have minimum values of 256.
Vertical Timing
The vertical timing is controlled by the FORMAT control
register, which dictates the field and line sequence.
524
525
FIELDS 1 AND 3
22
HSOUT
VSOUT
(TOUT = 1)
VSOUT
(TOUT = 0)
HSOUT
VSOUT
(TOUT = 1)
VSOUT
(TOUT = 0)
UVV
UVVEEEEEEEEESEESSSSSSSSSESSEEEEEEEEEBEE
262
UVV
263
26412652266326742685269627072718272
UVE
FIELDS 2 AND 4
Figure 12. NTSC Vertical Interval
•••
9
UBB
27310•••
UBB
19
UBB
UBB
282
UBB
UVV
UVV
2832028421285
UVV
UBV
UVV
65-6294-15
REV. 1.0.0 8/13/0315
TMC2192PRODUCT SPECIFICATION
Table 8. NTSC Field/Line Sequence and Identification
-VEHalf-line video, half-line equalization pulse, color burst suppressed.
UBVhalf-line black, half-line video
20REV. 1.0.0 8/13/03
PRODUCT SPECIFICATIONTMC2192
Chrominance Processor
Control registers for this section:
AddressBit(s)Name
0x067-6FORMAT
0x065-3MODE
0x075DDSRST
0x117DRSSEL
0x186GLKCTL1
0x185GLKCTL0
0x3F3GAUSS_BYP
0x407-0FREQL
0x417-0FREQ3
0x427-0FREQ2
0x437-0FREQM
0x447-4SYSPHL
0x453-0SYSPHM
0x467-4BURPHL
0x473-0BURPHM
0x487-4BRSTFULL
0x493-0BRST1
0x4A7-4BRST2
Subcarrier Programming
The color subcarrier is produced by an internal 32 bit digital
frequency synthesizer which is completely programmable in
frequency and phase. Separate registers, FREQx, SYSPHx,
BSTPHx, are provided for phase adjustment of the color
burst and of the active video, permitting external delay compensation, color adjustment, etc. FREQx is the subcarrier
phase step per pixel and SYSPHx is phase offset at field 1,
line 1 (line 4 for NTSC), pixel 1.
NTSC Subcarrier
For NTSC encoding, the subcarrier synthesizer frequency
has a simple relationship to the pixel clock period, repeating
over 2 lines: The decimal value for the subcarrier phase step
is:
455 2⁄
FREQx
Where the number of pixels/line is:
pixels line⁄
This value must be converted to binary and split into four 8
bit registers, FREQM, FREQ2, FREQ3, and FREQL.
PAL Subcarrier
The PAL relationship is more complex, repeating only once
in 8 fields (the well-known 25 Hz offset):
FREQx
This value must be converted to binary and split as described
previously for NTSC. The number of pixels/line is found as
in NTSC.
PAL-M Subcarrier
FREQ
SYSPHx establishes the appropriate phase relationship
between the internal synthesizer and the chroma modulator.
The nominal value for SYSPHx is zero.
Other values for SYSPHx must be converted to binary and
split into two 8 bit registers, SYSPHM and SYSPHL.
Burst Phase (BURPHx) sets up the correct relative NTSC
modulation angle. The value for BURPH is: