
TMC22x5yA
PRODUCT SPECIFICATION
56
If either of the error signals indicates that the magnitude 
difference is above the programmed threshold, or if ANEN is 
LOW, the adaptive notch filter is bypassed. The output of the 
adaptive notch filter is rounded to 8 or 10 bits, or the luma 
data that bypasses the coring filter is truncated to 8 or 10 bits 
depending upon the CORO register bit.
Figure 25. Adaptive Notch Filters
Luma Notch Filter
The simple luma notch filter is centered at 0.25 of normal-
ized frequency, it therefore intended for use only in the sub-
carrier mode (4 * fSC) and for limited use with 13.5MHz 
NTSC as the subcarrier sits at 0.265 of normalized fre-
quency. The notch filter is enabled by setting the NOTCH 
register bit HIGH.
Figure 26. Luminance Notch Filter
Matrix
The magnitude of the decoded luminance and color differ-
ence signals will vary, not only with the standard, but also 
with the input mode. For this reason the output matrix 
contains programmable multipliers, and not just fixed 
scaling factors. The following sub sections explain the differ-
ent scalar in the output matrix. The gain term in the Y, mixed 
sync, U and V scalar is the same - only the weighting makes 
them different. The scalar are capable of independently 
providing 6dB of gain if required.
Programmable U Scalar
The U scalar (UGx) provides the weighting required to 
produce (B-Y) or C
B
 from the demodulated U signal.
hence
(B-Y) = UGx * U
where UGx = gain / 0.493, and
C
B
 = UGx * U
where UGx = (gain * 448) / Umax
UGx has a scaling range of 0 to (2047/256).
Programmable V Scalar
The V scalar (VGx) provides the weighting required to 
produce (R-Y) or C
R
 from the demodulated V signal.
hence
(R-Y) = VGx * V
where VGx = gain / 0.877, and
C
R
 = VGx * V
where VGx = (gain * 448) / Vmax
VGx has a scaling range of 0 to (1023/256).
Programmable Y Scalar
The Y scalar (YGx) provides the scaling for the luminance 
signal if the output is YC
B
C
R
, or controls the magnitude of 
the RGB output along with the U scalar and V scalar. It is not 
possible to control the magnitude of the RGB signals inde-
pendently.
YGx has a scaling range of 0 to (1023/256).
Programmable MS Scalar
The sync scalar (SGx) provides the scaling for the sync 
signal if the output requires sync on RGB. The programmed 
sync scaling factor is used during the horizontal and vertical 
burst blanking periods. During the active lines, the luma 
scaling factor is used to allow scaling of “super blacks” etc., 
which will be passed down the mixed sync path because they 
fall below the clamp level.
SGx has a scaling range of 0 to (1023/256).
Fixed (B-Y) and (R-Y) Scalars
These two scalars are zero when the output is YC
B
C
R
 and 
provide the (B-Y) and (R-Y) weighting when the output is 
RGB. These are fixed scaling factors and are derived from 
the following equations.
(G-Y) = - [(0.299/0.587) * (R-Y)] 
- [(0.114/0.587) * (B-Y)]
or
(G-Y) = - [(1043/2048) * (R-Y)] 
- [(398/2048) * (B-Y)]
-70
-60
-50
-40
-30
-20
-10
0
0
0
0
0
0
0
Normalized Frequency
A
6
Adaptive Notch
Filter 1
Adaptive Notch
Filter 2
Adaptive Notch
Filter 3
2
-70
-60
-50
-40
-30
-20
-10
0
0
0
0
0
0
0
Normalized Frequency
A
6