HMP8117 Datasheet, PDF (8/45 Pages) Intersil Corporation – NTSC/PAL Video Decoder

HMP8117
remove high-frequency Y data. This may make a noisy image
more pleasing to the user, although softer.
Coring of the high-frequency Y data may be done to reduce
low-level high frequency noise.
Coring of the Y data may also be done to reduce low-level
noise around black. This forces Y data with the following
values to a value of 0:
coring = 1: ± 1
coring = 2: ± 1, ± 2
coring = 3: ± 1, ± 2. ± 3
High-frequency components of the luminance signal may be
“peaked” to control the sharpness of the image. Maximum
gain may be selected to occur at either 2.6MHz or the color
subcarrier frequency. This may be used to make the
displayed image more pleasing to the user. It should not be
used if the output video will be compressed, as the circuit
introduces high-frequency components that will reduce the
compression ratio.
The brightness control adds or subtracts a user-specified
DC offset to the Y data. The contrast control multiplies the Y
data by a user-specified amount. These may be used to
make the displayed image more pleasing to the user.
Finally, a value of 16 is added to generate a nominal range of
16 (black) to 235 (white).
CbCr Processing
The CbCr data is low-pass filtered to either 0.85MHz or
1.5MHz.
Coring of the CbCr data may be done to reduce low-level
noise around zero. This forces CbCr data with the following
values to a value of 128.
coring = 1: 127, 129
coring = 2: 126, 127, 129, 130
coring = 3: 125, 126, 127, 129, 130, 131
The saturation control multiplies the CbCr data by a user-
specified amount. This may be used to make the displayed
image more pleasing to the user. The CbCr data may also
be optionally multiplied by the contrast value to avoid color
shifts when changing contrast.
The hue control provides a user-specified phase offset to the
color subcarrier during decoding. This may be used to
correct slight hue errors due to transmission.
YCbCr Output Format Processing
Y has a nominal range of 16 to 235. Cb and Cr have a
nominal range of 16 to 240, with 128 corresponding to zero.
Values less than 1 are made 1 and values greater than 254
are made 254.
While BLANK is asserted, Y is forced to have a value of 16,
with Cb and Cr forced to have a value of 128, unless VBI
data is present.
RGB Output Format Processing
The 4:2:2 YCbCr data is converted to 4:4:4 YCbCr data and
then converted to either 15-bit or 16-bit gamma-corrected
RGB (R′G′B′) data. While BLANK is asserted, RGB data is
forced to a value of 0.
15-Bit R′G′B′
The following YCbCr to R′G′B′ equations are used to
maintain the proper black and white levels:
R′ = 0.142(Y - 16) + 0.194(Cr - 128)
G′ = 0.142(Y - 16) - 0.099(Cr - 128) - 0.048(Cb - 128)
B′ = 0.142(Y - 16) + 0.245(Cb - 128)
The resulting 15-bit R′G′B′ data has a range of 0 to 31.
Values less than 0 are made 0 and values greater than 31
are made 31.
The 15-bit R′G′B′ data may be converted to 15-bit linear
RGB, using the following equations. Although the PAL
specifications specify a gamma of 2.8, a gamma of 2.2 is
normally used. The HMP8117 allows the selection of the
gamma to be either 2.2 or 2.8, independent of the video
standard.
for gamma = 2.2:
for R′G′B′ < 0.0812*31
R = (31)((R′/31)/4.5)
G = (31)((G′/31)/4.5)
B = (31)((B′/31)/4.5)
for R′G′B′ >= 0.0812*31
R = (31)(((R′/31) + 0.099)/1.099)^2.2
G = (31)(((G′/31) + 0.099)/1.099)^2.2
B = (31)(((B′/31) + 0.099)/1.099)^2.2
for gamma = 2.8:
R = (31)(R′/31)^2.8
G = (31)(G′/31)^2.8
B = (31)(B′/31)^2.8
16-Bit R′G′B′
The following YCbCr to R′G′B′ equations are used to
maintain the proper black and white levels:
R′ = 0.142(Y - 16) + 0.194(Cr - 128)
G′ = 0.288(Y - 16) - 0.201(Cr - 128) - 0.097(Cb - 128)
B′ = 0.142(Y - 16) + 0.245(Cb - 128)
The resulting 16-bit R′G′B′ data has a range of 0 to 31 for R′
and B′, and a range of 0 to 63 for G′. Values less than 0 are
made 0; R′ and B′ values greater than 31 are made 31, G′
values greater than 63 are made 63.
The 16-bit R′G′B′ data may be converted to 16-bit linear RGB,
using the following equations. Although the PAL specifications
specify a gamma of 2.8, a gamma of 2.2 is normally used.
The HMP8117 allows the selection of the gamma to be either
2.2 or 2.8, independent of the video standard.
8