RY
11-bit (signed)
RCb
11-bit (signed)
RCr
11-bit (signed)
Y
8-bit (unsigned)
Cb
8-bit (unsigned)
Cr
8-bit (unsigned)
Adder
20-bit (signed)
Shifted value (right by 8)
12-bit (signed)
Clipping to [0:255]
8-bit (unsigned)
R component
8-bit (unsigned)
–128
–128
–16 or 0
dss-065
=
−
7
)
(
)
(
)
(
2
>>
+
Φ
=
∑
=
i
i
n
Rin
Ci
n
Rout
2
i
(
)
x
dss-E066
Public Version
Display Subsystem Functional Description
www.ti.com
Figure 7-75. Color Space Conversion Macro-Architecture
7.4.2.3.3 Hardware Cursor
The video layer can be used to display the hardware cursor. The encoded pixel data for the cursor image
are in RGB12, RGB16 or RGB24 formats and the color space conversion block is bypassed. The
transparency color key can be used when a non rectangle shape is used.
The alpha blending can be used to show a partial transparent cursor. When the alpha blender is enabled,
the graphics layer is on top of the video layers. The cursor uses the graphics layer. The pixel alpha
blending or the transparency color key can be used.
7.4.2.3.4 Up-/Down-Sampling
The video layer has a dedicated resizing block to upsample and downsample the video-encoded pixels.
The supported input formats from memory are RGB24, RGB16, and YUV4:2:2
(RGB12 and all the alpha formats like ARGB and RGBA are not supported)
Users must set the right size and position of the original video before resizing for the
upsampled/downsampled video to be inside the display screen boundaries.
The filtering applies on each component independently R, G, and B).
For the horizontal up/downsampling, the equation is R component with five taps):
(7)
For the vertical up/downsampling, the equation is R component with three taps):
1644
Display Subsystem
SWPU177N – December 2009 – Revised November 2010
Copyright © 2009–2010, Texas Instruments Incorporated