LM9830VJDX NSC [National Semiconductor], LM9830VJDX Datasheet - Page 24

• Manufacturer Part Number: LM9830VJDX
• Description: LM9830 36-Bit Color Document Scanner
• Manufacturer: NSC [National Semiconductor]

This equation also applies to the divide by 1.5 function.
The output of this stage is sent through the gamma and pixel
packing stages, resulting in output data formatted as shown in
Figure 15.
2.4.2 Pixel Processing 10/30 Bit Mode
Scanning in 10 bit mode (30 bit color) supports the Horizontal
DPI Divider function as well as the pixel rate shading and offset
functions. The output data is formatted as shown in Figure 12
(X=mask out in software).
The software on the host PC must perform any gamma correction
desired.
There are two variations on the 10 and 12 bit output modes: Full
Duplex and Half Duplex, determined by bit 5 of Configuration
Register 43. In the Full Duplex mode, there are 6 SRAM opera-
tions per pixel: offset data read, gain data read, pixel MSB write,
pixel LSB write, pixel MSB read, pixel LSB read). Since there are
8 MCLKs per pixel, the writes take 2 MCLK periods and the reads
take 1 MCLK period. This mode is preferred because it permits
faster scanning, but it requires fast SRAM access.
The Half Duplex mode accommodates slower SRAM. In Half
Duplex mode, the data in the SRAM can not be read by the host
PC until the buffer is full. Therefore there are two phases to scan-
ning data in the Half Duplex mode. The first is writing pixel data to
SRAM using 4 operations/pixel (offset data read, gain data read,
pixel MSB write, pixel LSB write). In this mode the read and write
cycles will all be 2 MCLKs long. The second phase is reading the
contents of the SRAM and sending them to the host PC. This
read operation uses 2 MCLK read cycles/byte.
2.4.3 Pixel Processing: 12/36 Bit Mode
Scanning in 12 bit mode (36 bit color) can only be done at the
optical resolution of the sensor, and the shading and offset func-
tions can not be used. The Horizontal DPI Divider function must
be set to 1 (register 09, bits 0-2 = 0) to get 12 bit data. The shad-
ing and offset functions must also be disabled (registers 3E, 3F,
40, and 41 all set to 0). The 10 bit pixel output from the multiplier
is combined with the 2 12 bit LSBs from the ADC to recreate the
12 bit pixel. The output data is formatted as shown in Figure 13
X
5
7
Divide
1.5
By
12
Figure 11: Decreasing Horizontal Resolution
1
2
3
4
6
8
X
4
6
Figure 12: 10 Bit Mode Pixel Data Format
X
3
5
(600 DPI sensor)
X
2
4
9
1
DPI
600
400
300
200
150
100
3
75
50
8
0
2
7
X
1
6
X
0
(300 DPI sensor)
Second Byte
First Byte
37.5
DPI
300
200
150
100
75
50
25
Order
24
(X=mask out in software).
The software on the host PC must perform all offset and shading
correction, horizontal resolution adjustment, and gamma correc-
tion.
The 12 bit mode uses the same Full or Half Duplex options
described in 2.4.2 Pixel Processing 10/30 Bit Mode
2.5 Gamma Correction Tables
There are 3 gamma lookup tables for R, G, and B. The input to
the table is the 10 bit pixel data coming from the previous stage
(2.4 Pixel Processing Block). The output is the 8 bit gamma cor-
rected pixel data. The tables are therefore 1K bytes x 8 bits in
size. Each gamma table (red, green, and blue) can be loaded with
any arbitrary user-defined transfer curve.
The gamma tables are loaded through the dataport (see 5.1 The
DataPort: Reading and Writing to Gamma, Offset, and Gain
Memory). In most LM9830 modes, the DataPort selects which
color (Red, Green or Blue) gamma table will be read from or writ-
ten to. In 1 Channel Mode A, the only gamma table that can be
accessed is the gamma table for the 1 Channel Mode A color
selected by bits 3 and 4 of register 26.
2.6 Pixel Packing/Thresholding Block
Some scans require only one bit per pixel (“line art” mode), others
may need only 2 or 4 bits/pixel. To increase scanning speed for
lower pixel depths, the LM9830 packs the desired MSBs of multi-
ple pixels together, increasing the transmission speed to the host
by a factor of 2, 4, or 8. Figure 15 shows how the pixels are
packed together for 8, 4, 2, and 1 bit pixel depths. In Figure 15,
“b” indicates the bit position (b7 = the most significant and b0 =
Depth
Pixel
255
8
4
2
1
0
X
Figure 15: Packing Multiple Pixels Into One Byte
7
7
0
b7 p
b7 p
b7 p
b7 p
X
6
Figure 13: 12 Bit Mode Pixel Data Format
6
7
0
0
0
0
X
5
5
b6 p
b6 p
b6 p
b7 p1 b7 p
Figure 14: Gamma Table
6
X
4
4
10MSBs of 12 bit Output
0
0
0
b5 p
b5 p
b7 p
11 10
3
3
5
0
0
1
2
2
2
b4 p
b4 p
b6 p
b7 p
4
9
1
0
0
1
3
1
b3 p
b7 p
b7 p
b7 p
3
8
0
0
0
1
2
4
http://www.national.com
b2 p
b6 p
b6 p
b7 p
Second Byte
2
First Byte
0
1
2
5
Order
b1 p
b5 p
b7 p
b7 p
1
0
1
3
6
b0 p
b4 p
b6 p
b7 p
1023
0
0
1
3
7