GF9105A 데이터 시트보기 (PDF) - Unspecified

부품명

상세내역

제조사

GF9105A

Component Digital Transcoder

Unspecified 

The BYPASS_F control bit (See Host Programming Section) can be used to bypass the interpolation filters. When this bit is

set low, the filters are enabled and normal operation occurs. When this bit is set high, the filters are bypassed and the data is

passed through the filter section unmodified. FIL_RND should be set low when BYPASS_F is set high. Total latency through

the filter is independent of the BYPASS_F control signal. Note that after changing the state of BYPASS_F, an initialization

period corresponding to the device’s latency is required before valid data is available at the output of the device.

BYPASS_F CONTROL BIT

BYPASS_F

DESCRIPTION

0

Filters are enabled. Data is sample-rate converted from 4:2:2:4 to 4:4:4:4 data.

1

Filters are disabled. Data is passed through the filter section unmodified.

3X3 COLOR SPACE CONVERTER AND KEY SCALER

In this section, a 3X3 matrix multiplication (color space conversion) may be performed on the Y/G, CB/B and CR/R data. The

3X3 matrix multiplier has 13-bit two’s complement coefficients and maintains full precision throughout the 3X3 calculation.

The nine 13-bit coefficients (See Host Programming Section) used in this 3X3 calculation determine the color space

conversion that the GF9105A will perform. These coefficients are referred to as CMij, where i refers to the row and j refers to

the column in which CMij is found. The matrix multiplication can be shown as:

Y/G O U T

CB/BOUT =

C R /RO U T

CM11 CM12 CM13

CM21 CM22 CM23

CM31 CM32 CM33

Y/GIN

C B /BI N

C R /RI N

The nine matrix coefficients have 13-bit two’s complement resolution and cover a range from -4 to +3.9990234375. Bit

weighting for the coefficients is as follows:

Coefficient Bit

Weighting

b12

b11

b10

b9

b8

b7

b6

b5

b4

b3

b2

b1

b0

-22

21

20

2-1

2-2

2-3

2-4

2-5

2-6

2-7

2-8

2-9

2-10

Matrix bypassing can be accomplished by setting FIL_RND high and loading an identity matrix, by setting CM11, CM22 and

CM33 to unity and setting the remaining six coefficients to zero. In this mode, gain through the matrix stage is 1.000.

Typical examples of matrix coefficients that will provide full range RGB to YCBCR, and YCBCR to full range RGB conversions

are:

Y

CB =

CR

0.5027

–0.2899

–0.3633

0.0976

0.4376

–0.0711

G

B=

R

1.1677 –0.3931

1.1677 2.0248

1.1677

0

0.2561 G

–0.1477 B

0.4374 R

–0.8164 Y

0 CB

1.6025 CR

KEY signals may also be scaled by a programmable scaling factor. The KEY scaling coefficient (See HOST Programming

Section) has the same resolution and bit weighting as the nine 3X3 matrix multiplier coefficients.

Typical examples of KEY scaler values that could be used are:

KEY scaler = 0.8563 for full range RGB to YCBCR conversions.

KEY scaler = 1.1677 for YCBCR to full range RGB conversions.

521 - 88 - 03

14

Share Link: