Document Feedback

CSG14K_CSG8K

Functional Description

Datasheet

• PUBLIC

DS000571

• v1-00 • 2020-Dec-15

112

│ 89

quality of the resulting image), which means the level at the output for black pixels will be around 30

DN.

BLACK_LEVELADC_OUT is the black level at the output of the ADC. Its value is set differently depending

on how the black level correction is done. This is detailed in the next two sections.

On-Chip OBC

Setting EOB_BYPASS to 0 enables the automatic on-chip black level correction (clamping). In this

mode, BLACK_LEVELADC_OUT is calculated on-chip, based on OB pixels, for every row allowing to

correct for row and frame noise in the image.

This is the recommended mode of operation.

Off-Chip OBC

Setting EOB_BYPASS to 1 disables the automatic OBC. BLACK_LEVELADC_OUT is now specified with

register EOB_BYPASS_VALUE. This value is constant and cannot correct for row noise in the image.

Equation 14:

������������������������������������������_������������������ = ������������������������[������������������������������������_������������������ − ������������������_������������������������������������_������������������������������ + ������������������_������������������������������������]

The theoretical value for register EOB_BYPASS_VALUE is shown in the table below. The parameter

tCLK_PIX can be derived from Figure 84.

Figure 111:

Theoretical Values for Register EOB_BYPASS_VALUE

EOB_BYPASS_VALUE

12.5*tCLK_PIX

Further external or off-chip OBC can be achieved by reading out the OB pixel data (refer to section

7.8.7).

7.7.2

Digital Gain

Digital gain can be specified for the 4 color channels individually (even/odd columns, even/odd rows).

E.g. GDIG_RE_CO relates to Rows Even and Columns Odd.

The digital gain for each channel is defined by a similar equation:

Equation 15:

������������������������������������������ ������������������������ (������������_������������) = (������������������������_������������_������������ + 1)/16