Public Version
Camera ISP Basic Programming Model
www.ti.com
The 32-bit destination addresses of the pixel data are set by the
and
registers.
NOTE:
The destination address must be aligned on a 32-byte boundary; the address 5 LSBs are
ignored. The pixel data lines are packed together at the destination address.
It is possible to perform double-buffering (ping-ponging) at the destination by setting different addresses in
the
and
registers. It is possible to disable
double-buffering by setting up the same address in both registers. The
[17]
PING_PONG status bit must be used by the software to determine which address contains the latest
frame.
A destination pitch controls the address jump between the address of the first pixel of the previous line
and the address of the first pixel of the current line. The destination pitch is set in bytes with the
register. It applies for
and
.
NOTE:
The destination pitch must be a multiple of 32 bytes; the address 5 LSBs are ignored.
The use of
is limited to the following destination data formats:
•
YUV422 little endian
•
YUV422 big endian
•
EXP16
•
RGB565
•
EXP32
For all other data formats, the
register is ignored (equivalent to
= 0x0).
The destination data format is set with the
[7:2] FORMAT bit field.
For the PING frame:
•
@Line0 =
•
@Line1 = @Line0 +
•
@Line2 = @Line1 +
For the PONG frame:
•
@Line0 =
•
@Line1 = @Line0 +
•
@Line2 = @Line1 +
When
= 0x0, the lines are written contiguously in memory. The destination pitch
enables 2D transfers; it is required to write the pixel data directly in the frame buffer, for instance.
In such cases,
= FRAME_BUFFER_WIDTH.
shows the pixel data
settings.
1254
Camera Image Signal Processor
SWPU177N – December 2009 – Revised November 2010
Copyright © 2009–2010, Texas Instruments Incorporated