
Philips Semiconductors
PNX15xx Series
Volume 1 of 1
Chapter 28: Pixel Formats
PNX15XX_SER_3
Koninklijke Philips Electronics N.V. 2006. All rights reserved.
Product data sheet
Rev. 3 — 17 March 2006
28-12
pixel format, which implies the unit size(s)
origin pointer—the (byte) address of the rst unit of the image
line pitch—the address difference between a pixel on a line and a pixel directly
below it
width W, in number of pixels
height H, in number of lines
Note that for indexed formats, each unit contains one or more pixels. For the packed
formats, a unit is a pixel, with the exception of packed YUV 4:2:2 where two units are
needed to describe a pixel pair.
Figure 13 shows how images are stored in memory
8.
System Endian Mode
The PNX15xx Series is designed to run either little-endian or big-endian software.
The entire system always operates in a single endian mode—i.e. the CPU and all
hardware subsystems run either little or big-endian. This is determined by a global
endian mode ag.
The endian mode determines how a multi-byte value is stored to/loaded from memory
byte addresses.
For the native pixel formats,
Section 3. always shows two elements in the gures: the
layout of a ‘unit’, which is always 8, 16 or 32 bits, and the mapping of adjacent units to
memory byte addresses. These two elements are always maintained, independent of
system endian mode.
What this implies is that each hardware subsystem needs to map a unit to memory
byte addresses in an endian mode-dependent manner. The rules are as follows:
Storing a 16-bit unit to address ‘A’ results in modifying memory bytes ‘A’ and
‘A+1’
Storing a 32-bit unit to memory address ‘A’ results in modifying memory bytes ‘A’
to ‘A+3’
Figure 13: Image Storage Format
origin
unit1
unit2
unit3
. . . .
unitk
. . . .
linepitch
1st line
2nd line
. . . .
last line
1unit
W pixels
H lines
a
a+s
a+2s
a+(k-1)s
unit1
unit2
unit3
unitk
unit1
unit2
unit3
unitk
(a)