This is the interface to let devices provide their own color mapping and dithering routines, that is, override the graphic system's defaults or provide routines for formats not supported by the graphics system (for example, a 32-bit 10R 10G 10B format for instance).
Notes
A device has the option to dither at external to internal conversion time (this is the only option for color conversion) or at logical to physical.
Note: A device can specify KEEP_EXTERNAL_FORMAT, in which case an external to physical conversion could take place. This should not matter to the conversion-handling routine because all the information in the image information structure should suffice in performing the conversion.
You can:
Selecting Dithering
Dithering is selected via:
Add the appropriate GDM_DITHER flag into the fExt2IntSup and/or fLog2PhysSup fields (dither could be done at both levels).
Modifying System Dithering Routines
You can change just the dithering defaults for that requested type of dithering without providing funtion. To accomplish that, provide either a pExt2IntDI or a pLog2PhysDI structure.
Note: The whole structure must be filled out.
Below are two currently used system dither structures:
MATRIXDITHERINFO System Default Values
ULONG ulLength sizeof(MATRIXDITHERINFO)
ULONG ulType GDM_MATRIX_DITHER
ULONG fOptions 00
ULONG ulIntensity 00
BYTE bRedWt 25
BYTE bGreenWt 60
BYTE bBlueWt 15
BYTE bPad 00
SIZEL szMatrix N/A
BYTE bHalftone[1] N/A
DIFFUSIONDITHERINFO
ULONG ulLength sizeof(DIFFUSIONDITHERINFO)
ULONG ulType GDM_ERRORDIF_DITHER
ULONG fOptions JJN_DIF
ULONG ulIntensity 00
BYTE bRedWt 25
BYTE bGreenWt 60
BYTE bBlueWt 15
BYTE bPad 00
PBYTE pBuffer N/A
SIZEL szFilter N/A
BYTE bFilterArry[1] N/A
JJN_DIF Currently only one supported
Providing Your Own Dithering Routines
To provide your own dither routines, place the GDM_USERDEF_DITHER in either the fExt2IntSup and/or fLog2PhysSup fields. Then provide a function for pfnDither:
pfnDither(PDITHERINFO pdi, PIMAGEINFO pii, ULONG ulFlags);
Where pdi = either pExt2IntDI or pLog2PhysDI. This should be your own allocated structure. For pii, see the detailed information about the IMAGEINFOS structure which follows.
Providing Support for Unsupported System Formats
To provide your own color conversion routines, place the GDM_COLOR_CONVERT in the fExt2IntSup field, and provide the three required functions in the COLORCONVERTINFO structure:
pfnRGB2ToPhyIndex (ULONG ulClr) pfnPhyIndexToRGB2 (ULONG ulClr) pfnConvert (PIMAGEINFO pimginfo)
Data Structures
COLORCONVERTINFO ULONG ulLength sizeof(COLORCONVERTINFO) ULONG ulType GDM_COLOR_CONVERT ULONG ulOptions 00 ULONG (* APIENTRY pfnRGB2ToPhyIndex) () N/A - Mapping from RGB to Phy index ULONG (* APIENTRY pfnPhyIndexToRGB2) () N/A - Mapping from Phy index to RGB ULONG (* APIENTRY pfnConvert) () N/A - Actual Convert Bits Function IMAGEINFOS PPOINTL pptlSrc PPOINTL pptlDst ULONG ulSrcBpp Source image bits per pixel ULONG ulSrcBytesPerLine Count of bytes per scan line (includes padding) PBYTE pbSrcBits Pointer to source data ULONG ulcSrcClrs Count of source colors ULONG ulSrcClrType Source color format see below PRGB2 pargb2Src Pointer to source color table ULONG ulTrgBpp Target image bits per pixel ULONG ulTrgBytesPerLine Count of bytes per scan line (includes padding) PBYTE pbTrgBits Pointer to target data ULONG ulcTrgClrs Count of target colors ULONG ulTrgClrType Source target format see below PRGB2 pargb2Trg Pointer to target color table ULONG ulOptions see below
IMAGEINFOS Color Types
When ulBpp <= 8:
COLOR_RGB2 Color is in RGB2 (default) COLOR_RGB Color is in RGB
When ulBpp = 24:
COLOR_24BIT_BGR 1st byte=blue, 2nd byte=green 3rd byte=red (default) COLOR_24BIT_RGB 1st byte=red, 2nd byte=green 3rd byte=blue
When ulBpp = 16:
COLOR_16BIT_565_BGR blue - 5 bits, green - 6 bits, red - 5 bits (default)
---------------------------------
|b|b|b|b|b|g|g|g|g|g|g|r|r|r|r|r|
---------------------------------
COLOR_16BIT_565_RGB red - 5 bits, green - 6 bits, blue - 5 bits
---------------------------------
|r|r|r|r|r|g|g|g|g|g|g|b|b|b|b|b|
---------------------------------
COLOR_16BIT_1555_BGR overlay - 1, red - 5 bits, green - 5 bits, blue - 5 bits
---------------------------------
|o|b|b|b|b|b|g|g|g|g|g|r|r|r|r|r|
---------------------------------
COLOR_16BIT_1555_RGB overlay - 1, red - 5 bits, green - 5 bits, blue - 5 bits
---------------------------------
|o|r|r|r|r|r|g|g|g|g|g|b|b|b|b|b|
---------------------------------
When ulBpp = 32:
COLOR_32BIT_BGR 1st byte=blue, 2nd byte=green 3rd byte=red, 4th byte=not used COLOR_32BIT_RGB 1st byte=red, 2nd byte=green 3rd byte=blue, 4th byte=not used
IMAGEINFO Options
IMAGEINFO_CLR_TO_MONO_BG_EXACT Src pels that are trg BG clr -> 1
all other pels -> 0
IMAGEINFO_CLR_TO_MONO_BG_EXACT_INV Src pels that are trg BG clr -> 0
all other pels -> 1
Nearest color match.
Source pels that are close to target BG
color become target BG color.
All other pels become target FG color.
IMAGEINFO_CLR_TO_MONO_BG_MAP Src pels that are close to trg BG clr -> 1
all other pels -> 0
IMAGEINFO_CLR_TO_MONO_BG_MAP_INV Src pels that are close to trg BG clr -> 0
all other pels -> 1
IMAGEINFO_SRC_Y_FLIP Source is going Y negative direction
i.e. second line = first line - ulBytesPerLine.
IMAGEINFO_TRG_Y_FLIP Target is going Y negative direction