🔒 Locked

Easy Question

Started by RDK1LL Jul 21, 2006 at 11:49 PM 7 replies 1.6k views

Original Post

Jul 21, 2006 at 11:49 PM

I've got a game which runs in 256(DIRECTX)color mode, which is'nt quite finished yet. But when Im running windows in 32bit(true color) the bitmap's color palette loads wrong (only a few colors), however it loads sweet as when running in 16bit mode. Is there a logical explanation for this?

DevLiquidKnight

Jul 22, 2006 at 1:08 AM

I suppose it depends on how your initializing DirectX. Could be more specific to the code you are using to initialize it.

pulpfist

Jul 22, 2006 at 11:06 AM

AFAIK there is no such thing as a color palette in 16/24/32 bit images.
What I did when I changed my 8 bit code to 32 bit was to skip and forget about the color palette all together.

The RGB macros is able to describe any 32 bit color, so there is no need to map speciffic colors to speciffic palette indexes.

RDK1LL

Jul 24, 2006 at 12:00 AM

Quote:
Original post by pulpfist
AFAIK there is no such thing as a color palette in 16/24/32 bit images.
What I did when I changed my 8 bit code to 32 bit was to skip and forget about the color palette all together.

The RGB macros is able to describe any 32 bit color, so there is no need to map speciffic colors to speciffic palette indexes.

Windows is running in 32bit mode 1024*768, and when I start my game it initializes DirectX in 8bit mode 640*480. This is when some of the palette gets loaded wrong. It does happen in 16bit mode as well, only if I've just had a paint editor running.
I just don't want it to happen when its finished, you know...

Aardvajk

Jul 24, 2006 at 3:44 AM

This must be a problem with the way you are initialising your palette. I never had a problem like this running 8-bit DDraw apps with windows in 32 bit mode.

Sounds to me (throwaway theory based on pure speculation) as if you are not initialising the palette properly, and you are getting "ghost" values from whatever was in your cards palette registers from before your program ran.

Could you post your palette loading code please.

RDK1LL

Jul 29, 2006 at 2:27 AM

The code I'm usingt is out of Andre Le'moths Tricks for game programming gurus.
I'm just using his T3DXENGINE with his own palette loading code.
I't just loads the palette from the bitmap file;

FILE *fp_file;
if((fp_file = fopen(filename, "r"))==NULL)
return(0);

for(int index=0; index<256; index++)
{
fscanf(fp_file, "%d %d %d %d, &palette[index].peRed,
&palette[index].peGreen,
&palette[index].peBlue,
&palette[index].peFlags);
}
fclose(fp_file);

After this the palette is just set using primary set palette function.

Aardvajk

Jul 29, 2006 at 4:45 AM

That code appears to be loading from a file containing nothing but a palette. The palette data in a bitmap file is located after the BITMAPHEADER information.

Are you sure Le'Mothe's example is supposed to be loading the palette from the bitmap directly? Some art editors let you export just the palette as a file in the format that code would work with.

RDK1LL

Jul 29, 2006 at 4:54 AM

Thanks I'll give that a try.

Aardvajk

Jul 29, 2006 at 5:07 AM

The following is pretty awful code I have slung together very quickly but should give you an idea of loading a palette from a .bmp file. I haven't tested it:

bool load_palette(const char *BmpFile){    FILE *fp=fopen(BmpFile,"rb"); if(!fp) return false;    BITMAPFILEHEADER Fh;    BITMAPINFOHEADER Ih;    fread(&Fh,sizeof(BITMAPFILEHEADER),1,fp);    fread(&Ih,sizeof(BITMAPINFOHEADER),1,fp);    if(Ih.biBitCount!=8){ fclose(fp); return false; }    fread(palette,sizeof(RGBQUAD),256,fp);    fclose(fp);    // do whatever with palette    return true;}

This could, of course, be much improved by using C++ fstreams and making the error checking more robust.

HTH

Topic Locked

This topic has been locked by a moderator. New replies are not allowed.