Hello, I wrote a function which takes 3 bitmaps as parimeters and blends the first 2 putting the result into the third. To do so I read the rgb value from each pixel of each image, add em up and divide by 2. It works pretty well but it sure is slow, I timed it at 320x200x32 and it was like 4 blends per second. I use getpixel to get the r g b values. Is there a faster way to do this, I use allegro and djgpp.
thanks
Acos is good
anyone?
//From Govtcheez: Don't bump threads - any more and I'll just start deleting your bumping posts... (before you ask, I'm sure anyone at fd would do the same)
Hi
if the two source bitmaps have the same structure (bitcount, plains, colordepth,etc.) and size, you can simply treat them as byte arrays, just adding and dividing each pair of bytes to form the resulted bitmap.
damyan
So that would be faster? Ill try it out when I get home tonight.
thanks
oh and govt, you said that so I wouldnt B*itch about cprog right?
> oh and govt, you said that so I wouldnt B*itch about cprog right?
Better believe it...
-Govtcheez
[email protected]
> if the two source bitmaps have the same structure
> (bitcount, plains, colordepth,etc.) and size, you can
> simply treat them as byte arrays, just adding and
> dividing each pair of bytes to form the resulted bitmap.
Actually, it's more complex than that. In an indexed color model, such as you have described, each pixel contains a value that actually refers to an RGB triplet in a color table.
Simply adding/subtracting pixels will not work unless the color table is organized to facility that.
They are truecolor bitmaps so it should work.
