Thread: creating image from pixels stored in file

I guess I would start by assuming the 16 bits were encoded as 5 bits for each red, green and blue. Since I dont know the dimensions or the start point of the picture, I would probably start with the first pixel in the bottom left corner, (as with bitmaps), and draw a reasonable number of pixels then the same number on the line above and so on until I'd drawn all the pixels, then look at what I'd got.

If the encoding is wrong, the picture will be garbage, so try a different encoding method. If the picture has diagonal lines, then vary the pixels per line and see if they start to line up into something.

The obvious suggestion is to find out more about the format of the file, or the file type and look it up at wotsit or somewhere.

I don't use MFC. What I would do is mask out the least significant 5 bits and store that in a byte, then the next 5 bits and store that in a byte, then the next and store that. Then build a COLORREF structure from these three bytes, and use the SetPixel() API routine to put it onto my windows DC. (COLORREF and SetPixel() are both describes in the help).

If that looked wrong, then I'd start at the other end, mask of the most significant 5 bits, etc...

If your confortable with MFC, you may see something in that rambling which gives you a steer.