FADEIN App Uses Palette Animation & Identity Palettes (149855)

MORE INFORMATION

The following file is available for download from the Microsoft Download Center:
For additional information about how to download Microsoft Support files, click the following article number to view the article in the Microsoft Knowledge Base: Microsoft scanned this file for viruses. Microsoft used the most current virus-detection software that was available on the date that the file was posted. The file is stored on security-enhanced servers that help to prevent any unauthorized changes to the file.

This sample application in FADEIN.EXE contains code that performs the following tasks:

• Creates a DIBSection from a bitmap stored as a resource
• Chooses the optimal 236 colors from a 256 color table
• Reduces a 256 color DIBSection to 236 colors
• Creates an identity palette
• Performs palette animation to fade from black to color

To perform the fade in the opposite direction, starting with all black and fading into color, requires some additional ingenuity. The original BitBlt() that gets the image's bits on the screen must be done without color matching. Because all of the pixels are intended to be black, if color matching were used, the pixels would be matched to the same palette entry. A subsequent palette animation would result in a single-colored rectangle.

You can take advantage of an optimization in Windows to work around this issue. When Windows is performing a BitBlt and it detects that an identity palette is being used, no color matching is performed. The source bits are moved to the destination unchanged. To perform a fade from black to color, use an identity palette to get the bits onto the screen and then fade the palette from all black to the original bitmap's desired colors.

To prepare a bitmap to use an identity palette, you must reduce the image from 256 colors to 236. An identity palette must contain the system entries in its first and last ten entries. The code in the sample below that reduces the image to 236 colors must first calculate the optimal 236 colors from a 256 color table. It then determines how close the colors are to each other, and discards colors that have close matches while keeping colors that have no close matches. Once it determines the best 236 colors, BitBlt() maps the 256 color image to 236 colors. Everything is then moved up ten to allow for the system colors in the identity palette.