[Editor's note: This VisualBasic script from David Viljoen, Geological Survey of Canada, for ArcGIS solves a transparency flattening issue when trying to blend color into a grayscale shaded relief image. Often the colors become distorted during the merge. This tool preserves the color (hue) and moderates the saturation and value instead. I've used HSV color adjustment layers in Photoshop with relief masks to accomplish the same thing, nice to know it's available in ArcMap, too. Aileen mentioned it at NACIS Sacramento earlier this month.]
I developed the SatValMod (SVM) method to address the problems associated with traditional methods of integrating color with gray-scale raster data (e.g. layer transparency, multiplying color by gray-scale values, etc.).
The main problem with traditional methods is color loss or corruption. SVM does not change the original hue and modulates the saturation and value so that the final output has the same rich colors of your input data.
SVM does not require Spatial Analyst. It supports Grid, BIL, and TIF formats for input. It outputs a BIL file.
The SVM method involves a pixel-by-pixel transformation of RGB color coordinates to HSV space, modulation of the saturation and value color components, and transformation of the orginal hue and modulated saturation and value components back to RGB space. More details are available in the PowerPoint slide show included in the ZIP.
This technique can work with rasterized polygon layers. You will need to create a CLR file that relates pixel values to the polygon colors.
I hope you find SVM useful in creating your color/gray-scale image integration products.