Original Post
I''ve search google and learnt about the idea in my graphics class. In general, I can read in a black and white image, with each color cooresponding to a different change-in-normal. AKA, white is greatest change, black is least change. Given this, at render time we look up the proper value in this texture, add that to the normal for rendering purposes. I''m familiar with texture shading and know how to implement it, and it seems like this is something similar. I''m more confused about how to actually specifiy the amount of change based on the colors in the image. Do i specifiy a constant amount of change, and based on the color, multiply that perterbation amount?