Quote Originally Posted by jfrazierjr View Post
That's a fairly easy fix as long as you have the rivers on their own layer. Just like with the mountain and forest masks you made earlier, you make a river mask, so whatever layers you add the mask to will "cut" out for the rivers. For maximum flexibility in the future create both a mask AND a direct layer (which you will hide). You can do this easily by selecting your rivers, create a new layer, fill "not" river with black, and river with white, then save as mask. You now have a B/W layer and a mask version of the same. The reason to make it a separate layer in addition to just a mask is so if you decide to add any effects such as a bevel or something like that, you can regenerate the mask from the layer and then apply the new mask where ever the old one was(assuming you had not merged down the layer mask previously.

Joe

Yeah, so basically, how I handled this was I applied a mask copy of the rivers to the forests, except I enlarged the mask by a few pixels (so the white part that represents the river on the mask is a few pixels wider than the actual blue of the river) and I blurred the edge by a few pixels. This allows the forest to thin out around the edges of the river.

For swamp, the idea I'm considering using (also haven't done one yet, but it's on my list) is an approach similar to the one used for forests, but using a different color scheme. Most swamps, I figure, are forested areas (at least they are in the bayou and Okeefenokee area). I figured I'd tone down the green used with a heavy dose of gray. Your idea to use the galaxy brush is also good, but I agree it needs something a little more to make it pop and appear distinct from the regular grasslands.

All in all, your first attempt looks a lot like my first attempt, especially around the mountains and forests. I played around with the mountains using some other techniques to get what I wanted, but I'm still trying to refine that technique. Dan posted a revised technique for forests that he used that produces results that I think are an incremental improvement on the original technique RobA details.