We start with a simple example. Our Source Image is our Arches National Park photo that we have loaded into a Texture Slot. In our Slot 1 we load the 4-Color Gradient Effect and with Output Color 1 set to Ignore, Output Color 2 set to Replace, and we set the LL Color to black, the UR Color to white, the UL and LR Colors to a medium grey. The result is the image on the right below:|
In Slot 2 we load our Texture Displace with default values. We get this as a result:
First, because we had a grey gradient for Color 2, the Source Image mapped the same in the Red, Green, and Blue Channels. That kept the result more cohesive (just wait, you'll see in a moment what happens when we don't use a grey Color 2).
Again, Color 2 is a map of how the pixels in the Source Image are repositioned on the screen. Since there are many of the same color in Color 2, some pixels in the Source Image will be referenced multiple times for a single point.
You can see that the result was highly distorted and to the left of the Canvas. We can use the Horizontal Offset option to move it to the right, exposing more of the Effect:
So what happens if our Color 2 is not a grey? Consider the following Source Image and Color 2, which in this case is the a 4-Color Gradient:
Here is the result using a default Texture Displace:
Unlike the grey gradient, which kept the image more cohesive (using the term very loosely), using a color image, in our case a 4-Color Gradient, split each color channel in the Source Texture. The Red, Green, and Blue channels of the photo were all mapped independently.
What happens if we switch our Source Texture and Color 2? Let's try it with both the grey gradient and color gradient:
With our Source Texture the grey gradient and our Color 2 the photo Texture, we get this:
Now that is cohesive. But why? It is because we were using a gradient. The map is now the photo, so all we are doing is reassigning the position of the gradient to new areas. This recolors the gradient and repositions the gradient, giving a brand new photo.
Here is the same thing with the Source Texture our 4-Color colorful gradient:
And this combination results in this:
Again, because our Source Texture is a color gradient it remains cohesive. The resulting image is colored completely by the Source Texture (no color is coming from the Photo in Color 2). However, the colors are no longer the same as the gradient because the different R, G and B values in Color 2 reassigned the R, G, and B values of the gradient.
Let's examine why the gradient made all the difference. What happens if we have both the Source Texture and Color 2 the same Photo:
Here is the highly non-cohesive result:
An interesting, though messy result. But also a good example of the difference of a gradient versus a photo.
In our final example we use our photo as the Source Texture and for Color 2 we use Clouds, Marble, and Granite:
This gives us a very interesting result, almost a stone-like pattern: