How about some blood.
It is almost Halloween time.

Good results on the blood, Rod. Did you use an environment map for that?

No just a channel bump map and lighting effects. Basically this tutorial.
http://www.gimpusers.com/tutorials/smel ... loody-text

But i also gave it my own twist by duplicating the visible result, inverting the color and setting the layer mode to grain extract. That made it sharper without blowing out the highlights.

Ah, OK. The cool thing about GIMP is there are so many different techniques that can be used to get similar results.

I think I'm about burned out on environment maps. I've gotten better results this time around so I guess that's progress.

More results using emaps...



A quick attempt bloody text...

So Tux when you use an environment map you use one much larger than the image you are creating right?

Yep. When using a pattern or an image, I create the emap at least 4x the site of the object I want to map the environment onto. The effect from the environment map stretches horizontally, so what gets mapped to the object, depends on the distance of the object from the surrounding environment. Here are the results of my tests.

These squares are mapped from an emap 4x the size of the original image, filled with a money pattern. The shiny value in the materials tab has a significant affect on the overall brightness. I didn't specify a bump map.

At a distance of .295


At a distance of .395


At a distance of .495


At a distance of .595


When I add a bump map for an edge bevel in the lighting filter, I get this result. This is from a Gauss blur of 15. Distance of .595 with shiny set to 50. This is the effect I get when the environment is bent (reflected) around the beveled edge.



All of this applies to an environment map of an image or pattern. When using a gradient or an emap like the ones below, and due to the stretching that occurs, the emap size requirements are different and they produce different results.



This is in all contrast to the simply bumping a pattern for height, then bumping for an edge bevel and adding light, which produces a totally different effect.

After a bit more testing, it seems that gradients should be at least 2x the height of the original image for smoother results. The width of the gradient doesn't seem to matter much. Here is my result with a horizon style gradient at 30x800 .



The size of what I'll refer to (for lack of a better term) as a topographical (satin?) gradient, should be at least large enough to produce smooth gradients when mapped. The standard 400px square seems fine in my tests. Although, increasing the size might give smoother results, depending on how the topographical/satin gradient is created. The random blend of dark and light gradients give different lighting and shading when mapped.

I started with standard gold map, then colored, transformed and iwarped it to get this...



..which produced this kind of shading when emapped. I'm calling it Blue Aluminum.



I got the above results using the blue emap, with a 15px Gauss applied to the duped text layer for my bump map (sinusodial - depth 0.01). Materials shiny set to ~.50 & distance at .550. I then re-ran the lighting filter on the results of the emapped & bmapped text, unchecking the emap and leaving the bump in place, trail & error adjusting the light and materials settings in the lighting filter to get the final result.

science !

Fantastic results Tux!
Pretty awesome that you need only open any pattern as an image and correct the height or width for different env map results. Science is wonderful.

Thanks Tux for all your testing with this. It is nice to finally see what can be done and how to accomplish it with this filter. It does give some outstanding results used correctly. Great stuff. I can not wait for your script results. I am betting they will be phenomenal.

whats up with this artifact ?

i had the same thing at, more or less the same point, several times in the past
is it some bumpmap related bug ?

I hadn't really paid much attention to that Esper, but it appears to be an artifact from the emap section of the lighting plug-in. The artifact occurs exactly at the midpoint of the image and is only noticeable on a diagonal line. I suspect it is due to the stretching that occurs from the midpoint (left and right outward) when the emap is actually mapped to the target object. It is only noticeable on a diagonal section. The fix is to ensure a diagonal section doesn't fall in the exact middle of the image.



@Rod - For patterns, I've just been creating an empty map image 4x the size of the original with no alpha, then filling with the selected pattern. I'm not sure if emapping is better than the traditional method of separate bumping and lighting, it's just a little different effect.

its still strange that i got the same artifact on my image - and i made that with no emap, but an ordinary bumpmap





thats the best result i have seen so far !
looks almost like car paint

Sometimes, I get some strange artifacts using bumpmaps too, especially with certain curves, like spherical. Do you remember the settings you used on that bumpmap? i wasn't able duplicate the artifact on my end.

yes i think so, i took the image from the bumpmap tutorial i wrote for Akros
http://gimp-science-labs.blogspot.de/20 ... pping.html

Depth 10, Azimuth 140 - all the other values on default
the bumpmap was a selection filled with white on a transparent layer, blurred by 5

i attached the xcf if that helps

I think it's just a bad bumpmap. I tried one with a Gauss blur of 10 and it doesn't have that artifact.

These are some excellent results Tux, must have a play soon, is it possible for you to start a dedicated tutorial thread on this as I'm getting lost with all the posts and settings.

Thanks, He4rty. I'll start a tut thread on these results.

Excellent work tux.