Creating caves in POV-Ray
by Mauro Pierluigi

This tutorial will show us a simple but successfull method to create caves and grottos, by manipulating height field objects in POV-Ray.

Note: This tutorial is not for beginners and it assumes that we know already the basics of POV-Ray rendering system, coordinates and scripting.

An height field is a fast and efficent way to generally create mountains or other complex surfaces. It needs a particular image map. The rendering engine reads every single pixel of the image map and accoring to the color of the pixels it draws triangles different in height. The more a pixel is near to white the more the triangle is high, the more a pixel is near to black the more the triangle is low, and so on with all the color scales.

So, let's start with creating the image maps. You will need a program like Photoshop, Photopaint Paint Shop Pro, etc.

Create a black square, let's say about 500x500 pixels, but the size doesen't matter.

With a gradient fill tool we have to create many spots like it is shown on picture 1 and we have to fill a large area paying attention to left the border of the image completely black. Then save it as *.tga or *.gif file. This will be the upper part of the cave.

Now we have to create the lower part, to do that we have nothing else to do than invert the first image using some inverting function of your preferred graphics application. Save it also as *.tga or *.gif image.

In picture 2 are shown the two images map that I created for the two height field objects.

This two pictures will create two opposite and simmetric height field objects, the first will develope in an upper direction and the second in a lower direction.

Now we are going to write the code to see how the height field will look after the rendering.

 

#include "colors.inc"

camera {
   location <5, 8, -6>
   look_at <5, 3, 0>
}

light_source { <0, 50, -50> color White }

height_field { tga "map1.tga"
   pigment { Yellow }
   scale <10, 1, 10>
}

The result of this rendering is shown in picture 3, to look at the other height field we just have to change the name of the image file from map1.tga to map2.tga or whatever we called it. The result of the second is shown in picture 4.

Now we have to join the two objects togheter, the flat areas shown in the pictures must be in the same Y position. So the code for that is the following:

#include "colors.inc"

camera {
   location <5, 1, -8>
   look_at <5, 1, 0>
}

light_source { <0, 50, -50> color White shadowless }

union {
   height_field { tga "map2.tga"
      scale <1, 5, 1>
   }
   height_field { tga "map1.tga"
      translate -3*y
   }
   scale <10, 1, 10>
   pigment { Yellow }
}

The result of the rendering is shown on picture 5, as we see they are the upper part and the lower part of a cave and they are perfectly joining each other in the shape. We scaled the upper part to make it look more high and give it a more real look. If we change the value of the translation from -3*y to -1*y we see that the two flat areas joins perfectly how shown in picture 6.

Now if we shift the camera position inside this object we will see like we are inside a cave. The result of the rendering is shown on picture 7 and we can see how the shape of the two objects are perfectly joining each other.

The result of the scene with textures and some objects inside is shown on picture 8.

Mauro Pierluigi is a graphic designer and programmer and collaborated for years with some of the most important italian ISP. He owns a company which creates multimedia products, animations, web design, etc, find out more about him at www.pmworld.net


Editors Note: For more information on POVRAY, try http://www.povray.org/