Integrating CG into photos - Part I: Creating environment maps from photographs
Bonjour and welcome to the first of my two part (possibly 3, who knows maybe even FOUR!) tutorial on integrating computer generated imagery into photographs. This obviously can be extended to video, which I might cover in another tutorial.
Whilst these tutorials are demonstrated using Blender, the techniques should be easily expanded to other 3D packages - so don’t be afraid to try this in SUPER 3D CREATOR MAKE ACE STUFF WITH ONE BUTTON PRO v2, or whatever software you use…
So I got a bit bored the other day, so I decided to attempt to create an Environment Map from a set of photographs… and it WORKED!
Summary of what photos you need to take:
- 6 photos for cubic mapping - front, back, left, right, top, bottom
- 1 (or more if you want to try out different angles) perspective photo (has to be taken in the same/similar lighting to the other 6 photos for the best results).
So basically what I did was to take 6 photos one for each face of a cube (front, back, left, right, top, bottom), as an environment map is usually (if not always) a cubic map:
I then had to use the batch tool in GIMP to crop the images so that they were square (clearly since these are getting mapped to a cube, the images must be square!).
The next step was to combine all these photos into one “block” - this will be the environment map. To check what this should look like I used the EnvMap texture function in Blender, and saved the outputted environment map of a simple 3D scene (just a bunch of cubes). I did this mainly to ensure that the environment map I created would be compatible with Blender, and here’s the Blender-generator environment map:
As you can see (well maybe not so clearly, sorry about that!) it is a rectangle 3 squares across by 2 squares down.
So! I used the alignment tools in Gimp to align my 6 photographs into the blender-compatible format, giving me this:
The images are in the following order (left to right): Left, Back, Right, Bottom, Top, Front.
Okay so all that’s left to do is to bring this image into Blender (or your 3D package of choice!) as an environment map… which I’ll cover in Part II.
As a little teaser for Part II (this is a bit of a WIP image, so it’s not perfect by any means):
