You can create 3D pictures not only with a camera or special software - you can use a pen or a graphical editor as well. In this article we will explain, how to paint an anaglyph and a few important rules for creating anaglyphs generally.

   As you know (from previous issues of 3DJournal), the right eye sees things shifted a little bit more to the left than the left one. This shift is bigger when the thing is near to the observer and smaller when it's far from him. It's important to remember this fact before we start our painting.
   There are more ways how to paint your own 3D pictures. The simplest one is this: Paint a 2D picture and then create it's variants for the left eye and for the right eye. And there is one more important thing to know: the quality of the picture depends not only on the precise shift between the eyes but on the chosen colors as well. Let's see it on an example.
  
Let's go
   You can see a sky on this picture, hills and two faces. The picture is flat - 2D - but you can probably imagine which objects are far and which of them are nearer to the observer. To create a 3D picture we have to create 2 different pictures - one for the left eye and one for the right one.
   Let's decide this first picture is the left eye picture. To create the picture for the right eye we have to shift the objects, which are nearer to us, a little bit to the left (we know the right eye sees things shifted a little bit more to the left than the left one). Nearer objects are shifted more, farther ones less.
   The best way to make a shift is to use a graphical editor. If you can use layers then it's a good idea to place different objects to different layers and then shift layers. If you (or your editor) can't work with layers, it's necessary to select the item you want to shift (by a selection tool) and move it. It would be necessary to fill the background of the item after shifting it with the right color.
   On the second image you can see the view of the right eye. At it's right border you can see a gap - it came into existence by shifting the objects. Load both pictures into 3DJournal software and create a 3D picture. You can see here what you get. Try to use 3D glasses. The picture isn't 3D - is it? Did we do a mistake? What?
  
Possible mistakes
   The first possible mistake is the shift - have we shifted the objects too much or too little? Sure, we should use a few mathematics, but for now we can estimate the shift.
   The second possible mistake could be the selection of colors. Red-cyan glasses change colors and filter some of them. Try to insert our pictures to the 3DJournal software (you can save them from the menu called by a right mouse button), create a 3D picture and then try to shift them. You can see that one of the pictures is too bright and the second too dark. And that's the problem.
   If you check the black and white checkbox the situation changes. You can see the 3D picture. (If you make your own experiments we recommend to use larger pictures to see the effects better).
  
Better image
   Try to make one more experiment: Watch the black and white picture and move your head from - and back to - your computer monitor. In certain distance is the picture the best (it should be around 1 meter). Some objects look like they are before the monitor, some are behind it. It's good to know the shift between the objects (on the left and right image) depends on the distance of the observer from the monitor (or paper with a printed image) as well.
   And one more experiment: Try to shift left and right pictures (in the 3DJournal software) to place the faces on the left picture exactly on the same place as the faces on the right one. Maybe you remember we shifted (in our first picture) the objects in the front - and now they are on the same position and shifted are the objects behind them. Now you can see (with 3D glasses) that the faces are on the same level as the screen and all other objects are behind it.
  
One more improvement
   Here you can see the picture after one more change - we have changed the colors. Their red and cyan components are better weighted. The 3D effect is much better now (sure, the black and white picture is still better).
   And now watch the last black and white picture. It's more 3D than the other ones. Why? Because there are no big single color areas. We have used textures instead. If you shift the big single color area most of the points are still the same - single color. But if you shift the texture then there is much more changes - every part of the texture shifts. And that's the trick. Textures are much better for 3D than single color areas.
   And one more thing: If an object on the pictures should be in front of the screen it mustn't intersect its border.
   Don't forget these rules. It will help you to create the best 3D pictures possible.
(3DJournal)