Any Video To Anaglyph 3D Nodes

Support Compositing and Post Processing
1

Here is a trick I came up with a long time ago to convert any video to Ana glyph 3d ( I used to convert Star Wars movies into 3d) :D. This is the kind where you have to wear the glasses to see the effect. With this node setup you can make any video 3d. However this is more of a hack type of 3d. Firstly the images will project their depth, into the screen, not out from it. Secondly either the camera or objects on screen have to be moving for there to be a 3d effect also, (luckily both of these criteria are usually present in most videos). The advantage of this technique is that you dont have to render 2 images, one for the right eye and one for the left eye. Also as I said you can use any video you already have.

I have been able to get the full 3d effect with a new node, the model now pops completely out of the screen> the model is also viewable from a wide viewing angle.

This video has been rendered with the new node, keep in mind that it has been compressed which lessens the 3d effect.

Any questions or comments are welcomed.
Also a rendered frame from an animation :

http://i705.photobucket.com/albums/ww58/Bluefurryyspider/3dultimate.jpg
3dultimate.jpg600×600

2

Can you tell how you did it?

3

You really need the image displacement and parallax of two camera/eye views to get a successful stereo image. In your example above there is some perception of “depth” but it’s very clearly not tied to the actual form of the subject, and some parts just seem to float separately with no connection where there should be one (the sensors on the front of Fat Man, for example). I appreciate your effort, but depth perception depends on more than flat image-plane displacement.

4

Yep pulling z-depth from a 2d image would be kinda amazing.