So, the idea is quite simple, let's make a 3D video with a regular camera!
In order to create a 3D illusion, our brain must get the same scene from both eyes, with almost the same perspective but different enough, in the same way we would if we actually were where the lens are when the camera is shooting/recording.
There are several ways in which we can accomplish this task. Key words to explore alternatives are:
- Anaglyph
- Stereographic
- Polarized 3D
- Active Shutter 3D
- 3D gif (not really what we're trying to do)
A quick search will give you all the information about these methods and many more.
As anaglyph googles are easy to find and quite cheap, I have chosen this method to create the 3D effect by means of red (left eye), cian (right-eye) googles.
Yes, there are!
A digital image is usually stored as a NxMx3 matrix. The number 3 stands for the Red, Green and Blue channels. Thus, we have three matrices that contain N by M elements. These elements are usually integers withing the range: 0-255 (8bits).
First you need to have two images, one for each eye. Furthermore, you need to know what image belongs to each perspective.
Then, as you wont see the red channel with your right eye, you need to take the right image and drop the red layer. In its place, we are going to insert the information of the left image, but not just the red channel. Instead, we are going to combine the three layers of the left image, as if we were converting it to grey scale, and then attach it to the previous image where the red channel used to be.
If the perspective of both images was accurate enough, we should be able to already see the 3D effect.
Doesn't this seem too simple?
The trick here is taking the images with the right distance in between, and maintain the same horizon in both of them.
Trial and error. Grab a camera and test your skills! In the beginning I thought I needed more distance between pictures, but our eyes are not 6cm appart by chance.
...to be continued...