The reason I want to start a blog post series about stereoscopy and autostereoscopy is that it has been my field of research for the past 2 years due to an investigation scholarship granted by Universitat Jaume I (Castellón, Spain).
Stereoscopy is well known around game development community, especially since the rise of 3D Head mounted displays(Oculus, Htc Vive, etc). But not so many people know the concept of autostereoscopy. This first post will try to clarify such concept.
Autostereoscopy refers to the technique that allows humans to perceive 3D from a 2D source without the need of Head Mounted Display to do so. A.K.A watching 3D without glasses. Autostereoscopy has existed for quite a while, however, due to its nature, it has not arrived quite to our homes yet. Autostereoscopy has some drawbacks that are slowing the process of becoming a multiple purpose standard display. The content creation is hard and meticulous. Also, it may cause some discomfort during and after the viewing process to some people. And finally, not everyone has stereo vision.
Autostereoscopy is possible thanks to the human vision system or binocular vision. When we see something, our brain processes 2 different images in the visual cortex to extract depth and other cues that allow us to see in 3D and have depth perception of things. Stereo displays use that principle to use one source scene, render it from 2 different perspectives and send each render to one eye. To send those images to one or other eye multiple techniques have been developed: anaglyph or polarized glasses, head mounted displays with special lenses and a large etc.
On one hand, stereo displays use that principle to use one source scene, render it from 2 different perspectives and send each render to one eye. To send those images to one or other eye multiple techniques have been developed: anaglyph or polarized glasses, head mounted displays with special lenses and a large etc.
On the auto-stereo displays, use precise math to make the viewer receive different images to each eye through light refraction. Generally speaking, autostereoscopic displays create 3D through a combination of both hardware techniques and software techniques. Firstly, hardware techniques, like a barrier in front of the LCD/plasma panel, are used to refract or neglect the light that is being emitted from the display. Secondly, on the software side, information about several images is reorganized into one single image in such a way that the pixel arrangement creates a pattern to match the type of hardware that is simulating the 3D to the viewer. For instance, if a display used lenticular lenses, it is needed that each N pixels of the LCD are within a lenticular lens to refract the image. Another example: for the 3D in the 3DS to work the image must be rendered forming a pattern such that some columns of pixels are blocked by the barrier while others are not.