Since the beginning of television there has been the dream of broadcast
in three-dimentions. There is an irony in that the 3-D movie craze of the
1950's was driven, at least in part, by the Hollywood's need to win back
the audience lost to the new upstart television. Proponents of 3-D TV
are fond of pointing out how first color then stereophonic sound have
become normal, accepted parts of the viewing and listening experience.
By extension, we are supposed to believe that 3-D will soon become part
of TV. It is always coming soon, just over the horizon or possible and pr
actical now if people would just purchase and wear some new gadget.
After 20 years of working with almost every method of 3-D imaging my sense
is that 3-D will not and need not become universal, mainstream or even simple.
It is a very special form of imaging and it will always be a special option with
every medium. What is important is that when it is done it is done well.
Let's start by looking at what 3-D is and isn't. The world out there exists in at
least three dimensions: height width and DEPTH. We (humans) visually perceive
the world in depth by a variety of means. One of these means is Stereoscopic or
binocular vision in which the brain analyzes the difference between what the two
eyes see. This is not the only or even the dominant way that we experience depth
but it can be the amazing. For the purposes of our discussion, 3-D Television is
any method of viewing where two different images are presented to each of the
viewer's eyes making stereoscopic vision possible. All 2-D media from drawing
to photography to film to television have ways of conveying depth but binocular,
stereoscopic imagery really is something else and that is really what we mean
when we say 3-D. There is one popular and important thing that is called 3-D
but isn't 3-D in our sense. When a computer is used to render a flat 2-D image
based on three dimensions of information it has stored about a scene, that is
also called 3-D Graphics although it is rare that the computer is asked to render
stereoscopic pairs of images.
It is over one hundred years since Louis DuCos du Hauron proposed that the two
images needed for 3-D be presented in two complementary colors and the viewing
public be given spectacles of those two opposite colors for viewing. One of the
two colors is red and the other color is either green or a blue-green mix. This
anaglyph process can work well for printing on paper or photographic film and
works great on a RGB computer screen but, unfortunately, it generally fails the
test of our NTSC color system.
Over & Under and Side by Side
Clever inventors have come up with every possible combination of putting the
two images next to each other or above each other or opposite and then viewing
the pair with the aid of lenses, mirrors, prisms and the like. Often one of these
arrangements will be perfect for one specific application. (ie. a 3-D arcade game)
One of the earliest methods of viewing 3-D with moving images has been to
alternate left & right images rapidly and to provide the viewer with glasses
that alternately shutter so that each eye gets the correct image only. In the
1920's the shutter was a rotating disk mounted in front of each seat in the
movie theater. The idea did not catch on. Recent developments in the area of
liquid crystals have made possible lightweight electronic shutter glasses and
some promising work is starting to be seen. It is useful to divide liquid crystal
shutter systems into two groups: high-speed or double frame rate and field-sequential. The high-speed or double frame rate systems alternate the left & right images very rapidly and the the resulting 3-D can look great. The disadvantage of this system is that the electronics and monitors are non-standard; meaning expensive. The field- sequential systems alternate left & right views in the two fields that make up our standard interlaced television system. This makes this system eminently practical but the reader will need to judge for him/her self whether the amount of flicker is acceptable.
In 1922, a German physicist named Carl Pulfrich discovered that if an object
moves horizontally across your field of view while you have a dark filter over
one eye, the object will be seen as forward or back in depth depending on the
speed and direction of the object. The reason for this is a time delay for the
eye wearing the filter. It is not known whether Pulfrich had any idea that his
discovery would lead to a method of 3-D television. In the late 1960šs a
device called "TV Stereo Spex" was widely sold that claimed to offer 3-D
from your existing TV. The device enabled you to darken one of your eyes by
rotating one of two polarizing filters. It also had blue and yellow filters to
heighten the effect. Inspired by this device, a number of filmmakers produced
films that exploited this effect. I started using the Pulfrich effect in 1975
and have developed my own proprietary incarnation of the process, PullTime 3-D.
Designing for this process is a tricky choreography job since direction and speed
of movement is almost the sole determinant of depth. Still, for the right subject
matter (ie. music video) it is possible to create 3-D television that is 100%
compatible with all normal TV and does not distort color in any way. At this
point, aprox. 20,000,000 of my PullTime 3-D glasses have been distributed.
3-D Without Glasses
Obviously, it would be wonderful to have 3-D television that does not require
any kind of 3-D glasses or special viewing device. Some very interesting research
is being done, but, don't hold your breath. The MIT MediaLab has created prototype
holographic television. Japanese researchers have created 3-D TV without glasses
using a barrier strip in front of a flat screen HDTV display. It looks great until
you move your head. A couple of American groups have developed systems that
insert subtile or not-so-subtile movement into what can become a conventional
TV image with the claim that this makes the picture more 3-D.
When 3-D movies are projected in a theater, two strips of film are projected, in
sync, through two polarizing filters that are aligned at right angles to each other.
The audience wears corresponding polarizing glasses. The same thing can be done
with video projection although the lower brightness of video projectors this less
Gerald Marks 1995
American Cinematographer Video Manual
back to previous page