As reported in Daily Variety Jeffrey Katzenberg was recently heard decrying the lack of live-action stereoscopic features.  He exhorted the industry to correct this situation.  He was also quoted as musing about his motivation for going beyond the mandate of his own particular self interest, animation, by taking on the live-action cause.  But if there were more live action stereoscopic features in the theaters it will also be good for people who make feature-length animated films, like Katezenberg. 

The stereoscopic cinema, as it has developed over the past four years, consists primarily of stereoscopic computer-generated animated features or, in plain English, 3-D cartoons.  There are a couple of reasons for this:  In terms of content this resurrection of the stereoscopic cinema is different from that which occurred in the past.  In the last half-century or so there have been two stereoscopic resurgences of note:  In the early ’50s 60 features were made, and they were all live-action shows.  (There were also a few animated shorts using cell animation.)  You are probably familiar with titles such as Creature from the Black Lagoon, Kiss Me Kate, Hondo, and Dial M for Murder.  

Then in the early ‘80s, some 20 years later, theaters projected in a single-strip 35mm projection method identical to the recent Iglourious Technicolor system.  What occurred in the ‘50s used interlocked 35mm projectors, repurposed from changeover, which were pressed into stereoscopic service.  But the 80’s single projector system, in an attempt to overcome the shortcomings of the two projector system, used prints with two subframes arranged above and below each other with each frame having a 2.4:1 aspect ratio. Several kinds of optical devices were added to the projector or its lens to combined and polarized the subframe images to project movies like Amityville 3-D, Comin’ At Ya!, and Metalstorm .  Although far fewer titles than were released in the ‘50s, once again these were live-action shows.  

Now a little over 20 years later in 2004,  we had Real D in about 89 theaters in North America showing Chicken Little – a bad film that did well because of Disney marketing and because it was in 3-D.  It was brought into existence by the vision of Disney, who has been an innovator in everything from the stereoscopic cinema, to stereoscopic sound and theme parks.  

We can expect the great majority of CG animated features to be stereoscopic because those recently released this way have made money – more money in 2-D than in 3-D.  There will be a handful of such shows that aren’t, but those efforts may be financially ill-advised because the ticket price for a 3-D movie is higher, so the exhibitors and the studios can potentially make more money – which is one way this whole thing gets driven.  

This cycle there have been a relative handful of live-action stereoscopic features.   Initially we had kids’ films, which are animated cartoons, and now we’ve got horror movies: Scar, a movie that was released in Europe but not in the United States; My Bloody Valentine; and Final Destination 3; soon we will have Piranha 3-D, which is set for release; and another horror film called The Hole.  If horror movies follow the same pattern as CG animation we will have another genre that has succumbed to the stereoscopic cinema. We also had Journey to the Center of the Earth a family film, or a teen sci-fi movie I think supports the noting that the demographic is going up – from little kids to older kids. 

The other night at an Industry diner I heard people saying that Avatar was a make or break for the 3-D cinema. I don’t think that’s true, but people are eagerly awaking the film which will be a combination of live action, CG and performance capture.  Avatar can probably best be classified as a live-action film with a lot of CG.  I think you would classify King Kong as a live-action film with a lot of CG, so why not Avatar?  However it’s classified it raises the age of those who will be interested in the film.  In fact Avatar has a very broad demographic. 

It’s too much to say that the fate of live-action stereoscopic cinema hangs in the balance, but if Avatar does well it’s surely going to be a good thing for the 3-D cinema.  And it can help inspire another genre on its way to becoming stereoscopic:  science-fiction.

 The Bruckheimer-Disney film G-Force is another sci-fi live-action film, albeit juvenal, with lots of CG effects. The technology behind G-Force is interesting because it wasn’t shot stereoscopically, and neither were two other live-action films, Alice in Wonderland, and Piranha 3-D.  These movies are being converted to 3-D from 2-D cinematography.  Alice is a big-budget major feature, and the produiction opted to shoot in 2-D.  The director of Piranha 3-D, Alex Aja, wanted to shoot in 35mm anamorphic.  That more or less ruled out shooting stereoscopically.  Conversion was also used for the successful rereleases of Nightmare Before Christmas

So we’ve got a handful of stereoscopic live action narrative feature films (the fantasy and horror elements are prevalent) and three concert films that I know of: the superb U2 film; the Hannah Montana film and the Jonas Brothers movie that bumped Coraline out of many theaters – Coraline being an interesting example of stop-motion animation, and quite an eye-dazzling film at that. 

The question remains unanswered, why aren’t there more live-action narrative stereoscopic movies?  There is no simple answer, but part of the answer may have to do with technology and part may have to do with aesthetics and filmmakers’ choices.  Paramount’s decision to make the latest Star Trek movie in 2-D may be a clue. For awhile the studio considered it to be a candidate for a stereoscopic release but Paramount decided not to, and it was a good decision.  It’s a strong franchise, I’m not sure they needed the stereoscopic medium, and at the time that the decision was made there were few stereoscopic theaters.  In fact, there still are comparatively few stereoscopic theaters; but when the decision was made there were hundreds of stereoscopic theaters, and now there are a few thousand – still not enough for a financially satisfying 3-D release of a major feature, and certainly not enough theaters if there are two major features coming out in the same time period. 

Could production technical limitations play a role? Let’s take a quick look at the state-of-the-art of production and post-production.  The current camera rigs are derived from a design by Floyd Ramsdell in the late 1940s.  A major issue with stereoscopic cinematography is that the cameras’ lenses have to be brought very close together – in fact, closer than the eyes are apart – to do a lot of successful stereoscopic cinematography.  That means you either have to use tiny cameras (which are not up to studio quality) or you have to find some clever way to get two big cameras optically closer together.   

The blimped studio cameras at the time that Ramsdell did his work were gigantic compared to today’s Arriflex and Panavision cameras, for example; but the digital cameras that are around today are not necessarily smaller than film cameras.  In any event, Ramsdell had the idea of placing two cameras at right angles to each other on a horizontal surface shooting through what is called a beamsplitter, or a pellicle, or a semi-silvered mirror.  In this way one camera sees a reflected image and the other camera sees through the semi-silvered mirror so the camera lens axes are brought close together.  

In the early 1950s Raymond and Nigel Spottiswoode had an idea for making the design more compact, and placed one camera on the top and the other on the bottom, the top camera shooting through a semi-silvered mirror and the bottom camera looking straight ahead.  That is the design used in the vast majority of extant stereoscopic camera rigs.  

There are three major vendors of these rigs.  I don’t want to sell the other guys short because there are many other people around like Kerner Optical, and Jason Goodman  at 21st Century 3-D, who also offer rigs, but the three major vendors are 3ality led by Steve Schklair, Pace Technology led by Vince Pace, and Paradise FX led by Max Penner.  These outfits are in the San Fernando Valley.  The camera rigs that these gentlemen have produced have been used to great effect for feature films.  When you see the result on the screen it looks good.  

But these are not easy cameras to use.  They’re complex rigs, they don’t look like normal movie cameras, and I know from talking to cinematographers that they are not completely convinced that they can use these instruments.  They are, as I said, based on a design that was cooked up by Raymond and Nigel Spottiswoode, two very important people in the history of stereoscopic cinematography because of their work in the Festival of Britain and also because of their invention of floating windows.  

Without going into all of the ins and outs of the differences between the various camera designs, as I said, they are all basically capable of good work.  They use two cameras that need to be coordinated, and making two cameras work like one camera – focusing and zooming and setting the stereoscopic controls – is not a trivial matter. They also, as far as I have been told, require extensive post-production correction.  But post-production sweetening or correction is the order of the day for producing digital intermediates, so this may not be so much out of line with what is the accepted practice. 

Fortunately, two post-production instruments have been placed in the hands of filmmakers in the last couple of years.  One is the Avid, which has the capability of allowing filmmakers to view stereoscopically.  This offline editing system has the capability of allowing footage to be viewed without it having been color timed without it having been stereo timed.  “Stereo timing” is the analogous function of color timing, in which the stereo parameters are set so that within a scene there is a good 3-d flow from shot to shot.  

The other tool is the Pablo by Quantel which provides for the actual stereoscopic sweetening itself.  The Pablo has gone a long way toward allowing filmmakers to control the image, but more work needs to be done. 

Why do I say that more work needs to be done?  Because the guys who are creating stereoscopic CG animation are masters and the camera people are just catching up.  The CG people have done fantastic work.  They are teachers, because the live-action work now has to be as good as what the CG stereo supervisor have accomplished in terms of images that are easy to view, are beautiful, and have the ability to further the story.  Stereoscopic CG animation is, in a sense, much easier than camera-derived images because the animators totally control the space.  They control the parameters of the virtual camera, and they can do all kinds of things that a live-action filmmaker can’t easily do, including changing the distance between foreground and background, and even using different interaxial separations for different objects in the scene that are different distances from the camera.  

These supervisors include Phil McNally (also known as “Captain 3-D”), of DreamWorks Animation, Robert Neuman at Disney, and Rob Engle at Sony Pictures Imageworks.  Also, Jayme Wilkinson at Blue Sky did a fantabulous job with Ice Age: Dawn of the Dinosaurs, which I enjoyed greatly. 

So I don’t have a simple answer to Jeffrey Katzenberg’s plaint, but I do think we’re going to see a genre-by-genre transition, the first being kids’ CG animated films and then live action horror films. I think science-fiction and action movies are next, and I we may see every genre fall to 3-D – or rise, as the case may be.  The reason for this is economic.  If you have a multiplex and 3-D movies are making, on average, $3 more per seat than 2-D movies, the exhibitor, the studios and the filmmakers take note. 

It’s hard for me to believe that all live action features are going to be synthesized or converted from 2-D.  That process certainly has its place. But once the effects producer breaks down the show shot-by-shot (and that’s the right person to do the job because 3-D is an effect) a decision will be made to determine which shots or scenes will be performance capture, which will be CG, which will be converted, and which will require stereoscopic cameras.  The reason will have to make sense and pass the usual effects tests:  Is this the way to do the shot that will look the best and cost the least? Most of the time, with the right instrument, the answer for live action features will be to use a 3-D camera. 

Full disclosure: My daughter calls me Poppy, and I’m working on a stereoscopic camera system.


  1. 3atltl Says:

    Tim Sassoon at SMPTE said that 3D live action won’t really take off until we have monocular stereoscopic systems….what do you think??

    • lennylipton Says:

      I honestly don’t know what that means since a stereoscopic image requires two perspective views.

    • lennylipton Says:

      We need cameras that are more convenient to use and more nearly resemble the kinds of cameras that DPs feel comfortable with.

  2. 3atltl Says:

    presumably you could do this with one lense but two sensors….Sony is trying….

    • lennylipton Says:

      Looks like their announcement was premature. Can’t get an adequate interaxial separation, except for extreme closeups, out of one lens. Also going to a high field rate with a single sensor just does not work becasue the iamges have to be captured essentially simultaneously. Electronic giants like Sony need to learn a lot more about stereo 3D before they can offer good products.

  3. fredfred27 Says:

    It seems to me that the correct procedure for live-action stereoscopy (as opposed to 2D to 3D conversion) will be:

    1. Photography of foreground characters/subjects in front of green/blue screens with specific inter-axials.

    2. Photography of the background with its own specific inter-axials.

    If the camera moves in the shot, a motion controled camera can be used.

    This places a heavy burden on inter-axial calculation on the set, but the costs of motion control and on-set 3D expertise should far outweigh the post-production costs for such shots.

    And on a slightly unrelated matter…

    I was intrigued by your discussions on the non-Euclidian aspects in the literature concerning stereopsis cited in “Foundations…”, do you have any more up to date references on the subject?

    • lennylipton Says:

      Your idea is a good one and the concerns that lead to your solution make conversion from planar to stereo 3D attractive.

      The discussion about non-Euclidean space representing how we preceive the visual world is intersting but I don’t think this is currently a hot topic in psychology. The position I now take is that the mind is a black box and stereo compostion is best determined empirically and subjectively.

    • treehaus1 Says:

      Hi fredfred27,

      that is a very interesting proposition.
      But what do you mean by different interaxials? Bigger for the FG or bigger for the BG?


      Simon Sieverts

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: