AO RESEARCHERS FIND SOLUTIONS TO LENS DISTORTION

From The News

Wednesday, October 16, 1985

 

TODD -AO Southbridge’s Role in the Movies - Third in a Series

 

Dr. Brian O’Brien was well aware of the problems Dr. Robert Hopkins faced in trying to create a movie lens at American Optical Co. for Hollywood producer Mike Todd.

O’Brien and Hopkins, rather than trying to create an impossibly perfect lens, elected to accept some distortions from the lens - distortions they figured could be taken out in the process of printing the film.

The distortion they chose to accept in the lens was “barrel” distortion.

Fortunately, this type of distortion in the negative is exactly what is needed if the image is to be distortionless when projected on a curved screen.

Notice how the image on the film is curved. This is a deliberate curve to compensate for the cure caused when a normal image is projected on a deeply curved screen and viewed from beneath the beam.

We have to be concerned, however, with more than the distortions by the wide angle camera lens. To give the illusion of depth through the allowance of peripheral (side) vision, we have to have a deeply curved screen. Further in a theatre the projector is usually mounted high above the heads of the audience.

First the effect of the curved screen on the image projected perpendicularly onto the screen is to produce a pronounced droop in the horizontal lines.

Try it yourself by shining a home slide or movie projector straight onto a piece of curved cardboard.

If we now mount the projector above the plane perpendicular to the screen, we now produce a keystone effect in the vertical lines in addition to the droop in the horizontal lines.

 

 

Researcher Brian O’Brien Jr. also noted that if you think about this further you will realize that unless something is done about it, normally more or less round faces at the ends of deeply curved screens are going to be distorted into ovals and look “pig-faced.”

This was one more problem for the optical designers who wanted to use the regular projection booth.

We can now see, however, why the type of barrel distortion left in the lens was desirable as it counteracted the barrel distortion introduced by the use of the deeply curved screen.

Dr. Walter Siegmund said there were three corrections made in the projection printing process:

A-correction for barrel distortion in the vertical lines; the B-correction for droop, and the K-correction for keystone. The result should be a distortionless, perfect image when the picture is properly projected in a theatre.

Arthur Kavanagh, now retired from AO Southbridge, designed the printing illumination optics while Henry Cole and Ian Crawford of the machine shop were deeply involved with the complex mechanical printing mechanisms.

Dr. O’Brien, no slouch himself when it comes to machine design, suggested deliberately deforming the film onto a curved, slanted shoe with a vacuum back to further correct for the residual horizontal and vertical distortion in the printing lenses. This metallic show had to be of optical finish quality and this job fell to Colin Yates, then head of the precision optical shop in the research division.

To achieve the ultimate corrections, it was necessary to have different corrections for each taking lens. O’Brien Jr. suggested using a line-by-line instead of a frame-by-frame printing with a cam mechanism to effect the desired changed in correction line by line.

The AO projection printer was probably the most exotic film printer that had been developed for the motion picture industry.

The Camera Body

It was one thing to make an exotic optical system that would function beautifully in the clean, secure environment of an optical laboratory. It was quite another thing to make a camera that could be used reliably in the field.

In fact, the camera got perhaps its roughest test when it was hauled all over the world - up onto roofs, down in boats, in cars, and trucks when Todd made his “Around the World in 80 Days.”

The supervision of the camera construction fell on Henry Cole now working with Siegmund at Reichert Fiber Optics in Southbridge. To add to Cole’s troubles, Dr. O’Brien had stipulated that to get all the detail he wanted on the film plus to get the desired amount of light through the film gate, the film had to be 70mm wide - twice as wide as the conventional 35mm film. 65mm was to be used for image and pull down holes and 5mm for the six magnetic sound tracks.

Problems! Problems! There were no commercially available cameras like this to experiment with. Eventually Cole found that the Thomas Camera Co. had tried to develop a 65mm film camera but it had never been successful. O’Brien Jr., ever the successful ferret, did manage to scrounge up parts of several old Thomas cameras. These were turned over to the Mitchell Camera Co. of Hollywood, Calif., and were revamped under Cole’s supervision.

One of the problems, however, was the clatter involved in the mechanism reaching up and yanking down the 65mm film 30 times a second. Thirty times a second was the speed rather than the conventional 24 times a second to reduce flicker. The mechanism was bad enough but the noisy motor added to the din.

Mitchell eventually solved many of these problems but even so he had to completely surround the camera with an elaborate sound-deadening enclosure appropriately called a “blimp.” This is normally done with movie cameras but was particularly necessary for the more complex Todd-AO film.

A standard Mitchell camera housing cost then about $15,000 but the Todd-AO camera with blimp cost $25,000. Consider that when a camera crew is shooting they always have at least one camera on standby, you can begin to see where the high costs of making movies comes in.

The Projector

Faced with the virtually impossible time schedule and the fact that there were no projectors for 70mm film, Henry Cole and Hobart French, the latter then of AO Buffalo, found some old Ehrneman projectors at Paramount. These were designed, however, for 65mm film. To accommodate the 5mm soundtrack, Cole and French came up with the idea of using a separate normal 35mm projector in addition to the Ehrneman and synchronizing the pair.

John David was involved with the projector as well as with the optical computing. He tells of the problem of trying to install arc carbons 16mm in diameter in cooled silver jaws capable of handling 400 amperes at 60V. Under normal times this is a problem but trying to do it with the frequent power outages caused by the Flood of 1955 in Southbridge made it even worse.

One time during the flood a hose broke loose and 400 amps of electricity was essentially on the loose. Those were trying times.

Eventually contact was made with N.V. Philips of The Netherlands and an agreement reached for them to design and produce a projector to meet the specific Todd-AO design. This projector was used in the world premiere of “Oklahoma!” at the Rivoli Theatre.

Just how close this whole showing came to failure was related by Norman Powers. Powers worked with the movie people in the printing of the film.

The printing was done at Fort Lee, N.J. It was Powers’ responsibility to set up a laboratory where they could compare the film being shot in “Oklahoma!” with that accepted as commercial quality. It was fortunate that such a lab was set up as Powers caught what would have been some way off-color prints.

Powers also told about the convoy (Colin Yates was in it) that left Fort Lee so late that they delivered the last reels of “Oklahoma!” while the first reels were still in the projector!

The final technical problem was that of the screen. The deeply curved screen that was planned for the Todd-AO system would require the projector’s light to be spread out over almost twice the area of an ordinary theatre screen. Thus even with the special 16mm carbons Davis used, Dr. O’Brien felt there would not be sufficient screen brightness and directed that a new screen be designed.

Ed Moon was in general charge of the project and Harry Crandon was the chemical engineer. The screen surface would consist of resin with plasticizer and loaded with aluminum flake.

In theory the screen was to be embossed with a concave spherical imprint whose axis would change as you moved away from center screen. There were several reasons for this.

One was to make the screen uniformly bright and another was to orient the spherical imprints so that light from the right side of the screen would not be reflected onto the left side of the screen with an obvious washout effect.

The embossing of the special angled, spherical reflectors was great - in principle. In practice this was almost impossible to achieve well. The result of the best efforts that could be mustered in the time available were not satisfactory, according to Crandon.

The experimental screens were streaky and there was an unsatisfactory spillover of light. Recourse had to be made to a specially selected commercial screen material suspended from an aluminum frame and tautly secured.

Mike Todd finally had his system - on time as promised by the AO and “it all came out one hole!”