In the meantime Hoppy was working on the camera optics. We arranged for a teletype line from the A.O. computer room to his home near Rochester, New York. The computer in those days was an IBM card Programmed Calculator (CPC), but it was better than the hand punched Marchant desk calculators that I tried to learn lens design on. Our lens design crew worked from the Southbridge end with the computer and teletyped results to Hoppy who would then make changes and the computer would go through another iteration.

At the center of the screen a circle is reproduced as a circle, but at the edge a circle will be stretched horizontally. Therefore, on the film it is necessary that a circle at the edge of the field is horizontally compressed, and progressively less compressed as it moves to the center of the field. The correct law for this turns out to be that the compression should progress as the tangent of the angle off from the axis. Barrel distortion of a lens can be made to follow this law quite well. Since getting rid of all barrel distortion in a wide angle lens is very difficult anyway so our requirement was quite serendipitous. We now have produced circular faces all over the screen. However, buildings, trees, etc. at the edge of the screen will be bowed inward at top and bottom like a barrel stave (hence the term “barrel” distortion). This we called “Type A” distortion, and I will discuss it later when discussing the distortion correction printers.

Remember that we are projecting from above the audience from the regular projection booth, so that we have two more distortions to consider for people seated below the line of projection. The “Keystone” distortion that is present in the conventional movie theatre is also present here. In addition since we are projection from above onto a deeply curved screen, horizontal lines will appear bent upward at the edges to anyone seated below the axis of the projector. This we called “Type B” distortion or “Droop”.

Now back to camera optics. The “Bugeye” lens was being finished up and it produced the type of image described above. At the same time the number one camera was being finished, Kodak had produced 65mm Eastman Color negative, and first tests were started. After initial static tests Mike took the number one camera to the Far Rockaway roller coaster, the canals of Venice etc to produce a demonstration film to help convince Rodgers & Hammerstein and others of the power of the process. R&H had never before allowed any of their properties to be made into a movie, and they were dubious about Todd-AO. Mike Todd’s “friends” and partners, namely George Skouras, Lee Schubert etc. were not in sympathy with Mike shooting a demonstration film, and moreover if Mike was out of the picture, there would be more gravy for them. So, while Mike was in Europe shooting they refused payment for more film from Kodak for him in an effort to shoot him down. Fortunately Mike was a survivor, and managed to eak out enough film to finish his shooting and this was the demo film shown in Amsterdam [The 70mm Newsletter issue 31]. I did not realize that a copy of it was still in existence. An interesting sidelight related to that film was a press showing of it at Stage 2 at MGM which was rigged as a review room for the process. At this showing the New York Times reviewer got so sea sick during the Venice sequence that he had to leave. We were sure that he would give it a bad review, but on the contrary he raved about the process.

We will next take up the screen problems. If we project an image on to a deeply curved standard white diffusing screen, the scattered light from one edge of the screen will fall on the rest of the screen with the obvious result: light from a bright area on one side of the screen will wash out the dark areas on the other side. Cinerama had this problem, and their solution was to fabricate a screen of vertical strips of regular screen material, each one perpendicular to the incident light, like a Venetian blind. While this solved the spill light problem it was very fragile and expensive to produce. For the few Cinerama theaters it did not matter, but we were aiming for the general release theaters as well.

What we wanted was a screen design that could be hung on a standard type of screen frame (although deeply curved), and also with as much reflective gain as possible to help produce the high brightness image we were after. This indicated an embossed lenticular aluminum type of screen material. The aluminised surface would not depolarise light in case polarized stereoscopic projection was ever wanted (remember, back then the stereo “3D” movies were fashionable.)

What was needed was a material coated with a thermoplastic loaded with aluminium flake that could be embossed into small spherical mirrors. These spheres had to be oriented differently for different parts of the screen to prevent the cross-illumination problem mentioned earlier. The lines from the vertex of each sphere to its center for each lenticule ideally would be parallel and pointed out towards the audience. Obviously with lenticules of the order of 2 millimeters in size a compromise was needed. It turns out that five different orientations on either side of center was perfectly satisfactory in eliminating the cross illumination. The question was how to make the embossing rolls to produce this result?

Swaging individual bumps on a steel roll was, obviously out of the question. We gave the problem to our chief metallurgist, George Granitsas and he came up with a very elegant solution. George had, over the years, developed processes for rolling patterns into the eye wires and temples of gold filled eyeglass frames, so he said “why not roll the spherical bumps on to one face of square wire and then wrap the wire on to a steel embossing drum?”.

We made up two steel embossing rolls, one with groove around it and the other with a raised portion around it that fit in the groove of the first one, but did not reach quite to the bottom. These were about 75mm in diameter. The bottom of the groove on the female roll was embossed with many spherical dimples around the roll, while the top of the raised portion of the male roll was smooth. If you now visualize these two rolls mated there will be a space between the groove and the “land” or raised portion of the other roll. If a ductile metal wire such as phosphor bronze is fed between these rotating rolls it will be swaged into square wire with spherical bumps on one side. This wire is now wound tightly around the full length of a steel drum about five feet long and one foot in diameter, and we have an embossing roll.

Remember now we wanted the little spherical mirrors to be at different angles on different parts of the screen so they would essentially “point” at the audience, and not spread light over other areas of the screen. We figured that panels with five different angles would suffice. If the top of the land on the male roll is machined at an angle to the axis of the roll, the final wire will have a back that is tilted (and hence will no longer be square). By selecting five different angles we make embossing rolls with the bumps oriented at five different angles, and the problem is solved.

My dad gave Ed Moon, one of our plastics and fabrication experts, the job of producing screens. After much experimentation, Ed came up with the proper formulation of vinyl coated cloth, and with a fabrication company to do the embossing, and they began to produce screens. With the flexibility that the above process gave us, Ed could build screens to fit any theatre on a custom basis.