Todd-AO Description of Todd-AO Mark III Printer Principle

Side view of the Mark III printer showing the supply reels at the top and the take-up reels at the bottom. The closed reels are for the raw stock and the open reels are for the negative. The circular pulley in the center is the heavy flywheel on the printing drive to insure smooth motion.

After our discussion of the Mark III Printer on July 11 [1957], I promised to write up a description of the principle which distinguished it from the earlier Mark II model. The incentive for developing the Mark III was primarily to get increased printer speed and hence print output.

The only ways to accomplish this were:

1) Increase light source brightness,
2) Increase lens speed,
3) increase printer slit width.

The light source problem was being considered but tungsten lamps were the only practical solution at the time and were already being used at maximum brightness levels. The printer lens speed was already at the borderline of adequate image sharpness and no substantial increase seemed feasible. This left only the printer slit width. Printer speed being directly proportional to slit width  indicated it would be desirable to find some way to increase this by a factor of two or more.

In the Mark III the slit width is limited by two things: 1) for anything but zero "droop" correction, the image of the negative does not coincide with the print film at all points but rather it intersects the print film at an angle which increases with increasing "droop" and which is equal to twice the bridge angle (i.e. angle of incidence of light on print film). This means that the image is in focus on the print film at only a single line and thus a given point on the negative is in focus only during the instant it crosses this line. On either side of this line it goes out of focus to a degree depending on the distance and on the angle. For a droop corresponding to the requirements of a theater like the Rivoli, the bridge angle is about 45° and the image makes an angle of 90° with the print film. In this case a slit width of about 0.010 inch is all that can be tolerated, before image blur becomes objectionable.

The other factor that limits the slit width is the amount of transverse motion of any given image point (especially near the edge of the picture) due to the change in magnification resulting from the oscillating lens. This change in magnification is necessary to produce the corrections for keystoning in the theatre (K-correction) and for vertical line curvature from the camera lens distortion (A-1 correction). For the case of K-correction alone, the transverse motion at the edge of the film occurs at a rate of almost 1/5 that of the longitudinal motion picture and hence a slit width of 0.010 inch allows a transverse blur of about 0.002 inch multiplied by an additional factor because the effective slit width is increased by the oblique incidence of light onto the film print. The addition of A-1 correction aggravates this transverse blur still further, particularly in the upper corners of the frame.