The decoder has a complementary weighting curve to the encoder. The decoder level sensor weighting is identical to the encoder network, the result being an encode-decode response which is flat over the audio band.

One necessary property of this system is the ability to decode monophonic signals summed from four individually companded tracks. Listening tests have demonstrated that a single expander can satisfactorily decode combined multitrack sound with only minor crossmodulation of signal envelopes, the final result being a significant improvement over the limited dynamic range currently reproduced from motion-picture soundtracks.

The Control Track

A narrow track is provided in the middle of the soundtrack group (Fig. 4). This track has four functions:

1. A position reference to provide in­formation to be used to correct film weave. Film-weave errors are primarily present below 100 Hz and can be quite large in the 5 to 30 Hz band. Since each track is unilateral, correction for weave reduces low-frequency noise and avoids wasting amplifier power and loudspeaker voice coil travel with subsonic noise signals.

2. A means of monitoring image spread to allow exposure and developing of soundtracks to be controlled optimally at each step through measurement and control of track width.

3. A command and data track to control ancillary effects in the theater and establish correct playback speaker format for each film automatically.

4. A means to control any other necessary functions such as smells, lights, curtains, theater equipment and to carry identification information.

This control track is typically 0.002 in wide. It is modulated in snake-track fashion, that is, from side to side, with constant line width. Modulation is ±0.002 in in amplitude using signals above 300 Hz. This track acts as an accurate position reference for errors below 100 Hz, even while it is being modulated.

One design requirement of this system is for compatible monophonic playback. The coupling of this control signal into mono playback is therefore of concern. The modulation of the control track is 30 dB under the combined level of the four variable-area tracks. Furthermore, a snake track couples into a light-level sensitive sound system only at the level of the push-pull crossmodulation products. Normal practice holds the single-ended crossmodulation products at about —40 dB. The push-pull products are usually 20 dB lower. Hence, the overall level of control tones present in mono playback is 90 dB below peak program level. This level is sufficiently low to be imperceptible under monophonic theater listening conditions. With the stereo playback system, each track is electrically separated by a position-sensing commutator system, and thus no crosstalk is present from the control track.

For convenience and economy the existing telephone touch-tone codes are used as command signals. A brief review of the touch-tone system is in order, to give a basis for understanding how this system is applied to motion-picture projection use.

The standard touch-tone keyboard has four rows and three columns arranged in a matrix. Each row and each column has a unique tone associated with it. Each button depressed results in two tones, one unique to each row and the other unique to each column. Note that the bottom row has two symbols other than "O," one marked with a star (*) and the other with a number symbol (#). Tones between 697 Hz and 1477 Hz are used to convey the information.

A large number of unique codes may be required for picture and print identification as well as for elaborate effects work, and yet, it may sometimes be necessary to give a series of closely spaced commands. A two-tier system has therefore been selected, a two-digit system for up to 100 rapid or commonly used commands, and a five-digit system for up to 100,000 commands, identification numbers, etc.

Since these commands may be manually generated with rather uneven push-button actuation, an actuate code must be used at the end of each word. This "actuate" command also indicates to the decoding circuit the number of characters to be expected in the word. The * indicates that a two-digit word has been transmitted and the # indicates that a five-digit word has been transmitted.

The frequency band above 1500 Hz is available for additional codes and other features such as automatic cue and cross­over actuation.

Replay Modes and Compatibilities
Read-out Modes of the Colortek Track. Basically, three read-out modes of the Colortek track are possible. They are: (1) read-out by the CCD scanner, as described; (2) read-out by a standard or customary optical sound reproducer; (3) read-out by means of a solar cell.

The video output of the CCD scanner is arranged, as described, to determine the electrical equivalent of the image edge, which is then sliced to ensure its accurate retention. After dc restoration or squaring, the circuitry deals with pulse trains at approximately four times the scanning rate. At this point, a monophonic output is available, being the electrical sum of all channels. (The Veritone low-frequency system mentioned below and an Academy filter are fed from this point, for replaying conventional soundtracks.) Thereafter, for stereophonic reproduction, logic is performed to unscramble the composite signal into four discrete outputs. These outputs go to four expander circuits and balanced line drivers to feed the main fader. At this juncture, one has a direct replacement for conventional four-track magnetic prints. Although the present reproducer is arranged to commutate four discrete tracks, a small alteration to this section of the circuitry will enable it to reproduce n tracks, where n could be 6, 8 or more.

When two additional channels are required, as is the case when it is desired to play 6-track Todd-AO 70mm film reduced to the 35mm format with a Colortek track and using the CCD scanner, this is made possible by the use of a 4-2-4 matrix, such as the Sansui QS System. Under these circumstances, the encoder is introduced before the recorder, and the decoder after the reproducer. This method has the ad­vantage of being able to play the same print as either 1, 4, 5, or 6 channels of sound, thereby satisfying the single inventory requirement. This technique uses the left and right tracks (1 & 3) for the matrix, leaving the center and surround (2 & 4) for discrete use, thereby assuring positional integrity. It has been found to be mandatory to retain the solid discrete center, in order to guarantee satisfactory reproduction for the vastly variable acoustical environments that prevail in the theaters.

Other than using the CCD scanner, it is possible to play Colortek recordings through a standard reproducer with the Academy rolloff incorporated, but this will not give an ideal reproduction. Therefore, an alternative simple overall read-out unit has been developed, which has a 1-mil optical slit and a solar cell as part of the sound head. The electronics contain one pre­amplifier per projector, mounted in the sound head, and one central control box. This consists of a power supply, expander, switching logic for the system selection (i.e., Academy or Colortek), projector selection and line amplifier with balanced transformer output to feed the main fader. By this method, the wide dynamic range and frequency response of Colortek are available as "Hi-Fi Mono."

Reproduction of Academy Tracks. Standard variable-area and variable-density tracks can be easily reproduced by the Colortek CCD scanner. Any pulse train from the center slicing comparator yields an average voltage, which is proportional to the transparent area of the track. This voltage is derived from a scan-rate synchronous low-pass filter, similar to those connected to the multi-channel commutator. Its output is taken through an Academy filter. Variable-density data are taken from the same filter connected to the input of the center slicer comparator. The CCD delivers an overall window envelope, with amplitude proportional to cell illumination. Both formats, variable-area and variable-density, together with the Colortek multi-channel track, can be logic recognized and correctly presented to the overall theater sound system without operator intervention.