Moon Col-R-Tel used its fields in a different way than Earth TV did. Moon video painted a color picture in six fields. But Earth TV painted that color picture in only two fields. For this reason, video coming from the Moon required conversion before we could view it.

Video 101. We need to define a few terms: A frame is one picture in a series that appears as moving video. Earth televison systems built each video frame from two fields. A field was half of a video frame. Each field contained alternate scan lines: Odd lines in one field. Even lines in the other. The scan lines ran left to right, and drew the picture on the screen. Back to the two video fields. Over time, two fields meshed, reducing flicker effects for the viewer.

Moon vs. Earth Col-R-Tel. The Moon color system was Earth Col-R-Tel in reverse. Recall that Earth Col-R-Tel transformed NTSC television signals into field-sequential signals. Video from the Moon was another matter. It was already field-sequential. The Earth downlink station translated this video to NTSC.

Standard TV sets could reproduce the downlink station's output. Downlink adapters made viewable pictures for home sets across the world. Now let's take a deeper look at the video arriving from the Moon.

The mooncam process complemented downlink station processes: The mooncam selected one of three colors per field and transmitted that color to the downlink station. Filters in the color wheel caused the camera to ignore or skip the other two colors. Skipping colors reduced the video bandwidth. Due to color-skipping, each color in a mooncam video only updated once per every three video fields. (Wetmore 20) Note: The output of an Earth Col-R-Tel adapter followed the same, lossy mooncam process.

Making data is tougher than discarding data. Just so, converting Moon Col-R-Tel to NTSC was more difficult than conversion going the other way. While Earth Col-R-Tel worked by losing color data, Moon Col-R-Tel (the downlink) worked by making color data. That is, Earth Col-R-Tel discarded two colors per NTSC video field. At the downlink station, Moon Col-R-Tel borrowed data from adjacent frames, and actually built new frames. The field-sequential to NTSC color conversion equipment was a form of mechanical television.

Time Delays. The downlink conversion process occurred during Moon telecasts, but delayed the output video. (Spacecraft Films, “TV Transmission 33:59 GET”) The Doppler correction alone caused about a 12-second delay. This delay resulted from a tape loop between machines. (Wetmore 20) Additional delays occurred during transmission from downlink stations to synchronous satellites in Earth orbit. Such transmissions occurred between downlink stations and NASA-Houston, and between NASA and other countries. In Europe and Asia, coaxial cable and TV standard conversion caused further delays.

Compensating for Doppler Shift

Doppler shift. Before the downlink station could begin its conversion, it had to correct for Doppler shift in the Moon signal. The shift varied depending on the relative motion of the Earth and the spacecraft. Standard TV sync circuits couldn't reliably lock a Doppler-shifted picture. Without Doppler shift correction, the pictures would tear and flip. The correction process was partly mechanical and partly electronic.

Video Backup. Before sending the Mooncam signals to Houston, downlink stations in California, Spain, and Australia made backup copies of the signals. For the local backups, the downlinks used Ampex VR-660B, video tape recorders. These were early helical VTRs for monochrome signals, not color. Yet because a mooncam didn't send a chroma signal, a monochrome recorder could still capture pictures in color! The VTRs used two-inch tape on reels. (Wood 11, 12)

Two Machines. Houston corrected the signals for Doppler shift. NASA used two, quadruplex, 2-inch videotape machines. One machine recorded. We'll call this machine the recorder. The other machine played back. We'll call this machine the player. (Wood, 7) The two machines operated as an electromechanical time base corrector.

Tape from the recorder ran to the player. Between the recorder and player, the tape was nearly slack. This slackness allowed one machine to operate faster than the other without breaking the tape. There was another complication: Normally when tape tension slackens, a tape sensor shuts down a tape machine. The answer to this problem was a counterweight between the players. The counterweight maintained just enough tape tension. This constant tension kept the tape sensors from stopping either machine. (Wetmore, 20)

Conversion from Field-Sequential to NTSC

Downlink station color conversion. Time for a little review. On the Moon, cameras scanned one color video field at a time. After six color fields, a full color frame resulted. This was field-sequential scanning. On Earth, color TV sets scanned all three colors simultaneously. Every Earth field contained three colors. Two Earth fields made up a frame. The three primary colors arrived at an Earth receiver over a chroma signal. Chroma was separate from the monochrome (luminance) signal. The Earth set combined chroma and luminance and voilà: Color TV! But the mooncam didn't provide chroma. Instead, the mooncam mixed color and monochrome data. And that's why NASA needed a downlink converter. The downlink station had to convert the Moon video to Earth video. The conversion process repeated each Moon color field three times, as part of three Earth fields. (Repeats began after the first two Moon fields.)

Distributing data. Despite the data duplication, the downlink equipment never invented new picture elements. Instead of dreaming up the missing data, the process involved distributing received data. For example, Moon fields 1-2-3 became Earth Field 1. Moon fields 2-3-4 became Earth Field 2. Moon fields 3-4-5 became Earth Field 3. Westinghouse called this field-combination method the moving window. The moving window preserved the natural flow of motion over time.

Sheared frames. The downlink process also sheared a half line from the top of many Moon fields. Other downlink circuitry added normal luminance, chrominance, burst, and sound to the Moon signal.

The downlink picture conversion process was analog. The idea was to record Moon fields serially and play them back in parallel. To store, repeat and combine Moon video fields, the Apollo missions used magnetic disc recorders. Ampex developed this type of recorder for use in sporting events. Starting in 1967, such recorders made possible the instant replays that football fans love. (Abramson, 117-118) Westinghouse modified Ampex HS-100 recorders for use in NASA downlink stations.

The stock HS-100 recorder had two discs, and recorded on both sides of each one. This recorder contained 1,800 tracks (450 tracks per side, times two discs). Each track could record one video field. Alternate fields recorded on alternate disc sides. The HS-100 disc rotated at 3,600 rpm. In one rotation, each side of each disc could record or play back one video field. The next video field would record on the next side. There were four record-play-erase heads, one for each side of the two discs. The four heads could be active at once, with each head operating on a separate track. (Ennes, 485, 488-489)

Probable Modification. Downlink operations required six tracks. (Specifications called for activity on five tracks, with the sixth track remaining unchanged.) The need for six tracks (five active at once) would have been beyond the HS-100's capacity. With four heads, the stock machine could only support four active tracks. Likely Westinghouse satisfied its requirements by adding a third disc and two more heads to the recorder. (Or maybe Westinghouse used two disc recorders.)

Disc operations. The recorder performed a different task on each track. On each successive disc rotation, this task switched by one operation. Playback (read) tasks occurred simultaneously and built the output video field. Each output field included two matching Moon fields and one that didn't match. For example, one even and two odd fields, or two even and one odd Moon fields. Normal field interlacing caused this mismatch. (Source of the mismatch: Standard odd fields began scanning at the top-left corner of the screen. Even fields began scanning a half line away, at the top-middle of the screen.)

The disc recorder corrected the mismatch by introducing a brief delay. A half-line delay before starting the record head sheared off half a line. After the delay, the resulting Moon field matched the other two Moon fields. Leaving the non-matching field as-is would start scanning at the wrong screen location. A delay line in the disc recorder achieved the shearing effect. (Ennes 493, 495) The delay line was a passive, analog device of the quartz (not RCL) type. The delay before scanning was 31.5 microseconds. (Wood, 6)

Video disc operation. The “Loading Moon Fields” tables below describe operation of the video disc. These tables derive from page 5-4, Figure 27 of this book: Apollo Color Television Subsystem Operation and Training Manual (by Westinghouse, or “WEC”). Each new disc rotation began immediately at the end of the previous rotation. (After Disc Rotation 6, the process at Disc Rotation 1 repeated, only with new data.) There was no time for heads to move radially. That is, “between disc rotations,” the six heads remained stationary. This mode of operation differed markedly from typical operation. Ennes (490) describes typical operation: One or more heads would step to a new track while another head recorded.

Error in Record-Playback Sequence. This page corrects an error in the manual's disc color sequence. The book gave the sequence as R-G-B (red-green-blue). A private email from the late Stanley Lebar clearly and emphatically states that the sequence was R-B-G. The basis of this sequence was the color wheel that revolved before the Moon camera lens. This wheel adhered to Peter Goldmark's original CBS specifications. Not surprisingly, the Col-R-Tel scanning disc also followed Goldmark's R-B-G scanning order.

RCA camera. According to the RCA camera manual, the RCA camera also followed the R-B-G sequence. (RCA 1-12) This sequence is why the same downlink equipment and program would work for both the WEC and RCA cameras.

There was no way to specify which color field was the “first one” to come from the Moon. (Those who wish to view the R-G-B order from the WEC manual may use the color-change buttons. These buttons are beneath each table below.)

First usable field. The disc had to contain three stored Moon fields. Otherwise, the disc couldn't play back enough data for one Earth field. The first four disc rotations recorded Moon fields into tracks 6, 1, 2, and 3. But what the disc played back was incomplete, and no usable Earth field resulted. Usable Earth fields didn't emerge from the system until Rotation 5. Each rotation took 16.7 milliseconds, so Rotation 5 wasn't complete until the 83-millisecond point. For that reason, there was a fleeting delay before the system could process the first Earth field.

The R-B-G colors for each track are examples from Westinghouse. (Apollo Color Television Subsystem Operation and Training Manual, Figure 27, 5-4.) For instance, in the first table: “Track 1: Red-Even.”

• A luminance signal (merged RBG fields)

• In-phase-I and quadrature-Q, amplitude-modulated (QAM), suppressed carrier chroma. The I and Q signals defined the hue (phase)and saturation (intensity) for each primary color in the picture. The chroma bandwidth was one-eighth as detailed as the luminance (black-and-white) bandwidth.

• The color burst signal, during horizontal blanking.

Phasor diagram: Chroma signals, NTSC (US, Mexico, Canada, Japan, etc.) and PAL TV (Australia, most of Europe, etc.)

Fringe “Benefit.” Apollo Moon pictures often had color fringes. Such fringes are a flaw of field-sequential scanning. Fringes resulted from unmatched content within neighboring Moon video fields. See photo, right. A mismatch would occur anytime an astronaut moved quickly. The color fringes persisted even after conversion to simultaneous color television (NTSC/PAL/SECAM).

What Viewers Saw. During motion sequences, such as an astronaut waving his arm, viewers noticed colored “trails.” For example, a moving arm would leave a momentary trail of arms: A red, a blue, and a green arm. A moment later, the “arms” would merge again, and appear as a single arm of natural color.

Earth Col-R-Tel TVs could have picked up Moon Col-R-Tel pictures. This reception method would require no NTSC conversion. Consider that downlink stations included Conrac studio monitors that engineers had modified for field-sequential operation. These Col-R-Tel-like monitors could directly receive the Moon, field-sequential telecast. (Wood 11)

Experiment. Someone could prove my statement about Col-R-Tel Moon pictures. The experiment involves transmitting Moon pictures on a standard TV carrier. Then the experimenter would reproduce the unconverted Moon signal. The picture would appear on a Col-R-Tel unit without hue and saturation conversion electronics. Both the Moon and Earth units would maintain their disc-sychronization systems.

Cliff Benham has now performed this experiment, and it works! Cliff's homemade mooncam has reproduced color TV pictures on an original Col-R-Tel set. Since the camera pictures were field sequential, the Col-R-Tel converter worked barefoot! That is, without the adapter electronics. See... Homemade mooncam.

Simpler Conversion. If Stanley Lebar had chosen a two-color system, he could have avoided using the disc-based converter. Two-color, field-sequential pictures transmit at the same frequency as normal NTSC does. That is, 30 frames per second. In terms of speed, this rate is a huge improvement over the three-color, field-sequential system that Apollo actually used. Recall that the three-color pictures transmitted at only 10 frames per second.

Still not live. Addition of NTSC signals such as the monochrome signal, chroma, and burst would still occur at the downlink. Also, Doppler correction would still be a requirement. Doppler correction involves a slight tape delay between two video recorders. This delay would remain. Yet two-color pictures could have been slightly closer to “live.”

Why go with three colors, then? Because by Newtonian theory, a three-color system has a broader gamut than does a two-color system. That is, the three-color system can display pictures with more colors and better color fidelity. By the way, Edwin Land would contest this statement. (Land, Experiments; Retinex) About the same time as the color Apollo flights, the Spectrac^® TV color converter reinforced Land's point. (Benrey 45-47)

Resolution. Another disadvantage of the two-color system is that frames transmit at half resolution. There is no odd and even field for each color. Instead, for each color, only one field transmits. As with Spectrac, probably nobody would notice the difference. (Topping patent 1967)

Which 2 colors? A typical two-color system displays orange and teal fields. A standard CRT makes the orange by turning on its red and green guns. To get teal, the CRT activates its blue and green guns. The result is much like the standard, NTSC chrominance I-signal. RCA engineers invented the I-signal because it is in the optimum color range for accurate flesh tones. The eye is more sensitive to flesh tones than to any other color.

Two-Color Mode. Early color TVs included a two-color mode using the high frequencies of the I-signal. Today, few if any TV sets can decode this mode. (Color Television: Simplified Theory 44-45)

One challenge for a two-color mooncam would be the U.S. flags that Apollo astronauts planted on the moon. Reproducing the white stripes and stars would be easy for such a two-color camera: Adding 100 percent orange and 100 percent teal produces white. But the camera's orange and teal axis is far away from either red or blue. What the camera would record would instead be dark red-orange and dark teal. From a reasonable distance, these colors would be fairly good stand-ins for red and blue. Take a look at our two-color flag (above). What do you think?

For more about two-color television, click TV in two colors.