The lost NASA tapes: Restoring lunar images after 40 years in the vault
Liquid nitrogen, vegetable steamers, Macintosh workstations and old, refrigerator-size tape drives. These are just some of the tools a new breed of Space Age archeologists is using to sift through the digital debris from the early days of NASA, mining the information in ways unimaginable when it was first gathered four decades ago.
At stake is data that could show Earth's risk of an asteroid strike, shed light on global warming and -- perhaps -- even satisfy those who think the moon landings were a hoax.
The most visible of the archeologists is arguably Dennis Wingo, head of Skycorp Inc., a small aerospace engineering firm in Huntsville, Ala. He's the driving force behind the Lunar Orbiter Image Recovery Project, operating out of a decommissioned McDonald's (since dubbed McMoon's) at NASA's Ames Research Center in Mountain View, Calif. The project's goal is to recover and enhance as many of the original lunar landing images as possible.
Between 1966 and 1967, five unmanned probes were sent into lunar orbit to map possible landing sites within the moon's equatorial regions at one-meter resolution and to map the rest of the surface at a resolution of 40 meters or better, Wingo explains. Those probes, known as Lunar Orbiters, sent back about 1,800 images that modern technology should be able to greatly improve.
The project's great scientific value to NASA is in enabling a comparison between the lunar surface as mapped by the Lunar Reconnaissance Orbiter, launched on June 18, with the lunar surface as it appeared 43 years ago, according to Wingo. The goal is to "get a fix on how many meteor impacts have occurred in the meantime," by cataloging the new craters.
"If we know the changes, we can establish the risk of working on the moon and even determine the small-body asteroid population of the inner solar system," Wingo says. Another valuable contribution: the ability to plot the possible risk to Earth of the impact of an asteroid.
The original black-and-white images were shot on 70mm film that was automatically developed and scanned within the robot spacecraft. The signal from the scanner was sent to Earth and was then displayed as partial frames on a monitor. Each monitor image was then captured with a film camera. These pictures were fit together, and then another picture was taken of the finished mosaic. Each step imposed a certain amount of image degradation.
The resulting Lunar Orbiter images are the basis of a digital lunar atlas. But Wingo figured that if he could process the tapes of the original signals, he could improve the dynamic range of the images by a factor of four, revealing far more surface features.
Although this theory has proved correct, the path has been challenging. Wingo first had to acquire the tapes, then reconstruct drives to read them and finally perform the actual processing.
It turns out that the original 2-in. tapes were available. Around 1986, NASA archivist Nancy Evans, who is now retired, was contacted by a federal records center asking what to do with them. Feeling that the data should not be discarded, she persuaded the Jet Propulsion Laboratory (JPL) in Pasadena, Calif., to put them into climate-controlled storage.
However, the tapes were useless without compatible tape drives -- in this case, analog Ampex FR-900 reel-to-reel units. Weighing half a ton and resembling refrigerators, the drives were formerly used by the U.S. Air Force to record radar data but have not been manufactured since 1975. "There were probably thousands of them at one time, but as the radar stations refitted with new drives, most [of the old ones] were dumped in the ocean to make coral reefs," Evans says. There are "thousands" of the old drives off Kwajalein -- an atoll that's part of the Marshall Islands -- and Florida, she says.
She finally got a call from an Air Force base that had four of the old drives. She stored them, along with documentation and spare parts, at her home in Sun Valley, Calif., and tried to get funding to restore the tapes. None was forthcoming, so the machines gathered dust for two decades.
By 2006, the tapes -- still in JPL storage -- fell under a new NASA edict that no planetary data should ever be destroyed, Evans explains. However, by then she needed the storage area occupied by the tape drives for the veterinarian practice she and her daughter maintained. In an effort to preserve the drives, she submitted a white paper about the tapes and drives at a Lunar and Planetary Institute conference. After seeing the white paper in a blog post, Wingo contacted her and arranged to have the drives, and later the tapes, transported to Ames in rented trucks.
Then Wingo obtained a grant of $250,000 from NASA to get started. His largely volunteer crew was able to restore two of the drives using pieces from the other two, plus off-the-shelf parts and additional components that had to be custom-made.
"We had to pay big bucks to get the bearings replaced, the motors rebuilt and rubber parts cast. We had to dip the motors in liquid nitrogen to get the bearings off," he recalls.
So far, all the tapes have proved usable. The data is read into a quad-processor Macintosh Pro workstation with 13GB of RAM and 4TB of storage. Data acquisition is done through a PCI Express card from Canadian firm AlazarTech that can read 180 million samples per second, although only 10 million are needed, Wingo says.
After capture, the images are processed with Adobe Photoshop and Igor Pro analysis software from WaveMetrics Inc. But the new plan is to move to a custom application written in C, largely because of its ability to take advantage of Mac OS X 10.6 (Snow Leopard). With Igor Pro and Photoshop, processing takes an hour for a high-resolution image and 20 minutes for a medium-resolution image. But after the switch to the C program, processing with the Snow Leopard version should be almost immediate, based on the testing that's been conducted, Wingo says.
With an additional $600,000 budget, Wingo hopes to have all the files processed by February, producing a moon atlas with a resolution higher than anything previously seen. Most of this new funding is again from NASA, with about 10% from private donors.
However, Wingo's "deliverable" to NASA is not the images themselves, but the raw data extracted from the tapes. "They would rather have the raw data so that someone even a thousand years from now could do their own processing," he says.
The lost Apollo 11 tapes
The NASA edict against data destruction was issued after the space agency's 2006 admission that it couldn't locate the original tapes of the Apollo 11 live slow-scan TV broadcast from the moon. The agency then initiated a search for the tapes, which remains ongoing, as is the Internet furor the admission generated among conspiracy theorists, who believe the landings were staged.
The data is assumed to be on 1-in. tapes, but, based on period photos, Wingo thinks they should be on 2-in. tapes like the Lunar Orbiter data. He is conducting his own search.
Begging to differ is Richard Nafzger, senior engineer at the Goddard Space Flight Center in Greenbelt, Md., who's been working for NASA since 1968 and was involved in television support and voice communications for the Apollo moon missions.
"Despite how old you get, there are certain things you don't forget, and we recorded all slow-scan images on 1-in. tapes that were 15 in. in diameter, and I have spent the last three years tracking them," he says. "I am certain that there was no slow-scan ever recorded on the Ampex 900." The video feed was one of 12 tracks of telemetry that were recorded on each tape, Nafzger explains.
Due to the low wattage of the transmitter on the lunar lander, they had only 500 kHz bandwidth to use for video, as opposed to the 4.5 MHz that was standard at the time for broadcast analog TV. So NASA used a slow-scan, black-and-white transmission at 10 frames per second with 320 lines per screen, Nafzger says. U.S. broadcast TV used 30 frames per second with 525 lines per screen. The conversion was made at each ground site with a device that basically pointed a broadcast TV camera at a special monitor displaying the slow-scan image.
The slow-scan monitor had persistent phosphor to make up for the slower scan rate, and as a result the movement of the astronauts looked ghostly and jerky, he explains. (Later moon landings used a more conventional TV broadcast system.)
The Apollo 11 TV signal was captured at NASA ground stations with 85-foot antennas in Spain, Australia and the Mojave Desert. NASA also borrowed a 210-foot radio astronomy antenna in Australia for the occasion. The signals were converted to broadcast format on-site and sent to Houston for redistribution to the TV networks. Both the slow-scan feed and the broadcast format were recorded on-site in case the live broadcast failed. The converted signals were routed through a single point in Houston so that NASA could cut off the signal if there were an "incident," Nafzger explains.
But that was the least of his worries.
"The night we landed and did the moonwalk, that is when I became scared," he recalls. Before that point, there hadn't been as much pressure to broadcast the proceedings in real time. But after the safe landing, "they were saying that they had better be able to see this on TV, and 600 million people were watching. Something as simple as plugging a wrong patch or pushing a wrong button would mean that no one would see it," Nafzger says.
Indeed, the camera had been installed on the lander upside down, Nafzger recalls. The TV technicians heard of this at the last minute and scrambled to install converters at the ground stations. The first few seconds of broadcast were upside down because the operator at the Mojave Desert ground station who understood the converter had left for the day, Nafzger recalls.
If the original tapes could be found, he estimates that they would appear three times clearer than the broadcast images. "Taking the clean data and extracting it in a digital high-definition format would let you go frame-by-frame and remove the noise, smearing, contrast problems and other things that were man-made, mostly by the original conversion. The tapes are worth getting just for that reason -- absolutely," Nafzger says.
He and others have been trying to do just that. But NASA has had at least 220,000 tapes of that variety in storage at some time, of which only about 15 might be the lost Apollo 11 tapes, he notes.
"We have gone through landfills on the tops of mountains. I have looked through rooms the size of two or three football fields, filled with rows of shelves going up 30 feet, and we have looked on every shelf that might contain the right tapes," Nafzger says. Tapes that were suspected of being the right ones were heated for hours in dry vegetable steamers to make sure the oxide was fixed to the substrate before Nafzger's team attempted to read them. Goddard has preserved the necessary 1-in. tape drives, so Nafzger did not have the refurbishing task that Wingo faced.
Nafzger is currently preparing a report on the results of the search and cannot discuss them until NASA releases the report, the date of which is uncertain. "But since I am not running down the street waving a flag and shouting 'Eureka!' you can draw your own conclusions. The big picture is that there is an explanation for everything," he says.
Meanwhile, in Las Vegas, Karen Person, head of the Renaissance Entertainment & Media Group, is not waiting for Nafzger's results. She says she has acquired one of the original 2-in. NASA recordings of the broadcast video and is using it as the basis of a documentary titled July Moon, which she hopes to have in theaters for the 40th anniversary of the moon landing on July 20. The video has been transferred to MPEG-4 format and parts have been enhanced, she says.
"They are about 200% clearer than anything you would have seen, and Walter Cronkite is not talking over them," she says. In fact, she showed clips to Apollo 11 astronaut Buzz Aldrin, and, according to her, he said he saw things that he had not previously remembered.
She claims she procured the tapes -- for an amount she would not disclose -- from a man who bought them at a government surplus property auction in 1976 while he was a NASA engineering intern. He reportedly paid $217.77 for a batch of 1,150 assorted tapes.
For his part, Wingo has received a grant from the National Oceanic and Atmospheric Administration to locate early Nimbus weather satellite tapes. Data from the satellites, first launched in 1964, was stored on tapes like those used with the Lunar Orbiters.
"Those images would push our knowledge of Arctic and Antarctic ice packs 14 years further into the past," he says.
Lamont Wood is a freelance writer in San Antonio. He can be reached at email@example.com .