Category: LOIRP

For information regarding the Lunar Orbiter Image Recovery Project (LOIRP) contact:
Keith Cowing
[email protected]
703-787-6567

‘Digital Dark Age’ may doom some data
10/27/08
Phil Ciciora, News Editor
217-333-2177;[email protected]
CHAMPAIGN, Ill. — What stands a better chance of surviving 50 years from now, a framed photograph or a 10-megabyte digital photo file on your computer’s hard drive?
The framed photograph will inevitably fade and yellow over time, but the digital photo file may be unreadable to future computers – an unintended consequence of our rapidly digitizing world that may ultimately lead to a “digital dark age,” says Jerome P. McDonough, assistant professor in the Graduate School of Library and Information Science at the University of Illinois at Urbana-Champaign.
According to McDonough, the issue of a looming digital dark age originates from the mass of data spawned by our ever-growing information economy – at last count, 369 exabytes worth of data, including electronic records, tax files, e-mail, music and photos, for starters. (An exabyte is 1 quintillion bytes; a quintillion is the number 1 followed by 18 zeroes.)
The concern for archivists and information scientists like McDonough is that, with ever-shifting platforms and file formats, much of the data we produce today could eventually fall into a black hole of inaccessibility.
“If we can’t keep today’s information alive for future generations,” McDonough said, “we will lose a lot of our culture.”
Contrary to popular belief, electronic data has proven to be much more ephemeral than books, journals or pieces of plastic art. After all, when was the last time you opened a WordPerfect file or tried to read an 8-inch floppy disk?
“Even over the course of 10 years, you can have a rapid enough evolution in the ways people store digital information and the programs they use to access it that file formats can fall out of date,” McDonough said.
Magnetic tape, which stores most of the world’s computer backups, can degrade within a decade. According to the National Archives Web site by the mid-1970s, only two machines could read the data from the 1960 U.S. Census: One was in Japan, the other in the Smithsonian Institution. Some of the data collected from NASA’s 1976 Viking landing on Mars is unreadable and lost forever.
From a cultural perspective, McDonough said there’s a “huge amount” of content that’s only being developed or is available in a digital-only format.
“E-mail is a classic example of that,” he said. “It runs both the modern business world and government. If that information is lost, you’ve lost the archive of what has actually happened in the modern world. We’ve seen a couple of examples of this so far.”
McDonough cited the missing White House e-mail archive from the run-up to the Iraq War, a violation of the Presidential Records Act.
“With the current state of the technology, data is vulnerable to both accidental and deliberate erasure,” he said. “What we would like to see is an environment where we can make sure that data does not die due to accidents, malicious intent or even benign neglect.”
McDonough also cited Barack Obama’s political advertising inside the latest editions of the popular videogames “Burnout Paradise” and “NBA Live” as an example of something that ought to be preserved for future generations but could possibly be lost because of the proprietary nature of videogames and videogame platforms.
“It’s not a matter of just preserving the game itself. There are whole parts of popular and political culture that we won’t be able to preserve if we can’t preserve what’s going on inside the gaming world.”
McDonough believes there would also be an economic effect to the loss of data from a digital dark age.
“We would essentially be burning money because we would lose the huge economic investment libraries and archives have made digitizing materials to make them accessible,” he said. “Governments are likewise investing huge sums to make documents available to the public in electronic form.”
To avoid a digital dark age, McDonough says that we need to figure out the best way to keep valuable data alive and accessible by using a multi-prong approach of migrating data to new formats, devising methods of getting old software to work on existing platforms, using open-source file formats and software, and creating data that’s “media-independent.”
“Reliance on open standards is certainly a huge part, but it’s not the only part,” he said. “If we want information to survive, we really need to avoid formats that depend on a particular media type. Commercial DVDs that employ protection schemes make it impossible for libraries to legally transfer the content to new media. When the old media dies, the information dies with it.”
Enthusiasm for switching from proprietary software such as Microsoft’s Office suite to open-source software such as OpenOffice has only recently begun to gather momentum outside of information technology circles.
“Software companies have seen the benefits of locking people into a platform and have been very resistant to change,” McDonough said. “Now we are actually starting to see some market mandates in the open direction.”
McDonough cites Brazil, the Netherlands and Norway as examples of countries that have mandated the use of non-proprietary file formats for government business.
“There has been quite a movement, particularly among governments, to say: ‘We’re not going to buy software that uses proprietary file formats exclusively. You’re going to have to provide an open format so we can escape from the platform,’ ” he said. “With that market demand, you really did see some more pressure on vendors to move to something open.”
Editor’s note: To contact Jerome McDonough, call 217-244-5916; e-mail [email protected]

Destination Moon: A History of the Lunar Orbiter Program, NASA TM X-3487
Of all the pictures which Lunar Orbiter I made, one of the most spectacular was the first photograph of the Earth taken from the vicinity of the Moon. This picture was not included in the original mission plan, and it required that the spacecraft’s attitude in relation to the lunar surface be changed so that the camera’s lenses were pointing away from the Moon. Such maneuvering meant a calculated risk and, coming early in the flight, the unplanned photograph of Earth raised some doubts among Boeing management about the safety of the spacecraft.
Robert J. Helberg, Boeing’s Program Manager for Lunar Orbiter, opposed such a hazardous unnecessary risk. The spacecraft would be pointed away from the Moon so that [242] the camera’s lenses could catch a quick view of Earth tangential to the lunar surface. Then, once the pictures were made (flight controllers would execute two photo sequences on two different orbits), Lunar Orbiter I would disappear behind the Moon where it would not be in communication with ground control. If, for some reason ground control failed to reestablish communications with it, the Apollo-oriented mission photography would probably remain undone, Moreover, Boeing had an incentive riding on the performance of the spacecraft, and Heiberg did not think it prudent to commit the spacecraft to a series of maneuvers for which no plans had been made. 30
The understandably conservative Boeing stance was changed through a series of meetings between top NASA program officials, including Dr. Floyd L. Thompson, Clifford H. Nelson, and Lee R. Scherer. They convinced Heiberg that the picture was worth the risk and that NASA would make compensation in the event of an unexpected mishap with the spacecraft. After agreement had been reached, Lunar Orbiter flight controllers executed the necessary maneuvers to point the spacecraft’s camera away from the lunar surface, and on two different orbits (16 and 26) it recorded two unprecedented, very useful photographs.
[243] The Earth-Moon pictures proved valuable for their oblique perspective of the lunar surface. Until these two photographs, all pictures had been taken along axes perpendicular or nearly perpendicular to the Moon’s surface. On subsequent Lunar Orbiter missions oblique photography was planned and used more often. 31
30. Taback interview. See also Transcript of Proceedings–Discussion between Nicks, et al., and members of National Academy of Public Administration, pp. 111-112.
31. For a detailed technical description of the Earth-Moon photographs refer to Lunar Orbiter I Photography, NASA CR-847, prepared by Boeing Company, Seattle, Washington, for the Langley Research Center, August 1967, pp. 64-71.

---

The Orbital Period for LO-1 was 3 hrs 26 minutes and 21 seconds – that’s 34.3 hours between orbit 16 and 26 …. you could have had orbit 16 on 23 Aug and orbit 26 on 25 Aug …

---

1. If you listen to this audio [ http://images.spaceref.com/news/2008/loirptest5.mp3] from a data tape you will hear the technician state that it is an analog copy of an earlier tape – but that it is only a copy of the portion with the Earth and moon on it and that it was originally recorded on GMT date 237.
2. In 1966 day 237 was 25 August. Lunar Orbiter 1 was launched on 10 August 1966 and imaged the moon from 18-29 August 1966.
So, NASA NSSDC says that this image was taken on 23 August yet the tech’s voice says day 237 – this must be the date when the tape was being played back from the spacecraft to Earth, yes

Keith Cowing: Here is the document that shows that there is a SECOND demodulator system and even outlines how the data is written to the tape drives.
The Lunar Orbiter Telecommunications System, NASA LaRC, 1965

The first image you see here is the explanation from the original documentation showing what a single line of lunar orbiter video looks like.  The second, is the machine fully locked up and reproducing the correct structure.  The image is inverted as it gets flipped with a jumper select on the board that we just have another board that has the jumper but we weren’t using it.

If you look at the oscilloscope, you can see the fiducial marks in their proper position before and after the sync pulse.  If you look at the lower right part of the screen, you will see the notation (200us).  That is the amount of time per division.  If you count the divisions to the next sync pulse, you will see that it exactly matches the number for a single scan line below.  The voltage scale is correct as well except that the drawing below is wrong.  The signal is 1 volt, not five volts.  We have confirmation that 1 volt is the correct number from the audio tapes where the engineers recorded the voltage levels on the tape.  It is these kinds of discrepancies that we have had to research and overcome.  We have reached that magic milestone.  We now have to put those lines together into an image and we are working that now.  However, the two critical questions that were in everyones mind when we started this have been answered.  Are the tapes good?  Yes.  Can the drive be brought back to operational status? Yes.  
We are at a crossroads here and I will put together a formal report and seek input on where to go from here.  As a result of our work we have found a few things that were not in the documentation that effects which way we go forward from here.  Our original plan for digitizing the images is no longer tenable and we are doing a near term work around for the image milestone.

Keith Cowing: We have a milestone to report tonight. We put a real LO-II tape on the drive with the known good head and was able to get audio and the test video patterns off the machine. We did not dare go far enough yet to get the LO image but here for your listening pleasure and verification of milestone, is a voice that has not been heard since November 30, 1966. Sorry but if you have a PC you may not be able to hear it but on the Mac it is great! [Audio file] Also a pic to go along with it.
We are closing in on the prize.
[Audio files of voice track: PC (.wav) Mac
“30 November 1966”
Here is a MP3 version (thanks Ken!)

A lot of mundane yet important milestones.
Dennis Wingo: We have now confirmed that the timing system on the drive for the servos and just about everything else, is back into original specifications. We found in one of the manuals that we have, the procedures for aligning the various systems of the drive. Though the other three FR-901/902’s are a little different, it is close enough that we now have verified this most critical system. We tested and confirmed that the oven controlled temperature stable clock reference (two out of three that we have tried) is within 6 x 10-8 Hertz, or better than one part in one million. This was a critical thing to find as if the system clock standards had been out of tolerance (after tweaking and 24 hour test), we would have had to have found an alternate means of providing a stable signal to the machine(s).
Fortunately as we were concerned about the possibility, we asked the ever intrepid Mark Newfield if he knew were a ultrastable 1 Mhz oscillator might be found. After a search with Ken Zin, one was located, from about the same era as the drives! We brought it over and it is hooked up and up to temperature and operating fine as a backup to the in machine standard.
Ken went through all of the rest of the timing system, and after replacing a bunch of transistors, the entire timing chain is within specification. This has had a beneficial effect on the servo system and even with the known bad bearings the servos that control the reel motors are within specification for testing at least. We still have out the reel motors and the capstan motor for bearing refurbishing. We have found more documentation in the manuals with details about the mechanical specifications of the servo system and its motors. We are anxiously awaiting the aperture cards that we are getting copied to see what other documentation that we now have from the former Ampex head of field engineering.
We have been digging into the software issues associated with the data acquisition card. At this time I can get it to digitize at 1.8 million samples per second, not good but better than before. It is going to take a lot of digging into and creating our custom software for the data acquisition and storage task. I am still looking at an alternative that I may pull the trigger on this week.
We have reorganized the work space and created an area for the computer that will be connected to the drive to reside, close enough for the hook up. With all the work that we have done and with this much verified, and knowing that the likelyhood of a tape being damaged was nil, we put a Lunar Orbiter tape on the drive Friday. There is a very good reason for this even before the system is fully back within operating specifications. The reason is that the servo system needs a tape with a 500 kHz pilot tone on it to come fully into lock. To make a new one means that the record and playback system has to be fully back into specification, which it is not as of yet.
So we took a LO tape and put it on the drive to examine the test data at the front of the LO tapes. With this we can get the best idea of how far we have to go (a ways still) before the drive is 100% back into its original specification. We will get back our rollers for the tapes this week. We can mount them and get the mechanics of the system fully back within original specification. The Capstan and Reel motors are still being worked on.
When we put the tape on, we did get video but the head that we were using is not the best one. There are still some problems in the time base corrector (for the tape, not the servo system) and other electronics but we have actually read multiburst off the oscilliscope (this does not mean that the demodulator is the correct one) from the tape. Pictures to follow and the video will be up on the facebook page in the morning.
We are also sad this week to lose our two students from San Jose state (Kenneth Willians and Austin Epps) this week as the fall term is restarting. They will be here part time (20 hours a week) to help out but their dedication and hard work this summer has really helped us with doing all the things we have had to do to make the progress that we have. Good job gentlemen! I am getting both Austin and Kenneth to talk to their professors about getting elective credit for working on the project. We did over 300 hours of this on my SEDSAT project at UAH for undergrads and would like to have the students do this here as well. I would posit at least one graduate level degree could come from this project as well, so we look forward to getting images soon!
We have also been monitoring the progress to get the $50k from Doug Comstock obligated and to get the head refurbishment going. We have found out that VMI has just had a big order come in so we might be a bit slow in getting our head back. Will know more this week

Dennis Wingo: A bit delayed but we had a lot happen over the past few days, not all of it good.
We have had a bit of a setback. We have received our interim computer from Apple and it is almost the same as the final one, which alerted us to the problem, which is that the I/O bus is incompatible with the National Instruments data acquisition card that we have purchased, even though their website said it was supported. I should have looked closer and verified this. We can get a replacement card that will work but since the hardware is being shipped from Hungary, it would cause a 5-10 business day delay. This is unacceptable at this point so Austin Epps, one of our engineering students found a work around with a PCI-Express to PCI adapter (the new bus to the old bus architecture) so we have purchased this and I am putting him on a plane to San Diego to pick it up and bring it back. This is cheaper than it seems as the price for expediting this is $150 dollars shipping and handling and having him fly down there is only twice that and saves two days. So, this evening we will be able to put the new computer and the data acquisition card together. Kenneth Williams is loading the software on the computer today.
On the good news front, $50k of additional funding has been provided through Doug Comstock from the NASA IPP. This money will be used to pay for the tooling to refurbish our heads (which are somewhat different than the standard commercial heads) and to get a single head completely refurbished. To save money we are doing all of the pre-refurbish work ourselves here as the first thing that the company that is doing the refurb did when contacted, was to call Kenneth Zin, our technical lead for help!
However, this $50k brings its challenges. The head refurbishing company demands a check for 2/3rds of the cost of the tooling to begin work. Due to the way that NASA procurement works, this means that we would have a 30 day delay before being able to move forward with the head refurb. This is unacceptable to me so I made a command decision to cancel the purchase of the high end MacIntosh computer and we will keep the existing interim computer and do the initial image processing with it.
This is an easy decision to make, which saves us $10k of expenditure that we can use to pay the head refurb company right away and get reimbursed when the NASA money comes in for this task. Our image processing may happen slower but since our goal at this time is only a few images, this is not a big sacrifice. It will be ok for our other image processing as well for now. We can order the new computer after we are successful in getting our images and moving into production (should funding become available). The interim computer can then become the secondary computer to run the second drive. The head refurb is by far the highest priority
We have made a lot of progress on the machine as well this past week.
We have started the process of refurbishing the mechanical systems of the primary drive. Since the mechanics are exactly the same for all four machines, we disassembled the parts donor drives for their tape guides (pictures enclosed), capstan motor, and the reel motors. All of the bearings are 40 years old and have to be replaced. We can’t do this ourselves so we took them to a house in San Francisco that does this type of work as they have the tooling to do so. We took some of the simpler parts, along with some bearings that Kenneth Zin already had, to a place in South San Jose to be replaced and we will have those back this week and placed on the machine. Since there is setup time and charges to be considered here we have taken the step of taking two sets of all of the mechanical hardware to be done at once. One will be spares and if we are funded to go forward we will just put the spare hardware on the second machine. It is critical to get the mechanics properly aligned so that the tape transport system is back to the original specifications.
We have also shipped out additional rubber parts to redco in Nevada for refurbishing/replacement. One of these is easy, as it is the rubber pads on the tape reel assemblies that the tapes rest against. The second is not so easy, as we have to get the rubber interior surface of a seamless nylon set of belts that run the motors refurbished. We sent the belts off of the parts donor machines to get them refurbed first and then when we put the new stuff on, we will send the other belts to be refurbed. This is a major concern but doable as these belts are also critical to the proper operation of the entire servo mechanical system.
We have also had success the past week (after Fedex ground’s inability to find our building for four days) in obtaining a lot of schematics and technical data for the FR-900 drives. This came from a labyrinth of contacts that led us finally to the retired head of Ampex’s field engineering force! He happens to have most of the information that we need on the drives in terms of schematics, parts lists, and other details. We are still missing some procedures but we are much closer to having everything that we need and worst case we can get the drive completely back to specifications with what we have. This is truly an adventure in technoarcheology in finding this data. One wonders how different our civilization would be today if we had all of the documents related to the masterful feats of Greek mechanical engineering and Roman civil engineering today. Just think of the problems that NASA is having in recreating a 40 year old technology to return to the Moon!
We have also been testing in detail the 16 heads that we have. It looks like that we have at least one additional head that can be used as a test head for our alignment procedures of the drive. We have tested and proven the known good head but do not want to use it until we are ready to put a real Lunar Orbiter tape on the drive. Kenneth Williams and Ken Zin have built test fixtures that allow the heads to be tested and the 43 year old relays on the head module to be “cleaned” by exercising them with a signal input.
We have also visited the NASA Ames surplus and have found some useful test equipment, including a rig that allows us to splice tape. We found some 19″ racks for test equipment as well and some oscilliscopes. We also picked up an older MacIntosh to use as a server for the printer and our local kludged internet connection so it has been a good week on the scrounging front. Another device that we found is a reader for microfilm. We will use it to compare the images that I have on my microfilm of the Moon to guarantee a match for the analog data vs the film.
Andrew Gold, the CEO of one of the companies in the research park is going to set us up with a fat pipe internet connection gratis this week so this solves our issue of having to pay a NASA contractor $6k for what we were able to do for an $80 dollar piece of hardware.
Conclusion
So basically we are moving forward and working through our challenges! Lots and lots of work in a detailed manner that is necessary to bring the drives back to full operational status and get them back into their original specifications.
Pictures follow.

After being away for a week and coming back I am very happy to report our current progress. The guys have done a marvelous job in turning the McDonald’s into a working laboratory. The fake planter in the middle of the room has now been turned into a nice working table for the testing of the hardware on the drives. They have also put together the wiring for power in the area in a manner that is safe and efficient for working. Pictures follow of this.
On the technical side, a lot of progress was made there as well. We have the second drive brought to functional status. Since we don’t have the money to replace all the parts we are not going much further but there are things that we can do with the drive to test parts and subsystems while the primary drive is being optimized for full operation. I am including an abbreviated video here but a longer one will be on the Facebook page of both drives operating at once.
Kenneth Williams has been testing all of our heads and it may be that we have more than the one good head. There are electronics that have degraded that may need replacing but there is no reason that we can’t get more than one head operational. We have verified that the known good head is in very good shape so we are confident of it when we actually put a real tape on the drive.
We have ordered and the parts are on the way for the Data Acquisition system as well as our image analysis software. We are doing an interim solution for our computer by renting one from the ODIN system here at Ames in order to maintain our schedule and our progress rate.
We are still on track to have a real Lunar Orbiter tape on the drive by the end of the month and we will continue to update you all on our progress. I cannot say enough good about the competence of Ken Zin and the enthusiastic participation and hard work of our students. We are going to be talking to their faculty advisors about them obtaining engineering credit for their work as it has been a very good lesson in overall engineering design and real world experience.