Magnetic Tape Drives Are Back In Style

Magnetic tape to the rescue, Economist
“The need for mass storage is reviving a technology which, only a few years ago, seemed destined for the scrapheap: magnetic tape. Tape is the oldest computer storage medium still in use. It was first put to work on a UNIVAC computer in 1951. But although tape sales have been falling since 2008 and dropped by 14% in 2012, according to the Santa Clara Consulting Group, tape’s decline has now gone into reverse: sales grew by 1% in the last quarter of 2012 and a 3% rise is expected this year. Alberto Pace, head of data and storage at CERN, says that tape has four advantages over hard disks for the long-term preservation of data. The first is speed. Although it takes about 40 seconds for an archive robot to select the right tape and put it in a reader, once it has loaded, extracting data from that tape is about four times as fast as reading from a hard disk.”

Reassembling a Third FR-900 Tape Drive


Dennis Wingo: This is Jacob Gold’s last days with us and one of the projects that we have wanted to get done is to reassemble the third FR-900 tape drive. We are not going to try and get it working as it is a record only machine and we don’t really need to record anything! However, we can use it to possibly raise money. We want to get another head and if we can get $10k of it that would be great. NOTE. This thing weighs 900 lbs when reassembled and thus we are NOT going to ship it out of the Bay area. We MIGHT ship it to LA as will fit in the back of my truck. Just a thought. And here are a couple of pictures of Jacob working on the machine today.

Priority vs Final Readout of Lunar Orbiter Images


Dennis WIngo: This image (click on image to enlarge) shows the sequence of images that were read out during what is termed “priority” readout vs the “final readout”. The priority readout was an opportunistic scanning of processed photos on the lunar orbiter before all of the images were taken. The photo process with the 70mm film began with an image being simultaneously taken by the 610 mm high resolution camera and by the 80 mm medium resolution camera. In a process remarkably similar to the old polaroid dry process instamatic cameras, the film was dry processed by a “bimat” dry processor. The bimat would separate from the film (most of the time) but would sometimes due to the timing would leave artifacts on the image, which are readily identified on the film.
The film would then be fed into the readout looper where it could be scanned and the images sent back to the Earth. During the mission when photographs were still being taken the film would run one direction through the looper. After all of the images were taken a command would be sent to cut the bimat and then the film could be read in the opposite direction.
Thus when we start with a low numbered tape, the first images that come off are from the priority readout in ascending order. However, the ascending order is not linear, jumping because images are still being taken and the film advancing while the spacecraft cannot transmit. The image of the day today shows what images are taken in ascending sequence (orbit) format. You can see the jumps. If you are really good you can see what images we are recording and then predict what the next images will be … At least during the priority readout.

LOIRP Status Report 22 July 2013


Dennis Wingo: We are running tape today. Here is a scan of a document that I am using to plan the sequence of tape runs. at the top of the page it shows the proportion of the time one ground station was in range of Lunar Orbiter V. Woomera had the majority of the time so we are starting with it. Then Goldstone, and then finally Madrid. There was some duplicate scans during the final readout that overlapped the priority readout data. Thus it is our hope that we will get most of our scans from Woomera and Goldstone and will have to run only a few Madrid tapes. Time will tell. Click on image to enlarge.

Video: Oscilloscope Trace of Lunar Orbiter Image Playback


Dennis Wingo: This is a two minute video that shows, about every 42 seconds, a shift in the video. This occurs when the line is scanned to the end of a framelet, over the calibration features that were pre recorded in the tape. Keith has posted images here that show the relationship between the scope trace and the dynamic range of the tapes. This gives us a qualitative measure of the performance of the tape drive. Even in its sub optimized state today, the scope trace almost exactly matches the original transfer function of the spacecraft film, thus giving us a means to evaluate the quality of the analog data derived from the tape.

LOIRP Status Report 20 June 2013

Dennis Wingo: WE ARE DONE WITH LUNAR ORBITER II PRIMARY CAPTURES!. All we have to do is some clean up of a few gakked framelets.
Here are the final tapes and images captured.
M2-089, partial capture LOII-2049H
M2-090, partial capture LOII-2047M, complete capture 2048H
M2-091, partial capture LOII-2046M, complete capture 2047H
M2-092, partial capture LOII-2045M, partial capture 2046H (completes image)
M2-093, Not Used, redundant with Woomera Tape
M2-094, partial capture LOII-2045H (Completes image)
M2-095, partial capture LOII-2043M, complete capture 2044H, 2042M
M2-096, complete capture LOII-2043H
M2-097, partial capture LOII-2041M, partial capture 2042H (completes image)
It will take several days to get everything into the back end queue and figure out what our final capture rate is, but it looks like we only had five bad tapes out of over 300 LOII tapes for about a 99% capture percentage, which is astounding considering how long it had been since these had been originally recorded.
More next week when we start on LO-V!

LOIRP Status Report 19 June 2013

Dennis Wingo: Mid afternoon 6-19-13. We almost had a catastrophe yesterday. These tapes sometimes shed large chunks of oxide and this happened on tape M-91, one of the Scotch tapes. We had been observing a lot of shedding of material and that had clogged up the vacuum system, which Ken cleaned out. However, there was a large chunk wedged between the control track head (which is a separate head used to read the servo information) and the tape guide. This disrupted the geometry of the tape and contributed to getting a short length of tape sliced in two. It also created a terrible head clog (which kills all the signal from the tape from a single head of the four).
Ken disassembled the head, found the foreign matter and removed it. He also had to re-align the control track head and use a strong solvent to remove the foreign matter from one of the head tips. We are back up and running as of this afternoon as I had to go to the DMV today and in California that takes forever.
We did two tapes yesterday before the problems and I am on one today right now. It is late in the day but we should be able to finish our Madrid Scotch tape captures before I head off on a short vacation to go backpacking in the high Sierras.

LOIRP Status Report 18 June 2013

Dennis WIngo: Status early afternoon: The FR-900 tape driveis down for the want of a 2″ long 11/64th ID piece of tubing. This is inside of the head and routes the vacuum from the vacuum pump that pulls the tape up against the female guide on the tape machine. This vacuum keeps the tape in a proper shape for the disk shaped head assembly that is spinning at several thousand RPM to not cut the tape to ribbons. This small part failed after two tapes this morning. Ken Zin is going to Grainger and can hopefully get the piece of hose without having to buy a whole 50′ roll! Seven more tapes to go.

LOIRP Status Report 17 June 2013

Dennis Wingo: As those who have followed these statuses know I am now running the Scotch tapes from the Lunar Orbiter II archive. Things are going ok so far, here is our status as of the end of the day today. There have been significant overlaps with both Madrid and Goldstone that have cut down the amount of tape we have had to run by at least 40%. We have had to run each tape, but for not as long as we would have had to otherwise. This is good.

LOIRP Status Report 12 June 2013

Dennis Wingo: We ran two more Goldstone tapes today that we had captured in 2010 but for some reason did not have today. These were G2-093 and G2-095.
We also figured out that some of our images that we did not have in our print out that we actually do have so we have more images complete that we thought. I am going through in detail now and comparing times and dates to see what overlaps there are between Madrid tapes and Woomera and Goldstone so that we can minimize the number of these Scotch tapes we have to run. I will have that done in the morning…
So now I can confidently declare that we have captured all of our Goldstone and Woomera tapes from LOII and just have the 18 Madrid tapes to go or some subset of them.

Our Tape Drive Now Displays The Stardate


Dennis Wingo: This is way cool for us. Notice that in the upper left there is a LED readout. That readout is accurately reading the proper date in 1966 for this particular tape. After several years of messing around trying to recreate a NASA time code reader we found one for $50 on Ebay. Jacob Gold and Ken Zin got it to work.


Keith Cowing: We used to tease people and refer to our tape drives as the “Stargate”, the “Reactor”, The “Time Machine”, or the “Warp Drive” on Twitter when we were testing (and also to enhance Pete Worden’s reputation for cutting edge stuff). Now we can show what the Star Date was when a certain tape was recorded. In this case the tape being played back was originally made on day 340 of 1966 – or 6 Dec 1966 – at 11:57:43 GMT. This tape was created when the images from Lunar Orbiter II were being sent back to Earth.
With regard to Stardates and time machines, we were especially happy that William Shatner had these nice words to say about us during our crowdfunding project earlier this year. You see, we at LOIRP and McMoons have a favorite Star Trek Episode: “City on the Edge of Forever”.
As you will recall Kirk, Spock, and McCoy go back in time to the 1930s. At one point Spock has to hack his 23rd century tricorder with early 20th century electronics to get data off of it. In our case, we used early 21st century electronics to hack mid 20th century electronics to gain access to 45 year old data. Like Spock and Kirk, we had to do some dumpster diving to get parts. We like to make old things work to explore space and go back in time.


Of course we also like the episode “Tomorrow is Yesterday” when the Enterprise travels back in time (they do that a lot) and have to grab magnetic tape and film footage that shows the Enterprise (i.e. a UFO) in the skies of the 1960s. They also have a fight in the tape storage room (we don’t).

Click on images to enlarge