Hacker News new | past | comments | ask | show | jobs | submit login
Expert help needed to decode rare piece of space history: Lunokhod data tapes (attivissimo.blogspot.com)
148 points by pamoroso on Aug 21, 2021 | hide | past | favorite | 45 comments



You should be aiming to play the tapes once only. If you have something that can physically play the tapes without damaging them, make the highest fidelity PCM (digital) recording possible of the analogue signal, without any regard to further decoding. The tapes should then be placed in a safe place and never played again (unless you later find something wrong with your PCM capture of the analogue signal).

Make a backup of your digital/PCM recording and put it in a safe place. All further decoding work is then done using a copy of your high-fidelity PCM/digital recording of the analogue signal.


Perhaps reference what preservation standard this follows?


Common sense?


Just from what I hear on the video, I would say a few things:

1. It sounds like it's around 9600 baud.

2. It sounds sparse.

3. At that bit rate, that amount of data doesn't contain an image.

My guess would be that one track is commands being sent or telemetry being received, and the other track is maybe an operator explaining what he's doing or commenting on the decoded telemetry.

Edit: I had the Russian guy in the next room translate some of it, and it sounds to me like there's a burst of data, then a comment about things like "device is on", something about lights, etc., so I'm going with the return telemetry one. Maybe they were also getting video, but it's on a different channel.


Actually there is a very long block of data in that sample. If your assumption of 9600 baud is correct it will equivalent to over 300kB of data - enough to send an image or part of it.


What was the start time?


> Between 1970 and 1973, the Soviet space program achieved an amazing feat: landing and driving the first remote-controlled roving robots on the Moon. These two vehicles were called Lunokhod-1 and -2, and they gathered invaluable scientific data as well as images of the surface of the Moon during their unprecedented, and still unrivaled, total of 52 kilometers of remotely controlled lunar exploration.

Wow. TIL.

Wish I had more signal processing know-how to help decode that footage. I'm old enough to remember programs being stored on cassette tapes, but not old enough to have any working experience with that.


Interesting side note : the people who built the Lunokhod moon rover were later called in to help with the Chernobyl cleanup. They had experience with remote-controlled radiation-hardened rovers.

There's an interesting documentary on Curiositystream about the Lunokhod program too.


As for results in the cleanup of Chernobyl, they did not fare that well. About 10% of the roof was cleared before radioactivity made it impossible to continue.

Though that effort probably saved lives, so mission accomplished I'd say.

https://spacecentre.co.uk/blog-post/the-moon-rovers-that-hel...


What's more impressive that this is 1970, so remote driving capabilities on that kind of distance is mindblowing. I was in NASA KSP tours, man, this tech from 1960 and 1970 looks weird and not functional today.


Incredible. At first I assumed that was a lunar record, but Opportunity was in action for 15 years and 'only' drove 45km


Wow even at moon distances the speed of light starts to play a bit. Remote control at that distance is impressive.


This sounds exactly like a scenario that openfly gave a Defcon talk on a few years ago -- about old NASA data storage technology that is on the verge of becoming unreadable and how we have to think about the formats that we currently use and preservation into the future.


For the last few decades, the best practice for data archiving has been to include a machine capable of reading the data inside the box with the data itself.

E.g. Inside your big crate of tapes, you include a tape reader, with printed instructions and schematics for how to read the data.


Do you have a reference for that talk? (or about the "lost" nasa data). I have a kink for bizarre file formats and this got me excited...



Hi. Native Russian here. Just took a quick glance, it's a very common type of tape, called svema (свема). Should be totally possible to make a digital copy. I can contact museums etc if needed. I also have folks who upscale old videos on YouTube with neutral nets, starting with first Lumiere videos to old SF tapes


Consider contacting some technical experts at sound and film archives, they often keep one or two machines in running order and have knowledge in identifying material. Source: worked for Australia’s ABC Archives doing the back-end for analog tape digitisation project. Step 1 was “get working machines.”


Sounds like a job for Mike Stewart [1], or rather a quick hack over his lunch break, based on his work on the AGC.

[1]: https://blog.samtec.com/post/restoring-agc-part1/


A lot of interesting details (and photos) about the onboard TV-photo system of the Lunokhod-1 could be found on [1].

Btw, it looks like the tapes indeed may contain the photos - the session dates seem to coincide with taking of panoramic shots, according to the official news coverage of the mission (mentioned earlier). Moreover, the red "Ст." marks on the tape cover might just point to the "Stereoscopic" shots, which are produced by the cameras 2 and 4 (left-right pan). This also explains the "even" remark on the tape cover.

Also, the narrator mentions that the recording was done directly from FT console (Photo-Television), likely post-factum (??for lab analysis/enhancement).

Wikipedia has a nice photo of the Control indication panel [2], the "ФТ" panel lights are on the right edge. There you could also spot SKU lights ("СИСТЕМА КУРСОУКАЗАНИЯ", Course Guidance System) on the left bottom. Lights for RIFMA were likely on the bottom edge in Scientific Instruments group (covered in the photo by a description plaque).

Fun to learn about this. Thank you for the "journey"! It was truly a great achievement not just for people of USSR, but for the whole humankind. Hopefully, given the recent declassifying of some of the Soviet space-programs, it might be possible that there could be already available the images of the very panoramas possibly contained on the tapes (matching by the session dates).

Anybody knows of any online archives of images taken by Lunokhod-1? ...Well, answering it by myself: [3] is an excellent selection of historic shots of Moon from Soviet exploration efforts, lots of technical info about cameras and signals.

[1]:https://www.kik-sssr.ru/Lunohod-1.htm (in Russian)

[2]:https://ru.m.wikipedia.org/wiki/%D0%9B%D1%83%D0%BD%D0%BE%D1%...

[3]:http://mentallandscape.com/C_CatalogMoon.htm


> ...it might be possible that there could be already available the images of the very panoramas possibly contained on the tapes (matching by the session dates).

Indeed, an exhaustive set of scans of Lunokhod-1 panoramas has been published recently by the Laboratory for Comparative Planetology. Lunar mission Day 5 (Lunation 5) panoramas are at [1].

There were reported total 16 panoramas made during Lunar mission Day 5 (March/9-20/1971, 12 days); last one - during the session 8 (see [2], Lunation 5. Session 8. Panorama 16).

Assuming the session/seance number coding is [Lunar-day][session], session-code 508 corresponds to Lunar-day:5, session:8. That is possibly no more reported panoramas in the remaining sessions of the Lunar day.

One of the tape in question contains sessions 509, 512, 515. Thus there might be no more reported/published panoramas on that tape.

However, the audio fragment given (?session 509), annotates it as "continuation of session 508". Perhaps it may have some additional panorama data for the 8th session panorama.

As for the other tape, it's during Lunar mission day 7 (May/7-15/1971, 9days). Total number of panoramas reported for this period is 21; last made during session 8.

So, similarly, the other tape (for the session 709) may just have either no more panoramas or some additional data for the panorama taken during session 708.

Anyways, hopefully the "buzzing" as heard on the sample will be decoded to reveal something interesting.

Good luck!

P.S. Btw, the reason the tape sample sounded in reverse may be because it was played on the wrong side. Perhaps it should have been rewound prior to the replaying. Meanwhile, the other tape has a remark on it "Перемотана" (Rewound).

[1]:https://www.planetology.ru/panoramas/lunokhod1.php?page=18&l...

[2]:https://www.planetology.ru/panoramas/lunokhod1.php?page=21&l...


>... However, the audio fragment given (?session 509), annotates it as "continuation of session 508". Perhaps it may have some additional panorama data for the 8th session panorama.

Relistening to the sample, it appears that narrator announces turning ON the RIFMA just before the supposed data signal begins. I would rather interpret this data as relating to the spectral analysis of the Lunar soil, not the panorama data stream.

Both feeds supposedly use the same transmission system. RIFMA signal is modulated in the same way as the signal from cameras 1 and 4 (which could be sent simultaneously on separate sub-carriers).

On reception the signal is filtered, demodulated, and then recorded (including onto the magnetic tape). This means that the signal we hear on the sample has been already demodulated and probably represents some kind of bit-stream. The panoramas were also recorded onto thermal paper, akin Fax, not necessarily with the same protocol though.

Panorama data is known to be transmitted at 4 lines/sec, each (vertical) line is 500 pixels. So the bit rate could be around 500 x pixelbits x 4=2000 x pixelbits per sec. There should likely be some control bits.

The 4 minute sample chunk then would contain around 240 x 2000 x pixelbits of data (960 lines). Assuming 8-bit pixels, this yields about 450KB. However, it as well could be 1-bit black or white.

If indeed it's RIFMA data, then it's hard to guess the data format. In case of panorama data, at least there's a chance of simply interpreting it as a stream of intensities.

This also leaves open the question of determining the 0/1 encodinng in the recorded "audio" signal (? Freq/delay modulation)

Also the sample appears to be essentially mono doubled to both channels, instead of data and narration on separate tracks as supposedly is on the tape. This bleeds the hum and buzz ?from the narrator's mic onto the data feed.

Perhaps you could re-record the tracks/channels separarely for better chance at the data track analysis.


>... One of the tapes in question contains sessions 509, 512, 515. Thus there might be no more reported/published panoramas on that tape.

The notes sheet in Fig. 10 of your blog page are obviously related to the session 512 (mentioned in the notes and also visibly marked 512 on the back side).

According to the outlined status for the beginning of the session 512, it's stern away from/off Sun, without SKU, without TM, without panorama, with RT.

That is, the RT (Roentgen telescope) is on. So possibly the contents of the session 512 is relating to the reception of RT imaging data.


> ...“Transmission 109 of 13 May 1971. It has two parts, first part with TM, second part without TM. Length: 40' [probably minutes]”.

Looking at the handwriting for reel 1, it appears that the transmission session number is rather 709, not 109.

Note, that the handwriting for the second reel shows the session 509 in March of the same year. The "first" reel dates the session in May.

P.S. Hard to judge about origins of the tapes, but by the looks, the reels seem more like some personal use dubs, not much an "official" storage. No reg numbers, names, very non-official.

Additionally, the edge remarks on the second reel mention that some parts were dubbed on "someone-else's" recorder. This is likely noted in case the head alignment needs adjustment for better read.


> ...Looking at the handwriting for reel 1, it appears that the transmission session number is rather 709, not 109.

Perhaps, I'm mistaken, however in support of this reading speaks also the mission's Lunar day numbering [1]

This places the March session recordings, as marked on the reel (509,...) on the 5th Lunar mission day, then May session recordings could correspond to the 7th Lunar mission day (thus 709).

Again, just a hunch. Maybe the narrator announces the number clearly on that May tape.

[1]:https://ru.m.wikipedia.org/wiki/%D0%9B%D1%83%D0%BD%D0%BE%D1%...

Lunar-mission-day, Date-range-1971

1st, Nov/17-24/1970

2nd, Dec/8-23/1970

3rd, Jan/7-21

4th, Feb/7-20

5th, Mar/7-20

6th, Apr/6-20

7th, May/6-20

8th, Jun/4-11

9th, Jul/3-17

10th, Aug/2-16

11th, Aug/31-Sep/15

P.S. Also an interesting read is the mission PR in the state newspapers [2]. Gives an outlook at what notable was achieved on which date. So this might provide some hints about possible kind of "graphic" content on the tapes...if any, of course.

[2]: http://epizodsspace.airbase.ru/bibl/osvoen-kosm-pr-sssr/1971...


In Russian handwriting, the "1" has a hooked top and the "7" has a strike through the middle. This is most likely a "1". You can see a "7" in "1971". Looks like a "1" to me but I'm not a native speaker.


If you look closely, 1 in 1971 is a lot closer than 7 that written in 109. Also as mentioned below, defining stroke in the middle with 7, and how it is written. For number 1, writing starts with upper left and goes down, that makes sense for translation from 1 to 0. For 7 you do same but middle stroke takes your hand off paper so there is no direct and smooth translation to next number. That's my reasoning for interpreting it as 109, rather than 709.


I just extracted the audio from the sample mp4. Each data record seems to be a lead-in tone with a frequency of 3.6kHz, followed by something (data?) followed by a lead-out tone with a frequency of 3.6kHz.

The spectrum of the "something" is broadened, suggesting a modulated signal, though it does have a strong peak at approximately 5943Hz.


The IBM tape lab in Tucson was used to recover the Space Shuttle Challenger tapes. They may be able to recover this data as well.


Asked a Russian friend of mine. The image of the page with lots of Russian text on it is about how to store magnetic tape.


Why is it Lunokhod? It should be Lunohod (MoonWalker). Russian writing Луноход. I hope author of the blog answers my email.


There are many systems for romanising Russian. Most of them use kh for х (including those used by US/UK for Russian place names, names in Russian passports, road signs in Russia).

Wikipedia has a handy table: https://en.wikipedia.org/wiki/Romanization_of_Russian#Transl...

As to why it is that way, I don't know. Some closely related languages (Belarusian, Bulgarian) follow different conventions. For example, the national system in Belarus transliterates х as ch and г as h.


Hello, I'm Paolo Attivissimo, the poster of the original article. I am using the "Lunokhod" spelling because this is the spelling used by NASA and many other reference sources, including Anatoly Zak's RussianSpaceWeb http://www.russianspaceweb.com/spacecraft_planetary_lunar.ht...


Because х is romanised as kh.


Maybe in english.


It's an English-language website so why wouldn't they use English transliteration?


Because 'h' and 'kh' usually denote different sounds, the latter encodes https://en.wikipedia.org/wiki/Voiceless_velar_fricative.


Thanks all for the suggestions and feedback. Paolo Attivissimo, the author of the linked post, is monitoring this thread.


is this legal?


That is a good question. The legal status of these data is unclear. That is one of the reasons for publishing only a sample of the data.

Personally, as a matter of opinion, I presume that these data can be published legally, based on other similar cases. But if the owner or myself are contacted by lawyers, I'll update the article.


It's not legal in the USSR. But I don't think they have the resources to go after you these days.


If you think this is funny, your kind of humor is perfect for reddit


Didn't think anyone knew about that place, that's where I built my experience!


> The owner has not disclosed the full list of contents of the tapes to me so far.

Why not? Why would someone want to help without an explanation for that? Seems pretty suspect.


The owner is very protective of these tapes and for personal reasons has found it very difficult to trust reporters.

Personally, I cannot guarantee the authenticity of these tapes, although everything I have seen so far seems to confirm it. Ascertaining the status of these tapes is the first step. Let's see where it takes us.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: