IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Digitizing NASA-ESA Taped Imagery, project to digitize 60s-70s data
apollo16uvc
post Feb 28 2019, 10:11 PM
Post #1


Newbie
*

Group: Members
Posts: 12
Joined: 27-August 18
Member No.: 8441



Hello dear readers,

Here are some satellite telemetry tapes I have acquired. They are from the 60's to 70's.





These tapes contain the raw signal received from satellites at NASA tracking stations like GFORKS, STIAGO and WINKFIELD.

It seems that there 2 telemetry tracks, 4 misc tracks (Including a reference track) and one voice track.

I am working on getting a recorder to play these tapes. If you know the location of any 7-track instrumentation recorders (Ampex FR-100, FR-600) that would help.

Here is a sample of tape 3141/2N003 played back on a 1/4 inch 4-track Akai at 7.5 I.P.S: https://www.dropbox.com/s/k2jbyoka6n50yhh/3...168-2.flac?dl=0
The tape number on the boxes and documentation is 3141/2N003. The number on the reel itself is 10786-16-8.

Satellite: 1963-014A & B (ERS5)
Recorder: FR-100
Speed: 15 I.P.S
Station Name: GFORKS

Here is an other sample from a different tape: https://forum.nasaspaceflight.com/index.php...1683;sess=54302

Here is a video showing the tracks: https://youtu.be/I85-aZuupxM
With this we have proven that something is on the tape, and it can be picked up with a sound head. This is the case with most of my tapes.

I am currently talking with someone who has several 1/2 8-track studio tape recorders, and we are looking into getting an Otari MX-5050 8 unit working. This unit is most useful because it can play at both 7.5 I.P.S and 15 I.P.S, which is what we need. All mechanical functions seem to work fine, and the input VU meters register when the build-in tone generator is switched on. But... there is no sound from his tape! (His own audio, not a NASA tape)

This recorder is currently undergoing repairs.

Other tapes I have acquired are ESA tapes from the same era.
I have tested out three of the five ESA tapes with a magnetic viewing solution, and all three clearly showed 7 tracks like the NASA tapes. This means they have not been degaused or overwritten with an audio recorder. The tracks look like raw telemetry, not computer tapes. One tape has a label that clearly says it came from a tracking station. I think we should be able to digitize these too eventually. The tracks are very clear.

The tapes I have tested are:
TD-1 (Tape ID: 1117-09-08- cool.gif
ESRO 1A (Tape ID: 800 645 08 10B)
HEOS A2 (Tape ID: 1115 06 11B)

I have made two videos on it in dutch.
First, a tutorial on how to make your town magnetic viewing solution.
https://www.youtube.com/watch?v=j_kA0cnkBLI


And finally, a video where I visualize the magnetic tracks on 3 tapes:
https://www.youtube.com/watch?v=LAjU2AHIksA

I have attached some photos of the tracks to this message. I promise I will publish a big archive with detailed scans and photos of all ESA and NASA tapes currently in my possession.

Macro photos (Large!)
https://imgur.com/a/rnlJH9P


So what are we looking for?
I am interested to know if its possible to determine the frequency of a track with the track photos.Is it possible to replace the 8-track head in an 1/2 inch Otari tape recorder with a 7-track head, perhaps from a computer tape drive?Is it possible to decode the digitized raw telemetry data into numbers, perhaps a spreadsheet or interactive database?Would people be interested to crowdfund some of the greater expenses that may be required?
Best regards,Niels
Go to the top of the page
 
+Quote Post
apollo16uvc
post Mar 1 2019, 11:58 AM
Post #2


Newbie
*

Group: Members
Posts: 12
Joined: 27-August 18
Member No.: 8441



I have finally finished v1.0 of the archive for the NASA satellite tapes.

Sat-53123114313-Version1.0 (Object photos, text file.)

Sat-GFORKS-314N003-Version1.0 (Object photos, documentation, Akai RTR sample, visualized tracks)

Sat-GFORKS-314N079-Version1.0 (Object photos, documentation, Akai RTR sample)

Sat-GFORKS-330N100-Version1.0 (Object photos, documentation, Akai RTR sample)

Sat-SNTAGO-120J827-Version1.0 (Object photos, documentation, visualized tracks)

Sat-SNTAGO-314J019-Version1.0 (Object photos, documentation)

Sat-WINKFIELD-330P001-Version1.0 (Object photos)
Go to the top of the page
 
+Quote Post
nprev
post Mar 2 2019, 12:55 AM
Post #3


Senior Member
****

Group: Admin
Posts: 8404
Joined: 8-December 05
From: Los Angeles
Member No.: 602



Topic title edited for clarity and generalization.


--------------------
A few will take this knowledge and use this power of a dream realized as a force for change, an impetus for further discovery to make less ancient dreams real.
Go to the top of the page
 
+Quote Post
apollo16uvc
post Mar 8 2019, 11:44 PM
Post #4


Newbie
*

Group: Members
Posts: 12
Joined: 27-August 18
Member No.: 8441



Some interesting new finds, did we play telemetry?

Last week I took the time to play some ESA tapes (1/2 inch 7-track) on my Akai X201D (1/4 inch 4-track)

The tapes played:
1.
SAT: ESRO 1A
TAPE ID: 680841-292-230
ESOC/Section TLM: 13496
DATE: 24 JULY 70

2.
SAT: 720,141
TAPE ID: 1135 05 10A
ESOC/Section TLM: 21554
DATE:

3.
SAT: TD-1
TAPE ID: 1117 09 08 B
ESOC/Section TLM: 16837
DATE: "Day 089"

4.
SAT: TD-1A
TAPE ID: 1118 07 09 A
ESOC/Section TLM: 16672
DATE:


To give you an idea of how satellites sounded in the 60's and 70's check out this website with recordings.


I made a video where I play the tapes and show it on an oscilliscope:
ESA satellite tape playback - YouTube

Some remarkable details:
ESRO 1A has a lot of activity at the beginning, it looks like a reference signal that is being adjusted. There pitch changes and there are periods of noise. Eventually we receive a stable signal which is certainly more complex than a simple sine wave.

ESRO 1A:
Oscilliscope:


Spectrogram: seems to show a kind of square wave, would this be satellite data?


Signal played at 20% original speed, sounds like morse code.


The space between the signals is similar to track 6 of the ESRO 1A tape:


TD-1
Oscilliscope : The wave of this signal swells up and comes down again.



Spectrogram : And here you can see that too.


TD-1A
Oscilliscope:
This signal has two harmonic waves:


And when we zoom out, it has a kind of block pattern:


But when it is very interesting to delay the signal, it sounds like a morse code again.
Spectrogram:


Signal played at 15% original speed, sounds like morse code again.

A lot of new information that will take some time to process.

It seems to me quite possible that this is the received data. If we find documents from the relevant satellite with information about telemetry, should it be possible to create a program or circuit that processes the signal?
A program could convert it to a spreadsheet. How much volts the battery outputs every second for example.

I do not know anything about it, but the ESA recordings do not seem to be FM-modulated, since such a wave looks very different.
The NASA recordings are usually not, so apparently AM and FM modulation was not common in recordings from this time.
The NASA documentation usually also has "Direct" recordings and not "FM"

I am looking for people who may be able to help with the relevant satellites, and who are more acquainted with this kind of work.

Niels
Go to the top of the page
 
+Quote Post
monty python
post Mar 10 2019, 09:19 PM
Post #5


Member
***

Group: Members
Posts: 172
Joined: 2-March 06
Member No.: 692



I love that you are interested in the fine details of early sat comm, preserving the nuances of actual comm that could be lost with time.
Go to the top of the page
 
+Quote Post
Gerald
post Mar 11 2019, 03:10 AM
Post #6


Senior Member
****

Group: Members
Posts: 2264
Joined: 7-December 12
Member No.: 6780



QUOTE (apollo16uvc @ Mar 9 2019, 12:44 AM) *
But when it is very interesting to delay the signal, it sounds like a morse code again.
Spectrogram:

...rtfturlunnvaana...
if interpreted as Morse code. No idea, what that might mean.
Some of the tape sounds like data tapes widely used until about the early 1980s, or similar to modems until the early 1990s, with some carrier signal followed by data.
So, it might be, that a code typically transmitted by RS-232 interfaces, like Baudot codes, BCD, EBCDIC, or ASCII would result in more meaningful interpretations.
Then, you'd have to find out, which the start and stop bits are, and translate the enclosed character.
But it's long ago, and has been rare, that I needed to work on that very fundamental level.
Go to the top of the page
 
+Quote Post
Gerald
post Mar 12 2019, 06:07 PM
Post #7


Senior Member
****

Group: Members
Posts: 2264
Joined: 7-December 12
Member No.: 6780



Here is an idea of how to convert the diagram into a bit stream:

Attached Image

When coding the upper potential with 1, and the lower potential with 0, with a more or less fixed clock rate, we get pairs 01 or 10, never 00 or 11. So, such a pair, or transition, appears to encode a bit.
It should hence be fairly straightforward to convert the signal into a bit stream.
Go to the top of the page
 
+Quote Post
apollo16uvc
post Mar 16 2019, 10:11 PM
Post #8


Newbie
*

Group: Members
Posts: 12
Joined: 27-August 18
Member No.: 8441



QUOTE (Gerald @ Mar 12 2019, 07:07 PM) *
Here is an idea of how to convert the diagram into a bit stream:

Attached Image

When coding the upper potential with 1, and the lower potential with 0, with a more or less fixed clock rate, we get pairs 01 or 10, never 00 or 11. So, such a pair, or transition, appears to encode a bit.
It should hence be fairly straightforward to convert the signal into a bit stream.


Interesting observation Gerald, thanks for posting. I don't think it would be too hard for us to write a program to process an audio file in this manner. Will ask some people what they think about this.

Voice found on beginning of NASA tape

On satellite tape 'Sat-SNTAGO-120J827' a voice can be heard briefly at the beginning. The voice probably tells us the recording time. "The time is 11 5 AM"

The satellite is Upsilon 61, the recording date was Sept 7, 1961

I had to play the recordings backwards.
Here the fragment as an MP3 file.
Go to the top of the page
 
+Quote Post
g4ayu
post Mar 17 2019, 07:42 AM
Post #9


Newbie
*

Group: Members
Posts: 8
Joined: 23-June 09
From: Leyland U.K.
Member No.: 4834



Also between the two voice recordings, "1605" is sent in CW (morse code)
Go to the top of the page
 
+Quote Post
apollo16uvc
post Mar 19 2019, 07:04 PM
Post #10


Newbie
*

Group: Members
Posts: 12
Joined: 27-August 18
Member No.: 8441



QUOTE (Gerald @ Mar 12 2019, 07:07 PM) *
Here is an idea of how to convert the diagram into a bit stream:

Attached Image

When coding the upper potential with 1, and the lower potential with 0, with a more or less fixed clock rate, we get pairs 01 or 10, never 00 or 11. So, such a pair, or transition, appears to encode a bit.
It should hence be fairly straightforward to convert the signal into a bit stream.


Hello Gerald,
Thanks for checking up on the signal! looks pretty good. I shared this with other people, and they think the signal looks like manchester or NRZ coding.See the following diagrams that

look very much like our signals:I think it loos most like Manchester. But which one?
kaq9 from the apollohoax forum said the following:
"The 01 and 10 (never 00 or 11) sequences is call "Manchester coding", and it was a very popular line code at one time. It still is, though there are better alternatives now. It eliminates any DC component from the data and pushes the spectrum away from zero so it can be transformer or capacitor coupled or recorded on magnetic media using old-style electromagnetic-type heads. It's also self-clocking."
Hopefully this helps. Once we got more of the signal processed into a bit-stream i can send a sample to the american with the computer tapes, maybe he will recognize something.
I noticed the compression of the MP3 files I provided corrupt the signal quite a bit.
These lossless FLACs come right off the video recordings made with my Nikon camera. So no weird compression artifacts or bitrate limits are happening.
The spectogram you got your code from was from TD-1A-1118-07-09-A. I made the spectogram with Audacity.
ESRO-1A-680841-292-230

TD-1-1117-09-08-B

TD-1A-1118-07-09-A

TD-1A-1135-05-10A
Go to the top of the page
 
+Quote Post
Gerald
post Mar 20 2019, 01:50 PM
Post #11


Senior Member
****

Group: Members
Posts: 2264
Joined: 7-December 12
Member No.: 6780



A data character is presumably stored on the tracks in parallel. So, it will be difficult to make a decision on the basis of a single track. The 8th track might encode a parity bit. So, testing the outcome of several assumptions may resolve the question about the applied code. It's some Manchester, rather obviously. My first choice would be the differential code, since it's more resilent against polarity errors. But on the basis of a character stream, some higher-level encoding, like Hamming or Gray codes may still be possible.
A first milestone would be retrieving a stream of characters like they are on the tape on the most basic level in (polar) NRZL. Any higher-level interpretation, like Manchester, can be derived from the NRZL character stream.
Go to the top of the page
 
+Quote Post
apollo16uvc
post Mar 21 2019, 07:44 PM
Post #12


Newbie
*

Group: Members
Posts: 12
Joined: 27-August 18
Member No.: 8441



QUOTE (Gerald @ Mar 20 2019, 02:50 PM) *
A data character is presumably stored on the tracks in parallel. So, it will be difficult to make a decision on the basis of a single track. The 8th track might encode a parity bit. So, testing the outcome of several assumptions may resolve the question about the applied code. It's some Manchester, rather obviously. My first choice would be the differential code, since it's more resilent against polarity errors. But on the basis of a character stream, some higher-level encoding, like Hamming or Gray codes may still be possible.
A first milestone would be retrieving a stream of characters like they are on the tape on the most basic level in (polar) NRZL. Any higher-level interpretation, like Manchester, can be derived from the NRZL character stream.


The satellite tapes are all 7 tracks. And I don't think they are computer tapes, but original tapes recorded at a satellite tracking station with instrumentational recorders. My NASA satellite tapes clearly have documentation that states this. Unfortunately no documentation came with the ESA tapes, but one of the 5 tapes has a label stating its from a tracking station, and recorded at 7.5ips. Given the labels on this ESA tape match the ones on other ESA tapes, all are from a tracking station. I think computer tapes are variable speed, so they would not have 7.5 or 15 ips written on them.

(It needs to be clarified that I also have NASA COMPUTER tapes, but the tapes posted in this thread are NOT.

I have actually visualized the tracks on NASA and ESA tapes with a DIY magnetic viewing solution, see here a ESA TD-1A tape (Not sure which one)
Attached Image

And here from NASA tape Sat-SNTAGO-120J827:
Attached Image


Compare this to a computer tape, clearly different:
Attached Image

And the documentation that comes with most of the NASA tapes looks like this:
Attached Image


What I think has happened, is that the signal from TD-1A was processed into a bitstream, and this bitstream was recorded on tape with an instrumentational recorder.
So not a computer/digital tape per se, as it contains analog tracks with voice, timing, etc. So I don't think there will be a parity track, none stated anywhere.
I think the tracking station processed the satellite signal, encoded the telemetry bitstream as NRZL or manchester and this is what we are now reproducing. Of course it is all possible we are reproducing the wrong track and its actually a timing code, but I doubt that. The problem is that we got only 1 of the total 7 tracks. Looking for an instrumentation recorder, or just the 7-track head if possible. We don't even need a working one, as long as the head is in good shape I can transplant the 7-track head in a 1/2 inch Otari MX5050 III studio recorder for proper playback. The Otari is under repairs.

I thought manchester and NRZL were two separated methods of encoding, and not just two different levels?I think the signal we get with the tape recorder is about as low as we can go!
Go to the top of the page
 
+Quote Post
Gerald
post Mar 21 2019, 11:56 PM
Post #13


Senior Member
****

Group: Members
Posts: 2264
Joined: 7-December 12
Member No.: 6780



Thanks for constraining this further down!
A 100101, for example, in NRZL can be interpreted as a 01 in differential Manchester BP-M, with the initial 10 in (10)0101 as clock signal, the 00 in 1(00)101 encoding a 0, then 01 in 10(01)01 the next clock signal, 10 in 100(10)1 encoding a 1, then 01 1001(01) clock, and so forth. In BP-S, the same NRZL sequence would encode 10, if read from left to right.
So, you can just digitize the NRZL, and decide later, how to translate it into a Manchester bit stream of various possible types. If the tape is recording a sequential bit stream, it's possible that the direction of the recording is alternating with the tracks, meaning that, in principle, only one head is necessary to write and read the track. After reading the first track, the head is moved to the next track, and the tape is running backward. That's to be considered for decoding.

The next step would then look for a header or preamble in the recording, in order to see, whether it's defining some type of higher-level convention. But I'm not sure, which standards have been defined back in 1963.
Go to the top of the page
 
+Quote Post
apollo16uvc
post May 7 2019, 07:43 PM
Post #14


Newbie
*

Group: Members
Posts: 12
Joined: 27-August 18
Member No.: 8441



QUOTE (Gerald @ Mar 22 2019, 12:56 AM) *
Thanks for constraining this further down!
A 100101, for example, in NRZL can be interpreted as a 01 in differential Manchester BP-M, with the initial 10 in (10)0101 as clock signal, the 00 in 1(00)101 encoding a 0, then 01 in 10(01)01 the next clock signal, 10 in 100(10)1 encoding a 1, then 01 1001(01) clock, and so forth. In BP-S, the same NRZL sequence would encode 10, if read from left to right.
So, you can just digitize the NRZL, and decide later, how to translate it into a Manchester bit stream of various possible types. If the tape is recording a sequential bit stream, it's possible that the direction of the recording is alternating with the tracks, meaning that, in principle, only one head is necessary to write and read the track. After reading the first track, the head is moved to the next track, and the tape is running backward. That's to be considered for decoding.

The next step would then look for a header or preamble in the recording, in order to see, whether it's defining some type of higher-level convention. But I'm not sure, which standards have been defined back in 1963.


Due to these being telemetry recordings made on instrumentational recorders, its likely they are compliant with the IRIG standard.

See the following IRIG timecodes: https://en.wikipedia.org/wiki/IRIG_timecode

Modulation type
  1. (DCLS) Direct Current Level Shift (width coded)
  2. Sine wave carrier (amplitude modulated)
  3. Manchester modulated
Carrier frequency
  1. No carrier (DCLS)
  2. 100 Hz (10 ms resolution)
  3. 1 kHz (1 ms resolution)
  4. 10 kHz (100 Ás resolution)
  5. 100 kHz (10 Ás resolution)
  6. 1 MHz (1 Ás resolution)
The signal received on TD-A1-3 is 5Khz or 10Khz depending on the playback speed. I hope I can achieve the required recording resolution.


Some useful links:
http://irig.org/

http://www.irig106.org/
Go to the top of the page
 
+Quote Post
apollo16uvc
post Jul 30 2019, 08:17 PM
Post #15


Newbie
*

Group: Members
Posts: 12
Joined: 27-August 18
Member No.: 8441



Finally acquired some 7-track IRIG tape recorder heads, from an Ampex FR-600 no less! two 4-track 1/2 inch heads setup for interlaced 7-track. The 7-track IRIG standard allows for a single 7-track head, or two heads for interlaced 7-track.

Unfortunately not much luck with the Otari MX5050 III. Peter tried an other reel that he knows has audio recorded on it, and the Otari's VU meters do move. But there is still no audio coming out...
Attached Image
Attached Image
Attached Image
Attached Image

Go to the top of the page
 
+Quote Post

Reply to this topicStart new topic

 



RSS Lo-Fi Version Time is now: 22nd November 2019 - 07:15 PM
RULES AND GUIDELINES
Please read the Forum Rules and Guidelines before posting.

IMAGE COPYRIGHT
Images posted on UnmannedSpaceflight.com may be copyrighted. Do not reproduce without permission. Read here for further information on space images and copyright.

OPINIONS AND MODERATION
Opinions expressed on UnmannedSpaceflight.com are those of the individual posters and do not necessarily reflect the opinions of UnmannedSpaceflight.com or The Planetary Society. The all-volunteer UnmannedSpaceflight.com moderation team is wholly independent of The Planetary Society. The Planetary Society has no influence over decisions made by the UnmannedSpaceflight.com moderators.
SUPPORT THE FORUM
Unmannedspaceflight.com is a project of the Planetary Society and is funded by donations from visitors and members. Help keep this forum up and running by contributing here.