Cameras with quartz date/data backs going past 2020

[MattKing]

I have a data back (i think it is Oly) for my OM1n.
But i have never used it and do not know how high it counts.
I am quoting you because i know you use Olympus a lot.
Perhaps you have used it.........

Nope - but I'm pretty good at locating Olympus manuals.
EDIT: There were 6 different data backs.
The first 4 were compatible with various versions of the single digit OM bodies.
None of them were designed to deal with dates after 2009, and the first two ended before that.
http://omesif.moosemystic.net/om-sif/phototechnicalgroup.htm

 

[CMoore]

Nope - but I'm pretty good at locating Olympus manuals.
EDIT: There were 6 different data backs.
The first 4 were compatible with various versions of the single digit OM bodies.
None of them were designed to deal with dates after 2009, and the first two ended before that.
http://omesif.moosemystic.net/om-sif/phototechnicalgroup.htm

Oh Wow..... OK.
I am pretty sure it is one of the first two..... with the "wheels" on the back.
I can try to sell it, but I doubt there is much demand.
If you have any interest, I could send it for shipping, probably not worth it to you.
I will verify what I have and probably throw it into the Free Classifieds.

Thanks For The Info

 

[MattKing]

Data backs are useful for a lot of other things besides recording a date on a frame.
If you have ever used a camera to document a number of similar items, you will understand what I am talking about.
As an example, see the very manual system I used to document the film batches in this classifieds thread (all film is sold!)
https://www.photrio.com/forum/threa...duction-now-canadian-dollars-all-sold.165783/

 

[Alexander K]

My FUJI Fujifilm GA645 Zi Pro has back and it's still working fine.
year/month/day and exposure parameters, plus lens focal length

 

[AgX]

Unfortunately, aside from a helpful article on JCH's site, finding out the year ranges of quartz date backs has been quite a pain to be honest, ....

And that article even starts with calling the Canodate E the first Quartz date camera, whereas it only yields manual setting of these by twisting transparency wheels.

 

[M-88]

It always bummed me out how Olympus Recordata backs wouldn't go beyond 2009. Those were made for Olympus OM System, which was advertised as a professional-grade tool. Their latest - Recordata Back 4 was made in 1984 or so, surely it would have lasted way beyond those 25 years that manufacturers allowed it to record.

 

[AgX]

But the idea of manufacturers is to supersede models. So why should they contemplate even about something as far ahead as 25 years?

 

[Helge]

To me on-film data imprinting is basically a nice cyberpunk/grandma-punk ornament, that is a hundred percent only for the fun of it.
I usually just set the year to the earliest possible or the corresponding year in the last decade the function worked.

Anyone else feel that way?

Between frame or sprocket hole imprint is another story entirely.

 

[xkaes]

You can add the Yashica Samurai 35mm half-frame film cameras to the list of cameras that have data backs but don't go past 2020 -- the X3.0, X4.0, Z, and Z2.

Seagull made several modified Minolta 35mm SLR cameras that have databacks -- AFTER Minolta got out of the manual-focus SLR business. Some have the SEAGULL label, some don't. Some of these might go past 2020, but I have not tested the ones I have.

See -- www.subclub.org/minchin/

 

[AgX]

I don't see how "quartz" backs could be upgraded. But the manual predecessors, with transparency disks, can be.

 

[Nicholas Lindan]

The two date backs I have, a Yashica T4S/T5 and a Nikon MF-14 (F3), have two digit years that should go on in perpetuity. They count '00 as a leap year, so will have to be manually adjusted from 02 29 00 to 03 01 00 when 2100 rolls around.

How do the backs that stop at 2009 or 2019 work - do they only allow adjusting from, say, '80 to '19 or do they have 4-digit years? Seems pretty dumb, it takes less microprocessor code to let them run the gamut of any century.

 

[grat]

I don't see how "quartz" backs could be upgraded. But the manual predecessors, with transparency disks, can be.

From a programming viewpoint, it's trivial-- reset the date of "0". Quartz doesn't know the time, it just knows how to vibrate at 1/32768th of a second. The EOS 5/QD starts on May 1st, 1998 and runs for 21 years-- so pick a new starting date in the recent past, and go from there.

From a hardware point of view, the downside is, this probably requires burning a new eeprom or equivalent.

 

[reddesert]

Date range limitations often have to do with the internal representation of time. For example, if you represent time as a 32 bit signed integer counting in seconds, the longest interval from 0 that you can represent is (2^31 -1) seconds, which is about 68 years. Many UNIX systems represented time as seconds since 1/1/1970, which will overflow a 32-bit signed integer in 2038 - the "Year 2038 problem." Obviously the system clock runs much faster than 1 Hz, but this representation was used in times like file modification dates. Newer systems use a bigger integer.

In the case of date backs, I don't know the internal algorithms, but the available processing power and storage was very small, so it's likely that they cheaped out on the representation, and then restricted the available range to prevent overflowing or crashing the algorithm.

 

[Nicholas Lindan]

Well, the chances are it's a 4-bit microprocessor. BCD* math is the usual representation. Count the time & date in decimal, then send each digit through a binary->7segment table and drop in in the right I/O location for display. The whole thing rolls over every hundred years. They could also be using a compiled clock chip with new silicon source 'software.'

*BCD, for those who don't have to worry about such things - Binary Coded Decimal where a 0-9 digit is encoded in 4 binary bits (or 'nibble'). Nibbles can be strung next to each other to any desired length - 2 nibbles counts to 99, 10 nibbles counts to 9,999,999,999. Cheap and simple applications like this use 4 bit microprocessors, costing about a nickle. Memory is tight, say 8 bytes when scraping the bottom of the barrel, and so a mm/dd/yy representation only takes 3 bytes. Letting it all roll over every hundred years keeps the system from having to check for an end date.

Of course, somewhere, there is a fool making a date back with Unix or Windows and your data back is going to be infected with ransomwear.

 

[AgX]

From a hardware point of view, the downside is, this probably requires burning a new eeprom or equivalent.

I was thinking instead of an Eprom, which I have not seen in such backs.

From a programming viewpoint, it's trivial-- reset the date of "0". Quartz doesn't know the time, it just knows how to vibrate at 1/32768th of a second. The EOS 5/QD starts on May 1st, 1998 and runs for 21 years-- so pick a new starting date in the recent past, and go from there.

How to pick a new starting date? Or are you referring to that EEprom?

 

[grat]

I was thinking instead of an Eprom, which I have not seen in such backs.

True. Both would be rare, and eeprom even rarer. But it would be useful.

How to pick a new starting date? Or are you referring to that EEprom?

My guess is that somewhere, the epoch is declared within the code-- for instance, all linux machines recognize "0" as "00:00 hours, January 1st, 1970". So a timestamp of "3600" would be 01:00 January 1st, 1970". Due to limitations of code, some of the older versions of linux are going to bug out on Jan. 19th, 2038 at 03:14:08 am-- because the 32 bit integer being used to track time rolls over at that time. Most systems have been upgraded to 64 bit time, which I think is somewhere around the heat death of the universe.

I assume the QD backs have a similar limit. Mine seems to be good from 1998 to 2019. So in theory, somewhere in there is a "0 equals May 1st, 1998" bit of code. On the other hand, it *is* possible, instead of doing date calculations, someone just embedded 21 years of calendar data in a lookup table. If the chip was actually fabricated with that start date, then... well... it was a nice idea.

Alternatively look for a year that closely matches this one, and ignore the year-- for example, Aug. 2016 starts on a Monday, and ends on a Wednesday, just like August 2022.

 

[M-88]

But the idea of manufacturers is to supersede models. So why should they contemplate even about something as far ahead as 25 years?

Unfortunately I'm neither Olympus, nor Nikon worker, so I'm not qualified to give an answer as to why one company have data backs longevity of 25 years, while other has 100+.

 

[Halka]

Well, the chances are it's a 4-bit microprocessor. BCD* math is the usual representation. Count the time & date in decimal, then send each digit through a binary->7segment table and drop in in the right I/O location for display. The whole thing rolls over every hundred years. They could also be using a compiled clock chip with new silicon source 'software.'

*BCD, for those who don't have to worry about such things - Binary Coded Decimal where a 0-9 digit is encoded in 4 binary bits (or 'nibble'). Nibbles can be strung next to each other to any desired length - 2 nibbles counts to 99, 10 nibbles counts to 9,999,999,999. Cheap and simple applications like this use 4 bit microprocessors, costing about a nickle. Memory is tight, say 8 bytes when scraping the bottom of the barrel, and so a mm/dd/yy representation only takes 3 bytes. Letting it all roll over every hundred years keeps the system from having to check for an end date.

Of course, somewhere, there is a fool making a date back with Unix or Windows and your data back is going to be infected with ransomwear.

Yes, but will it run Doom?

(Thanks for the write-up BTW, I learned something new today)


Another one for the future-proof list is the Contax Aria with the D-9 data back, which allows the year to be set anywhere between 00-99. This one does not imprint any data on the frame itself, instead printing it between the frames. Unless you flip a tiny switch on the inside labeled "EXP.DATA IMPRINT" to the on position, which will cause the film counter to start at 3 for the newly inserted roll of film. After finishing said roll, the camera will print the exposure data line-by-line when rewinding (which will take about 30 seconds) on the first two frames.

 

[Stephen Furley]

Mine are in store at the moment, so I can't check the ending dates, but I do know that all of my Nikon backs work. They are still very much in use; I like to put a date stamp on the first frame of a roll, and on the first frame of each session if I use a roll over several days. The backs I have are:

MF-12 (All models of the FM/FE/FA family except the FM-10 and FE-10)
MF-16 (All models of the FM/FE/FA family except the original FM and FE, FM-10 and FE-10)
MF-15 (FG)
MF-19 (F-301 and F501)
MF 26 (F90 and F90x)
As of yesterday I also have the mF-29 for the F100, which I believe works, but I haven't managed to get the battery into it yet.

The only Nikon camera I have with a built-in non removable back, the F50D, doesn't work.

The Canon Data Back A is weird, in contains no clock, you set the date manually with three dials on the back. It only has two dogits for the date and is limited to certain ranges which seem to be :79-90, 0-9, roman from i-x, a-g and what looks like a square.this fits the ae-1, ae-1 program, at-1 and a1, but not the av-1 or al-1. it has no built in contacts to connect to the camera, but is triggered by a short jumper cable to the camera's pc flash connector, like the Nikon MF-12. This is one area where Nikon did a better job than Canon!

 

[Huss]

Nikon F6 has a built in data back that works correctly. Imprints between the images - exposure data and time/date stamp.

 

[grat]

Of course, somewhere, there is a fool making a date back with Unix or Windows and your data back is going to be infected with ransomwear.

Entirely possibly, although A) most such applications would have a network service to exploit and B) most ransomware wouldn't run on such a highly specialized embedded device.

 

[Fredrixxon]

My Nikon MF-26 data back still counting with F90 (F90X)

 

[Lachlan Young]

The Konica Hexar AF data back starts counting from 1987 - 6 years before they hit the market...

 

[Huss]

The Konica Hexar AF data back starts counting from 1987 - 6 years before they hit the market...

It was a camera past its time.

 

[staviss]

Do you have any sample images of what the date looks like imprinted on pictures taken with your Nikon?? I have an FE2 and looking to buy an M-16 for the date imprint abilities but can't seem to find any sample images of what it will look like...