Sensor module for old film cameras

Hi Zone8, the real point is that digital sensors perform best paired with telecentric lenses. Old lenses are mostly non-telecentric. IQ would be poor, unusable, except as curiosity in most cases, IMO.
Just for curiosiity, here is a poor, unusable pic taken with a DSLR using a non-telecentric, manual focus lens whose design dates back to the film era (1981):





Lens: Nikkor AI-S 24mm 1:2.0 (photo from http://www.photosynthesis.co.nz ):



BTW: Olympus dropped the "telecentricity" marketing hype. Micro-FT lenses, with their extremely reduced 'register', show the exact contrary of telecentric design.
--
Rapick
Jalbum supporter
http://www.pbase.com/rapick
http://rapick.jalbum.net/
 
Here's my proposal.

Use a LiDE or any scanner using a CIS (Contact Image Sensor). This gives you a sensor array which will cover the whole width on the image plane. Also, because there is no internal objective lens, you can use the scanner untouched and just put the platen where the film would go, focussing on the scanner. You probably would want to disconnect the LED arrays on the scan head.

A good scanner would be the CanoScan 700F. This has a slide mode, which I guess turns off the scanner LEDs in favour of the slide unit LED's. It's available from Amazon for about £90 in the UK, probably cheaper off Ebay. It also scans at 9600 dpi, which means from your 5x4 camera you'll get 1.8 GP, an impressive enough pixel count to perhaps not worry about the per pixel DR too much.

Since I'm in edit having mis-hit a control key, I'll tell you how to do colour next.
--
Bob
 
OK, you need a colour wheel (in front or behind lens as you prefer. This is simply a coloured disc attached to a motor, maybe constructed from Cokin filters or similar. The Canoscan 700F has a slide unit, with RGB LED's. Taking the driver signal for this, and putting a phase locking servo on the motor, the correct RGB could be provided with an RGB wheel, provided the motor could run fast enough. If not, then you'd need a wheel like this (minus my drafting errors):





This modulates the red through two cycles per wheel revolution and the blue through three. After capturing the image, a software PLL needs to be used to recover the resultant chroma subchannels, which can then be used to recover a colour image. Given that we've got more than a gigapixel in the image, acceptable results should be recoverable.
--
Bob
 
A lot of people had some hopes for it but IIRC there were heat issues with the sensor or some of the other related components. But, they did have a prototype that was out there it just never went anywhere.

Probably because the major camera companies started making DSLRs that were comparable if not better. Plus, since no two camera manufacturers had the same film back, you would have to make a back for each type of camera. The sensor and electronics could be the same but it would still almost be a custom application.

Something you could look into was that first generation "digital" Nikon SLR. It had a digital back on a N90 or something.

In the past I have attempted grafting cheap digital compacts to back of my SLRs but its a little more complicated if you actually want to have real control circuitry that works WITH the camera. Not setting the camera to bulb mode to hold the shutter open, then activating the compact to take the picture. That's how I did it. Did not work out so well. I also tried various reducers for the image. Overall, poor quality.

--

Arthur Dent hoped and prayed there wasn't an afterlife. Then, realizing the contradiction, he merely hoped there wasn't an afterlife.
  • Douglas Adams, In Literature/Douglas Adams
 
Hi Zone8, the real point is that digital sensors perform best paired with telecentric lenses. Old lenses are mostly non-telecentric. IQ would be poor, unusable, except as curiosity in most cases, IMO.
Not true in practice. I use film-camera lenses on both my digi systems. On my Sigma DSLRs SD9 and SD14, I still use the older full-frame 15-30EX and 24-70EX lenses and they give superb results. I also use a couple of old M42 screw-fit film-camera lenses with adapters, again with excellent results (70-210 Tamron with Adaptall Pentax-K fitting that fits the Sigmas and a 30mm Lydith - superb quality lens that a dealer wouldchuck in the nearest bin. On my Kodak FF DCS 14n I use a Tamron 28-200, a Vivitat 28-100 and a Nikon 55mm Micro.Nikor (older f3.5). Only problem with the Tamron is for some subjects, it suffers from the Italian Flag syndrome.

--
Zone8

The photograph isolates and perpetuates a moment of time: an important and revealing moment, or an unimportant and meaningless one, depending upon the photographer's understanding of his subject and mastery of his process. -Edward Weston
http://www.photosnowdonia.co.uk/ZPS
 
Hi Zone8, the real point is that digital sensors perform best paired with telecentric lenses. Old lenses are mostly non-telecentric. IQ would be poor, unusable, except as curiosity in most cases, IMO.
Hi John.

As two others pointed out, this is simply not true, and even Olymous, the people who raised the telecentric lens argument in the first place, no longer mention it.

A slight offset in the microlenses on FF sensors is enough to achieve optimal performance on existing film lenses over the range from ultrawide to telephoto. Nikon and Canon employ such an offset. No offset, at all, is necessary for smaller formats such as APS or four thirds when using lenses designed for SLRs (exit pupil location no closer to the sensor than 1.33x the mirror swing, relative to FF.

A greater offset allows exit pupils closer to the sensor, as used in both micro-four thirds cameras and Leica rangefinders. The Leica M9 and its associated very, very non-telecentric lenses leads the industry for optical performance.

That's a big part of what led Oly to abandon telecentric lenses and adopt the micro four thirds system, the clear superiority of lenses that reduce the degree of retrofocus design (the opposite of telecentric design, which is typically more retrofocus than conventional "SLR lenses").

--
Rahon Klavanian 1912-2008.

Armenian genocide survivor, amazing cook, scrabble master, and loving grandmother. You will be missed.

Ciao! Joseph

http://www.swissarmyfork.com
 
A lot of people had some hopes for it but IIRC there were heat issues with the sensor or some of the other related components.
Nope. They didn't have a sensor to have heat issues. The only real issue was the "business case". The product made so little sense that they couldn't get the funds to build it.
But, they did have a prototype that was out there it just never went anywhere.
They had what we call a "mechanical prototype". If you dig through the paperwork on Silicon Film (Irvine Sensors, ImageK, Side, Voyager One, etc) you'll see that the only thing that they spent any serious money on was the services of a machine shop to make mockups out of solid metal and plastic.

The neat thing about the original concept, put a cartridge in a camera, shoot 100 pictures without being able to review them on an LCD, then pop the cartridge in the reader, download to the computer, and then get to see the pictures, is that you could...
  1. shoot 100 pictures of your "booth bunny" in a particular outfit against a particular backdrop with a competitor's camera.
  2. bring that booth bunny, in the same outfit, with the same backdrop and lights, to a trade show.
  3. shoot a hundred shots at the trade show
  4. show everyone the hundred shots on the computer
And no one would ever catch on that your product didn't actually work

Later, after several name changes, and the most fascinating accounting magic, where they sold their own company to itself, and tried for more funding, they developed a new concept, where a sensor board went inside a camera, but was connected by a thin ribbon cable to an electronics box under the camera (think Leica Modul-R, Kodak DCS-630, or other "digital back" for 35mm SLR solutions). They hired Applied Color Science to take a FillFactory sensor, bolt it on to the back of a camera (no cartridge) and bolt a rack (a "cage" with multiple circuit boards in it) 4x the size of the camera to an SLR, they "demonstrated" the concept. But they still had no business plan for reducing electronics that were bigger than the camera down to a unit that could go inside
Probably because the major camera companies started making DSLRs that were comparable if not better.
That's what totally destroyed the business case that would get these things to sell in large enough quantities to make them affordable.
Plus, since no two camera manufacturers had the same film back, you would have to make a back for each type of camera. The sensor and electronics could be the same but it would still almost be a custom application.
I did a proposal for another company where the sensor and electronics pod "snapped" into a molded back. The expensive part, a little kit with sensor board and electronics pod, would get stocked by camera stores as one item, and backs for a variety of popular cameras got stocked as a much lower cost item.

And I told the client that, by the best math I could work out, the product was not marketable.
Something you could look into was that first generation "digital" Nikon SLR. It had a digital back on a N90 or something.
The first actual Nikon, the D100, was a very sophisticated camera that used a mix of F5 and F100 subsystems, a new cast chassis, and new clamshell and bottom plate units, and tight integration between F100 processor "top board" and D1 main board. That was 11 years ago.

You're thinking of one of the old DCS Kodaks, which Nikon helped a bit on, but were pretty much a Kodak effort. Kodak did them with the F3 (really), F4, F5, 8008, N90, and Pronea (honest, they did).
In the past I have attempted grafting cheap digital compacts to back of my SLRs but its a little more complicated if you actually want to have real control circuitry that works WITH the camera.
The 4x-6x crop factor would add a certain amount of suckage to the concept, too. The 20mm f2.8 becomes a portrait tele. A viewfinder mask with a tiny window in the center would make using it pretty miserable, too.
Not setting the camera to bulb mode to hold the shutter open, then activating the compact to take the picture. That's how I did it. Did not work out so well. I also tried various reducers for the image. Overall, poor quality.
Yup. reducers are tricky.

--
Rahon Klavanian 1912-2008.

Armenian genocide survivor, amazing cook, scrabble master, and loving grandmother. You will be missed.

Ciao! Joseph

http://www.swissarmyfork.com
 
I remember some reports of them actually having a working prototype. But I can see how easy it would be to fake it for venture capital.

I also remember them having entered some contest and scoring high. Maybe that was the second generation one with the external circuit board pack.

Originally, I tried a fiber optic reducer to shrink the image. It did not work out well.

Pretty cool to hear it from someone in the know.
--

Arthur Dent hoped and prayed there wasn't an afterlife. Then, realizing the contradiction, he merely hoped there wasn't an afterlife.
  • Douglas Adams, In Literature/Douglas Adams
 
There are no wikileaks from the inside of that company, so it's hard to prove a negative. All we really know, is that they never let anyone (like Phil) review one.

At the time (early 2000s) DSLRs were scarce, and the prospect of turning all the existing SLRs into digital cameras was appealing. If they had anything working, even at the 2 megapixel range with an APS sensor, they would probably get a favorable review from DPReview.

That they didn't suggests but does not prove that they didn't have anything worth checking out.
 
Thanks, I learned some things from your replies :)
--
JohnK
Take a picture, it'll last longer.
 
Suddenly started fantasising about cramming a CMOS sensor into the film compartment of my old Agfa
Hi Hakapikka, instead of cramming a digital sensor into your old Agfa, which does not seem feasible, why not cram the lens from your old Agfa onto a digital camera? The end result would be the same. The only requirement is that the focal distance (flange to film/digital sensor) of the lens be equal to or preferably greater than 44mm. I'll share with you a related hack I made. My first camera, back in the day, was a Baby Brownie Special. When I saw one in a flea market I got to wondering if I could put its lens on my 5D. So I did, using a body cap. I was even able to find two lens accessories online, an original close-up adapter, and an original bakelite lens hood. The focal distance came out perfectly and it works great either on aperture priority or manual. Of course it only uses the center part of the lens circle, but that's the sweet spot!



Here's a sample pic, has a very "Holga" quality to it, don't you think?



--
JohnK
Take a picture, it'll last longer.
 

Keyboard shortcuts

Back
Top