Mitchell Feinberg’s 8×10 Digital Capture Back

- - Photographers

When I talked to Mitchell Feinberg recently he mentioned that he owned the world’s largest non-scanning color sensor array, something he created so that he could continue to shoot 8×10 film without Polaroid. Normally I avoid anything to do with equipment but this sounded interesting.

APE: Tell me why you created the 8×10 digital capture back and how it works?

Mitchell: When I look back, three years ago, it was crazy that I even tried to do this: design and manufacture the world’s largest color capture back, large enough to cover the 8” x 10” format, so that I could continue to shoot like my glass plate-carrying brethren a hundred years ago.

It was a race against the clock, or, specifically, a race to see if my stocks of 8×10 Polaroid would run out before the back was completed. In the early months of investigation most of my evenings were spent on meandering Internet searches. Months more were spent deciphering unintelligible technical papers. The few companies with the right technical expertise were eventually identified, but it was extremely difficult to be taken seriously. The experience felt like working at a call center, making unsolicited calls for storm windows. Eventually, one firm was convinced, and, after over a year of difficult design work, a first prototype was delivered in February 2010. The first production model was delivered about 9 months later.

The Maxback, as it has been named (the Brontoback, Velocicaptor and Back Scratch Fever were rejected), is the largest non-scanning color digital back in the world, with a capture area of over 8×10 inches. The largest commercially available color digital camera backs are about 4.5 x 6cm in size. It attaches without modification to a Sinar, and delivers high quality interim captures in under 30 seconds.

I use the back while I am working. Once I am happy with a photo, I flip off the back and then shoot a couple of sheets. In this way, I have the quality of 8×10 film, and the immediacy of digital capture. Crazy, right?

Well, the idea is not that crazy but I’m guessing the cost was astronomical. I shouldn’t ask but can you give me an idea on that?

The development and production of two backs (I wanted to have a spare) was equal to the cost of a good size house – before the housing crash. I know it sounds insane, but the financials on it are not so bad: I used to shoot on average 7.5 Polaroids per photo, and I shoot between 400 to 500 images a year. That’s at least 3000 Polaroids. At 15 bucks a pop. Or about 50K per year, minimum. Polaroid was at one point my highest single cost. I am depreciating the back, charging clients for its use, and I was eligible for the technology investment credit. I also took out a loan based on the projected income from the back, so I did not have a huge hit on my bank account. It is certainly not a fantastic rate of return, but the back is designed to last a very long time, so it should generate a strong profit over the long term (And that is not including the all-important photo-related issue that my clients love receiving 8×10 film).

The engineers and I discussed selling them, but no one wanted to bother with customer sales and support. I think there are maybe a dozen of so photographers who might have the desire and resources to buy one or two (I have two, so that I have a spare handy). This means we would not sell enough to start a proper production line, and it would be tricky to order small quantities from the sensor foundry, not to mention the main boards and other critical parts. It’s straightforward to make prototypes and hundreds of units, but five is a difficult number from a production/manpower standpoint.

We never set a unit price, but it would be in the low six figures. Anyone purchasing a device for that kind of money would expect excellent tech support, which implies that we would need to have backup devices ready in case there was an equipment failure. That would be costly. If I had an order tomorrow for ten of them, we could probably move forward with it, but it does not make much financial sense to pursue sales on a one-off basis.

There Are 154 Comments On This Article.

    • Hacked2123

      if its 10MP, its going to be small no matter what…the only difference he gets is that light isn’t being focused through a lens (which would essentially shrink it) This lack of shrink my result in a slightly more in focus, slightly more vivid color range, and more precise 1 to 1 ratio with a traditional camera.

      (I’m not a camera/photography guy…I just know a lot of tech and science.) (Of course I could always be wrong…but not on the matter of the 1Gb thing.)

      • “if its 10MP, its going to be small no matter what…”

        thats the only correct statement in your post

    • James Locke

      I would have to suggest a higher file size as a scan from 6X6 neg using an imacon is just under 400mb, forgot file size of my 4X5 scans with the imacon.
      Second condsideration, where do you store file and what size with growth when photoshop’ed? What am I missing that I don’t understand about this capture device? 10MP? as per statement my Hacked 2123 … If this is a “camera” back, light has to travel through a lens, my file size with full frame (35mm) on average is 36MP ,,, what am I lacking to understand?
      james

    • Ha!! I totally agree with the Spinal Tap quote. I guess if you have a lot of time on your hands and money to burn an 8 x 10 back seems perfect.

      I’m currently working on a light strobe that delivers f 90 from over a mile away! Any takers?

  1. How cool is that?!?! One thing I really miss about true large format are the lenses and being able to manipulate the focus. The small backs just can’t give you the swings needed to achieve similar focus results. I’d love to see sample images and file size.

  2. I would think certain intelligence agencies would be interested… but I suppose they already are using their own custom large scale sensors.

  3. This makes total sense to me, but it sounds like he’s still shooting film for his final files and using this for his tests, replacing polaroids. I wonder if that means the image quality is not up to par for final work?

    I’d personally be happy with a true 6×7 sensor for the RZ, or hell even a true 60x45mm sensor. Can’t replicate the look of those longer normal lenses on 35mm or 40x50mm that we are all shooting with now.

    • Even with all of the advancements in technology, I don’t think there is anything out there that can beat the detail, tonalities, latitude, and gradient detail that you get from an 8×10 slide and a quality drum scan… Plus, since this was an original and no sensors have ever been as large, I’d wager that the quality of the image from the sensor would still have room to improve. Canon, Nikon, Leaf and Sinar all have HUGE R&D budgets and their backs improve at insane rates (so much so, that you have to replace a body or back every 1-3 years???).

      That said, his motivation for development was his limited supply of 8×10 Polaroid and likely a bit of innovative drive. The numbers make sense financially for his business and he gets some pretty cool bragging rights.

      Well done Mitchell.

  4. JP–645 film has an image size of 56mm x 41.5mm. 6cm x 4.5cm is the nominal size of each frame with the complete rebate included. Current MFDBs have a sensor size of approximately 53mm x 40mm. The difference between the two is inconsequential–unless you were meticulous in the dark room, the simple act of cropping out the rebate for a film shot would lose you at least the 3mm x 1.5mm difference in picture area.

    • Good point re: 645, thanks! But there’s still a gulf between that and using a 110mm normal lens on an RZ, hard to replace in my opinion.

      • I hear ya. I would sell everything I to buy for a 6×9 digital back with moderate resolution (say 1500 or 2000 pixels per linear cm) and a user replaceable IR filter..

  5. Spinal Tap indeed. Did I forgot to mention that the back focusses past infinity ?

    A number of people have emailed me regarding the back’s specs. The device creates images a bit over 10MP; when cropped to correspond to 8×10 the final image size is 3285 x 2611. The image is 16bit, in RGB. Quality is excellent, due in part to the large pixel pitch. I am currently on vacation, so I can not post any examples. The image quality is not exemplary, but similar to a very high quality amateur camera of similar resolution. I do not use the back for final art; it simply does not have enough pixels to go to print.

    – M

    • I miss 8×10 Polaroid as final media – especially the European stuff.
      While I know it does not provide the exact same optical rendering or camera adjustments, couldn’t a reducing back and 4×5 Pol or (creation of) custom 4×5 dig back be used at a fraction of the cost – especially if final media is still film? Interesting capture device regardless.

  6. Gordon Lafleur

    Hopefully the Chinese are working on an ebay knock-off for under a hundred bucks.

  7. Frank Petronio

    Heck you might just as well put in an film processing line and shoot real film as tests, like they did before Polaroid.

  8. Very cool – I totally get this – he can still have the 8×10 look with this back. Like many photographers, I would love to know the technical specifications so I can drool over them. What is the file size? What are the pixel dimensions? (16,000 x 20,000)?! Did they have to invent the processor system for it from scratch or did they modify an existing system? 16 bit capture? I’d love to know all that stuff. I’m a techie geek sometimes and I’m not ashamed.

    • Ohh I didn’t see Mitch’s response in the comments.. Hmmmm so this back is used just as a preview.. “digital Polaroid” so to speak. Neat nonetheless. So Mitch when does the version come out that produces final capture (instead of film)?

  9. @Darius Davis
    The film is better than digital argument is really getting old and lame now- cant believe people are still saying it in fact.

    @Ber Murphy
    “As a shooter of 8×10 film, just like Bartlett, I’m jealous!!! Knew it was coming though.”

    It never came, because it will never be released, this was just an experiment undertaken by one very motivated photographer, unless he sells his spare or he dies and both get sold(unlikely because they will more likely end up in museums)

    Only he will have one(or two as the case me be) :-)

    8×10 is way to niche, for something like this to become commercially viable.

    • I don’t believe that 8×10 is niche at all. What is happening in my opinion is that people don’t get into 8×10 because of the lack of support. True photographers have always pushed the limits. Folks that shoot 8×10 film have to have a lot of time and expertise to nail the exposures. This gets to be an expensive endeavor, but lets face it, the best landscape prints came from 8×10. The difficulty in doing it in film, and lack of a digital large format back, and laziness that digital (35 and 645) has given us brought us these conditions. I know plenty of pro’s that would sell the house to buy one of these, but in the end it may not be worth it, film is cheap and pro’s that shoot it know how to use it.

      In order to separate ourselves from the masses is difficult things, and this is just the thing that somebody will pounce on. Also in play is that people simply figured that this back was impossible to manufacture. Now that it is known that it isn’t, watch as companies trip over themselves trying to come up with a commercially viable one. I don’t consider the idea niche, but thought of as impossible.

      • Its not only lack of support that puts people of from using 8×10, its is the fact that using 8×10 is ridiculously expensive too :-)

    • Well I have only just started to say it, a while back I would of argued digital beats film. It depends what you want out of a picture, if it’s just plain resolution then digital wins. If you would rather have nice juicy colour and lots of tones, Kodak Portra scanned at 48 bit and saved as a dng wins.

  10. I wonder if a 4×5 version of this back would be more salable, as I, like the rest of us, am also running out of Polaroid. Kudos on taking the initiative to push the envelope when the experts say it cannot be done.

  11. Wow! So when is this going to hit the market ! I want one!
    congrats on taking matter into your own hands. looks like Necessity really is the mother of invention!!
    best to you …keep it in focus….

  12. Chaz Miller

    I want more details on how the sensor itself is made. Is it a normal silicon CMOS sensor, or is it printed on some kind of polymer, or is it organic, or what?

    Because 8×10 is so big you can’t make a normal silicon sensor even if you use an entire 30cm wafer.

    • Joseph S. Wisniewski

      Could be a normal silicon sensor, on a single 30cm wafer. The low resolution combined with the huge size just scream that it’s a medical imaging sensor, and they do make those that big. Actually, about 8.5 inch square, but with what he’s doing with it, that last little bit isn’t important.

  13. I love this! It’s Quixote meets Tesla…on a champaign budget. There is a physicality to working in 8×10 that is hard to give up – and I would have loved to have seen a sample. What really rocked me is his website, excellent, excellent understanding of light.

  14. Wow, Mitch has done it, made an 8 x 10 Digital Capture back. It CAN be done. This is much more significant than just having an expensive exposure meter or replacing Polaroid. It opens the path to using large format lenses and the movements of large cameras without dragging around film holders.
    This means that a 4 x 5 back is possible too. The prospect of having a digital back for a Graflex with an ƒ/2.5 Cooke lens is very interesting. No 35mm camera can make that image, not even with a Noctilux.
    How about a 6 x 8 cm back for a Fuji GX680? Now that would be neat, just 15 MB would be fine.
    Thank you Mitch for blazing the trail. You know why you did it. Your images tell us and I understand. Thanks also for showing us what it looks like.

  15. The next digital back advancement should be increasing the scanning back exposure speed. A focal plane shutter works on the same principal. The CCD could be more cost efficient than a large array.

  16. Kudos to Mitchell for the dedication and finding a creative… albeit pricey way to extend his stock of 8×10 polaroid film.

    For those going on about 4×5 digital back should be brought to market… should really check out http://www.luminous-landscape.com/ which has been covering digital backs for medium format photography for a number of years now. If you want to go MF, it’s here and available now. Just need to fork out the $$.

    I can appreciate the benefit of a 8×10 back: VERY selective depth of field and VERY wide angles possible. I cannot help but wonder, however, whether or not there could have been another way to do this in a more cost effective manner.

    All the same, hats off for a great project.

  17. it only does 10mp? i’ve read on other blogs.. And he only uses it for test schots.. How ridiculous is that ? get a 8×10 scan back from betterlight or whatever.. and it only costs 50 K i’v got a few… and they shoot at 800 mp’s …

  18. Now – if we can get one in 4×5 with a resolution up around 80-100mp – I’ll be a happy camper.

    Anyone manage to get the name of the engineering company he used? :D

  19. Ian Tester

    I don’t understand the need for this. Why was the use of Polaroid previews so necessary that he needed to make this? I thought they were used just to test exposure and maybe colour rendition. Why can’t he just use a ground glass and a good light meter? And bracket his shots.

    Kudos though, for continuing to shoot film.

  20. A few observations about the technology, economics, and marketing potential for this back:

    1. The cost of these backs is a function of yield: I.e., how many backs pass QC as “salable” compared to flawed units that go to the trash heap. If these backs are marketed as Polaroid replacements –rather than for final image capture– as Mitchell uses them, then it’s not essential that only “perfect” backs go out the door. If you’re only proofing your shot before committing to film for your final image, then who cares if there are a number of dead or stuck pixels? That increases yield, and that pushes the cost down. That can make the difference in seeing these become mass-produced for 4×5 & 8×10. We want (potential) manufacturers to realize this important economic consideration.

    2. While many were drooling over the prospect of 4×5 and 8×10 backs that deliver pixel densities rivaling that of their dSLRs, it’s not cost-effective to produce such a back. They would be, by necessity, final image backs, would have to be perfect, and the yield would therefore be SO low that they would “be equal to the cost of a good sized estate in Malibu”. (You’d also wait all afternoon to view your image post-capture, and the back would have to include its own hard drive & gigabit networking.) Mitchell did the smart thing: bankrolled a back that had just enough pixel density to effectively replace Polaroid. Film still reigns as king. :^)

    3. Night shooters should realize another advantage to Mitchell’s choice: Higher light sensitivity for capturing low-light proofs. Making large (~30 micron square raw) pixels has another advantage beyond higher manufacturing yields: Larger pixels are more light sensitive, and so should result in otherwise shorter exposures for capturing low-light / night shots.

    This back is a big deal… Well done & many thanks to you, Mitchell!

    I look forward to seeing how the industry reacts to it, and I hope they take the above points in consideration, as they all encourage mass-production from economic and marketing standpoints. The key thing is to start with just a Polaroid replacement –as this is– and progress beyond that later as view cameras stage their comeback. This technology will make that possible — and Photography itself will benefit for it.

  21. While a marvelous technical achievement, kudos, we need to get get some facts straight. First, this is a matrix array sensor, which mens you need to de-mosiac the image with an ASIC, which means 1/3 of your data is fake like all matrix array chips (I’m assuming Foveon didn’t build this). It’s why we are still using a tri-linear CCDs today — no fake data. Next is the micron gap between photo sites on the silicon; probably huge by todays standards, why does that matter? The bigger the gap, the softer the image. Then there is the issue of limited depth of field — a curious scientific conundrum of digital imaging that due to light hitting the sensor at an angle (all photo sites on the sensor but the center ones), you are subject to the cone of confusion being amplified. The bigger the sensor, the more of this unwanted effect you get. Unlike film which does not care what angle the photons of light arrive at to expose silver halide or other photosensitive materials. With a brief collaboration of my colleagues, we could go on for 2 months non-stop to discuss why this isn’t going to be good as film — assuming a Tango drum scan to digitize said 8×10 film. I know a few will disagree with me — you know who you are. At the end of the day, this guy could have hooked up an 8×10 flat bed scanner to the 8×10 rear standard and use it in transparency film mode treating the ground glass image as an illuminated transparency — that would have been cheaper and probably better. But I’m still impressed, none the less.

  22. Amazing. I would gladly go back to shooting large format with something like this.

    A question: I had heard somewhere, perhaps incorrectly, that digital sensors didn’t handle rear camera plane movement, as they couldn’t resolve light at an angle that well. Do you find that to be the case if you angle the back plane?

    Thanks!

  23. This is great news! film 8×10 will always be king!

    you all should check out le negatif mag (a FREE online magazine dedicated only to FILM photography)

    just google it