36 comments on “Scanning 35mm Black and White Negatives with the D800E

  1. Pingback: DPUG.ORG

  2. Pingback: google

  3. I’ve been using my D300 and 55/3.5 Micro-Nikkor for several years for “scanning” B&W, after my “real” film scanner went south. It works exceptionally well, and I’ve felt that I was near, but not quite up to, the resolution needed to get everything that’s there, so it’s great to see your examples. You’ve come to basically the same process I’ve arrived at, too, in most respects.

    I tried several enlarger lenses and didn’t get particularly good results and finally arrived at the Micro-Nikkor as the best thing available to me. The Rodagon-D has been on my list of possible things to get if I get a higher res camera, too.

    My results are most of the B&W on my flickr pages: http://flickr.com/mdarnton

    • Michael, it looks like you are getting good results from your setup (http://www.flickr.com/photos/mdarnton/7183241686/). I like your setup with the enlarger column and the trick of using the mirror to align things. I will have to try that. That should make my future attempts at using this for medium format scanning much easier. Thanks.

      I have also acquired a 60mm AFS micro lens. This won’t work on the bellows, but it will allow 1:1 reproduction. With a setup like yours I could easily test it. I have a feeling it may be as good as the Rodagon D, based on general use.

  4. I see in your APUG comments that you didn’t get good results with the Micro. I found it worked better if I shoot the emulsion side of the neg, letting any slight curvature of the neg work with the field curvature of the lens, rather than against it.

    • That’s an interesting idea. I think the bellows hold the film strips pretty flat, but the slides all have some flex to them. I hadn’t thought about checking if one way is sharper than the other.

  5. Pingback: Scanning negatives - which side up ? - Leica User Forum

  6. Pingback: DPUG.ORG

  7. Pingback: DPUG.ORG

  8. Hi,

    What an excellent, informative and encouraging site.

    I am about to embark on trying to scan B&W Acros 100 120 negative film, using a Nikon D800 minus its anti-aliasing (low pass) filter. I aim to grab as many of the grains my Hasselblad 500 ELX has recorded as I can, by digitising the 60x60mm 120 film to print on my A3+ HP PHOTOSMART PRO printer.

    Still waiting, as I am, for D800’s to drop further in price, with the advent of the D810, I am pondering the optimum choice of manual lens for the job. Thanks to your input I would plump for the PB5 Bellows also as anything else seems too flimsy.


    • Ernie, I haven’t done much work with medium format film and the D800E since I don’t have stable rig setup to hold the film. So I usually just go straight to the drum scanner with film larger than 35mm. My guess is the various Nikon Macro lenses will prove to be excellent for medium format film, but I haven’t tested them in any controlled manner.

  9. I have a Beseler Dual Mode slide duplicator with bellows, which would make zeroing out color negative bases (anybody know the starting filter pack of my moldy 1964 Kodacolor?) easier by doing a custom WB on clear film base, then dialing out before the shots.
    I also have a Polaroid MP4 with a tripod head; (usually hang a 2.25 graphlex off it to copy) and a Bowens Illumentron I can use for silver B&W. In all cases, the heads natively handle 6×9; and it’s easy to mount a 4×5 mixing chamber above, should I ever feel the need. With this rig, I usually also hang a very long dual track bellows from the old Chromega slide duplicator (just to keep a full mix of every ’60’s/70’s junk in the mix). The bellows handle 35mm and and lens combo I can think of. I can also directly mount a camera body with anything from 55mm f3.5 Nikon Macros, 2.8AF, 60AF, etc. If I don’t like those results I’ll try the closet full of old non moldy apos, Wollys, G-Companons, Tominons, Boyer Saphirs, enlarging lenses, printing Ektars… and only then maybe spring for a new Rodagon-D 1:1.
    I’ve decided to wait until after the holiday season before picking up a 810, both to see if the price drops, and to see if any traditional Nikon recalls come out on the early units… Or Pentax drops the 645 price in half… 🙂
    I’m planning on using Negatrans, glass inserts, and whatever else it takes to keep film under control.
    I’m going to first test off of 1951 and other glass targets for lens flatness and resolution.
    The competition is against my Plustek and Epson 750 flatbed for speed and quality workflow.
    I also want to see if multiple 810 shots, offset by small x:y movements on the base, could be merged for any needed greater resolution or dust control. I have some microscope stage hardware I’m looking over now, to see how fine a repeatable control I can maintain.
    Cropping is another thing I’ll try with the old Tech Pans…

    • Joe, if you are using a DSLR you can get to the starting filter pack quickly by setting the white balance to daylight and including the film edge in the shot. Then adjust the filters until all three peaks on the right of the histogram line up. Takes a few shots, unless your camera offers real time histograms in live view, but it’s quick to do.

      I tried all the lenses I could before springing for the Rodagon-D. The Rodagon D is by far the sharpest across the frame.

      I imagine you could get a bit more resolution with your x:y offset method, but it would require much more effort. I don’t know if Photoshop can handle that type of operation or if you will need to write some code to do it. For most 35mm film I don’t think it will be worth the effort. Tech Pan and the like, if shot with good technique, could certainly benefit from more resolution.

  10. Pingback: DPUG.ORG

  11. Thanks for the great information. I am using a D800E, 55mm f3.5 Micro Comp Aperture, PK-13, and ES-1 slide copier/holder to scan slides. Focusing with the ES-1 is difficult since the unit telescopes in and out. I think I will try a Bellows since I have wanted to acquire one for some time. I hope you are able to post some information on color negatives in the future since I am very interested in scanning and inverting color negatives. Thanks again.

  12. Pingback: Renaissance Two-fer | Devin S Harrison : visual

  13. This is brilliant, and you’re the first I’ve seen/read on the web that has
    pursued this DSLR + Rodenstock 1:1 copying arrangement.

    After mucking about with various enlarger lenses I decided to take
    a risk and bought the Rodenstock APO-D 75mm f4 1:1 copy lens.

    My setup is similar, but not optimal like yours, I used (and still use)
    a DX-sensored Nikon D3200 (24MP), Pentax Bellows, and the APO-D
    75mm f4 set between f4-f5.6 aperture.

    I had to get a M39-M42 adapter made for the job, and was thrilled with
    the results that this produced. Even though the magnification ratio is not
    optimal, but somewhere around 0.63, from memory.

    At the back of the neg-holder I’ve taken off the opaque diffuser and then placed
    a light-box of diffused light, and when I move this further away, the light becomes
    a little more collimated, than diffused.

    Owing to the less-than-optimal copy ratio, the lens probably projects a slightly
    contrast-flattened image, and added to that, the LPF of the D3200 also adds
    to the softening of the copied negative. However, all it rather subtle and can be
    corrected at the 24MP-copied stage by small radius sharpening, then down-sampling
    to 10MP-14MP for most of my needs.

    A question to you if I may – I am intrigued. What apertures do you have to stop-down
    to, for your FX sensor to get near-perfect flatness of field?

    Kind regards


    • That’s good to hear the lens works well with the DX sensor. I wondered if it would.

      I use f/8. I don’t recall if I did careful testing to find the optimal aperture.

      • Hi Larry,

        Thanks for the reply.

        With having a full-framed sensor, you could create the DX equivalent scenario
        by adjusting the sensor-lens, and the lens-neg distance to approximately a 0.67
        ratio. The focal length of the 75mm should be okay. You should have enough
        variability in your set up.

        Some points that have come to mind in general ..

        (1) Diffused light helps reduce non-emulsion scratches, the more diffused the
        more productive the scratch reduction is.

        (2) The diffused light also flattens the contrast, I’ve yet to have a problem with
        copying the negative’s dynamic range.

        (3) I use NEF to record the images, as JPG capture has a limited quantization
        of 2^8, ie eight f-stops. So if we are not able to exploit this, ie the dynamic range
        of capture is less, then the quantization (“tonal gradation”?) is reduced to less
        than 8 f-stops.

        I suppose one drawback to a machine like the D800 is the vast NEF files,
        and the computational time involved for ViewNX2 (or others) to decode
        this very large file!?

        Many thanks for your informative and intelligent article.


  14. Hi Eric,

    I had thought about testing the setup using the DX crop, but it would be hard to tell how much the loss of sharpness was from using fewer pixels vs. using the lens at it’s non optimal magnification. I guess to test that out you’d get a 24MP DX and FX camera and see, but there are other factors at play such as the presence or strength of the loss pass filter. If you are trying to fully optimize the idea it would be worth testing, but my main goal was to have a system that was quicker than a CCD film scanner with roughly equal results. The fact that it turned out better is just a nice bonus.

    I’m not sure for negatives flattening the contrast is ideal. With negatives only having a density range of about 4 to 5 stops we’ll get more tonal resolution with more contrasty light. But the scratch mitigation is probably a bigger benefit. In any event I think with my use of a large diffusion enlarger head and the bellows diffusion screen it’s about as diffuse as you can get.

    I use NEF exclusively. With Lightroom, on my system, the 36MP images aren’t seemingly slow to decode. If I use the Nikon software things seem to grind to a halt.

    With JPG being 8bits per color it’s really going to be an issue. To stretch the 4 or 5 stops in a typical negative out to provide a “correct” positive you are going to be heavily manipulating the curves. Without adjusting curves in camera (using a more contrasty setting for example) you will be throwing the least significant 4 bits away. This alone will leave you with much less room for creative tonal manipulations. But once the image is inverted the loss of tonal resolution will be in the highlights and not the shadows, making it even more noticeable.

    With either NEF or JPG you want to expose to the right in order to capture as much tonal resolution as possible. Just being one stop under exposed will cut the tonal resolution in half.

    Thanks for raising these issues.

  15. Thank you very much for this D810 “scanner” technique!
    Have you tried the same technique on color negatives and if so, what are the differences?

    • I don’t shoot much 35mm color negative. However I’ve tested color negatives a few times and I haven’t been 100% happy with the results. For the most part they seem correct, but some colors seem off after processing.

      I started using the color head as a way of dialing out the orange mask. It seems to work very well for this if you align the histogram peaks in the highlights (actually the film base) then a straight inversion gets the color roughly right. But in the few tests some colors didn’t come out correct, even when others seemed right. I wonder if this is the filters used in the sensor, issues with the film I was testing, or just simply my color balancing skills. Note that I sometimes have trouble getting color right when scanning with the drum scanner or printing in the darkroom. Without more testing I’m not prepared to say if it works well or not.

  16. Re Color Copying.

    My findings from memory (I have written all this down somewhere!) …..

    There is a distinct lack of effective blue content in color negs, so the blue channels needs more exposure or amplification; whatever’s available.

    Try this simple technique – under White Balance within your DSLR, go to Preset Manual and from there it will allow you to photograph a test color negative. It should recognize the bias away from blue or lack of blue in the image. Once you take a Measured shot of the test color neg, internally this ought to self correct to get a color balance. Normally, if you inspect your histogram, there’s a blue shift with low brightness intensity, but this will be corrected using your test or Measured shot, ie Blue gets amplified internally?

    Color correct in Photoshop or whatever as normal. This works exceptionally well for me.

    Check out my flickr pages and look for those shots in color, and see what you think.

    I have written material on this, but cannot find it at the moment, so this is all from memory.

    Best Wishes


    • Using the camera’s white balance doesn’t really change the sensitivity of the color sensors. Because the orange mask so heavily skews the color I think you would be better off trying to neutralize it during the exposure. That’s what I do using an old color enlarger head with its filters. It seems to work better than just using the camera’s white balance. If done right you only need to invert in Photoshop and then adjust the curves.

  17. Forgot to mention,

    finally, I assume you know how to invert a color neg in software. 😉


  18. Hi.
    It would be a couple of questions for you.
    But also I send pictures.
    Write your e-mail address, if you have time to help.
    Thank you in advance

  19. I have been using my D800 to capture BW negatives for a while. I use a 105 macro lens with the camera on a copy stand. The negatives are then photographed over an LED light box. I have several marks on the copy stand stem depending on the format 35mm to 6×6. The results are fantastic. I also have a preset in LR5 to invert the tone curve and convert to BW. I can capture a whole roll of film and be working on it in LR5 in less than 20 mins with no compromise with image quality. Some of my results can be seen on the link below..


    I have tried using a V850 scanner but to be honest the results with direct DSLR capture are so good I have ditched the scanner for good.

    Interestingly I also find that dust is less of an issue with the DSLR approach.

    • Glad to see your results mirror mine. I still find drum scanning worth it for larger film sizes, but for 35mm and 6×6 I’m usually satisfied with the DSLR. I too have noticed dust is less of an issue than with the flatbed scanners. I’m not entirely sure why, but I assume it’s related to narrow depth of field and uncollimated light.

  20. Thanks for this great article. I have been experimenting with several approaches for copying b&w negatives. I started with an Illumitran using various enlarging lenses. Results were really good, but it was a tedious process. The lens had to be opened manually to focus each negative and then closed down again (using an old Schneider enlarging lens that only opened to f/3.5 or so.) It’s difficult to mount my D610 on the Illumitran’s bellows unit. Also, I didn’t like the idea of exposing the sensor to so much dust from the bellows.

    Then I bought an ES slide adapter for my D610 and coupled that with a 60mm f/2.8 macro. I set the camera on a copy stand, put an LED lightpad under the ES-1, popped in an AC adapter for the camera and tethered it to Lightroom. I used an old scanner negative carrier that easily slides into the ES-1 to hold the film flat.

    I made my first big print (9×13.5″) print from this setup a few weeks ago and it looks great. One huge plus about doing it this way is that there was the huge reduction in dust and scratch repair needed vs. the Nikon 5000 scans. This method saves hours of work for b&w.

    I don’t like working with the images in Lightroom though, because even after inverting the image, all of the tone adjustments are backwards. I adjusted my scans to a point where everything looked fairly decent, then sent the image to Photoshop and saved it back as a new file. Then the sliders worked as the do in a normal image.

    This is still a work in progress for me. I’m not sure I will try it with CN or slides, because the Coolscan does a great job on the dust and scratches on those. I did a few CN “scans,” but they were difficult to process. Using a color head and doing the balancing tweaks mentioned above sound like a good way to go.

    Thanks again for this great article.


  21. Fantastic.

    Thanks so much for this brilliant write up. I’ve spent several evenings looking for something like this across the web, to no avail (until now). I have a D7000, and a bellows / slide carrier (PB-5 and PS-5) on the way in the mail, with the explicit purpose of digital scanning negs and some old slides.

    I have the 55 micro 3.5, and was planning to use that, but have been wondering if an enlarger lens wouldn’t be better (although the Rodagon-D is out of my budget range). What was your issue with using the 55 micro? And any advice for digital scanning with the DX camera?

    • If I remember correctly the main issue I found with the Micro lenses and the D800E was very slight softening of the corners. I no longer have a DX camera. However I remember the D7000 with the 55 f/3.5 Micro and 55 f/2.8 Micro both worked well, just at a lower total resolution from the D800E.

      The Rodagon-D was designed for the 1:1 reproduction only, so it won’t be ideal on a DX camera. You might get even better results with the Micro lenses, though I don’t remember if I tested that.

      • Hi Larry,

        Can you recall how ‘flat field’ the 55m f3.5 or f2.8 lens was? Could you afford to copy at say f4 or f5.6?

        BTW, and off-topic slightly, I have noticed that when copying colour negatives (with inherent orange base of course) I get a much better colour rendition and colour control if I convert my NEF files (from the DX Nikon D3200) to DNG using Adobe’s DNG Converter. Not sure why, unless it is me!?, but the whole experience is better. Perhaps there is a ‘stripping out’ of post processing camera settings by Adobe?

        Eric J.

        • Eric, I didn’t test the lens for field flatness. I know the negatives don’t lay totally flat in the copy attachment. If I remember correctly I found f/5.6 was where I started to to happy with the results, and I think f/8 was better, at the expense of overall diffraction.

          I’ve never tested NEF vs DNG. I’d expect the results to be the same given they’d go through the same Adobe program, but as you say some camera settings may be lost/ignored in the conversion. I’ll have to try it. I use the color enlarger head to remove the orange mask, so that also may make a difference. I’m almost positive it gives better results since you aren’t throwing away a lot in the conversion.

          • Hi Larry, thanks for the feedback.

            “I use the color enlarger head to remove the orange mask, so that also may make a difference. I’m almost positive it gives better results since you aren’t throwing away a lot in the conversion.”

            I agree with you totally, although I am happy with the results I get without enlarger intervention. Of interest too – Nikon and Canon differ in their rendition of RGB colors, so there’s another variable for us experiemnetors to consider!

            Slightly off topic, I did an experiment the other week, and inserted a colour neg into my Leitz V35. Just observing the orange mask, I dialed-out the orange mask as best I could to get a near greyish background. So a better (grey-ish) white balance can be roughly achieved from the start as you’ve probably experienced.

            Thanks and best regards

  22. Thanks for your reply. I received the bellows in the mail, and after a number of tries, found that the reversed 105 micro works extremely well with the dx sensors.
    Thanks again and all the best.

Leave a Reply

Your email address will not be published. Required fields are marked *