slide copier vs. scanner

Discussion in 'Digital Darkroom' started by ptkeam, Apr 12, 2007.

  1. Hi I have been looking at getting a slide copying attachment for my D70. If you've used one of these, please give me some feedback. How does this compare with a film/slide scanner???? Thanks
  2. jtk


    Browse P.N on this as there have been comparisons. Having done a tremendous amount of slide duplication onto film, including several duplication films and pre-flashed Kodachrome 25, using the ultimate optics and systems (Sickles duplication camera with S-Planar), a good duplicate *to film* can be indistinguishable from the original, and can often be an "improvement," with no perceptible loss of detail. However the D70 chip doesn't resolve as highly as film, and it's unlikely that your optic will be ideal. A copy stand, light box, and an old Micro Nikkor prime lens would probably perform better than any duplication attachment. Scanning with a Nikon scanner, which would also remove dust in most chromes (issues around Kodachrome) will IMO resolve even better than that Sickles/Planar setup. I do a lot of it and am astounded. The advantage to duplicating is speed, although you'll need to take slide dusting time into consideration. Actual scan time on my lesser Nikon V is 1.5 minutes with Ice...Ice takes about 20 seconds of that.
  3. [​IMG]
    Cotton Cafe, Hendersonville 1999
    ©2007 by Godfrey DiGiorgi
    Minox EC
    Agfa APX100 @ ISO 200, processed in XTOL 1:1, scanned with Nikon LS-40

    Click the image above for the full-resolution version in a separate window.

    I've been rendering film images into digital for about a decade and a little bit now, tried every combination of optical and scanner capture to do it. Several thousand 35mm slides and negatives at least, as well as other formats.

    While you can get decent results with a slide copier and camera, a film scanner is simply more reiiable and easier to setup, obtain consistently excellent results.

    The photo at the top of this response is captured via a 2900 ppi Nikon Coolscan IV scanner from a Minox subminiature negative, 8 x 11 mm in size. No image processing adjustments to the full resolution image were made (other than adding my watermark and border).

    While I can set up the a 10Mpixel camera with a macro lens and light box to do the job and produce a higher resolution digital image, it is nowhere near as consistent and effective to do the job. With a full-frame 35mm film image, the total pixels from the same scanner would be 10.7 Mpixels worth of data (this one from a tiny negative is about 1.1 Mpixels). The used Nikon Coolscan IV I bought that produced this was about $200, much less than the cost of a good macro lens, light source, etc required to do the job right optically.

  4. Here is a copied slide using a Pentax DL from 1963:
  5. John, I can't tell anything from a web rez photo. I need to see a 1:1 pixel copy, which is why I added that as a link from my scan above. G
  6. As already noted above, desktop film scanners have far more resolution as exhibited in these comparisons of results using a Coolscan 5000 at 4000dpi, 8MP Canon 20D with the 100mm f2.8 macro (optimal aperture, tripod, etc) and my 3200dpi Canon C8400 flatbed of the same frame of Fuji RVP. [​IMG]
    Click the image for full res version. It can't be emphasized enough that even though I shot the Canon 20D tethered to my PC, the setup, handling, image transfer and subsequent processing that needs to be applied to it take far longer then the Coolscan's less then 50 second per frame scan with ICE in full auto mode especially for more challenging frames of film - dirty, scratched, dark and/or dense.
  7. Re: slide copying, as you can see from Les's image the scanned slide resolves more detail than the 20D. You can use a digital camera as a copier but shouldn't expect the same quality- your slide will lose resolution that the digital camera can't resolve. You will also have to contend with removing dust from the slide shots manually. "It can't be emphasized enough that even though I shot the Canon 20D tethered to my PC, the setup, handling, image transfer and subsequent processing that needs to be applied to it take far longer then the Coolscan's less then 50 second per frame scan with ICE in full auto mode especially for more challenging frames of film - dirty, scratched, dark and/or dense." After only shooting and scanning film, I just bought a 20D so my comments are preliminary. Using Adobe Lightroom to process RAW (I don't shoot jpeg) 20D files, this route is far faster than film scanning. The transfer time per image is only a few seconds from my highspeed memory card via USB. Lightroom lets you apply the same edits to a series of files instantaneously. The "auto" button also works surprisingly well for many scenes. Scanning film introduces the variable of the film's exposure. You can create scanner-exposure dependent IT8 profiles and try to match them to slides based on their density (i.e. scan darker slides with more hardware exposure and try to get the histogram right) but this certainly takes more than 50 seconds per frame. Then add time to crop correctly, finish dust-busting, and then get into editing the image color. Les- if I recall, you had to manually change the color balance on your slide film map slides as they were far from the digital and the actual color, right? For mortals not using a scanner as fast as the LS-5000, each frame can be between 1-4 minutes just for the scan, let alone post processing. However the quality should be higher on a 4000dpi scanner than with a D70.
  8. Roger, you can confirm my numbers, but shooting the Canon 20D tethered to my PC takes about 15 seconds to transfer each RAW file - faster if batched per CF. Auto batch mode - in this case levels and sharpening, take another few seconds for about 20seconds per frame. However, if there are any blemish - scratch and dust, then cloning each of those will push these well over the 50 seconds it takes for my Coolscan 5000 to return a scan with ICE. Incidentally, adjusting color does not add any scanning time. Also, simply feeding in a frame strip of film in batch mode saves handling time as well. As for changing color balance of all the samples in my Film and Digital Relative Resolution Album, I believe all films have their own color characteristics but that doesn't mean the scan is inaccurate. As you can see, my Canon 20D results were not all that accurate without processing even with studio flash. I have found the scans straight out of the Coolscan 5000 in batch mode, no color corrections, are as accurate as the optical prints I get from the same frame. You can view the results using this scanner in my FILM2 Album which are - unless stated, mostly full, res "neutral" scans using only auto expose/focus ICE(not on b&w only) crop and rotation - all other settings OFF or neutral.
  9. Before I got the Coolscan 5000, I used to send my slides to be optically printed but they first had to copy it to Fuji interneg as they were not chrome prints. Below is the result of one such transfer. [​IMG]
    Click the image for full res version. The top inset is a 100% crop of the original and the second is a 100% crop of the scan of the interneg.
  10. jtk


    Peter..out of curiosity, why do you want to have digital versions of all those negs and slides? I ask myself that. My current thinking, fwiw, is that my time is better spent making new images. I have of course digitized the most interesting of my family pictures, printing archivally, thinking some family member will care in the future, the way I have with hundred-year-old family photos... family's unlikely to be interested in slide shows or negs in the future! But my own decision has been to edit harshly, and only digitize the best or most necessary. In my case thats 1% or less.

Share This Page