Is scanning at higher than 2700 dpi worth the cost?

Discussion in 'Digital Darkroom' started by vuthy_chrun, Aug 11, 2003.

  1. I recently acquired a Minolta Dimage Scan Elite 5400. I did some
    tests (you can read my detailed review and see some samples at
    http://www.photo-direct.net/cameras/Scan_Elite_5400-text.htm) and
    found that 2700 dpi is good enough for a 35mm negative or
    transparency, and that there is not enough difference (image
    sharpness) to justify the storage of your scan images in 5400 dpi
    resolution (at least 100 MB per image at 5400 dpi).

    I am interested in learning other people's opinion about the merit
    of scanning a 35mm slide or negative at a higher resolution than
    2700 dpi. If you think that it is worthwhile to scan at higher
    resolution than 3700 dpi, then please substantiate your opinion by
    showing us some scanned images.
     
  2. If you're wanting to print larger without having to interpolate then yes.
     
  3. Generally, I find you can get almost 4000ppi of detail from a good slower speed film. Color negative materials don't contain as much, particularly at higher speeds. Going beyond 4000ppi only provides extra useful information in a few situations with certain films.

    The main benefit of the 5400 IMO isn't the extra resolution, although at 4000ppi and lower it should be quite good, but is the increased speed and density handling of the scanner. It's a much matured unit over scanners of years gone by but the extra resolution just helps sell the thing.
     
  4. Perhaps I'll hunt down or create some scans at 2700 for comparison to scans at 4000 later this evening, but yes, there is reason to scan at higher resolution. If (as already said), you want to print larger than 9 inches across at 300 dpi, it's better to take that information directly from the original than try to invent it based on the 2700 dpi file. While there are certainly cases where 2700 may approach all of the information in the original image, there are also certainly cases where it does not. Though the results vary somewhat, tests do show that well-exposed film, taken with good equipment, can resolve much more detail than 2700. [5400 on the other hand, may well exceed the information present in many 35mm images, but that's for another time.]
     
  5. Ijust purchased the 5400, and so far it seems superb, after a couple of 48 bit/5400 dpi scans at 200+mb I ran out and purchased a 160gb HD!
     
  6. stb

    stb

    I have ordered a 5400 an am waiting for it. I currently own the original Elite
    (2820dpi). I ordered it because there were things my enlarging lense would show that
    the Elite scanner wouldnot. For me, that shows that there is more in my B&W films
    than the 2820dpi elite is able to see.

    Colour films do have lower resolution than B&W. Now, you chose velvia, one of the
    highest resolution E6 films.

    I have downloaded your TIFF files. I do not agree that comparison should be made
    with downsizing the 5400dpi file, because by doing that you obviously negate any
    advantage it could have. So I upsampled your 2700dpi file. I agree with you that the
    differences are negligible.

    I do not agree, though, that it meens that 5400dpi is useless. To me, your scans show
    that the limiting factor of this shot is either the lens or the focusing precision of the
    shot. I think of the lens bescause colour fringes are very present, betraying a lense
    that maybe cannot fully exploit Velvia.
     
  7. I certainly agree with the comments above about newer and hence better technology (less noise, faster, cheaper) and more scanned pixels allowing smoother enlargements. Also you might get less grain aliasing and hence smoother tones.

    I think that 4000 DPI does make a slight difference in terms of detail if you use a fine grain film (Provia 100F) and good technique. Otherwise, it won't. I'll try and get some scans to confirm/disprove this statement. (But I'll have to do this later when I get home from work.)
     
  8. Vuthy: I will make one comment on your review. You state that you should always scan with USM on. Many people would disagree. I prefer to scan with USM off. I can then apply high pass sharpening in Photoshop. I find this gives me more flexibility. Of course at the end of the day this is a matter of taste.
     
  9. I can see stuff in a loupe that my 4000 dpi scanner misses. Of course, the fact that I'm shooting Tech Pan may have something to do with it :)

    If my scanner can't see it, how am I supposed to upload it for you??
     
  10. Leif,

    You are absolutely right. I do use USM in Photoshop, not with the Minolta-supplied software. I will update my text to reflect this.

    In fact, I now do the following routine religiously when I scan a negative or a slide:

    1) Minolta: manual focus adjustment.

    2) Photoshop: rotate and/or crop whenever necessary.

    3) Photoshop: Usharp Mask.

    4) Photoshop: Auto Levels.

    5) Photoshop: Auto Constrast.

    6) Photoshop: Auto Color.

    7) Photoshop: Manually color balance (if auto color is not satisfactory).

    8) Photoshop: tweak with the 'curves' if still not statisfied with the visual result.
     
  11. Because a few people had said that my tests could be flawed by the use of a low quality lens (Nikkor 200/4 circa. 1972), I did another series of tests using a Micro-Nikkor 55mm/3.5 on Velvia 50 film.
    Please click HERE to see the new set of tests.
     

Share This Page

1111