Page 2 of 2 FirstFirst 12
Results 11 to 14 of 14

Thread: Max scanning resolution

  1. #11

    Join Date
    Nov 2003
    Posts
    1,219

    Re: Max scanning resolution

    It should be kept in mind that the resolution resulting from multiple steps in the process is LESS THAN that of each of the steps alone. That means that even the best possible scanner will reduce the resolution you end up with, but that may be only very slightly if the scanner is good enough.

    There are two popular rules of thumb for calculating the combined resolution which are used in practice. If you start with the component resolutions about equal, then either rule tells you the resulting resolution will be significant reduced. When you scan a large format negative or transparency, there are at least three resolutions: that of the lens, that of the film, and that of the scanner. The first two will already reduce the resolution to something less than 50 lp/mm (perhaps much less). So you want the scanner to resolve significantly more than that if possible so that the scanning step makes only a small further reduction.

    4800 ppi is the same as about 189 pixels per mm. But you have to divide that in half because you want LINE PAIRS, each of which requires two pixels. That gets you down to about 94 lp/mm. That would be large enough not to affect the final result enough to matter, if the scanner actually delivered that. Unfortunately, there is one more complication. The scanner rating tells you just how many pixels it collects without telling you the quality of those pixels in terms of resolving detail. Thus, a 4800 ppi scanner in reality might only deliver say 40 lp/mm. Scanners differ significantly in how close they come to providing the theoretical resolution the scanning rate would suggest. Typically, for example, a 4000 ppi scanner for 35 mm film will do better, for example, than a 4800 ppi scanner which can do large format. That is because it is easier to mainatin optical quality over the smaller format.

    Unfortunately, if you scan at extremely high sampling rates, you get very large files, which require a lot of memory, a fast processor, and lots of storage space. So there are practical limits, as well as cost, in using a very high scanning rate. Also, you can improve the apparent sharpness of the image by various digital techniques. So unless you are making wall sized prints which you expect people to view from close up, you need not use such high resolution scanners. I am comfortable with what I've been able to get from my Epson 3200 scanner, at least for the present. I can produce a 16 x 20 print which will look sharp as long as the viewer stays at normal viewing distance. Some day I wlll get a 4800 ppi scanner, but I won't bother going beyond that.

    But, if cost is no object, and you can manage the enormous files you would end up with, then get the highest scanning rate you can get, and make sure the optical quality of the scanner is high enough that you get something close to that scanning rate in practice. If you do that, the resolution you end up with will be determined primarily by the lenses and film you use, and you will have eliminated the scanner as a limiting factor in your work.

    P.S. The high scanning rates provided by many consumer grade scanners are meant to be used with medium format or 35 mm. In practice you might not need to highest scanning rate for 4 x 5.

  2. #12

    Join Date
    Dec 1997
    Location
    Baraboo, Wisconsin
    Posts
    7,697

    Re: Max scanning resolution

    Quote Originally Posted by Bruce Watson View Post
    Oh, I dunno. How about Max scanning resolution? That'll turn up this thread anyway.

    There are few questions one can ask that haven't already been answered. Happens to us all. Search is your friend.
    I don't remember anyone asking what camera to buy for their first LF camera - not today that is.
    Brian Ellis
    Before you criticize someone, walk a mile in their shoes. That way when you do criticize them you'll be
    a mile away and you'll have their shoes.

  3. #13

    Join Date
    Sep 2005
    Location
    Germany, Aalen
    Posts
    849

    Re: Max scanning resolution

    In general I think it is rather simple. Even if you scan your film beyond expected resolution of your lens/film combination, it looks better in a large print as when you just extrapolate with some software. I will better add SOME information (even though it will be mostly film grain and random structures) than fully information-less extrapolation. I also think that scanner noise is easier to fight with if I do not have to enlarge it.
    Matus

  4. #14

    Join Date
    Aug 2008
    Location
    Berlin, Germany
    Posts
    28

    Re: Max scanning resolution

    There're several approaches, and some were already mentioned in this thread:

    - Use the highest optical resolution of the scanner to get the most information out of the film. But at very high end scanners it can lead to extremly large file sizes without any gain in image detail compared to a lower resolution scan. So this approach might not be very practical in some cases.

    - Calculate the resolution based on a 300-400 dpi print of the maximum size you think you'll ever want for this image. This is the most simple approach. Even if your image doesn't have that much information the print will look more natural with enlarged grain instead of interpolated pixels.

    - Choose the resolution by the image detail on film. If you know you shot this image at f/22 you know that no matter how good your lens is diffraction will not give you more than 70 lp/mm. Knowing that the scanner and in fact all raster media resolve the image better in scanning direction than diagonal (45°) to that direction you should chose a resolution to let you capture the potential 70 lp/mm in diagonal resolution. For 70 lp/mm in scanning direction you need at least 3556 ppi, multiply this by the squareroot of 2 to make up for the diagonal loss and you have your input scanning resolution of 5029 spi. You might want to add a couple of spi to be on the safe side when it comes to sampling artifacts. An easier way to calculate a safe ball park figure is to take the resolution limit of your lens and film combination for that image in lp/mm multiply it by 75. In the diffraction limited f/22 example you would get 5250 spi. It's a large file, and you'll need a scanner that can optically resolve such detail but you can be sure there's no need to scan that film again for image detail. In many cases it's overkill compared to what might be really captured on film. For f/32 the number would be 3750 dpi and for f/16 7125 spi in case of diffraction limited performance.
    The best large format lenses I tested could resolve just short of 100 lp/mm in the center on film at f-stops wider than 11 but at the expense of very soft corners.

    -Dominique

Similar Threads

  1. Estimating resolution lost during scanning ?
    By Ron Marshall in forum Digital Hardware
    Replies: 36
    Last Post: 22-Mar-2009, 23:02
  2. Scanning negatives resolution
    By bounty in forum Digital Hardware
    Replies: 17
    Last Post: 4-Dec-2007, 20:18
  3. Sprintscan 45 Ultra Max Resolution - 5412 dpi??
    By Scott Rosenberg in forum Digital Hardware
    Replies: 4
    Last Post: 21-Mar-2007, 13:19
  4. Scan at Maximum Optical or Stated Resolution?
    By Brian Ellis in forum Digital Processing
    Replies: 15
    Last Post: 11-Oct-2006, 07:55
  5. Resolution limits of prints
    By paulr in forum Digital Hardware
    Replies: 18
    Last Post: 27-Sep-2004, 11:20

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •