Page 1 of 2 12 LastLast
Results 1 to 10 of 16

Thread: Scan at Maximum Optical or Stated Resolution?

Hybrid View

  1. #1

    Join Date
    Dec 1997
    Location
    Baraboo, Wisconsin
    Posts
    7,742

    Scan at Maximum Optical or Stated Resolution?

    In a recent thread started by Michael Heald Ted Harris says: "I find that scanning at the maximul real world resolution of your scanner produces the best results for a "master" file that you can then manipulate as needed depending on the final image. In terms of all of the consumer scanners this will mean setting your scanner to sample at ~ 2400 spi (depending on the settings)." So Ted says to scan at the maximum optical resolution rather than the maximum resolution stated by the manufacturer.

    OTOH, in the same thread robc says: "I would scan at 4800ppi which gives you an image of approx 24000x19200 pixels."

    So one says to scan at the maximum optical resolution of the scanner, the other says to scan at the maximum resolution stated by the manufacturer (robc was talking about the Epson 4990, which Epson says has a maximum ppi of 4800).

    I had always understood that there was no point in scanning at a ppi higher than what the scanner could resolve, that while a scanner would create a file based on the maximum stated resolution there was no benefit in doing so if the scanner could only resolve a lesser number. However, robc's statement about scanning at the maximum spec resolution is the second time I've seen that suggestion here and both robc and the other person, whoever that was, seemed to be knowledgeable about scanning. I'm not.

    So I'd like to know whether there is any kind of a consensus about this. Do you scan at the maximum optical resolution or the maximum resolution stated by the manufacturer for best results, forgetting about what size print you want to make since either way you'll presumably downsample once you decide on that.
    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.

  2. #2
    Michael E. Gordon
    Join Date
    Jan 2005
    Location
    Southern California
    Posts
    460

    Re: Scan at Maximum Optical or Stated Resolution?

    Quote Originally Posted by Brian Ellis View Post
    Do you scan at the maximum optical resolution or the maximum resolution stated by the manufacturer for best results, forgetting about what size print you want to make since either way you'll presumably downsample once you decide on that.
    Yes. WCI's article on the topic is worth reading.

  3. #3

    Join Date
    Jul 2005
    Posts
    950

    Re: Scan at Maximum Optical or Stated Resolution?

    Firstly, I am not expert in scanning. I am just applying some reasoned arguments to the process.

    In my first post in this thread I asked a question to which there have been no responses.

    My theoretical thinking on this is based on having seen absolutely no evidence that scanning at lower than the hardware resolution produces a better quality scan than can be achieved by scanning at the hardware resolution.

    The argument goes that the resolvable resolution obtained is upto 2400ppi so there is no benefit in scanning at higher resolution. I disagree. The reason I disagree is because, as I have already stated, the scan software is post scan processing the data in an uncertain way. I.E. it is either dropping scan lines, downsizing or both. You have no direct control over those processes and cannot therefore optimise them.
    If the result of scanning at lower than hardware resolution produces a sharper scan then I say it is because the scan software is applying sharpening which you could be controlling yourself in PS. Scanning at the hardware resolution allows you to control the downsize and also to do it in steps applying as much or as little sharpening in steps as you like. Also it allows for the final downsize to be done as minus 2X factor which results in having to apply less or no sharpening to the final image which results in less image artifacts.

    There was recent thread on RIP's where it was claimed that with the correct RIP it was possible to see a marked improvement in print quality when printing at 1440dpi.
    A 20x16 print at 1440dpi requires a 28800x23040 pixel image. Scanning a 4x5 neg at 2400ppi produces a 12000 x 9600 pixel image. Obviuously that size image is nowhere near big enough to print a 1440dpi 20x16 image. Therefore I argue that the best scan to use as a base work image is at max hardware resolution. This leaves you with far more options for post processing and printing at different sizes and qualities. It also means you don't need to scan again at a later date for a bigger print. This is a moot point since at a later date a better quality scanner may be available thereby rendering your earlier scan redundant.

    If your scanner software is doing a good job of post processing the raw scan data and the output is big enough for the print size you want then by all means scan at lower than hardware resolution.

    Finally I think it is no surprise that it has been recommended that a scan of 2400ppi on an epson 4990 is optimum. I would suggest that it is not coincidental that 2400ppi is a minus 2x factor from 4800ppi which allows the scan software an exact dowsize of 2x (probably bicubic sharper) or dropping of alternate scan lines (which I don't think silverfast is doing but I can't be sure).

  4. #4

    Join Date
    Feb 2005
    Posts
    2,953

    Re: Scan at Maximum Optical or Stated Resolution?

    Comparison scans on Epson 4990 made at 4800, 2400 and 1200:

    http://www.olegnovikov.com/technical...pson4990.shtml

  5. #5
    Kirk Gittings's Avatar
    Join Date
    Mar 2004
    Location
    Albuquerque, Nuevo Mexico
    Posts
    8,330

    Re: Scan at Maximum Optical or Stated Resolution?

    I just ran a test on the V750 and found that scanning at 6400 and DS to 2400 gave slightly better detail and shodow separation than scanning at 2400.
    Thanks,
    Kirk

    "When did photography become a desk job?" Kirk Gittings 2009

    KIRK GITTINGS
    WEBSITE

    LIGHT+SPACE+STRUCTURE (blog)

  6. #6

    Re: Scan at Maximum Optical or Stated Resolution?

    I took a photograph of a air force test target at a distance of 40x the focal length of the lens that I was using.

    I then used that negative to test my scanner.

    I found that at an advertised 1600 optical and 3200 max resolution on my scanner I failed to gain anything after 1200.

    However, you almost have to have a microscope to determine how much information you have captured on the negative but with any photographic skill at all you will capture way more that you can scan with a flat bed scanner.

  7. #7

    Join Date
    Feb 2005
    Posts
    2,953

    Re: Scan at Maximum Optical or Stated Resolution?

    Quote Originally Posted by Neal Shields View Post
    .

    I found that at an advertised 1600 optical and 3200 max resolution on my scanner I failed to gain anything after 1200.
    Neal, what scanner were you using?

  8. #8
    Abuser of God's Sunlight
    Join Date
    Aug 2004
    Location
    brooklyn, nyc
    Posts
    5,401

    Re: Scan at Maximum Optical or Stated Resolution?

    scanning at the machine's sampling frequency (like 4800 ppi) has some advantages, even though it excedes the optical resolution.

    First of all, it lets you start with a file that hasn't been interpolated at all. Then, if you downsample by a factor of 2 (to 2400 ppi, which is still probably a bit higher than the optical resolution) you gain a noise reduction benefit. Each 4-pixel cluster is averaged down to one pixel, so the effects of random noise are theoretically reduced by 75%. I don't see that kind of reduction in real life, but it does seem to help a bit.

    Some scanning software will do this downsampling for you, so you don't have to add the extra step to your workflow.

  9. #9

    Join Date
    Jul 2005
    Posts
    950

    Re: Scan at Maximum Optical or Stated Resolution?

    Quote Originally Posted by paulr View Post
    scanning at the machine's sampling frequency (like 4800 ppi) has some advantages, even though it excedes the optical resolution.

    First of all, it lets you start with a file that hasn't been interpolated at all. Then, if you downsample by a factor of 2 (to 2400 ppi, which is still probably a bit higher than the optical resolution) you gain a noise reduction benefit. Each 4-pixel cluster is averaged down to one pixel, so the effects of random noise are theoretically reduced by 75%. I don't see that kind of reduction in real life, but it does seem to help a bit.

    Some scanning software will do this downsampling for you, so you don't have to add the extra step to your workflow.
    the argument of removing scan noise by downsampling is a moot point as far as I'm concerned. By doubling your print resolution to say 720dpi, you will achieve the same thing visually since the printed pixels will be so close that they will be visually averaged by the eye. But you will not have avergaed/altered other detail in the image which, as has been stated by those using a RIP, can been seen as an improvement in image quality when printing at higher print resolutions such as 1440dpi.
    Last edited by robc; 10-Oct-2006 at 12:27.

  10. #10

    Join Date
    Aug 2006
    Posts
    96

    Re: Scan at Maximum Optical or Stated Resolution?

    Quote Originally Posted by paulr View Post
    Each 4-pixel cluster is averaged down to one pixel, so the effects of random noise are theoretically reduced by 75%. I don't see that kind of reduction in real life, but it does seem to help a bit.
    Noise reduction is as the inverse of the square root of N. For four samples going to one you get a 50% reduction of noise which may be closer to what you actually see in real life ;-)

    Stew

Similar Threads

  1. Flatbed vs drum scan
    By Aaron Ng in forum Business
    Replies: 28
    Last Post: 3-Feb-2010, 20:02
  2. How do you scan for web-publishing?
    By Patrik Roseen in forum Digital Processing
    Replies: 4
    Last Post: 21-May-2006, 14:10
  3. Aliasing and scanning resolutions
    By Ed Richards in forum Digital Hardware
    Replies: 31
    Last Post: 20-Oct-2005, 22:35
  4. Paper Resolution
    By Patrick Troccolo in forum On Photography
    Replies: 37
    Last Post: 11-Dec-2004, 05:52
  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
  •