can a flatbed scanner match camera-macro lens quality

Discussion in 'Digital Darkroom' started by ed_okie, Mar 26, 2008.

  1. Copying old photographs: Can the output of a good desktop scanner (under $500?)
    approach that of direct photography using a quality camera and macro lens?
    I've compared my few-years-old HP scanjet 4570c set to its highest scan
    resolution... then photographed the same picture with a Canon 5D and 100mm macro
    lens, both sent to PhotoShop CS3.
    Scanned image is okay, but certainly doesn't match camera-macro lens output.
    Downside of the camera method is time consumption, maybe 30 minutes. Whereas,
    the scanner might only take 3 minutes and is direct to computer.
    Output quality is the objective. Restoration and preservation work.
  2. The tough thing that I have found in scanning some old photos is the texture of the photo paper. In some cases the paper has a pretty strong texture which is accented in the scanned image by the direct light. Perhaps this might be less apparent with the correct lighting with a digital camera and a copy lens.
  3. When you say it doesn't match the macro approach, do you mean in terms of sharpness, tonality or both? Any decent scanner should do an excellent job if its software is used correctly.

    I do quite a bit of photo restoration and giclee printing for artists and find scans to be at least as good as macro shots of paintings with a 4 X 5 view camera. There's plenty of detail and the color's right on. Does your scanner autofocus and/or permit spot focusing?
  4. I have to agree with Evan. If the paper has little or no texture, the scan should be MUCH better. But if the paper has much texture, and especially if you are willing to light your shots with two stroke, polarizers, etc., then that might be the better way to go.
  5. I have tried both methods (albeit with a 40D) and prefer the flatbed any time. The main issue is setting up the camera method so that the subject is flat, evenly illuminated and with out reflections. A good studio would help. If you are doing a lot or the original is not too large to fit into the scanner then use the scanning method.
  6. As the others have said, it depends ... ;-)

    Try scanning something with polished inset strips of silver or gold leaf. Almost impossible to
    get a scan that way, you need to light it very carefully so a camera will record it.

  7. Scanner wins hands down- well... in general. I admit I don't know anything about the scanner you mention but the cameras highest output is 4368 x 2912... you can buy (and I have) a scanner today for less than a hundred bucks that will scan at 4800 x 4800 dpi. If for example you have a photo that is 4x6 which is 24 square inches. In order for you to approximate what that scanner can do you would have to zoom in on the photograph and take 24 shots of each square inch and "sew" it together in photoshop using the Automate Photomerge function and even then you wouldn't have the same resolution becuase the camera's resolution isn't square, its rectangular. Not to mention you will have to deal with lighting - the ungodly horror of having to make sure all 24 shots are lit consistently, evenly and are taken at the identical angle.

    None of this is important if you are just trying to do 1 to 1 ratio reproductions. But if you are tyring to blow the image up.. get a new scanner at best buy and breath a sigh of relief.
  8. Christopher - your math is rather impressive, but you are assuming a few things that are not that easy to demonstrate:
    a) that your picture does contain enough details to fill in the 4800 digits per inch. I am not familiar with the chemistry of film, but that sounds a bit extreme to me.
    b) that your scanner does really resolve images at the promised resolution of 4800x4800 dpi. According to many lab tests even the Epson v700 (which is not a hundred bucks cheap) peaks at ~3000 dpi in real tests when it announces 4800 dpi and even an enhanced mode at 6400 dpi. And the v700 seems to be a top performer compared to many other flatbed scanners.
    Not that I am saying you're wrong, but I don't think it is as simple as you state it.

Share This Page