0
\$\begingroup\$

I'm a dentist trying to decide what scanner to buy and I dont understand what is better by pixel size.

\$\endgroup\$
2
  • \$\begingroup\$ A 25µm or 30 µm screen would both be incredibly small. I'm not sure it would make much of a difference. \$\endgroup\$
    – Michael C
    Commented Aug 30, 2020 at 10:30
  • 3
    \$\begingroup\$ Please describe the context of your problem, the items you are scanning, the scanner you are using, the screen you are currently using, and what the micrometer lengths are specifically referring to. \$\endgroup\$
    – xiota
    Commented Aug 31, 2020 at 7:15

2 Answers 2

2
\$\begingroup\$

I assume your intent is to scan physical media, perhaps film X-rays, and display the results on a monitor?

The issue boils down to resolution :

  • Resolution of the physical media
  • Resolution of the monitor (not accounting for zooming)
  • Resolution needed for the scan detail

Without a lot more information, a minimum cannot be determined. However in general, more resolution is better. Usually scanners are specified in resolution per inch terms instead of pixel size. In terms of pixel size, smaller pixels are generally better in a scanner.

As a point of reference, 25 µm pixels would be around 1000 pixels per inch (not accounting for inter pixel spacing). This would be 3 to 4 times better resolution than your monitor or that of a printed photo, but one presumes you want a bigger display than one-to-one, meaning a 1 inch film X-ray would be a 4 inch display on your monitor.

Don't forget that dynamic range in terms of bits per pixel is likely important as well.

I have no idea as to your actual requirements. If this is for Dental stuff, I suggest talking with other dentists.

\$\endgroup\$
3
  • \$\begingroup\$ A typical 23-24" FHD monitor has about 96ppi. A typical 32" 4K monitor has about 136ppi. \$\endgroup\$
    – Michael C
    Commented Aug 31, 2020 at 7:56
  • \$\begingroup\$ Yes and Apple Retina displays range from 264 ppi on newer iPads to over 400 ppi on newer iPhones. The individual device details don't effect the overall concept. \$\endgroup\$ Commented Aug 31, 2020 at 19:48
  • \$\begingroup\$ I've never been to a dentist's office where the staff views imaging on phones or even pads. My most recent visit had 27" Dell monitors in each exam room. \$\endgroup\$
    – Michael C
    Commented Sep 2, 2020 at 2:41
0
\$\begingroup\$

To add to what user10216038 said:

The optimum scanner resolution can be pretty easily determined... it should match the resolution of the camera as scaled to print size. Being that this is to convert film to digital we simply need to scale the minimum airy disk on film to the size of the media being scanned.

If a 35mm camera is being used to generate 10" media to be scanned; that is a scale factor of 7.2x. Being that this is dental (macro) photography, it is probably safe to assume apertures of f/11 or smaller will be used; and that puts the minimum airy disk at no smaller than 11um (and probably much larger). 11um x 7.2 = 79.2um details in the 10" media; and if we apply the nyquist limit of .5 then you need a scanner resolution of ~40um.

So either scanner is more than adequate for the requirement, and there is not a lot of benefit of going with a higher resolution scanner. There are some benefits from oversampling (using smaller pixels), but they also come w/ some costs (file size/expense/etc).

If this is for scanning x-ray film we can work from the CoC standard for image sharpness instead; which equates to 270um in a 10" print viewed from ~ 12", but that also equates to ~ 20/40 vision. You could reduce that to 135um for 20/20 vision; and you could also reduce that farther to 67um for 20" display viewed from 12" (2x). And still, both scanners are more than adequate.

Somewhere around 50um is the limit of the best human vision with the most critical of viewing... the 25um scanner is a better match for that. But there are almost certainly other more limiting factors involved.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.