0
\$\begingroup\$

How can we calculate an image's effective megapixel count at each aperture setting, specifically as loss relates to diffraction?

Assume that we already know the sensor size, its true megapixel count, and assume that the camera is paired with a perfect lens.

Improve this question
\$\endgroup\$
5
  • 1
    \$\begingroup\$ What is mega pixel loss? What is a perfect lens, is that a thing, do they exist ? \$\endgroup\$
    – Alaska Man
    Commented Feb 28, 2020 at 8:46
  • \$\begingroup\$ What's an "effective megapixel"? \$\endgroup\$
    – xiota
    Commented Feb 28, 2020 at 11:08
  • \$\begingroup\$ Not an answer, because it doesn't involve megapixels, but take a look at Wikipedia's entry on "Optical resolution" especially the "System resolution" section. \$\endgroup\$
    – David Rouse
    Commented Feb 28, 2020 at 13:34
  • \$\begingroup\$ You might look at DxOMark's lens evaluations to see how they construct a MP rating for a given lens-camera combo. In real world shooting the diffraction limit is the least of my concerns trying to get what I need in reasonable focus. \$\endgroup\$
    – StephenG - Help Ukraine
    Commented Feb 28, 2020 at 17:33
  • 1
    \$\begingroup\$ "Effective megapixel" is an invented single number way to rate a lens. It was invented not because lenses lend themselves to single number score comparisons, but rather because consumers want single number scores to either decide which thing is "better" than the other thing to to brag that their thing is "better" than someone else's thing. \$\endgroup\$
    – Michael C
    Commented Feb 29, 2020 at 2:00

1 Answer 1

Reset to default
2
\$\begingroup\$

Maximum recordable resolution based on sensor size and aperture (diffraction); for blue/green/red wavelengths (green being most important).

The sensor's resolution is (mostly) irrelevant unless it is less than the diffraction imposed limit. And a perfect lens doesn't matter if it resolves to a resolution greater than the sensor's capability.

maximum recordable resolution

Taken from this paper

Improve this answer
\$\endgroup\$

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