Why does small apertures lead to reduced overall contrast?

[torger]

Here's a question for the optics experts.

Say you have a large bright white test chart with large deep black square on it. Then you take a photo at f/5.6 of it. High contrast, sharp edges. Then you take a photo at f/22. Soft edges of course, diffraction. But you also note that the bright white is now duller and the black is grayer -- there's an overall reduction in contrast. You can also see that there's some "bleed" from the white into the black much wider than the Airy disk.

I haven't figured this out and it makes me crazy :-). Someone that can explain the phenomenon? Is it a side effect of diffraction, and if so how? Or is it something related to light loss or reflections in glass elements that become different with small apertures?

[Bart_van_der_Wolf]

Here's a question for the optics experts.

Say you have a large bright white test chart with large deep black square on it. Then you take a photo at f/5.6 of it. High contrast, sharp edges. Then you take a photo at f/22. Soft edges of course, diffraction. But you also note that the bright white is now duller and the black is grayer -- there's an overall reduction in contrast. You can also see that there's some "bleed" from the white into the black much wider than the Airy disk.

Hi,

And there you have it! The Airy disk doesn't stop at the first zero from the center of the diffraction pattern, it continues with more 'rings' that have a decreasing overall contribution as they go to infinity. The center contributes the most, so it's contribution to the blur (= lower contrast) dominates though.

I haven't figured this out and it makes me crazy :-). Someone that can explain the phenomenon? Is it a side effect of diffraction, and if so how?

Yes.  Diffraction blur affects all of the spatial frequencies in the MTF, even the lowest (DC, which is overall brightness).

Or is it something related to light loss or reflections in glass elements that become different with small apertures?

That may play a minor role, but it will affect the shadows more than the highlights due to internal reflections not being handled as should be by the lens construction (e.g. blackening the edges of lens elements). Diffraction blur affects all brightnesses.

Cheers,
Bart

[torger]

Thanks!

But why then doesn't typical deconvolution sharpening algorithm improve the global contrast? Is it some limitation of the PSF used?

[Bart_van_der_Wolf]

Thanks!

But why then doesn't typical deconvolution sharpening algorithm improve the global contrast? Is it some limitation of the PSF used?

Deconvolution sharpening should exactly address that issue, if the PSF is close to what it should be. However, diffraction requires a more complex and larger PSF to get it close to perfect, although a PSF based on a Gaussian can already help significantly.

The problem with f/22 though is that in addition to overall contrast loss, it also loses high contrast micro-detail resolution, so it is rather damaging to the overall image quality.

Cheers,
Bart