[bwDraco]

I'm wondering about the diffraction correction feature in the K-3 and K-3 II. I tested it and noticed an increase in fine detail at very small aperture settings, but how does it work? It's my understanding that diffraction involves a well-defined Airy disk and could be reversed with deconvolution (see my question on Photography Stack Exchange), so is the camera deconvolving the output in some manner or just applying some advanced USM or other sharpening algorithm?

—DragonLord

[wombat2go]

I wrote a script to do deconvolution in the complex frequency domain, using an adjustable inverse gassian.
It works but takes high computing power and memory and many minutes on the Intel I7 here, to do the Fourier transform, apply the decon, and then inverse Fourier.

I doubt if that could be done in the camera microprocessors. Most likely , a sharpening is done in the spatial domain, based on what is known about the lens and f/-

[Nass]

That'd be my guess too wombat2go - USM is less processor intensive. And, since it can only be used on Pentax lenses, it's most probably some lens-specific value because Pentax has measured and knows the characteristics of its own lenses. In fact this was one of the things I was planning to feed back to Pentax as a brand ambassador, that it might be a popular move to make things like this and the other Pentax-lens-only in-camera corrections a sliding low/medium/high setting that could also be applied to any 3rd party lens, in addition to the existing specific values for specific Pentax lenses.

[stevebrot]

Most likely , a sharpening is done in the spatial domain, based on what is known about the lens and f/-

That is my understanding as well. IMHO, it is a cludge that more appropriately would be applied by the user in PP.


Steve