Jack Hogan

I am afraid not Ted. Perhaps Iliah Borg, I remember him suggesting that LR does it after for speed.

They do it before and/or after, based on speed and convenience. In practice, whether results are better depends on the interpolation algorithm. I prefer to do it before demosaicing, in part ...

Aliasing in such conditions (same lens, same MP, different format) is mainly a function of lens 'sharpness' and effective pixel size, which here we will take to include the AA. So if the lens and ...

If this is a module in software (as opposed to in-camera) I would guess by the wording used that they are doing simple deconvolution with an Airy PSF, the size of which is probably based on ...

Worrisome: can you repeat 'person, woman, man, camera, TV'? ;-)

I think of green as being the fulcrum of a teeter totter, as the illuminant swings from low-B-high-R (say tungsten) to high-B-low-R (say a scene under the nordic sun).  Mark explained it well here . ...

That's an interesting post and thread Joofa.

:-) Maybe the difference is minimal with well corrected lenses, though it seems that something of the order of one p-v wavelength OPD for rho^4 aberrations is somewhat normal for such at lowish ...

Right. Without knowing much about this I would naively be partial to RMS wave, as a proxy to energy.  I guess RMS spot size would be next. :-)

Maximum energy would make more sense than maximum amplitude, which would correlate with MTF. I see that Alan has provided a list of available alternatives for criteria to be used.

Circle of Least Confusion, the precise figure used in further calculations when a range is available. How do you calculate it?

Bill, slightly OT. You know I am a raytracing noob and the first time I tried it recently I was flabbergasted at the amount of spherical aberration from a single lens with perfectly spherical ...

'light level' today is normally assumed to be a quantity with SI units that does not depend on ISO/ASA, but for argument's sake let's take it to mean Luminance (L) in cd/m^2. The formula for a ...

Are DOE elements the same as Nikon's Phase Fresnel, Bill? If so the bench would need a lot more detail on that front element. Jack

Interesting but not enough info there Trebor1. One can assume a randomized CFA to try to get around aliasing issues, Fuji style, and using the white pixels to stretch DR - but one would just be ...

Hello Erik, this can be considered to be true in an ideal single thin lens set up. A typical photographic lens, however, is made up of a number of thick lenses and stops, therefore the single thin ...

Excellent new functionality and post Frans, good work!

In that case you may want to check out the 1.5um pixel Raspberry Pi HQ Camera, $50 bucks (+ an adapter for any of your existing lenses) will do it.  Very flexible, great fun.

Yes your example is indeed correct. Brandon, though, has pupil magnification wrong at 2/3 instead of 2 in his example, as validated.  T herefore the distance between exit pupil and image plane ...

'Preciate it, I was working on it but had to step away for a while before finishing it off. Of course p = 2 also in his second example - so every calculation that follows in that post is wrong as ...