alanr0

Are you prepared to believe the numbers you reported here , namely 34 nit variation over a 400 nit screen?  366/400  is 0.13 stop, or roughly 1/8 stop. If you accept that 1/6 stop is close to ...

Good point, although on reflection one could argue that this is a necessary consequence of asymmetric entrance and exit pupil sizes, combined with a rectilinear projection. One way to define a ...

Sure, but entrance pupil at first principal plane was a simple and convenient starting point. Generalising the analysis to other entrance pupil positions is trivial. As Kerr points out on page 6, ...

<snip> Kerr (fig. 1, page 3) starts with the case of a thin lens with the entrance and exit pupils at the front and rear principal surfaces, which in this case are planar, and coincide with the ...

Not for the Doug Kerr analysis linked below.  Kerr derives the cos⁴ relationship using the entrance pupil, with distances and angles measured in the object space.  Propagation through the lens and ...

Bill has explained that any fall-off tends to be second order compared with vignetting More generally, It does depend on the lens, but need not be a fourth power of the cosine of the direction to ...

You and Dr Hartmann may have a point about F2 glass for solid endoscopes, but using SF57 as an example for UV-blue performance is not very convincing. The graph above suggests that over a 0.4 m ...

Short answer: it depends . Almost equally useless answer: yes, yes , (but with caveats), and it depends . The devil is in the detail of what you mean by " very different scenes ", and " the same ...

Sure, but you are asking for a subjective assessment of your display, where you have provided neither a sufficiently complete set of objective measurements, nor an accurate pictorial ...

Rather than absolute luminance difference in nits, it makes more sense to ask what is the minimum detectable contrast , typically expressed as Weber contrast or Michelson contrast. The process of a ...

That is not my take from the responses so far. You have not told us which camera you used, but modern sensors typically have a pretty linear response. However, that linearity is NOT preserved when ...

These numbers are not consistent. Did you mean something closer to: Center 225 , RHS 214 , ratio 214/225 = 0.95 I believe ACR applies an sRGB tone curve or similar. If raw level 2619 is 67% of full ...

If you are seeing 8-bit values between 0 and 255, this is almost certainly after a tone curve has been applied. With an accurate sRGB tone curve the higher intensity levels follow a power law with ...

Thanks Bill. The marginal ray half-angles now look consistent with your working f# values. <snip> OK.  My stated assumption was that the clear aperture of the close-up lens exceeded the entrance ...

Sure, but which angles are you reporting? They don't appear to be consistent with the angle subtended by the exit pupil at the image plane for your reported F-numbers. This is what I do not understand.

Bill, Can you explain what the angles you reported from your simulations represent? How are you calculating the effective F-number for the lens combination? The usual definition is the inverse of ...

Joseph W Goodman, "Introduction to Fourier Optics" 4th Edition (2017) The older 2nd Edition is readable on-line and downloadable.

Octave: https://www.gnu.org/software/octave/ Octave has similarities to Matlab, but is open source, so free to download and use. In outline, what the calculation does is: (1) Construct a 2D array ...

50 Mp: F/5.6 : MTF50=92.0 cycle/mm; 2207 cycle/ph; 24 × 36 × (2 × MTF)² = 29.2 Mpe 100 Mp F/7.1 : MTF50=83.3 cycle/mm; 2750 cycle/ph; 33 × 44 × (2 × MTF)² = 40.3 Mpe The higher pixel count of the ...

Perhaps you could clarify what you mean by effective MP . How sensitive are you to colour aliasing and de-mosaicing artefacts? If we ignore such factors, and use MTF50 as a measure of resolution, ...