Photographic reflector lenses are typically low contrast (internal reflections I guess), as well as low micro-contrast and resolution (aperture obstruction I guess) compared to refractors.
So my question is a rule of thumb of how much longer the focal length of the reflector needs to be before it would be better than just cropping an image from a shorter focal length refractor. I have a theory, but it risks just being one of those coincidences where flawed analysis comes up with a reasonable answer.
Based on the limits imposed by the central obstruction it would seem something like 50% longer than the refractor. This is one of many equivalent references I'm basing this on:
And most photographic mirror lenses seem to have an obstruction ratio around 33%. The third figure in the linked reference (also seen in some optics books) indicates that over the MTF spatial frequencies relevant in photographic optics we are seeing the curve cut to 66% of the unobstructed value. So it seems we could take an obstructed design of 2/3's the focal length and crop/magnify it to get about the same MTF.
Now this sort of seems to jive with experience, and by experience I mean "decent" mirror lenses and not some of the junk sold for a $100 these days (which would obviously be much worse). As a quick reference photozone tests the Tamron SP 500/8 vs. the Nikkor 300/4 at the center as 1550 vs. 2100 which is about the expected ratio from the analysis above.
So, obviously optics are way more complicated than this, but is this a rational explanation? The best mirror lens is only going to do about as well as a refractor that is 2/3's the focal length cropped to the same FoV?
Sorry for the slightly off-topic question, but it was inspired by alt-mirror lenses and this forum seems to have the most optics geeks wandering around!
kwalsh wrote:
So, obviously optics are way more complicated than this, but is this a rational explanation? The best mirror lens is only going to do about as well as a refractor that is 2/3's the focal length cropped to the same FoV?
I think your line of reasoning is correct, but only for diffraction-limited imaging. The resolution of a perfect imaging system is governed by the total area of the aperture. When you consider practical examples, the question arises how well aberrations are controlled in both the refractor and the reflector.
Mmmm I'm not any kind of "optics guru" but I think there's more to it. You said "Photographic reflector lenses are typically low contrast (internal reflections I guess), as well as low micro-contrast and resolution (aperture obstruction I guess) compared to refractors" but there's sharpness and stuff to consider in that formula as well. It seems maybe true that there is less micro-contrast I dunno. The images I took with my Tamron SP 500/8 did all look like the textures were a tad flat. But edge sharpness and color-definition/isolation were better than most 500mm lenses under $2,000 IMO. So the images just look a tad different and not really comparable to refractors in a straightforward way like I think you're wanting to do. At least if I understood your question correctly: "So my question is a rule of thumb of how much longer the focal length of the reflector needs to be before it would be better than just cropping an image from a shorter focal length refractor. "
When these other attributes are mixed with false detail added in by digital noise and some other things and then processed specifically for the strengths of a mirror lens there doesn't seem to be a good way of answering your question in any direct way. I'm not sure at all how the center obstruction would reduce micro-contrast or resolution and even if it did wouldn't the aperture limitations of the refractor be doing something similar?
I dunno, it is different somehow and surfaces which would normally show off a lens's micro-contrast do seem flatter but can we really say that cropped 300mm or whatever is going to match the image in some way? And which 300mm? There are awesome 300mm refractors, mediocre 300mm refractors, and terrible 300mm refractors. It seems to me again, like it's just different - because of all the other attributes and things going on with both lenses differently. With some subjects and lighting conditions this flatness when combined with the better color-defintion/isolation thing I mentioned, actually works to the overall advantage of the photograph:
Again, I dunno crap about the math involved here but just going from the results I get I think it's pretty hard to come up with a meaningful coefficient which can be used to predict or relate the results from a refractor and a reflector. It'll be interesting to read what people who are into this kinda stuff have to say. I'm looking forward to it.
@Toothwalker - Yeah, I was presuming diffraction limited under the assumption that at the speeds of most of these lenses (F/8 or so) that most quality refractors and reflectors would be approaching diffraction limited performance near the optical axis at least. Certainly it is more complicated than that in reality!
@Bifurcator - I've got the Tamron 500/8 as well (us it on m43) and it is a pretty darn good lens - actually amazing for the size and weight. Your observations about sharpness actually match the diffraction limited predictions - while an obstructed aperture has a "poor" MTF at the lower resolutions (poor "microcontrast") it is actually a bit better than the equivalent unobstructed aperture at the very highest resolutions (good "resolution"). And yes, it is all definitely more complicated than my very simple rule of thumb!
So what's lost in sensitometric small-structure contrast is gained in acutance. Kewl! I'll add that at least in the case of the Tamron 500/8, it's superior color contrast provides an extra measure of definition which produces a general sense of clarity I've not seen the exact likeness of from any of my pro or semi-pro 300mm refractors.
