Which ONE lens above 50mm for landscape/urbanscape?

[lct]

Advertisement (gone after registration)

As a compact lens, the Macro-Elmar 90/4 has no serious competitors but lenses with more flare (T-E 90/2.8, Elmar-C 90/4) or M-Rokkor 90/4 for Minolta CLE which has less flare but also less acutance than the Leica. Compared to the Rokkor, the Macro-Elmar has the advantage of shorter size when retracted, shorter MFD (0.8m vs 1m) and macro capabilities together with the Leica Macro-Adapter-M. Edit: The Rokkor does work with this adapter (link on the digital CL) but needs some extra sharpening in PP to compete with the Macro-Elmar + Macro-Adapter-M.

Edited April 27, 2022 by lct

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[Fedro]

22 hours ago, pedaes said:

Interesting to learn from a recent David Farkus (Red Dot Forum) video that diffraction starts much earlier with high mp sensors, and we should not go above f5.6/ f8 if you want zero degradation. I expect he is correct. 

HI Pedaes, that is fair enough, though my comment was camera/sensor agnostic,. In reality I would likely still use it at f8/9.5 or even 11 and benefit from the increased DoF if i need it. On a higher mp sensor, at f8 or f11, would a lens that can push to f22 still have less diffraction than one that can push to f16 max?

[pedaes]

1 hour ago, Fedro said:

would a lens that can push to f22 still have less diffraction than one that can push to f16 max?

Not sure the premise is correct, but hopefully @adan will give you the full  answer. 

Edited April 27, 2022 by pedaes

[TheGodParticle/Hari]

I highly recommend ONE copy of the APO75

ONE copy of the APO90

and ONE copy of the APO135

you will be well set 🍻

[adan]

4 hours ago, Fedro said:

HI Pedaes, that is fair enough, though my comment was camera/sensor agnostic,. In reality I would likely still use it at f8/9.5 or even 11 and benefit from the increased DoF if i need it. On a higher mp sensor, at f8 or f11, would a lens that can push to f22 still have less diffraction than one that can push to f16 max?

2 hours ago, pedaes said:

Not sure the premise is correct, but hopefully @adan will give you the full  answer. 

Probably not.

One issue I still have trouble with is that the published literature/math of diffraction is generally based on apertures measured and defined in absolute units - e.g. diameters in nanometers, or mm, or meters. Telescope apertures are usually defined in absolute values (e.g. Hubble, 2.4m aperture; Webb, 6.5m aperture; Arecibo Radio Telescope, 468m aperture, etc.)

Our photographic lenses are specced in relative units or f/stops - i.e. f(ocal length in mm)/x (absolute aperture in mm)

So a 100mm lens set to f/10 has an absolute aperture of diameter 10mm, whereas a 20mm lens set to f/10 has an absolute aperture of diameter 2mm.

Conversely, a 100mm lens set to f/22 will have the same absolute aperture (4.54mm diameter) as a 20mm lens set to f/4.4.

It is not at all clear to me that f/10 will actually produce the same amount of diffraction from a 100mm lens as from a 20mm lens. Or in Leica-M-world, and on-topic (lenses above 50mm), a 90mm may not be as diffraction-sensitive as a 21mm (or 50mm).

OTOH, as the Red Dot video says, resolution does not "fall off a cliff" at f/7.1 - it just peaks at that point (if using 47-60 Mpixels in 24 x 36 format), and then declines the more one stops down. And the differences (either good or bad) between a 10mm aperture or a 2mm aperture (or with a 50mm, a 5mm aperture) may not be all that significant, so using an "average worst case" just provides a safety margin.

I did a little studio experiment with a 75 APO-Summicron-M at 0.7m a few years back, and resolution (on 24 Mpixels) increased up to f/10, then flat-lined between f/10 and f/13, and then declined at f/16. Just detectable viewing at 100% pixels, but not visible at all in an A2/16"x20" print.

In any event, f/22 - all other things being equal (e.g. focal length) -  will never produce less diffraction than f/8-11-16 - always more.

[LarsHP]

 

Regarding diffraction, I have seen many fast, short telephoto lenses topping out in the center already at f/2.8 or f/4. This means the point where diffraction sets in is depending on the actual lens also.

 

@adanSince we are talking diffraction and way stopped down use: 

I have noticed that the aperture blades in some (maybe most, or even all?) Leica M lenses forms a circular opening at the minimum aperture. I have suspected that it is to minimize diffraction. Am I guessing right?

Edited April 28, 2022 by LarsHP

[adan]

Advertisement (gone after registration)

I don't think so. Diffraction is caused by light being bent by passing over an edge (i.e. the blade/air or pinhole/air knife-edge).

Whether the edge is curved or straight should have no effect.

For one thing, at the scale of a photon (smaller than an atomic nucleus**), any edge will "appear" just as straight and flat and uncurved as the ocean looks to a human sailing on it.

Or, more realistically, with an aperture blade cut by humans with normal tools and materials, as rough but uncurved as the ridgeline of the Himalayas.

A human looking into a lens at the aperture is like a human looking at the Earth from the Moon - we can see that the Earth is more-or-less round, and not a polygon, from way out there, but not when we are on the Earth's surface.

_____________________
** technically, of course, photons are treated as effectively dimensionless. All we can measure is the amplitude and length/frequency of their waveforms - and their interactions with other particles . Generally speaking, the "effective size" of a photon is assumed to be its wavelength - which for visible light is in the range of 380-740 nanometers, with green light in the middle (500-ish nanometers or 0.002 mm ).

That is why diffraction will soften an image a little but not blur it completely - only a tiny percent of the photons passing through the aperture get close enough (0.002 mm on average) to the aperture blades to be "bent" by interactions. Unless, of course, the lens can be stopped down to f/500 or so. The vast majority of photons breeze right through the aperture unaffected ("Aperture? What aperture?") - or are blocked completely to give us the exposure we expect.

https://briankoberlein.com/blog/thats-about-the-size/

It's also how diffraction gratings work, to separate the colors/wavelengths/"size" of photons by size/color - Big ol' fat ("740nM" effective) red photons interact (on average - probabilistic quantum effects) more strongly with the grating than puny little (380nM) violet photons, and get bent more. (Note that diffraction works backwards to Newtonian refraction, which bends violet light more)

https://www.edmundoptics.com/resource-page/application-notes/optics/all-about-diffraction-gratings/ 

And before someone asks - Airy disks from diffraction can therefore contain "chromatic aberration" or color separations in their structure, as well as rings. But one probably needs a sensor of several Gigapixels to see them - and maybe a second "aperture" of slits or a circle behind the regular lens aperture. And will still be very tiny in the scheme of things, and mostly lost in the background noise of regular lens CA, and subject colors.

https://www.cloudynights.com/topic/401491-why-no-color-fringing-in-airy-disk-patterns/

http://home.myfairpoint.net/vzeeg3o2/id5.html

Edited April 28, 2022 by adan

[adan]

In any event, never forget the reason W. Eugene Smith gave for destroying all his pre-WW2 pictures after the war.

"Great depth of field; no depth of feeling!"

[Jared]

Diffraction doesn’t “set in” at any particular aperture. It always increases as you stop down, starting from wide open. You always get the least affects from diffraction with the largest aperture, and softening from diffraction starts immediately.

However, lens aberrations tend to impact resolution much more than diffraction with lenses wide open—spherical aberration, chromatic aberration, astigmatism, coma, etc., so you don’t notice the softening from diffraction. These various aberrations tend to improve as you stop down while diffraction always worsens. At some point you reach a “tipping point” where lens aberrations are no longer reduced by stopping down, and diffraction starts to become noticeable. It was always there, but it didn’t dominate. As a general rule, the better the lens quality, the sooner your image peaks in quality. 

The same is true for megapixels/sampling rate. The sensor doesn’t affect diffraction, but the higher the megapixel count, the sooner you can detect it. 
 

To a first approximation, the focal length of the lens doesn’t affect diffraction—at least “kind of.” An 10mm f/2 lens has more softening from diffraction than a 100mm f/2 lens, but because it magnifies the image much less you end up seeing the same result. 10x less diffraction from the 100mm (because of the larger physical opening) but 10x the image scale. They cancel out. 
 

As to which lens I would choose for landscape work… More often than not one stops down a fair amount for landscape shots. There are exceptions, but most shots benefit from a fair amount of depth of field. At f/8, the performance differences between most of the Leica M Mount choices are minimal. They are all good. Choose the focal length that appeals to you, then pick the lens based on availability, budget, and handling. Actual images from a 75mm Summarit vs a 75mm Noctilux at f/8 or f/11 should be essentially indistinguishable, but the budget, weight, and handling sure aren’t. 
 

My favorite focal length for telephoto landscapes is probably 90mm, but that’s purely a matter of style, not a matter of “better” or “worse”. I would pick the 90 APO for its reasonable mix of cost, weight, ergonomics, and abilities to isolate subjects wide open when I’m NOT looking for depth of field.

[hdmesa]

On 4/29/2022 at 12:59 PM, Jared said:

Diffraction doesn’t “set in” at any particular aperture. It always increases as you stop down, starting from wide open. You always get the least affects from diffraction with the largest aperture, and softening from diffraction starts immediately.

...

The phrase "diffraction sets in at f/X" is usually meant to express the point at which diffraction becomes noticeable, as you described. They didn't say, "diffraction starts at...".

[Jared]

5 hours ago, hdmesa said:

The phrase "diffraction sets in at f/X" is usually meant to express the point at which diffraction becomes noticeable, as you described. They didn't say, "diffraction starts at...".

I agree with you, but I’m not sure everyone looks at it that way. A lot of people seem to think that diffraction is negligible at apertures below the point where it dominates, or that lenses can “out resolve” a given sensor, etc., when, in fact, all aspects of the imaging train contribute to the final result. It’s not a matter of the weakest link being the only important factor.

[hdmesa]

42 minutes ago, Jared said:

I agree with you, but I’m not sure everyone looks at it that way. A lot of people seem to think that diffraction is negligible at apertures below the point where it dominates, or that lenses can “out resolve” a given sensor, etc., when, in fact, all aspects of the imaging train contribute to the final result. It’s not a matter of the weakest link being the only important factor.

What would be your idea of a consequence of disregarding the effects of diffraction below say f/5.6 for an f/2 lens? And what would be a way to mitigate such a consequence outside of a wider aperture? Certainly you wouldn't expect me to focus stack at f/2 with an f/2 lens, right? Once the effects of diffraction have no effect on how I shoot (since there is nothing further I can realistically do about it), why should I worry about it? 

Edited May 1, 2022 by hdmesa

[Jared]

5 hours ago, hdmesa said:

What would be your idea of a consequence of disregarding the effects of diffraction below say f/5.6 for an f/2 lens? And what would be a way to mitigate such a consequence outside of a wider aperture? Certainly you wouldn't expect me to focus stack at f/2 with an f/2 lens, right? Once the effects of diffraction have no effect on how I shoot (since there is nothing further I can realistically do about it), why should I worry about it? 

The only consequence would be that any given lens design and sample might actually perform better at f4 or f/2.8, especially if the particular subject needs resolution primarily at the center of the field. If you look at Leica’s MTF data for some of their more recent lenses in particular you will find center sharpness is often decreasing by f/5.6. That’s all. Some people (and I didn’t mean you) seem to think diffraction doesn’t matter at all till you reach f/8 or even f/11. That’s not how it works.

[lct]

31 minutes ago, Jared said:

[...] Some people (and I didn’t mean you) seem to think diffraction doesn’t matter at all till you reach f/8 or even f/11.[...]

Count me in . I don't choose any aperture because of diffraction but because of light and DoF in the first place. It is when i hesitate between, say, f/8 or f/11 that i will think of diffraction, assuming i don't need the widest DoF. So yes, for me, diffraction never matters till i reach f/8, f/11 or even f/16 (macro). YMMV.

[hdmesa]

1 hour ago, Jared said:

The only consequence would be that any given lens design and sample might actually perform better at f4 or f/2.8, especially if the particular subject needs resolution primarily at the center of the field. If you look at Leica’s MTF data for some of their more recent lenses in particular you will find center sharpness is often decreasing by f/5.6. That’s all. Some people (and I didn’t mean you) seem to think diffraction doesn’t matter at all till you reach f/8 or even f/11. That’s not how it works.

👆👇 I just target f/5.6 on most of my lenses since it's usually a good balance of optimal color/contrast with corners that are acceptable (or excellent with Leica/CV APO 35/50). Then if the scene has to have f/8 or f/11, I'll consider it. But if a scene allows, I'll just take a two-shot focus stack at f/5.6 and manually blend in Photoshop.

16 minutes ago, lct said:

Count me in . I don't choose any aperture because of diffraction but because of light and DoF in the first place. It is when i hesitate between, say, f/8 or f/11 that i will think of diffraction, assuming i don't need the widest DoF. So yes, for me, diffraction never matters till i reach f/8, f/11 or even f/16 (macro). YMMV.