A simple (hopefully) lens question . . .

[xpatUSA]

Some of us buy lens that are capable of seriously low f-numbers - presumably in search of low light or high-speed photography, or perhaps low DOF, whatever. Some might just buy them no-expense-spared because they are generally "better".

I myself am more interested in sharpness, detail, good color and little less interested in aberrations like lateral CA.

So, say that I have two similar lenses but one is f/2.8, the other f/5.6.

It is often said that "stopping down a couple of stops" improves sharpness by getting into a "sweet spot" between aberrations and the so-called onset of diffraction.

The question is: can we regard the lens with f/5.6 (and set to that) as being as "good" as the one with f/2.8 stopped down to f/5.6?

. . . all other things being equal of course.

[Geoff F]

Generally speaking, most (almost all) lenses lose a bit of quality at their extreme ends. So a 2.8 lens at 5.6 will produce better results than a fully open 5.6 lens. Providing we are talking about roughly similar quality lenses.

[ktuli]

. . . all other things being equal of course.

Well, there's the rub, isn't it!

Manufacturers tend to use different glass, coatings, groupings of lens elements, focusing motors, etc, etc, etc in their different lenses - all of which will factor into those things you are interested in. If the only difference between a 2.8 lens and a 5.6 lens was how wide the aperture opened, why would anyone spend the extra money on a 2.8 unless they absolutely needed the shallower DoF - especially considering how large those price differences usually are.

You're unfortunately comparing apples to oranges here, so I doubt you'll get the precise answer you're looking for unless you have two specific lens models that you are comparing and can eliminate as many variables as possible.

- Bill

[William W]

. . . So, say that I have two similar lenses but one is f/2.8, the other f/5.6. It is often said that "stopping down a couple of stops" improves sharpness by getting into a "sweet spot" between aberrations and the so-called onset of diffraction. The question is: can we regard the lens with f/5.6 (and set to that) as being as "good" as the one with f/2.8 stopped down to f/5.6? . . . all other things being equal of course.

Simple and general answer: No.

The F/2.8 lens will be "better at F/5.6"

Rationale for answer: If we make a criterion of "similar lenses" generally meaning similar FL, then being such a large maximum aperture difference there will be a large price difference - but that price differential goes to other elements - such as overall design and optical quality.

More complex answer: we'll argue to eternity about the parameters of "similar lenses" and "all other things being equal".

WW

Quick example: any DSLR kit zoom lens 18 to 55ish F/3.5 to F/5.6 at 50mm at F/5.6 and any "similar lens" being a 17 to 50ish F/2.8 zoom ... see what I mean we can debate if they are "similar" or not. . .

[William W]

. . . thanks Ted

Also as a footnote: I read your question literally - i.e. in regard to the large difference of the maximum aperture.

My point being - if we have two "similar lenses" with a very small maximum aperture difference then it is a different ball game.

A couple examples from Canon:

> the EF50 F/1.0L and the EF50F/1.2L and then we compare both at (for example) F/4, the F/1.2 is probably better, I don't know but that's a reasonable guess having used the F/1.0L

> Similarly comparing the set of five 70 to 200L zooms all set at F/5.6 - they're be very similar and maybe one of the F/4 versions might be better than on eof the F/2.8 versions

happy new year
Bill

[Adzman808]

Geoff's answer is IMO about right in absolute terms (ie all else being equal)

But as others have mentioned the reality is more complex.

Lenses have many different components that constitute to their overall quality, on top of this you have to factor in what you want the lens for... So if a lens is a bit soft at the extreme edges, and you're planning to use it for portraiture probably not a huge drama, for landscape, it's a bit of a problem.

It also gets IMO complicated by the fact that a lens that's ace at sharpness charts and brickwalls, doesn't necessarily produce pleasing images, because things like contrast, character and brokeh are important too and vary between lenses

To further complicate the matter!! A lot of cameras and a great many raw convertors have inbuilt lens correction profiles that can alter the output of a lens quite significantly.

There can also be exceptions to the "stopped down a bit" rule in regards to zooms, SOME zooms can be great wide open at the short end, but not so much at the long end.

IMO it is true that all lenses have a sweet spot, but it's IMO not governed by a rule that you can apply to all.

And I haven't even mentioned the complexity of what size sensor said lens is being used on which changes things yet again

[Steaphany]

A parameter you failed to specify in your example is the diameter of the first, primary, optical element.

If you ever have taken a look at astronomical instruments such as telescopes, they will often have photographically poor apertures. I have a small 90mm Maksutov-Cassegrain that I use for astronomical and telephoto shooting with a fixed aperture of f/13.9. Other examples are the

• Meade 12" LX850 Advanced Coma-Free Telescope at f/8
• Meade 16" LightBridge Dobsonian Reflector Telescope at f/4.5 ( Fast for a Telescope )
• Vixen 115mm ED115S Telescope at f/7.7

Most photographers would be screaming as they ran out into the night if told they would be shooting through a lens offering f/7.7, f/8, or f/13.9.

The key difference is astronomical photography values sharpness over maximum aperture, and to achieve a greater sharpness, you need a larger diameter to the primary optical element. It turns out that the diffraction limit, or sharpness, of an optical system or photographic lens is directly proportional to the size of the primary.

With out looking at any specifications, by only comparing the size of the first lens element, you can easily know which lens has the potential to provide greater sharpness, in this case, bigger really is better.

See Angular resolution on Wikipedia for formulas

[rpcrowe]

If sharpness were the only criteria for which to judge the "quality" of a lens, we might all be shooting with macro lenses

There are other facets which differentiate the quality of a lens such as (but not restricted to): auto focus speed and accuracy, lack of, or minimal aberrations such as chromatic, lack of distortion such as pincushion or barrel, lack of vignetting, flare control, bokeh quality, build quality and so on...

Top-line lenses often produce top-line image quality even when shot wide open. As an example of two lenses which have excellent image quality throughout their f/range are the Canon 70-200mm f/4L IS lens and the Canon 400mm f/5.6 L lens. I own both of these lenses and have absolutely no qualms about shooting them wide open. However, I seriously question both the image quality and the autofocus of my Canon 50mm f/1.8 Mark-I (metal mount) lens when shot wide open. Stopping down to f/2.8 improves the image quality quite a bit but the autofocus is still a bit chancy in lower light levels...

This is why I have pretty much supplanted my 50mm f/1.8 Mk-1 with my 17-55mm f/2.8 IS lens. Better IQ, zoom capability and like the name indicates: image stabilization.

The image of the various Canon kit lenses also improves when the focal length is at neither extreme and when the lens is stopped down to f/8 or f/11. The image quality of these lenses is really quite good when tripod mounted and shot stopped down and with the focal length of the zoom away from the widest or longest lengths.

For a zoom range on a full frame camera, I do like the 28-135mm f/3.5-5.6 IS lens. However, at both ends of the range (widest and narrowest) the IQ of that lens is not really great and I would like an f/2.8 non-variable aperture.

If I were to envision a people lens for a full frame camera, I would select a 35-105mm f/2.8L IS lens with a constant aperture. Unfortunately, that lens doesn't exist! My rationale is that 105mm is a decent focal length for portraits on a full framer while the 35mm doesn't produce the distortion of people that is produced at 28mm or 24mm on the full frame camera. Limiting the focal range to 3x would allow the f/2.8 constant aperture with today's technology. The only problem is that I would probably be the only photographer in the world to purchase that lens

[dem]

A parameter you failed to specify in your example is the diameter of the first, primary, optical element.

This is interesting. I do not think that the diameter of the front element is relevant here, the f-ratio is all that matters. The diffraction limited angular resolution depends on the entrance pupil (virtual aperture) diameter, not the lens diameter.

See Angular resolution on Wikipedia for formulas

... as confirmed by the formulas you kindly provided. The wikipedia says "the value of D corresponds to the diameter of the light beam, not the lens".

Then I found this
http://photo.stackexchange.com/quest...-photo-quality

Note also that the diffraction limit calculator on CiC requires the f-ratio but not the size of the lens:
https://www.cambridgeincolour.com/tu...hotography.htm

I do not know much about telescopes, but it seems that the size of the mirror (because it acts as aperture) is indeed relevant to the resolution of reflecting telescopes like Hubble. The refracting telescopes (those that use a lens to focus light) do not seem to have aperture blades in the way the photographic lenses do. So technically their aperture is the same as their lens diameter, but this is not true for photo lenses.

[Manfred M]

Assuming the same fixed focal length, the main reason that the f/5.6 lens exists is because it is far less expensive to make than the f/2.8. The target audience is likely different so there is absolutely no guarantee that the less expensive lens will perform better and in fact will likely not perform as well.

As others have said, this is the old "apples and oranges" comparison. Nothing ever seems to be simple in photography.

[Brev00]

Well, you can always head over to Digitalpicture.com and look at some pairs of lenses head to head using their chart and comparison tool. I would think a corollary to your question would be which lens is better at good old f 8. Does giving the slower lens a chance to stop down even the playing field? If one always shoots at f 8, does it make sense to spend on specialized lenses? Of course, who always shoots at a particular f stop and do we always make sure we are using the sweet spot?

I think there are some lenses that are better wide open than others. My Tamron 70-300 is very good at f 5.6 and 300mm and I use it there all the time but my Sigma 17-70 is relatively weak at 70mm and f 4 and I typically stop down to f 5.6 at least.

[xpatUSA]

If sharpness were the only criteria for which to judge the "quality" of a lens, we might all be shooting with macro lenses

In the OP, I said "I myself am more interested in sharpness, detail . ."

Not forcing that criterion on everybody else, i.e. "we" . . .

[xpatUSA]

Well, you can always head over to Digitalpicture.com and look at some pairs of lenses head to head using their chart and comparison tool.

Couldn't find anything to do with lenses there. Do you have a more specific link?

[george013]

A parameter you failed to specify in your example is the diameter of the first, primary, optical element.

If you ever have taken a look at astronomical instruments such as telescopes, they will often have photographically poor apertures. I have a small 90mm Maksutov-Cassegrain that I use for astronomical and telephoto shooting with a fixed aperture of f/13.9. Other examples are the

• Meade 12" LX850 Advanced Coma-Free Telescope at f/8
• Meade 16" LightBridge Dobsonian Reflector Telescope at f/4.5 ( Fast for a Telescope )
• Vixen 115mm ED115S Telescope at f/7.7

Most photographers would be screaming as they ran out into the night if told they would be shooting through a lens offering f/7.7, f/8, or f/13.9.

The key difference is astronomical photography values sharpness over maximum aperture, and to achieve a greater sharpness, you need a larger diameter to the primary optical element. It turns out that the diffraction limit, or sharpness, of an optical system or photographic lens is directly proportional to the size of the primary.

With out looking at any specifications, by only comparing the size of the first lens element, you can easily know which lens has the potential to provide greater sharpness, in this case, bigger really is better.

See Angular resolution on Wikipedia for formulas

Are you sure about this? From your quote

The interplay between diffraction and aberration can be characterised by the point spread function (PSF). The narrower the aperture of a lens the more likely the PSF is dominated by diffraction. In that case, the angular resolution of an optical system can be estimated (from the diameter of the aperture and the wavelength of the light) by the Rayleigh criterion invented by Lord Rayleigh:

I don't see any reference to the diameter of the glass.

George

[xpatUSA]

A parameter you failed to specify in your example is the diameter of the first, primary, optical element.

Assuming "you" to mean the original poster, do pardon my inadequacy

With out looking at any specifications, by only comparing the size of the first lens element, you can easily know which lens has the potential to provide greater sharpness, in this case, bigger really is better.

See Angular resolution on Wikipedia for formulas

Steaphany, your statements imply that the front element diameter is a determinant of the diameter of the entrance pupil; the which, in turn, determines the angular diffraction for a given aperture setting, focal length and wavelength.

I am confused because, at f=50mm and f/2.8, I have a lens of 55mm front element diameter whose entrance pupil diameter is significantly greater than that of another lens of 77mm front element diameter.

See:

https://en.wikipedia.org/wiki/Diffra...imited_imaging

https://en.wikipedia.org/wiki/Entrance_pupil

Pardon me for posting links to Optics 101, no disrespect intended.

(my choice of f/5.6 was deliberate in order to avoid getting into aberration or diffraction, by the way; and, to avoid obfuscation by sensor pixel pitch, lets have big ones, maybe 10um).

[tezclarke]

As a relative newby to this, I am amazed by the really knowledgeable answers to the above question by Ted. Steaphany's answer is way above my level. Do we Really need to be so exacting?

Steaphany, your statements imply that the front element diameter is a determinant of the diameter of the entrance pupil; the which, in turn, determines the angular diffraction for a given aperture setting, focal length and wavelength.

I mean no offence, just interested that photography reaches such high levels.
Tez

[xpatUSA]

As a relative newby to this, I am amazed by the really knowledgeable answers to the above question by Ted. Steaphany's answer is way above my level. Do we Really need to be so exacting?

Steaphany, your statements imply that the front element diameter is a determinant of the diameter of the entrance pupil; the which, in turn, determines the angular diffraction for a given aperture setting, focal length and wavelength.

I mean no offence, just interested that photography reaches such high levels.
Tez

Tez, the statement in bold above is mine; could you please edit your post above by highlighting it and clicking on the 'quote' button? Or put "Ted said: " and add double quotes around the statement.

[Brev00]

Couldn't find anything to do with lenses there. Do you have a more specific link?

http://www.the-digital-picture.com/R...ple-Crops.aspx

I hope this helps. If it is any consolation, I have gone to the wrong link a few times myself.

[xpatUSA]

http://www.the-digital-picture.com/R...ple-Crops.aspx

I hope this helps. If it is any consolation, I have gone to the wrong link a few times myself.

Thanks! Your other post was missing the dashes, no biggie.

I compared two of my Sigma-mount lenses the 18-200mm f/3.5-6.3 DC OS and the 17-50mm f/2.8 DC OS.

Both at 28mm, f/5.6. The 18-200mm was horrible

No surprise at all, based on responses in this thread.

[george013]

Thanks! Your other post was missing the dashes, no biggie.

I compared two of my Sigma-mount lenses the 18-200mm f/3.5-6.3 DC OS and the 17-50mm f/2.8 DC OS.

Both at 28mm, f/5.6. The 18-200mm was horrible

No surprise at all, based on responses in this thread.

I just wonder how they did it. The same picture with a different look for different lenses, different diafragma's and different camera's.

George