the myth of DIFFRACTION

[tresise]

I have just received the email about - UPDATED: TUTORIAL ON DIFFRACTION & PHOTOGRAPHY

What a load of inaccurate information, there is no actually validation or references, just an expression of an opinion
such as:-

The exact value of the diffraction-limited aperture is often a contentious topic amongst photographers.

NO just read

The Manual of Photography and Digital Imaging
Elizabeth Allen , Sophie Triantaphillidou PhD (professor of photography)

or the earlier edition

Manual of Photography, Ninth Edition (Media Manual)
Ralph Jacobson PhD ASIS FRIC FBIPP CChem (Author), Sidney Ray BSc MSc ASIS HonFBIPP FMPA HonFRPS (Author), Geoffrey G Attridge Bsc PhD ASIS FRPS (Author), Norman Axford BSc (Author)

And the four authors listed are the lecturers I had when doing my BSc in Photography - the answer is there in the text book.

Search out and look at the images for the Nikon D800 and D800e - with and with out Anti-Aliasing - the difference is very noticeable.

The information is in these text books written by subject matter experts, not by amateurs or lay photographers as this site appears to be.

They sum up in a few worlds what is a very complex subject of lens design the different aberrations and diffraction. The key point is really that the different aberrations can be minimised but never totally corrected, in particular chromatic aberration.

These are always greater than the effect of diffraction in the real world. To put it in context and just considering the accuracy that a camera lens would need to be built too (to show diffraction). A lens would have to be built to a smaller margin of error than half the wavelength of light. If you say why half the wavelength? because of destructive interference.

The build accuracy, for a diffraction limited lens, has been suggested at a quarter the wave length of light (and remember the wave length of light is a range of wave lengths (some suggest a tenth)) and other errors even smaller.

Let me put it another way if you buy a lens taking into account the profit made by the shop, the distributor and the manufacturer how much money actually goes into making the components? lets say 30% is reasonable - do you really expect your lens of a few hundred £ or $ is built to that standard?

Certainly any zoom lens has to make a number of compromises to just to function at all and therefore it becomes impossible to conceive that aberrations are reduced below that of diffraction.

No manufacture is selling their lenses as being 'diffraction limited' or purporting to be of a design and build quality where optical aberrations are so reduced that the only factor influencing the quality is diffraction.

Plus if you are looking at your pictures as in Jpeg don't forget to add in those effects as well.

so the myth of 'Diffraction' - yes diffraction exists

when is it obvious? - when all other aberrations have been corrected for

are all aberrations corrected in the camera lenses available today - no because the level of build quality required plus the optical quality of the components would be prohibitively expensive if actually possible.

If you want correct information - sorry you need to go to the textbook written by subject matter experts, however they are text books so therefore not actually easy to understand.

I am sure this post will attract criticism but sorry the information is available in university text books which thanks to the web can be found second-hand. If you can not get the basics right how will you actually understand photography.
Particularly :- The exact value of the diffraction-limited aperture is often a contentious topic amongst photographers.

I was taught this 36 years ago why is there any debate about the best aperture? (it is in the text books) - because the web can proliferate non peer reviewed 'information' which is repeated so many times people believe it is fact.

Unfortunately for photography it suffers from many amateurs expressing opinions on matters they are ill equipped to understand. Or professionals in other fields taking basic theoretical facts and applying them to the real world ignoring the real world factors.

Now the only person who I would accept as being able to correct the above is Sidney Ray who has written the definitive text book on camera lenses, or a lens designer from say Nikon or Canon. Because unless you actually work in this field you wont have the specialist knowledge to correct any error I have made.

[Glenn NK]

Actual lens tests do show that there is a range where the resolution of a lens is best, typically not wide open nor stopped down to minimum aperture.

Starting with the lens wide open, the resolution typically increases somewhat as the aperture size is decreased, then falls off with further decrease in aperture size. This result is quite common in lens testing.

The loss of resolution of some lenses at very small apertures is quite dramatic and noticeable.

This one is typical: http://www.photozone.de/canon_eos_ff..._28_5d?start=1

Another: http://www.photozone.de/nikon_ff/499...f2vrff?start=1

I've often wondered why resolution would get better, then worse with decreased aperture size - do the authors provide an explanation of this?

Glenn

[Colin Southern]

so the myth of 'Diffraction' - yes diffraction exists

Does diffraction exist? Yes.

Does it have any relevance to real-world photography? None what-so-ever.

[McQ]

I'm hesitant to reply to this post, since it appears to be targeted at stirring things up. However, this is an important topic, so I think it deserves at least something in response. While the technicalities of diffraction don't really matter with real-world photography, I think that the high-level implications are hugely relevant -- almost as much as depth of field. Otherwise we'd all be advised to shoot our tripod-mounted landscapes at f/32 or higher.

First:

Because unless you actually work in this field you wont have the specialist knowledge to correct any error I have made.

Then why did you post this? I think anyone has the potential to make helpful contributions, regardless of their particular qualifications or formal education. Just for the record though, I work in the field, and did a PhD that dealt heavily with optics and imaging. By your benchmark I should therefore be able to comment. Either way, this particular article has also been referenced in patents on camera design, and was deferred to by the former Microsoft CTO in an online debate a while back:

I was going to launch into a full treatment of this topic in rebuttal to spell out what happens, but then I discovered the following site that has an excellent description, complete with examples https://www.cambridgeincolour.com/tu...hotography.htm

This includes a very cool applet built into it in the section called VISUAL EXAMPLE: APERTURE VS. PIXEL SIZE. I recommend that you run your mouse over various apertures in the Aperture column, and it will show you the Airy disk for each aperture. It will stay on whatever the last one you left it on. Then run your mouse over the cameras and it will show you the pixel grid. This shows utterly and clearly what the effect of diffraction is. Play with it and it will become clear what happens.

The Cambridge in Color page referenced above also has actual photographs taken with a Canon 20D at f/8, f/11 and f/22. Mouse over the various f/ numbers in the caption and you will see the difference. If you don’t believe the calculations, just look at those examples. It shows what happens to high resolution detail (some corduroy fabric) when you take the photo at various f-numbers.

Now, to where I think you are misinterpreting both this article and the authors you cite:

The key point is really that the different aberrations can be minimised but never totally corrected, in particular chromatic aberration. These are always greater than the effect of diffraction in the real world. To put it in context and just considering the accuracy that a camera lens would need to be built too (to show diffraction). A lens would have to be built to a smaller margin of error than half the wavelength of light.

Yes, at large apertures -- but at small apertures this in untrue (which is the focus of the diffraction tutorial), and the authors you cite agree. Also, just because a true diffraction-limited lens cannot be built does not mean that diffraction's effects are therefore insignificant. All this means is that a real-world lens won't ever fully reach the theoretical limit. I think the article is pretty up-front about that though:

[Diffraction] is normally negligible, since smaller apertures often improve sharpness by minimizing lens aberrations. However, for sufficiently small apertures, this strategy becomes counterproductive — at which point your camera is said to have become diffraction limited.
...
Furthermore, this limit is only a best-case scenario when using an otherwise perfect lens; real-world results may vary.

In practice though, many lenses still come surprisingly close at small apertures. And by close, I mean that diffraction will often limit resolution at an f/number which is less than a stop away from the theoretical limit. This website also makes a similar point in the more in-depth article on lens quality, MTF and resolution:

At large apertures, resolution and contrast are typically limited by light aberrations. An aberration is when imperfect lens design causes a point light source in the image not to converge onto a point on your camera's sensor.

At small apertures, resolution and contrast are typically limited by diffraction. Unlike aberrations, diffraction is a fundamental physical limit caused by the scattering of light, and is not necessarily any fault of the lens design.

In between these two extremes, there's often an optimal aperture, although this varies from lens to lens (or camera sensor to camera sensor).

I'm not going to go in-depth where you say the phrase "The exact value of the diffraction-limited aperture is often a contentious topic amongst photographers" is untrue, because I think that the very fact we're having this exchange proves otherwise...

[Cantab]

McQ, thank you for your response to the original post.

[Cheryl Davidson]

Thank you Sean. Well put.

[dje]

Sean, you may have been hesitant to respond to this post but I'm glad you did. "Tresise" may not be aware of the tremendous assistance you have provided to the countless number of people who have accessed your tutorials. He or she may also not be aware that most of the folk on this forum like to discuss things in a civil and courteous manner without any grandstanding, browbeating or condescension.

Most of us are not in a position to read and comprehend the high level texts and are looking for someone to translate them and other high level sources of infomation into something more attuned to our level of understanding. And that's what we have with your tutorials. Keep up the good work.

Dave

[jcuknz]

After my recent experience I say LOL and Colin has it right

I worked it out for myself without the answers* unhelpful people gave by simply telling me I was wrong without explaining why. I wonder if they really know but just regurgiitate the tutorial.
The key part of my question was alluded to with talk about circle of confusion in the CiC tutorial which I carefully re-read after coming to my conclusions .... making a given sized print from a quarter-plate negative requires less magnification than if from an APC-s sensor so any defect will not be so apparent to normal viewing. At f/64 the problem was worse but compensated by the reduced amount of magnification. Just as the airey cone gets bigger and smaller with focal length likewise the CoC with magnification.
*not those who quoted the tutorial.

[Glenn NK]

A special thanks to Sean for his post.

However I would still like an explanation as to why lens resolution obtained in controlled tests, falls off at the very small apertures.

I'm not going to suggest that in real life shooting that this (tested) loss of resolution has much impact (I've seen shots taken at f/22 and f/32 that still looked quite sharp).

Glenn

[McQ]

However I would still like an explanation as to why lens resolution obtained in controlled tests, falls off at the very small apertures.

I'm not going to suggest that in real life shooting that this (tested) loss of resolution has much impact (I've seen shots taken at f/22 and f/32 that still looked quite sharp).

Yes, smaller apertures typically improve lens resolution when stopping down from wide open, since this reduces most lens aberrations. Although diffraction is still present, over this range its effects are typically negligible compared to aberrations, so overall sharpness improves.

However, if one continues to stop down, at some point the benefits of reducing aberrations becomes offset by the disadvantages of increasing diffraction. After that point, at very small apertures, resolution will continue to decrease as the lens is stopped down further. With most SLR cameras this definitely happens after f/16 (although often sooner).



This loss of resolution can definitely have an impact. As a real-life example: I was recently trying to photograph the Venus solar transit using (an inadequate) 70-200mm lens and a 1.4X extender on a 5D Mk II. The size of Venus was therefore at the limits of what this camera setup could resolve. Even with a strong ND filter, I was still getting too much light and clipping the sun, so I had to use as small an aperture as possible. However, once I went up to the necessary f/32 or f/45, I could no longer resolve Venus since diffraction caused it to become blurred beyond detection. Luckily, once the sun dipped down near the horizon, the atmosphere blocked enough of the sun's light to allow me to use f/16 -- and voila, Venus re-appeared in the photos.

[xpatUSA]

What a load of inaccurate information, there is no actually validation or references, just an expression of an opinion
such as:-

The exact value of the diffraction-limited aperture is often a contentious topic amongst photographers.

[etc, etc]

Now the only person who I would accept as being able to correct the above is Sidney Ray who has written the definitive text book on camera lenses, or a lens designer from say Nikon or Canon.
Because unless you actually work in this field you wont have the specialist knowledge to correct any error I have made.

With grammar and punctuation like that in the OP, I would say that it is best ignored or, better yet, deleted. This person cannot be a professional of any standing and has certainly gained no respect from me.

Seriously. Amongst the many gaffes in the OP, the implication that amateurs have insufficient knowledge to "correct any error" is particularly irksome. The last gaffe where "the only person [singular] I would accept" turns out to be two or more people is laughable and typical of the entire post.

You asked for it. You got it.

Ted

[William W]

I teach this stuff – yes at a registered school – not an ad hoc website.

The letters after my name you can research for yourself which pretty easy to do and thereby decide for yourself whether or not my Doctorate, TVOCP, BOCP, AdDip. Photog and Dip. Photog meet your benchmark to accept my comment, which is:

The CiC Tutorials are bloody first rate and just what the doctor ordered as an easy read.
They are a concise, precise and easily understood resource for the beginner and advanced photographer alike.

These tutorials also serve as an adjunct to prescribed texts for Students of the Theory of Photography especially as a pre-study: because often, the one encapsulated and conversational document will open the door of reception for the Student to envelope the more detailed text such that the latter might be more easily, more quickly and better understood.

***

Specifically addressing:

“[TUTORIAL ON DIFFRACTION & PHOTOGRAPHY] - What a load of inaccurate information, there is no actually validation or references, just an expression of an opinion such as: - The exact value of the diffraction-limited aperture is often a contentious topic amongst photographers.”

There is accurate information in the CiC tutorials, which is quite obvious to any technically qualified person.

Moreover should any inaccuracies have existed they would have excised within days considering the requests for proof reading.

Notwithstanding that proof reading, the internet can be vicious and there are sour people just itching to nick-pick at technical misdemeanors and of course there are also the internet thugs who are just plain silly with their rants.

It is obvious to me (and seven other teachers whose opinions I have sought before writing) that the phase “the exact value of the diffraction-limited aperture” is indeed a generally a contentious issue amongst Photographers because obviously Photographers are interested in visual outputs which are premised upon Optics Theory.

Obviously for any particular shooting scenario whilst a Photographer worth his salt will ensure she is armed with a solid grasp of theory they might need to make a judgment call as to how to make the better image – and if that involves ‘contention’ (i.e. a different view), then so be it.

And with loud voice let us celebrate the ability to discuss the Practical and also the Theory behind the Practical: but mind let us do so robustly, but also let us do so with humility and a polite exchange . . .

WW

[musickna]

Personally, I found the updated tutorial on diffraction informative, clearly written and useful. If you are in the least bit interested in optical theory but not a specialist in the field (I am actually a medical scientist), it provides helpful background for those of us who approach optics from the purely practical photographer's point of view. What's not to like about it?

[Manfred M]

Thanks Sean - your resolution vs aperture graphs shows exactly what I would have expected. Lens aberations reduce resolution at wider apertures and as you stop down, resolution improves until diffraction effects start become the limiting factor.

The other interesting thing that your graph shows is that all the people that tell you to stop down by two or three stops to improve resolution are right, but when they go on to complain how diffraction is negatively impacting their image, the resolution values at f/22 corresponds approximately to 2-stops down from wide open and f/16 corresponds to 3-stops down from wide open.

[rhadorn]

What is limited by diffraction while stopping down is... the benefit drawn from stopping down: there is a point where resolution diminishes again, due to diffraction, all other factors being kept constant. This never meant that you should not take any picture with smaller apertures.

The impact of that fact varies a lot depending on the scope of the picture. If real world photography means putting pictures on Flickr, you can just take the picture with F256... If pictures have to be enlarged a lot, diffraction certainly matters, as does tripod mounting and some other blur-reducing strategies.

In the example given by Sean in his second post, the lens is as good at F16 at it would be at F4 - limited but not bad, with the benefit of increased deapth of field. This does not mean that you should not take any pictures with a smaller aperture than F11, just that you have to think of a draw-back as you do when you have to open the lens and take all the other aberrations into account. The tutorial as well as the discussion formulate this point very well.

The tutorial is very good at "softening" the argument about the sweet spot. It certainly helps me thinkig about using more often apertures beyond the sweet spot.

The discussion gives some hints about possible extentions to this tutorial:
- The graph in Sean's post should also be integrated into the tutorial
- This graph would be more explicit, if it did not show only the curve resulting from decreasing aberrations and increasing diffraction, but these curves also.
- An empirical estimate of the impact of camera shake at various speeds and focal lenghts would help evaluating how much the diffraction limit actually matters. Could be an interesting topic for some students interested in technic matters?

Reto

[arith]

I had not even heard of Diffraction or more accurately cannot remember seeing anything related to it until I read the tutorial. Therefore the tutorial is excellent and was useful to me for choosing camera + lens system, as well as the cost that is.

I think it impracticle to test your own lenses but there are plenty of sites that do tests for you.

http://www.photozone.de/lens-test-faq

I used the tutorial to explain why my images of flowers always used to look blured and with the LW/PH for the camera, choose a lens and range of apertures that could achieve highest resolution for an A2 size print, about 24" x 16", and then I no longer needed to remember the theory.

My ef 70-200 L type lens has such high resolution at all apertures and focal lengths that they exceed the ability of my camera, and I may have got something wrong but I think that means I wont notice any differences down to the diffraction limit, and diffraction limits are so high for some camera's, f16 for an SD1 I think, so that stopping down to f22 even isn't going to be noticeable.

I found the tutorial to be extremely helpful and has had considerable influence on my photography, as have the tutorials on DoF and CA ect.

Cheers

[Glenn NK]

Somewhat on topic:

In his book Audubon Society Guide to Landscape Photography, Tim Fitzharris states; "The image will be most arresting if it displays sharply from front to back. This can be accomplished by shooting at the smallest aperture to maximize depth of field, and by focusing about one third of the way into the picture space to center the in-focus zone over the framed area. Use your cameras' depth-of-field preview feature to check results in the viewfinder".

Elsewhere in the book, he says; "Use the depth of field preview to view the scene at shooting aperture. Make sure the viewfinder is well hooded, give your eyes a few moments to adjust to the dimness and examine the scene carefully. Start with focus at infinity and back off until the most distant features begin to lose sharpness. Reverse focus a smidge and then examine the foremost picture elements for adequate detail. If they are not sharp, adjust to a smaller aperture, or move the tripod back from the foreground features and repeat the procedure".

He also covers the topic of diffraction resolution: "To attain the best combination of image resolution and depth of field, you need to determine the sharpest aperture for the lens. Use this setting whenever it provides adequate DOF for your purposes. Sometimes the highest resolving power is only marginally better than an aperture two or three stops smaller. In such cases, opt for the lower resolution/greater depth of field aperture if it furthers your compositional goals."

For anyone wanting to improve their landscape photography skills, I highly recommend this book:

http://www.amazon.com/National-Audub...0645121&sr=1-4

Glenn

[Simon Garrett]

Unfortunately for photography it suffers from many amateurs expressing opinions on matters they are ill equipped to understand.

How true!

[Colin Southern]

If pictures have to be enlarged a lot, diffraction certainly matters

Even in that situation, I would suggest that it's not a big deal because as the print size increases so does the viewing distance - and of course the greater the viewing distance, the less detail that our eyes can resolve.

[arith]

How true!

I think maybe tresise is saying diffraction only exists in theory because of dispersion and other faults affecting resolution are too great to ignore, and nobody but those with degrees in photography have a right to express an opinion anyway.

But then what are all those examples about?

Why not leave testing of lenses to the experts but really, to say just because I haven't got a degree in photography called media studies around here, and normally associated by failing something at 'A' level, is about the same as I criticising someone for adding a couple of fractions bevause they haven't a degree in mathematics.

I think my photography has improved directly as a result of this site; I would have liked to take a degree in photography but I'm afraid it didn't come up as an option and now it is a bit late.