Diffraction...

[jaapv]

Advertisement (gone after registration)

Now that we have dug ourselves deep down a whole warren of rabbit holes, let’s step back and consider some practicalities Basically diffraction occurs when light touches a material edge and changes direction. . Always but the angle varies with the physical properties of the edge (and the frequency of the light but that is another rabbit hole).

This means that the question “ when does diffraction set in” is meaningless. The question should be : when does it become measurable or visible or a problem ?  
 

Then there is the fact that a sensor consists of pixels which will influence the recording of light and aberrations, blur, etc.

Then , avoiding the multiple rabbit holes of pixel number vs pixel size, Nyquiist frequencies, Airy disks and lp, and a whole bunch more, we can conclude that a higher resolving sensor can in general record aberrations amongst which diffraction better 

Which boils down to the practical question   Implied in the original post: what aperture is needed with which lens to get an optimal result in print?  To which the real life answer is: depending on the subject matter and photographic circumstances, f 8.0 is a good ballpark figure on a full frame system. 

[SrMi]

I consider that there are two situations:

a) You don't care about DOF and want the best possible sharpness. The optimal aperture varies with the lens.

b) You care about DOF, in which case "blurriness" caused by lack of DOF will always be worse than "blurriness" caused by diffraction. My suggestion is not to waste sharpness by using unnecessary large DOF, but do not blur the image so that the objects in the focal plane have the best sharpness.

[IkarusJohn]

1 hour ago, jaapv said:

Now that we have dug ourselves deep down a whole warren of rabbit holes , let’s step back and consider some practicalities Basically diffraction occurs when light touches a material edge and changes direction. . Always but the angle varies with the physical properties of the edge (and the frequency of the light but that is another rabbit hole).

This means that the question “ when does diffraction set in” is meaningless. The question should be : when does it become measurable or visible or a problem ?  
 

Then there is the fact that a sensor consists of pixels which will influence the recording of light and aberrations, blur, etc.

Then , avoiding the multiple rabbit holes of pixel number vs pixel size, Nyquiist frequencies, Airy disks and lp, and a whole bunch more, we can conclude that a higher resolving sensor can in general record aberrations amongst which diffraction better 

Which boils down to the practical question   Implied in the original post: what aperture is needed with which lens to get an optimal result in print?  To which the real life answer is: depending on the subject matter and photographic circumstances, f 8.0 is a good ballpark figure on a full frame system. 

I think we got all that.  I’m not sure that Andy’s conclusion that diffraction sets in at a more open aperture with a wide angle lens than with a telephoto lens has been answered.  Can you pull that rabbit out of your hat?

Edited April 26, 2024 by IkarusJohn

[jaapv]

I have not worked that one thoroughly through but the physical size of the aperture is of course larger on a longer lens. That is a nice simple rabbit to pull out of a hat. 

[IkarusJohn]

Well, it was an interesting discussion while it lasted, then the rabbit disappeared in a puff of smoke.

Never mind.

[01af]

vor 17 Stunden schrieb adan:

Seems to be mostly unsupported and disorganized "o-pee-nyuns posing as facts," ...

It is not very smart to dismiss facts as opinions. And unsupported they are not, unlike your alternative facts.

.

vor 17 Stunden schrieb adan:

"Yeah—the 'one Airy disk on one pixel' fallacy ..."

This was a reference to the article referred to by SrMi in post #7 above ... I thought you've read it.

.

vor 17 Stunden schrieb adan:

When you said "no matter how many pixels," you placed no limit on "many"—therefore your categorical statement as written applies for any number of pixels, from 1-1000000000..........0).

That's right. I said, "no matter how many"—and that includes "one."

.

vor 17 Stunden schrieb adan:

It is an incorrect statement if it fails in the case of a single, large pixel. Basic science. It falls into the category of "a beautiful theory slain by one ugly little fact." It also fails in the case of a small number of large pixels—if the artifact is not large enough to affect more than one of them.

Again, you're falling for the ubiquitous "one artifact on one pixel" fallacy. Even if camera shake streaks were shorter than the pixel pitch, the camera shake blur would still affect the sharpness of the image. Because the shake does not consist of one streak that falls on one pixel. Instead, it is a continuous field of streaks that fall everywhere, including pixel boundaries. If the length of the camera shake corresponds to half the pixel pitch, for example, then half of all image points will hit two pixels instead of just one—basic geometry—which of course will impair overall image sharpness.

And please don't bother us with this silly idea of a sensor consisting of exactly one huge pixel. We are talking about capable imaging sensors, not unreal absurdities.

.

vor 16 Stunden schrieb adan:

I do know, from "accidental" real-world personal experimentation, that an excellent 28 mm Elmarit-M v.4 produced quite clear across-the-image diffraction blur at f/16, quite a bit more than the much-longer 75 mm Apo-Summicron-M Asph at f/16. Both at close-focus limit ...

How much is "quite a bit more"?

You saw some blur—how do you know it was diffraction blur and nothing else? At minimum focus distance, the Elmarit-M 28 mm, being an asymmetrical wide-angle lens without floating elements, will take a greater loss of performance than the more modern Apo-Summicron-M 75 mm Asph with floating elements. Then the subject's details will be smaller in the image with the 28 mm from the same distance. So upon close inspection of the details, the same degree of diffraction blur will appear stronger to the scrutinizing eye. Furthermore, the 28 mm lens could have been slightly mis-focused, adding out-of-focus blur to the diffraction blur.

If your theory about diffraction blur depending on absolute aperture diameter (as opposed to relative aperture) was correct then you'd see approximately equal levels of diffraction blur with the 75 mm set to f/16 and the 28 mm lens set to f/6 (I guess f/5.6 would be close enough). It shouldn't be too hard to put this theory to test ... and then to rest. But don't test at minimum focus distance, to avoid floating versus non-floating elements effects at close range affecting the results.

Edited April 26, 2024 by 01af

[SrMi]

Advertisement (gone after registration)

For your enjoyment, here is another competent article:

Diffraction and Sensors by Jim Kasson

But the bottom line ..... is that increasing the resolution of the sensor with the same ratio of pitch to pixel aperture won’t make the effects of diffraction any worse in the capture at the same print size. In fact, it will make it better. Look at the top graph above. Consider the two curves as applying to two cameras with 5.3 and 3.76 um pitches, with pixel apertures equal to the pitch. Note that the combined blur circles are always smaller for the finer-pitch sensor, even though the differences are less significant at narrower apertures.

[IkarusJohn]

On 4/25/2024 at 10:04 AM, SrMi said:

One of the more authoritative articles on diffraction:

Diffraction Limited Pixels? Really?

Enjoy.

Excellent article, thank you.  The answer to my question (simplistically) is here (apologies for the length):

Dragging out a very old rule of thumb from the distant past, when it was considered a good trade-off combining both diffraction AND depth of field. It says: To limit excessive diffraction:

 

Those /4 limits:
600 mm	f/150
300 mm	f/75
200 mm	f/50
100 mm	f/25
50 mm	f/12.5
24 mm	f/6
12 mm	f/3
6 mm	f/1.5

Generally don't stop down to exceed f-stop number = focal length / 4.

Unless depth of field is more important. Just meaning, have a reason when you do. Diffraction is not good, but Depth of Field certainly can be more important in some situations. When it helps, go for it. If it does not turn out, no harm done, but you will often be very pleased about your try.

I don't know the origin of that old rule of thumb, but here is my suspicion: You may have read about Ansel Adam's Group f.64 in the 1930s which was an early purist photography group, promoting the art of the "clearness and definition of the photographic image", named for the greater depth of field of f/64 (remember, 8x10 inch view cameras were popular back then). That link is interesting reading, but here's a bit more: For Ansel's 8x10 inch view camera, a "normal" lens was around 300+ mm, but he also used longer lens. So f/64 really wasn't much of a stretch for this (other than exposure time of course, but their large cameras required a tripod anyway.) f/64 is the third full stop past f/22, i.e. f/22, f/32, f/45, f/64). They were seeking the greater depth of field, their longer lenses were limited at f/8. Note that Hyperfocal distance can play a big part of Depth of Field too.

Since f/stop number = focal length / aperture diameter, this FL/4 rule is technically just simply specifying at least a minimum 4 mm aperture diameter, so diffraction caused by the aperture edge won't excessively limit resolution. But focal length / 4 mm also defines the f/number. Later when 50 mm was the "normal" lens for the popular 35 mm film, we did hear that f/11 was about the limit to be less concerned with diffraction, which about matches this rule of thumb, and the later source probably just repeated it. We can't see diffraction in normal photos, we only see the damage of the blurring.

[01af]

vor 42 Minuten schrieb IkarusJohn:

[...]

Those /4 limits:
600 mm	f/150
300 mm	f/75
200 mm	f/50
100 mm	f/25
50 mm	f/12.5
24 mm	f/6
12 mm	f/3
6 mm	f/1.5

Generally don't stop down to exceed f-stop number = focal length / 4.

It's crucial to understand that this 'rule of thumb' holds for standard lenses. 300 mm, for example, in this table is not meant as a long telephoto lens for 35-mm format but as a standard lens for 8×10" format. In other words: these focal lengths indirectly specify image formats.

[IkarusJohn]

1 hour ago, 01af said:

It's crucial to understand that this 'rule of thumb' holds for standard lenses. 300 mm, for example, in this table is not meant as a long telephoto lens for 35-mm format but as a standard lens for 8×10" format. In other words: these focal lengths indirectly specify image formats.

Yes, that’s an interesting point.  For diffraction, pixels and sensor format doesn’t affect the difraction itself.  The diffraction is still there.  The format is simply a factor in how the image is captured (detail) and how visible the diffraction is in the final image.

It’s a useful rule of thumb nonetheless.

[BernardC]

13 hours ago, IkarusJohn said:

Yes, that’s an interesting point.  For diffraction, pixels and sensor format doesn’t affect the difraction itself.  The diffraction is still there.  The format is simply a factor in how the image is captured (detail) and how visible the diffraction is in the final image.

It’s a useful rule of thumb nonetheless.

That's a roundabout way of saying that magnification is more important than f-stop. An 8x10 negative will likely be contact-printed (magnification = 1x), or only slightly enlarged. However, if you use the same film size in aerial photography (technically it's a 24cm wide roll), f:75 would be infested with diffraction. Same thing if you use a 600mm telephoto on a 35mm camera: you'll just get blur at f:75 or f:150. In those cases, you'll need to use the other rule of thumb: diffraction starts to affect your images at f:5.6.

For extreme magnification, like photolithography, the limit might be f:0.7 or 0.5, but that's why those lenses are buildings that cost hundreds of millions of dollars.

[Stuart Richardson]

One option, which seems to have been lost in the details, is to take a picture at different apertures and see which is best. Film is expensive, but digital is free...at least per image.

You may find that stopping down gets softer not because of diffraction, but because of shutter shock, or that your f4 photo is tack sharp at the plane of focus, but does not have enough DOF. When DOF is not a factor, I have found that f4-5.6 is best on the SL2 with the APO Summicrons, but that f8-f11 is an improvement over wider stops on the 90-280. The camera/lens/tripod/stabilization/shutter/wind/subject movement/air disturbance/aperture/geological stability/jackhammer/delivery truck/etc/etc system is a bit too complex to boil down to a single number.