Fairy tales about 'diffraction'

[agorabasta]

Diffraction in a stopped-down photo lens never emphasises any spatial frequencies; that's just in your fantasy.

That's because it's not diffraction making the blur there. Had it been the diffraction, it definitely would, just like it so does in microscope images.

[agorabasta]

A more sane 'experiment' would be to move the sharpening radius slider in ACR while keeping the other parameters fixed; you'd see then the subjective overall sharpness increasing with an increased blur radius, although the finest detail is obliterated. Exactly that is done by diffraction as the primary trough radius increases.

Sorry, that's just bunkum. The radius parameter in the unsharp-mask algorithm does an entirely different thing that has nothing to do with diffraction.

The USM radius means the effective diameter of the image part averaged value around a given pixel that is then subtracted from the pixel value, thus increasing the individual pixel value difference from the neighbours.
The effect of the primary trough of diffraction pattern is very similar, it prevents the central peak from projecting itself onto the closest neighbours thus increasing the difference from the immediate neighbours; then the secondary peaks are still negligibly small.
Hence the effect of diffraction is similar to USM of different radii applied to the primary colour channels.

You really shouldn't declare something on the matters you don't understand. You'd better ask politely, so I could explain it to you in more understandable terms.

[agorabasta]

Now there's enough direct measurements data available on the net that blatantly contradicts the diffraction being the primary source of blur at tight apertures.
You may check the 3d blur plots at slrgear.com , a good example is here - http://www.slrgear.com/reviews/zproduct ... loader.htm
If you check that plot at 250mm f/40, it shows a very uneven non-uniform shape. Had the blur been caused by diffraction, that plot simply must have been pristinely smooth and uniform. This may mean only one thing - there's another mechanism creating that blur and it is considerably stronger than diffraction.
Then there surely are better quality lenses that show a much more uniform blur plots, but virtually none of them is perfectly monotonous at tightest apertures.

Here I would consider the case closed, but I still can clarify further, if necessary.

[01af]

Now you're confusing interference and diffraction.

It's exactly you who is supposed to know that diffraction is nothing more than interference

Oh ... I think I am beginning to understand where your delusions are coming from. Obviously you're having the famous two-slit diffraction experiment before your inner eye and now you're confusing cause and effect. Here, diffraction is the cause, and interference is the effect. But diffraction is only one possible cause for interference; there are many more, so they're not the same thing obviously. Moreover, please note that the aperture of a lens is no double slit ... neither it's a single slit because even a narrow aperture still is much wider the the light's wavelength. Instead, it basically acts as an edge.

If you're still willing to see diffraction and interference as the same thing then please note that ultimately, scattering also is nothing but a diffraction effect. If we, in our specific context, define scattering as the sum of diffractions occuring at all kinds of imperfections in the light path of a lens except the aperture then scattering sure does have an impact on image quality, too, but that's at least one, if not two or three orders of magnitude smaller than the effect of diffraction at the aperture's edge.

So ... before we get carried away about the details of diffraction and interference (which is a really broad and complex subject but doesn't add anything to our original topic), let's just keep one fact for the record: Small-aperture blur is caused by diffraction at the edge of the aperture; the maximum resolution of an ideal lens is ultimately limited by the width of the aperture.

[01af]

If you check that plot at 250 mm f/40, it shows a very uneven non-uniform shape. Had the blur been caused by diffraction, that plot simply must have been pristinely smooth and uniform. This may mean only one thing—there's another mechanism creating that blur and it is considerably stronger than diffraction.

Oh please! Don't be silly! It may mean many things ... as a matter of fact, it simply means there are more factors involved than just diffraction. My guess is: lens element decentering. It's a cheap lens after all. And even expensive lenses won't be 100 % perfect. So what you see in diagrams like that always is the sum of several factors. Diffraction causes the graph to basically go up with every f-stop beyond the optimum (here, higher is worse); other factors create the uneven shape.

Here I would consider the case closed ...

So do I.

... but I still can clarify further, if necessary.

No, please don't bother.

[agorabasta]

Oh ... I think I am beginning to understand where your delusions are coming from. Obviously you're having the famous two-slit diffraction experiment before your inner eye and now you're confusing cause and effect. Here, diffraction is the cause, and interference is the effect.

You have guessed quite wrong. The matter of fact is that diffraction is nothing other than interference of the different parts of the same wavefront. It happens over every sharp enough obstacle edge on the way of the wavefront.

You really cannot tell me what diffraction is. I'm a professional physicist experimentalist.

[agorabasta]

If you check that plot at 250 mm f/40, it shows a very uneven non-uniform shape. Had the blur been caused by diffraction, that plot simply must have been pristinely smooth and uniform. This may mean only one thing—there's another mechanism creating that blur and it is considerably stronger than diffraction.

Oh please! Don't be silly! It may mean many things ... as a matter of fact, it simply means there are more factors involved than just diffraction. My guess is: lens element decentering.

Another piece of typical bunk on your part.

That decentering would be a good reason at the wide apertures only, the smallest apertures cut out very small parts of the glass elements for the lightpath to a specific sensel, so decentering is quite irrelevant there. In fact, one may consider that at the tightest apertures every sensel is mapped to a very small area projected over the front element.

[agorabasta]

...the maximum resolution of an ideal lens is ultimately limited by the width of the aperture.

That's quite right. The problem is that 99% of lenses available are too far from ideal. And the blur has another cause that kicks in just a little earlier and masks the diffraction effects.

That was exactly the purpose why I started this thread - to advise of that non-trivial fact.

And there is yet another practical observation - tighter apertures increase the blur of the most distant objects more than they do so for the closest ones because the air medium in the light path from the object to lens becomes too narrow, making all the impurities/non-uniformities in the lightpath much more localised. And even you can check that for yourself - it's a simple fact that doesn't require too much knowledge/intellect.

[01af]

I'm a professional physicist experimentalist.

That doesn't necessarily mean you were good at what you're doing.

Another piece of typical bunk on your part.

Well ... while talking of bunk—let's go back to this thread's beginning. There you say; "The fact is that the blurring at tight apertures is not wavelength-dependent as the said blurring shows no colour fringes."

Colour fringes: bunkum! The apparent absence of colour fringes does not prove diffraction wasn't there. The blurring's dependency on wavelengths can manifest itself in many forms, not just fringes. As a matter of fact, diffraction blur pretty much is a distributed phenomenon. Red being blurred more than blue does not show up as obvious red fringes around high-contrast edges; instead it manifests itself as a ubiquitous veil of red. That doesn't necessarily mean the distribution of red was perfectly even across the whole image—it isn't; it just means it's not restricted to obvious local fringes.

Not wavelength-dependent: bunkum! As a matter of fact, the same details do show different colour renditions at smaller apertures—above-mentioned veil of red. So contradicting your ultimate precondition for this thread, the blurring at tight apertures is wavelength-dependent. Not too obviously so, admittedly (takes very careful observation to notice) ... but ultimately it is.

As I already said above—this pulls the carpet from under your feet. Your precondition just isn't. If you really are a physicist (can hardly believe that) then you're not a good one.

... the maximum resolution of an ideal lens is ultimately limited by the width of the aperture.

That's quite right. The problem is that 99 % of lenses available are too far from ideal. And the blur has another cause that kicks in just a little earlier and masks the diffraction effects.

That's just plain nonsense. If you really are a physicist (can hardly believe that) then you know the formula for the resolution liimit at a given aperture. Any half-decent lens easily resolves better at its optimal aperture than the theoretical diffraction limit at, say, f/16 or f/22. So the difference of effective resolutions at optimal aperture vs small aperture very obviously is caused by diffraction. That doesn't mean the resolution at apertures smaller than the optimum were always at 100 % the theoretical limit. Other factors join in so the effective resolution may only approach the theoretical limit but never actually reach it. Still, diffraction obviously is the primary factor in the limitation; other factors exist but are secondary. In other words—the effective resolution of a lens at any given aperture being less than ideal is never mono-causal. But at apertures smaller than the optimum, the primary resolution-limiting factor is diffraction; at apertures wider than the optimum, the primary resolution-limiting factors are anything but diffraction.

[agorabasta]

If you're still willing to see diffraction and interference as the same thing then please note that ultimately, scattering also is nothing but a diffraction effect.

Yet another piece of bunk...

Scattering may be of diffraction, refractive or reflective nature. The atmospheric scattering responsible for the blue sky colour is exactly refractive.

[agorabasta]

Red being blurred more than blue does not show up as obvious red fringes around high-contrast edges; instead it manifests itself as a ubiquitous veil of red.

The "veil of red" may be only caused by three mechanisms - first is your ill imagination, second is the incorrect black cutoff subtraction in the raw developer, the third is the reflection off the IR filter sitting just above the sensor and also off the sensor wiring. No other physical reason exists if the 'veil' is not present in the original view.

Like I told you above, there are a few lenses out there that do show coloured diffraction effects. One of them is the Sigma EX 30/1.4. Then 99% of the others don't.

[agorabasta]

Any half-decent lens easily resolves better at its optimal aperture than the theoretical diffraction limit at, say, f/16 or f/22. So the difference of effective resolutions at optimal aperture vs small aperture very obviously is caused by diffraction.

That's a quite expectedly indecent non-sequitur on your part.

What you say essentially boils down to 'if something causes the blur at aperture tightening, it must be nothing but diffraction'. Great logic, congrats on that!

[01af]

If you're still willing to see diffraction and interference as the same thing then please note that ultimately, scattering also is nothing but a diffraction effect.

Yet another piece of bunk ...

Scattering may be of diffraction, refractive or reflective nature. The atmospheric scattering responsible for the blue sky colour is exactly refractive.

Refractive, huh!? So there are billions of tiny lenses floating around in the sky, right? Well, let's see ... water droplets indeed can act as lenses. Poor-quality lenses with lots of spherical aberrations but lenses nonetheless.

But then just why is the sky eye-popping blue on dry sunny days but rather washed out on humid, sticky days ... and white (or grey) on foggy days? Why are clouds white, not blue? I'm afraid you still have a lot to learn about physics.

Anyway ... what does this have to do with small-aperture blur? Are you trying to say that 'scattering' in the lens (which allegedly is the primary reason for small-aperture blur, according to you) was of a refractive kind? If so then where are the blue colour fringes?

[01af]

What you say essentially boils down to 'if something causes the blur at aperture tightening, it must be nothing but diffraction'.

There must be something wrong with your reading comprehension. I didn't say 'nothing but'; I said 'primarily'.

Great logic, congrats on that!

Don't get me started on your logic

[agorabasta]

If you're still willing to see diffraction and interference as the same thing then please note that ultimately, scattering also is nothing but a diffraction effect.

Yet another piece of bunk ...

Scattering may be of diffraction, refractive or reflective nature. The atmospheric scattering responsible for the blue sky colour is exactly refractive.

Refractive, huh!? So there are billions of tiny lenses floating around in the sky, right? Well, let's see ... water droplets indeed can act as lenses. Poor-quality lenses with lots of spherical aberrations but lenses nonetheless.

But then just why is the sky eye-popping blue on dry sunny days but rather washed out on humid, sticky days ... and white (or grey) on foggy days? Why are clouds white, not blue? I'm afraid you still have a lot to learn about physics.

Anyway ... what does this have to do with small-aperture blur? Are you trying to say that 'scattering' in the lens (which allegedly is the primary reason for small-aperture blur, according to you) was of a refractive kind? If so then where are the blue colour fringes?

First of, the Rayleigh scattering and refraction in atmosphere are essentially the very same thing.
Second, there's also a very basic refractive scattering over the air density non-uniformities, and that is quite relevant to photography since that scattering causes extra blur in distant object images at tighter apertures as I clearly explained above in the thread.

But please feel free to laugh and blather revealing your true intellectual/cultural background...