[Jan67]

In other words, the sensor's ability to create an image when diffraction is apparent does not relate to the actual presence of diffraction.

I am sorry, I don't understand this sentence. Do you want to say, that we can't observe influence of diffraction at all?

I would say:
1) If we take enough good lens, which is not out-resolved by sensor, then we can observe diffraction from certain f stop (which depends on pixel pitch)
2) if we take a bad lens, which is out-resolved by sensor, than the image is more influenced by this lens imperfection

[bossa]

Here's my take on this subject: I could easily be dreaming this up though and I'm sure I'll be put right.

Pixel Size = PS
Airy Disk = AD

1. If PS:AD = 1:1 then Diffraction effects will be spread over immediately surrounding pixels due to interference and "contamination" from outer secondary diffraction rings of the AD.

2. If PS > AD then diffraction is localized to the larger pixel and averaged with less spillage.

3. If PS:AD =< 1:2 then I would expect the "contamination" to subside and image quality increase to the point where actual diffraction is being sampled more accurately and the sensor is not affecting the negative outcome as much as where PS:AD = 1:1.

4. The larger the aperture the smaller the AD, thus moderating adjacent pixel contamination. This is why, when the AD approaches PS, diffraction and "spillage" starts to be more noticeable at larger apertures (i.e f8 instead of f/11 etc) especially when higher and higher density sensors are being used.

[Cannikin]

It is a persistent myth that diffraction somehow makes a higher resolution sensor resolve less than a lower resolution one at the same aperture. Diffraction is always there and of equal size airy disks at a given aperture, regardless of resolution or even sensor size. In a way you can think of diffraction as a "detail" that the higher resolution sensor can resolve, but the lower resolution one cannot.

Here's an analogy:

Nowadays lots of actors/actresses/models and makeup artists fret over skin imperfections. In the days of SD and lower resolution, they didn't care so much because they couldn't be resolved. With the advent of HD (and even more so now with 4K), it becomes easier to see blemishes and such if you look close enough. Did HD video make those imperfections appear? No, they were always there. Is the HD video somehow worse than SD video because it is possible to see them? From a technical point of view, one would be crazy to think so (from an artistic point of view it's debateable).

The takeaway: with a given lens at a given aperture, a higher resolution sensor will always resolve at least as much real detail as a lower resolution one of the same size, usually more. No exceptions (ignoring technical limitations of the sensors such as noise, etc.). If you find a good aperture to use on the K-5, use the same on the K-3. Your picture will never be any less detailed as a result of diffraction.

[bdery]

Thanks for putting in your knowledge.

Except that things didn't get any easier.

So my lenses to use are all top quality with the FA31mm, DA*55mm and FA*85mm. Where I already somewhere stated that that the DA*55 beats the FA*85 and I already knew that the FA31mm was the poor child in this groupe.

DA*55mm | FA*85mm

That all with not knowing why.

Anything to say about those focal lengths without having the lenses in your hands to test?

Sorry I missed your question earlier.

The resolving power of a lens (its sharpness, if you will) is driven by many elements. Diffraction effects are part of that.

The principle of the diffraction limit can be taken to mean that when the aperture gets really small, resolving power decreases.

Why one lens is sharper will be influenced by diffraction limits, the lack of aberrations, mainly chromatic aberration, coma (dots look like teardrops), vignetting, a few other effects.

I am sorry, I don't understand this sentence. Do you want to say, that we can't observe influence of diffraction at all?

I would say:
1) If we take enough good lens, which is not out-resolved by sensor, then we can observe diffraction from certain f stop (which depends on pixel pitch)
2) if we take a bad lens, which is out-resolved by sensor, than the image is more influenced by this lens imperfection

No I did not mean that we cannot observe diffraction. Your two points are basically valid.

Here's my take on this subject: I could easily be dreaming this up though and I'm sure I'll be put right.

Pixel Size = PS
Airy Disk = AD

1. If PS:AD = 1:1 then Diffraction effects will be spread over immediately surrounding pixels due to interference and "contamination" from outer secondary diffraction rings of the AD.

2. If PS > AD then diffraction is localized to the larger pixel and averaged with less spillage.

3. If PS:AD =< 1:2 then I would expect the "contamination" to subside and image quality increase to the point where actual diffraction is being sampled more accurately and the sensor is not affecting the negative outcome as much as where PS:AD = 1:1.

4. The larger the aperture the smaller the AD, thus moderating adjacent pixel contamination. This is why, when the AD approaches PS, diffraction and "spillage" starts to be more noticeable at larger apertures (i.e f8 instead of f/11 etc) especially when higher and higher density sensors are being used.

Well put!

---

Last edited by bdery; 11-22-2013 at 06:11 AM.

[bossa]

It is a persistent myth that diffraction somehow makes a higher resolution sensor resolve less than a lower resolution one at the same aperture. Diffraction is always there and of equal size airy disks at a given aperture, regardless of resolution or even sensor size. In a way you can think of diffraction as a "detail" that the higher resolution sensor can resolve, but the lower resolution one cannot.

Here's an analogy:

Nowadays lots of actors/actresses/models and makeup artists fret over skin imperfections. In the days of SD and lower resolution, they didn't care so much because they couldn't be resolved. With the advent of HD (and even more so now with 4K), it becomes easier to see blemishes and such if you look close enough. Did HD video make those imperfections appear? No, they were always there. Is the HD video somehow worse than SD video because it is possible to see them? From a technical point of view, one would be crazy to think so (from an artistic point of view it's debateable).

The takeaway: with a given lens at a given aperture, a higher resolution sensor will always resolve at least as much real detail as a lower resolution one of the same size, usually more. No exceptions (ignoring technical limitations of the sensors such as noise, etc.). If you find a good aperture to use on the K-5, use the same on the K-3. Your picture will never be any less detailed as a result of diffraction.

If you are replying to me; I am not disagreeing as I understand the diffraction is always there it's just that pixels affect how that information is recorded.

[Cannikin]

If you are replying to me; I am not disagreeing as I understand the diffraction is always there it's just that pixels affect how that information is recorded.

I wasn't replying to you, but rather the thread in general (and the threads that endlessly show up on diffraction), as there is a recurring question going around here and elsewhere due to these so-called "diffraction limits": Is diffraction going to limit what apertures I can use on the K-3 compared to the K-5 due to resolution loss?

The clear and simple answer is no. Diffraction will never cause a higher resolution sensor to show less real detail than a lower resolution one with the same lens and same aperture.

[DeadJohn]

... If you find a good aperture to use on the K-5, use the same on the K-3. Your picture will never be any less detailed as a result of diffraction.

Excellent summation. Here's a quote from the link in post #1: Just because the diffraction limit has been reached (with large pixels) does not necessarily mean an image is any worse than if smaller pixels had been used (and the limit was surpassed); both scenarios still have the same total resolution (even though the smaller pixels produce a larger file).

In practical terms, f8 is a good default for the sharpest images because most K-mount lenses are sharpest around f8. That's where the K-3 should show the biggest resolution boost over the K-5. When you stop down to f11 and beyond for landscape or slow shutter speeds, diffraction starts to set in and the K-3 and K-5 will give more equal images when viewed or printed at the same size (if you zoom all the way in, the K-3 will seem less sharp because of its smaller pixels)

---

Last edited by DeadJohn; 11-22-2013 at 06:45 AM.

[bossa]

A 24.3MP APS-C is diffraction limited at f/8 where pixel size governs the maximum size of the AD.

Diffraction Limited Photography: Pixel Size, Aperture and Airy Disks

[Barry Pearson]

Diffraction is, in fact, why our lenses can create images. Diffraction is the phenomenon by which a light ray changes its direction when changing its environment (going from air to glass, to another type of glass, to water, etc). The principle of the diffraction limit can be taken to mean that when the aperture gets really small, resolving power decreases.

Diffraction isn't about changing environment in those ways, and isn't why lenses can create images. (You may mean "refraction"?)

Diffraction occurs even if there is just a hole in opaque material.

[bdery]

Diffraction isn't about changing environment in those ways, and isn't why lenses can create images. (You may mean "refraction"?)

Diffraction occurs even if there is just a hole in opaque material.

LOL you are right, of course, I replied while feeding my baby girl (I shouldn't do that) and mixed up concepts. I'm ashamed but thankful for the correction. I will edit my post.

[RonHendriks1966]

Wow I had really not much knowledge when starting this thread.

[Clavius]

Wow I had really not much knowledge when starting this thread.

I understood quite a bit before I started reading this topic.

[Joojoo2010]

All i undestand f8 is maybe good idea...But we did know this already yes? too much Theory makes my head hurt and theories are not much useful at moment when taking pics. Must admit this is intresting anyway...

[rawr]

My understanding is that the diffraction effect means only a progressive degradation of resolution past a certain peak, until you get into f22 or f32 territory. The impact of this effect on resolution is very dependent on lens quality and other factors. It doesn't mean that once you go past f8 on a K-3 with any lens, image quality will fall rapidly to zero, so that by f11 you have an unusable image.

To dispel any anxiety about how diffraction may somehow limit the sharpness of the K-3 relative to the K-5, dxomark.com lets us compare use of the same lens (Nikon 50mm f1.8 G) on a Nikon D7000 (16MP) and a D7100 (24MP). In general, sharpness of the lens doesn't suffer on the D7100, as you can see from the overall score. And if you look into the details, eg at the sharpness field map, even at 'diffraction limited' f8 or f11, the D7100 performs better than the D7000.

[Na Horuk]

You know, on the one hand, I keep hearing that diffraction can ruin a photo, make it look very soft, dull. But on the other hand, I've seen some amazing shots taken at f16 and f22.
I just did the pp challenge, and it reminded me of this thread. the photo is taken with 18-55mm at f22 and it seems underexposed. But with the proper processing, you can see so much detail. Lights on cars, on construction site, windows, sidewalk, girders on the bridge..