[bymy141]

Diffraction is an optical phenomenon that is linked to the lens and not the sensor. The diffraction limit is something used to decide what aperture to use on the same sensor, not to be compared against other sensors. I made a crappy graph here: http://imgur.com/9xtyR.png

The physical size of the airy disk that causes the softening is independent of the medium that it is being projected on to. A higher resolution sensor will capture more softening on a per pixel basis, but the overall image will not be softer, as the absolute physical size of the airy disk being captured will be the same on both sensors.

For more information, see the discussion here:
How do you find out the "sweet spot" of a lens - Photography - Stack Exchange

I understand the concept (I think).

But if you are right, how can the images taken by the D3 and the D3x of the same subject with the same sensor size, resized to the same smaller frame, look so very different?
It should then look the same, not?
How do you explain that?

Nevertheless, the lesson that I've learned here is:
1) At 16Mp your cropping will not only be limited by the sensor resolution and the lens shaprness. You should not work with apertures that are too small.
2) Perhaps I need to learn how to do focus stacking for macro photos...

- Bert

[bymy141]

Hi bert,

Life is full of compromises. . . .

What great learning experience to discover just how much there is yet to learn. . .

Scott

Yeah Scott, I agree with both...

[raz]

Changing the image output to 10Mp will not change the sensor pixel size.
So I do not see how that will change the physics here, or do I miss something?

- Bert

Chaning the image output in camera should resize the image taken at 16Megapixels to 10Megapixels inside camera (this is what I know - the camera takes a picture full size then resize it to the output desire size - it will not take the image with one pixel then the neighor one will be ignored). Someone correct me if I'm wrong.

But that isn't so important. The thing about difraction is that too many pixels are "batling" over a very small portion of glass.

So if you get difraction in a photo - wich basicly means a soft image, just resize the image to a small resolution and the image will "look" sharper on your screen.

But if you want to print it on A4 paper for example, even without resizing, a F/40 image taken from K-5 will look better or at least equal than a F/40 image taken by K10. Even if for you on the screen, pixel peeping, the K-5 will look softer.

[melander]

Still, stop looking at prints from 10 inch distances

[philbaum]

Here's a diffraction quote i posted here in Jan 2009:

Diffraction Article by Lloyd Chambers

--------------------------------------------------------------------------------

The Jan/Feb 2009 issue of Photo Techniques, Photo Techniques Magazine - The Magazine dedicated to professional photographers, has an excellent article on: Diffraction: Resolution Taxed to its Limits

The author explains:

" actual image detail is constrained by optical performance: overall sharpness and depth of field require stopping down, but stopping down too far degrades image quality due to diffraction, an optical effect that puts an upper bound on resolution. Even worse, well before the resolution limit is reached, the image has long since been declining in contrast (whites and blacks become grayish), which we perceive as lower resolution, and a loss of "snap"..."

He provides this table of recommended aperture limit for best contrast:

Pixel size: 8+ microns, Min aperture: f/11 Ex cameras: Nikon D3/D700, Canon EOS 5D

Pixel size: 5 to 8+ microns, Min aperture: f/8, Ex cameras Canon 1D, Nikon D2x, D300, any 10+MP DX/APS

Pixel size: 2-4 microns, Min aperture: f/4, Ex cameras: Numerous digicams with 8+ MPs
====================

The final Conclusions of the article are unusually blunt:

-"As megapixels increase, diffraction will become the dominant factor limiting image sharpness, unless and until improved optical designs allow near-diffraction limited imaging at f2, 2.8 and f4. Such lenses are feasible but will be larger, heavier and much more expensive than today's optics

-To paraphase an old maxim: f/8 and stop there. That simple rule will maintain optimal or near-optimal lens performance and image contrast resolution with today's DSLRs...stopping down to f/11 or f/16 is warranted with some subjects, but the contrast compromise should be kept in mind"

Notice that the table above is all about pixel size. As i remember the article from 18 months back, the problem with small photosite size is that as the airy disc gets larger and overlaps half way onto a photosite, its causing a flare like effect that washes out contrast and resolution, in that order. Doesn't matter whether one has downsized to 10 meg from 16 meg, the photosite size is still the same. (i think - this is far too complicated for me to be sure about it)

[kittykat46]

Here's a diffraction quote i posted here in Jan 2009:

Diffraction Article by Lloyd Chambers

--------------------------------------------------------------------------------

The Jan/Feb 2009 issue of Photo Techniques, Photo Techniques Magazine - The Magazine dedicated to professional photographers, has an excellent article on: Diffraction: Resolution Taxed to its Limits

The author explains:


Notice that the table above is all about pixel size. As i remember the article from 18 months back, the problem with small photosite size is that as the airy disc gets larger and overlaps half way onto a photosite, its causing a flare like effect that washes out contrast and resolution, in that order. Doesn't matter whether one has downsized to 10 meg from 16 meg, the photosite size is still the same. (i think - this is far too complicated for me to be sure about it)

Optical Sensor technology has moved on since that January 2009 article, which itself was based on limitations of even older sensor designs (i.e. cameras already in the market)
Take a look at the High-ISO performance of the latest K-R/K-5/D7000 cameras, and you can see that in action.
Sensor technology ultimately cannot break pass the laws of physics, of course, but the limits of what's currently possible are constantly being broken.

I cannot say often enough - Diffraction is an optical effect caused by the lens aperture specifically, accentuated by possibly other lens flaws.

All that happens on the sensor is that as sensor resolution gets higher and higher, those flaws, including diffraction get more visible.

Sure, were going to hit a limit with APS-C sized sensors soon, I think, because lens performance has not improved to the same extent.

I have a few El Cheapo old lenses which were OK on my 6MP K100D, rarely used on the 14MP K-7, and I think I will either throw them away or sell cheap on E-Bay when I move to the K-5.

For most users, my advise is - get a better quality lens.
Some lenses which were reasonably OK on the K10D may likely not cut it on the K-5.

[barondla]

Diffraction is easy to see. Shot some macro images looking down the length of a ruler. The idea was to demonstrate, for photo class, depth of field vs aperture. The tripod mounted K20D was focused at a certain distance and never refocused. Amazing how obvious the sharpness change is at the exact focus point. Shot from f2.8 to 22. Past about f8 the focused point starts getting softer. This was viewable on a Pana 27" tv (standard definition). The class was sitting at least 8 ft away. Everyone saw it. Try it.

thanks
barondla

Check out POINT & SHOOT CONTEST #34 in compact camera forum. Enter #35. Any brand camera. Enter now!

[Kirivon]

Hi,
There are several references to other Internet sources on the subject of diffraction.
One being: Diffraction Limited Photography: Pixel Size, Aperture and Airy Disks
It has a diffraction limitation calculator...

I filled in the diffraction calculator for my planned K-5 purchase. Assuming:

Maximum print size: 30 cm (appr. A4 paper size / my PC screen size)
Viewing distance: 25 cm
Resolution: 16 Mega pixels
Camera type: DSLR 1.5X

After trying several apertures, the calculator shows that the images will be diffraction limited starting from F16!!

You know that calculator doesn't use the megapixel value for anything, right? Put in 1mp then 1,000mp and see if you see a difference in diffraction limit. I don't see how the diffraction limit should be a surprise. I figured that one out a month after I transitioned to digital with the K100D and was shooting with film sunny 16 rules. All my photos were very soft and I wasn't sure why! IMO the practical limit on APS-C is f/11.

As for whether or not the smaller pixels are worse, here's an answer straight from the very page you linked:

"Are smaller pixels somehow worse? Not necessarily. Just because the diffraction limit has been reached with large pixels does not mean the final photo will be any worse than if there were instead smaller pixels and the limit was surpassed; both scenarios still have the same total resolution (although one will produce a larger file). Even though the resolution is the same, the camera with the smaller pixels will render the photo with fewer artifacts (such as color moiré and aliasing). Smaller pixels also provide the flexibility of having better resolution with larger apertures, in situations where the depth of field can be more shallow. When other factors such as noise and depth of field are considered, the answer as to which is better becomes more complicated."

[jpzk]

I' m late reading this thread ... What's going on?
Already found a problem with the K5?

I thought diffraction was an optical issue and not an electronic one ... what Am I missing from my physics courses?

Found this on the Web, just out of curiosity:
http://www.bobatkins.com/photography/technical/diffraction.html

And this ... Is this what you guys mean:

[Steelski]

My experience tells me that defraction eating away at sharpness does occur, but its not that heavy unless you are using extreme F stops (above F16) and are not printing very large. (A3 and below) and when you come to those sizes, you will almost certainly not be noticing because of how far aways from the picture you have to be to view it well.
For many its a none issue.

[newarts]

The diffraction spot's size depends on magnification. The equation is:

D=2.44*f(1+m)*wavelength... m is usually omitted because it is very small for normal photography.

For greenish light, the equation is easy to remember:

D_microns = (4/3)f(1+m)

ie. the diffraction spot diameter in micrometers is about 4/3 the f-stop.


Therefore, the maximum f-stop before diffraction becomes a problem is decreased by a factor of 1/(1+m) from non-macro conditions.

Say a camera- perfect lens combination is diffraction limited at f:16 for normal photography; at 1:2 magnification it will be diffraction limited at f:11, and at 1:1 magnification it will be diffraction limited at f:8.

Dave

[falconeye]

Bert,

the answers to your questions:

Would a K-5 with only 10Mpixels not have been better??

No

Or should we all have to move to FF for 16 Mpixels in order to be able to use F22??

No, F22 on FF has same DoF as F16 on APSC.

And thank you for giving us a new synonym for idiot: jewellery photographer

[Asahiflex]

Having seen the official Pentax fullsize K-5 samples (some of which were at f/11) I can say that the diffraction is only a problem in people's minds, but not in real life. At least at that f/stop! And I never go beyond f/8 unless I really have to.

[Rondec]

As others have said, the issue with more mega pixels is that you can see more clearly lens flaws when you pixel peep. Diffraction does start making your images softer. I seldom shoot above f11 on my K7 because in general sharpness doesn't improve and at f16 it starts getting softer. At the same time, diffraction isn't a wall that you just can't shoot beyond.

This reminds me of when the K10 came out and suddenly people felt like their old lenses weren't as good on the 10 megapixel sensor (particularly the old kit lens). Good lenses (say, the limiteds) will look good when pushed hard by a high megapixel sensor. Not so good lenses will suffer.

[Digitalis]

I don't seem to recall any Nikkor lenses that natively go to f/40. f/32 yes, but f/40 would imply he is using a teleconverter which would also add to the degradation of image quality.

I use f/16 on my k-7 diffraction be damned!, besides a little USM can help counteract it's effects, though for any f stops over f/16 I will DOF stack. For macro work we have to struggle to get as much DOF as we can for some subjects, so stacking has really helped in the digital age, DOF stacking really wasn't possible in the days of film cameras.

Falk called it: idiot.