[newarts]

Say one is displaying a Pentax K-x image at 300 mm width. The K-x's sensor is 23.5mm wide, so the sensor's image is enlarged by a factor of 300/23.5=12.8x.

The viewer's resolution is about 0.25mm on the display which corresponds to a spot size on the sensor of 0.25mm/12.8x=19.6 micrometers.

At a macro magnification of 1:1 the diffraction spot size on the sensor is related to the f-stop by:

19.6=(4/3)f(1+1), or f = 7.3 based on Airy Disk diameter for greenish light including magnification effect

This means that above f:7.3 the diffraction spot size on the display is larger that the viewer's resolution, hence the image loses sharpness.

Notice that the sensor's pixel count did not enter the discussion because the display size was fixed and the sensor's pixel count was more than needed to fill the display at the desired resolution.

Sensor pixel pitch would be controlling if we had specified a 100% crop for example - in that case the diffraction spot size would be compared to the sensor pixel size, not the display pixel size since for a 100% crop each sensor pixel is mapped to each display pixel. The K-x's pixel size is about 5.5 micrometers so a 1:1 macro's diffraction spot size equals the pixel size at about f:2!

Dave

PS The above is a simplified version of what's going on because of details of displaying digital images with digital devices.

More stringent display resolutions (like for a high quality print) would have required a smaller f-stop.

---

Last edited by newarts; 10-15-2010 at 04:54 AM.

[falconeye]

f/32 yes, but f/40

Well, f/40 is the effective F-stop at f/32 and 0.25x magnification (as ar as exposure and diffraction is concerned, not DoF tables though...). However, it would be an astounding observation for a jewellery photographer. Maybe, Nikon writes effective f-stops into EXIFs?

[RioRico]

Life is full of compromises. I have very few lenses that go beyond f22, and none that go past f32.

I just checked all my A+type zooms on my K20D. The results:

DA18-250 and Sigma DG170-500 go to f/45.
DA18-55 and FA100-300 both go to f/40.
DA10-17, F35-70 and A35-80 go to f/32.

Until I checked the DA10-17 I thought maybe there was a divide between F and FA lenses, but I see that ain't so. Anyway, check your A+type zooms and see how far past the diffraction limit you can go, and to what effect.

[yeatzee]

I haven't shot above F/8 in forever so I thought I'd take a quick shot just to remember why I do photo stacking....

Here's a crop of a photo where I stacked 5 images at F/5.6


Here's a photo where I did no stacking and shot at F/16


yes they both are in focus. These are not controlled tests, just an observation with decent examples The stacked crop is also cropped significantly more than the other

[WerTicus]

how do you stack focus? what photoshop function is that?

[Clarkey]

Tell your friend to get a 645D. Issue solved.
I regularly shoot at F22 on the Fuji GA645.

[yeatzee]

how do you stack focus? what photoshop function is that?

Basically you take multiple images with different focus points and combine them to give greater sharpness and dof. I do not own any form of photoshop so I'm not one to answer that

[reeftool]

Tell your friend to get a 645D. Issue solved.
I regularly shoot at F22 on the Fuji GA645.

Diffraction will occur at a different point when using larger format cameras and lenses. It becomes less of an issue the larger the camera and lens because they are just physically larger.

Here is an f/32 shot taken with my DA 10-17 and K10D. I played around with different apertures with this scene when shooting for the Single in September challenge last month. The shot I chose for my album was an f/22 shot. I tend to try stuff that people tell me not to do and I will stop my lenses down all the way from time to time. The first thing my grandfather told me 50 yrs ago when he gave me my first camera was never shoot into the sun. I didn't listen to that one either. This is also downsized and that will also affect sharpness.