UV, Diffraction and Sensor Size

[DaveO]

ADMIN NOTE: I have split Dave's question off into its own thread because it is interesting. I think the discussion will be useful. We need some facts - and any relevant user experiences. Please do stick to the Diffraction topic rather than veering off into which cameras are better for UV.

 

********

 

Hi Bjørn,

Many moons ago you told me I would probably one day want to get a Nikon to put my UV-Nikkor on to get infinity focus. Since you use both APS-C (D3200) and Full frame (D600) I wondered what the trade off was in using the APS-C sensor, does the increased risk of diffraction leading to softer images restrict what you can do with the D3200? I would look at the D7100 versus the D610.

 

Cheers,

 

Dave

[Shane]

Dave

 

diffraction in UV is less of an issue than at visible wavelengths.

 

Here are some approximate diffraction spot calculations:

550nm: f/5.6 7.5um

f/8 10.7um

f/11 14.8um

 

380nm: f/5.6 5.2um

f/8 7.4um

f/11 10.2um

 

The shorter the wavelength ~ the smaller the diffraction spot

 

(sorry about the formatting, not only did this post halfway through writing it but formatting went out the window too)

 

D7100 ~ 3.9um sensel

D610 ~ 5.9um sensel

 

The big debate is what factor times the sensel size does diffraction become an issue and the consensus varies from 1.4x sensel size up to almost 3x the sensel size, but whatever the x factor is, the D610 is less effected.

[Andrea B.]

Thank you, Shane.

Having some example numbers to show the difference between Visible and UV diffraction is very enlightening.

 

Can you also explain briefly for the general reader what sensor size has to do with lens diffraction ?

 

You list the disk at 3.9microns for the DX sensor size of the D7100.

And 5.9 microns for the FX sensor size of the D610.

So, for example, is this with same lens at equal distances?

Does these measurements have to do with the user's perception at certain distances or is it a raw measurement?

(I'm not sure I asked that correctly.)

 

The 3.9 microns number for the D7100 and the 5.8 microns number for the D610 were not labeled.

Now I think you were referring to the pixel (sensel?) size with those numbers, yes??

If you would please relabel what might have been lost, thank you. "-)

 

(BTW, I'm sorry our forum software is so bad at formatting.)

[Alex H]

The sensor size does not matter. The pixel density does. Both D610 and D7100 have 24 Mp sensors, but their absolute size is different. Hence, pixels on the D7100 sensor are arranged closer to each other - see the numbers that Shane gave. As a result, diffraction spot of any given size will "cover" more pixels of the D7100 sensor, thus affecting the sharpness of the image.

[Andrea B.]

That was very clear. Thank you.

 

Apropos of the D610 vs. D7100 reference - in other words, the FX vs. DX cameras, to borrow the common Nikon designation - I have found it quite useful to have a broadband camera of each type - in my case the D600 and K5, respectively, although brand is not important here.

 

Bjørn showed me some illustrations of improved detail in photos from a denser sensor DX camera. But of course that has to be balanced against the diffraction issues - as well as a huge host of other variables. In short, both FX and DX are useful. Fortunately the smaller sensor cameras tend to be less expensive, so it was not too painful to add one.

[Shane]

Interestingly enough the D200 and D600/610 sensels are the practically the same size.

 

I prefer to use sensel at the sensor level and pixel at the interpolated image level. Sure you still have the same number of pixels as sensels but they do not contain the same information due to interpolation. I never thought the distinction that important until I started looking at, CoC, diffraction diameter, defocus blur and pixel/sensel size.

[DaveO]

Thank you. I have agonized over the diffraction/Airy disc question ever since I first read of it in relation to digital images some years ago. I never came to any conclusion because it ultimately depends on what you are going to do with the image and so how large you will be viewing it, as a mere screen resolution image or a 30 x 40 cm print. I think I decided it was unlikely to be noticed, apart from a softening of the image, in my prints. I once did a series of shots with my Pentax K20D (14 MpPx) at 1:1 using my Zeiss Macro Planar (with visible flash) and sort of convinced myself that the texture of the fibres in the sheet of paper in my target were sharp down to f/11 but perceptibly softer at f/16. Of course that took me down the path of looking at stacked shots @ f/11 rather than trying for increased depth of field from f/16. It's on my to-do list to compare visible and UV light. I'm sure my UV shots with my quartz lens are sharper than my visible ones @ f/16 but I haven't really pixel peeped with the UV-Nikkor.

Dave

[nfoto]

The manifestation of diffraction also depends on the lens. Apart from the inevitable lowering of contrast some lenses hold up particularly well when stopped down while others turn the image into mush. The Coastal 60 mm f/4 APO is capable of making pretty sharp UV images even at f/45 if you adjust for the lower contrast in the processing later.

[Andrea B.]

It is very instructive to make yourself a foto series at all apertures with a particular lens and then pixel peep comparatively at 100% to look for the effects of diffraction. Then you can experiment with how sharpening in the editor can mitigate it - or not. And you can get a feel for how far to push a particular lens. And how your prints might fare.

 

The UV-Nikkor holds up rather well also.