Can Someone Explain: Diffraction Limits of Dslr BODIES?

[NikonUser]

Hi there,

In my travels around the net reading various reviews on the D200 and D2X I have read things about the bodies being 'Diffraction limited'

One such reviewer Thom Hogan if I remember correctly says that the D200 is diffraction limited at F13 and the D2X at F11. This I think has something to do with the pixel density of the CCD/CMOS

Can anyone explain to me:

1) If this is a real phenomenon
2) What does it mean
3) Does that mean that for more DOF stuff (eg the macro's I love to shoot) a lower Megapixel body would be more suitable? At (for example) F16 would a D70 make a more detailed and sharper picture than the D2X?

Thanks to anyone who can help clear this up for me

The D2X (or D3X by the time I'm ready) is looking like the camera for me but if the higher MP bodies means I can't shoot macros at more than F11 then that will really suck!!

Paul

[DaveB]

Does this link help?

[MattC]

The best explanation that I could come up with is somebody elses explanation.
http://www.cambridgeincolour.com/tutori ... graphy.htm

Cheers

[MattC]

Got to be quick around here

[moz]

the bodies being 'Diffraction limited'

I think that means is that the circle of confusion from diffraction is the same as the pixel pitch at that aperture. So you can take photos below that, you just won't see an increase in sharpness (instead it will decrease). The only solution is a bigger sensor with bigger pixels.

This is why Nikon is stuck at 10-12MP and Canon is stuck at 16MP - they've met or exceeded the resolution limits of the glass they make, and in some cases they're counting single photons to get the resolution asked for. I read a review the other days that said that even if the Canon 16MP sensor was 100% efficient, it could only get 16 bit res with very fast lenses... there just aren't enough photons hitting the sensor :) I think it was linked from Luminous Landscape.

)edit: sheesh, you do have to be fast on these things. And yes, points off for misleading use of CoC)

[DaveB]

Matt, looking at the timestamps you only just missed!

[NikonUser]

Yup that link and explanation helped. Thanks

Usinng the calculator at the bottom of the page:

The bigger the print the less effect the diffraction limit has (F16 on a 10 inch print viewed at 25cm is diffraction limited whereas a 24 inch print viewed at 1m isn't on a D2X)

So can anyone tell me if an F16 print on a D2x will be softer than the same print from a D70 at A4 and A3 size?

Paul

[Gordon]

Diffraction effects arent everything, amateur astronomers taking images of planets often work at up to f/100. Image scale is also important. A very small image of a planet at f/5 or f/10 will generally show less detail than a larger image at f/50, which is why eyepiece projection is often used to get huge focal lengths (200mm aperture@f/50 = 10000mm!)
Of course your telescope's tracking needs to be near perfect, and the atmosphere needs to be very steady (seeing under 1") to get good results, which is why this is a difficult area of astrophotography.

cheers, Gordon

[NikonUser]

I'm guessing you can't handhold at 10000mm then

How on earth do you keep that steady? A tripod of stone?

Paul

[Gordon]

Through a telescope
Telescope mounts are much more stable than any photographic tripod.
As an example, I can be sitting on my 45cm telescope mount while it tracks the sky

Gordon

[Steffen]

Diffraction effects arent everything, amateur astronomers taking images of planets often work at up to f/100. Image scale is also important. A very small image of a planet at f/5 or f/10 will generally show less detail than a larger image at f/50, which is why eyepiece projection is often used to get huge focal lengths (200mm aperture@f/50 = 10000mm!)

I find this to be a bit of a marketing spin, too. With telescopes "diffraction limited" is the term for ultimate optical quality - the resolution is only limited by diffraction at the aperture and not by optical aberrations or misalignments. Introducing this term into photography seems a bit sus.

At least until they make sensors that are "diffraction limited" at the best lenses' optimal apertures, like f/5.6 or f/8. Then it would mean something (that the image resolution is limited by the lens rather than the sensor's pixel density).

Cheers
Steffen.

[Gordon]

Yes, the diffraction limited claim is often BS. I should mean the mirror surface is within +/- 1/8 of a wavelength of its ideal shape (parabolic for a Newtonian), but a lot of mirrors aren't.
The theoretical limit is 115/aperture in mm, if my memory is working...often not the case when I have just got out of bed! So a 230mm aperture telescope should resolve 0".5 (half an arcsecond). You should be oversampling by about 2X in order to really get this resolution, ie about 0".25 pixels. In practice, the unsteady atmosphere will rarely let you get this resolution, but in some locations you often can, such as the high desert mountains in Chile, where a large number of telescopes are located. At Siding Spring typical seeing at the 1m telescope is more like 2-3". The 3.9m AAT people say they often get better than this, probably due to their telescope being a long way above the turbulence near ground level, but rarely is it as good as in Chile.
I could rave on for ages about this stuff, but I have to get the car loaded and drive to Newcastle area this morning for mountain bike racing today and tomorrow...

Gordon