diffraction limits, DX vs 35mm format

[Brooke A]

Yes, I was referring to the same F stop. It is interesting to read a lot about diffraction and f stops with the DX format. I believe the conversation is academic, and not so much practical. But really, as was pointed out earlier, the "f stop number" means nothing. ie.f/32/64 8x10, f/16 4x5, f/11 35mm, & f/8 DX. Some lenses a squeek betterr than others....

My point was, as a portrait photogrpher, I am going the other way, usually limiting my depth of field. In the DX format I have to use big heavy glass to get the maximum quality out of my D2X at the larger aperatures - f/2-f5.6 is usually my range. In the studio I can use f/8 quite a bit. I do try to carry as lite as possible when out on the road.

I suppose the format is only relative, as the format increases in size, we are stopping down more to gain depth of field, and the relative diffration limit for each format given equal depth of field seems to be about the same.

What I hope Nikon will provide with their addition of 35mm format in digital, with gains in better CCD/CMOS technology, would be more (larger) photo sensors, producing less noise and higher resolution.

As we have all seen, Megapixels is not the full sotry when it comes to outstanding digtal negative quality. I would take a 16 mega pixel DSLR with low noise, high usuable ISO's, and superior dynamic range, over a 22 mega pixel DSLR any day.

Doesn't craming smaller photo sites, more densely packed CCD's contribute to the defraction problem as well?

--
Brooke
D2x, F3, FA, & M3

 

[olhovivo]

I wish to step into digital with the D200 and need enough pixels to print (laser/lambda) up to 40x60cm with enough quality for an exhibition.

And should I first resample to 200dpi and then resize to the desired size?
--
Cheers,
Ricardo

 

[BJL]

My point was, as a portrait photogrpher, I am going the other way, usually limiting my depth of field.

That is another issue, where for some, the need for even lower minimum f-stops in DX format than 35mm can be an issue. I am one of many who are happy to do portraits at f/2.8 to f/4 in 35mm, so about f/1.8-2.5 gives the same shallow DOF in DX, but I know some want heavier background blur than even a 50mm f/1.2 could give in DX format.

But diffraction is irrelevant to image quality at that limit of large apertures and low DOF, and will be unless pixel counts get vastly higher than at present. For a portrait taken in DX format at f/2, diffraction will not effect resolution unless pixel counts reach about 200MP. I predict pixel counts will not reach 100MP in any DSLR format, including 645 medium format.

 

[BJL]

and the effects of uncorrected cubic terms to the linear approximation of sine makes sense.

 

[BJL]

But every now and then I do encounter a situation where I'm fighting between DOF and diffraction. Usually this occurs with telephoto or macro lenses, not wide angle.

Thanks for the reply. The closing jaws of DOF and diffraction limits on image sharpness and detail are exactly the reason for my prediction that 35mm or any smaller formats will never go very far beyond about 20MP worth of resolution, and certainly not to as far as 50MP. I am also very interested in macro and wildlife telephoto, places where, as you say, this hurts most.

Doubling linear resolution (from 12.5MP to 50MP) only pays off when one doubles enlargement, to view the extra detail. Adequate DOF as judged on those larger prints is likely to requires about halving the maximum acceptable aperture diameter (doubling lower f-stop limit) for a given shot. But doubling resolution also requires halving maximum acceptable diffraction spot size, or doubling the minimum acceptable aperture diameter (halving upper f-stop limit).

Changing formats scales both f-stop limits up roughly in proportion to focal length and format size, but leaves the aperture diameter limits about the same. So the latitude between apertures too big for adequate DOF and apertures too small for adequate resolution due to diffraction limits is likely to close by a total of two stops for each doubling of resolution (quadrupling of pixel count).

High resolution large format work often has to use camera motions to get enough of the subject in sharp focus while using apertures large enough to control diffraction (going beyond about f/22 in any format throws away a significant part of the resolution capability of modern films, and even more so with digital.) I expect 50MP+ digital work will likewise require a lot of use of camera motions, and so probably be limited to special tools like 645 format view camera systems.