DPReview TV: Understanding diffraction

The size of the aperture is related to the focal length and the f-stop. The aperture is FL/f-stop. So a 60m lens at f19 has an aperture diameter of 3.2mm.
A telescope of say FL=1200mm at f50 would have an aperture of 24mm, which is more like the 60mm lens at f2.4 (25mm)
So for long focal length lenses / telescopes you can use higher f numbers before seeing the effects of difraction.

@Grimstod, interesting that you say that the "science doesn't add up," when the article you cite says the following: "When it comes to diffraction at small apertures like f/16 or f/22, it is true that your IQ drops at those levels. Diffraction puts an upper limit on the resolving power of a lens, and each narrower aperture lowers that limit. Diffraction, however, blurs an image in a very even, very predictable way. That means softening caused by diffraction can fairly easily be corrected with some sharpening while post-processing." This fellow believes sharpening solves it, not that the science isn't accurate. Test it yourself. Do a series of photo with high-frequency information (i.e. blades of grass, needles on a fir tree, etc.) and run through various apertures, then look at them magnified (what a large print might look like), and then try to sharpen the softer images and see what you get. I have tried this and it is beyond reasonable question that diffraction is real.

Read this and get a much better perspective on the truth of diffraction. There’s very little validity at all in the above posted video.
https://jonrista.com/2013/03/24/the-diffraction-myth/

@MrBrightSide: “deconvolution … has made diffraction much less of a concern”

Correct — if you replace ”much less” by “a bit”. Diffraction is two-prong: at some (spacial) frequencies (or: “at some level of details”), it just weakens the signal. This decreases S/N ratio — which cannot be “undone” — but if your noise was very small, this may not matter.

However, at smaller spacial frequencies, the CUT-OFF IS ABSOLUTE. No signal passes at all. No matter how small your noise is, there is NOTHING TO RESTORE by deconvolution.

If you understand these two issues (the S/N ratio and the absolute cut-off), then indeed deconvolution may give a significant visual improvement (I would say, make things “about 2 times sharper”).

@stevevelvia50: interesting that in "difraction myth" article didn't mention magnification at all. Posted video primarily touch the theme of difraction in much more bigger (real) apertures than F22 or F32 - the biggest one that "myth article" know about.

Use F22 aperture and 10x magnification in real world (real aperture no. F242) and be happy with results from sharpening algorithms that maked up details which are not taken from results.

Instead of photographing it right way.

My concern is details of tiny insects. Deconvolution software can change dull pictures into crispy ones, but it cannot bring to light details that have been lost by using a too small aperture. Test your camera and lens combination and conclude what works for you.

I don't know which Panasonic camera you tried for auto focus bracketing, but even my old gx8 which only uses highly compressed 4k video can turn out really impressive results with good stacking software.
So the g9 which shoots much improved 6k should be capable of turning out stunning results!
But I moved to the Olympus em1-mk2 (before the g9 was launched) and it it shoots ful resolution RAW & jpg brackets with as fine or as coarse a step as required and you can program it up 999 shits.
Results are amazing. What's not to like :-)

@Don you got your math a bit wrong. A 20MP m43 sensor has the same pixel density as a 80MP FF sensor

Actually, we're both wrong!

MFT sensor: 18 x 13.5 = 243
Full Frame: 36 x 24 = 864
864 / 243 = 3.56

Some sensors are 17.3mm x 13mm, and that creates a number of 3.84x

I digress, I was thinking in one dimensional numbers, but there's your more accurate ones based on the actual sensor sizes. :)

Don I think you made an important mistake. When comparing across sensors you naturally want to compare at equivalent FL, and the 1/2x f number of M43 (or 2/3x for APSC) comes from the same physical aperture meaning different F numbers to different FL.
If you view different sensors as crops from the same lens, then the 80mp 3:2 FF sensor faces the same amount of diffraction as the 20 mp 3:2 M43 sensor, i.e. both start to see diffraction at f5.6 and you cannot get away with closing down further on FF.

You'd get the same amount of diffraction in either case. The light, regardless, still needs to pass through an opening of the same size.

HjVN If your F2 lens and F5.6 lens have the same focal length (say 50mm in this case) then both at F16 will have the identical diameter hole in the iris ...the hole in the iris is calculated by dividing the F number with the focal length
Now your F2 50mm lens wide open will have an iris diameter of 25mm ..so 50 divided by 2= 25mm your F5.6 50mm lens wide open will have an iris diameter of 8.9mm
now both the 50mm F2 lens and the 50mm F5.6 lens both stopped down to F16 BOTH will have an iris diameter of 3.1mm

i imagine a pin hole that small will completely upset the optical formula of the lens .it may work like another element in front of the lens
but the mathematical formula for calculating diffraction (i am not an expert on algebra i read the accompanying text) say its the ratio of the size of the light wave and the hole the light is going through that determines the amount of diffraction ..i am guessing that the fact the pin hole will not change the size of the the light wave and the 3.1mm iris is unaltered in size than the diffraction formula will still give the same answer, As the pinhole is smaller if its calculation is made it may have more diffraction and add to it
an afterthought as waves can cancel each other out if they fall wrong ..If the light went through 2 holes 1 after the other at a set distance apart but the spacing may be very critical like to a fraction of a light wavelength maybe it will undefract the light but what do i know i only went to school sometimes

@Don Komarechka

Diffraction is correlated to DoF. If you aim for the same DoF you will have the same amount of diffraction regardless of format. If you achieve that by a large aperture number on a large format or a small aperture number on a small format doesn't matter

For the same FoV and subject distance, DoF is directly correlated to diffraction

Ok ilza, how does one use what you wrote in the real world? ;)

"larger format film at smaller apertures will show less evidence of diffraction on a print because the physical area of the film is larger"

Not if the print is the same size.

The same goes for higher resolution. The diffraction is only more noticeable if you view the image at higher magnification/ print larger.

@Don
Not if you match the aperture number for DoF

The "effective aperture" is not just the amount of light required, this number also has a direct impact on the way diffraction occurs in the lens. The two factors (amount of light needed and diffraction) are tightly linked when you're dealing with an increase in magnification.

Easy to test this out practically as well - just shoot at 1x-5x @ F/8 and see the quality start dropping as the magnification increases. Diffraction at work. :)

Your test moves the aperture further from the film/sensor plane, you can obtain more magnification in other ways that don't move the aperture further from the sensor plane.

This is true - and in most scenarios the aperture will change position when your magnification increases well beyond 1:1 lifesize - such as adding extension tubes, etc. - it's all part of the process.

I'll be keeping in touch with you by PM's as when the covid lockdowns lift I'll be finishing a few photo-scientific projects you would be interested in. On another note I recall Venus Optics making the claim that wider angle lenses tend to have less diffraction, I suspect that is because the aperture is much closer to the film plane. Now by less I don't know how much less, I've never tried to find that out.

There are SO many factors to consider here. Pupil ratio is an element we didn't discuss in this video which can have a dramatic impact on the calculation of effective aperture and also impacts diffraction - this is also a factor to consider when using close-up filters. A scientific rabbit hole that would have been lost on the majority of the audience of the video. :)