Working theory to combat diffraction of tiny aperture.

[L0n3Gr3yW0lf]

Hello, recently I was thinking of what could help with the diffraction effect from working at f 11, f 16, f 22 and f 32 (very few lenses can close down that much). One solution is working with either manual or automatic focusing rail, which is more or less mandatory for extreme magnifications and tiny subjects but isn't always an option.
Using a high megapixel count has the benefit of giving you more "information" (however you want to look at it), like 36, 42, 45, 47, 50, 61, etc Megapixels) but these sensors push the optics of a lens into diffraction territory sooner (sometimes f 11 or even f 8) and are more affected by the degradation of sharpness. Going "down" to 24 or 20 MP isn't something some people might want either (it could mean buying a different camera) and shooting in smaller RAW files does not solve the problem because diffraction is related to the physical size of the pixel (photodiode), the smaller the "pixel" the more significant the impact. Also, smaller sensor size cameras reach diffraction quicker, f 8 for ASP-C and f 5.6 for Micro Four Thirds, and the higher the resolution you push the smaller sensors the greater the impact as well. So, while you do get more DoF with smaller sensors, your image quality will be impacted quicker and the advantages of smaller sensor size equal out no matter how you try to look at it.

One thing that could help is High Resolution stacking. This is a limited feature to only a few brands and even within the few brands is limited to specific models of cameras, the ones that I know of (and with a bit of research but you should look into it if you are interested yourself):
*Sony a7r III: 42 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Sony a7r IIIa: 42 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Sony a7r IV: 61 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Sony a7r IVa: 61 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Sony a7r V: 61 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Olympus OM-D E-M5 Mark II: 40 MP or 64 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Olympus OM-D E-M5 Mark III: 50 MP or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*OM System OM-5: 50 MP (Hand-Held & Tripod/Still Mode) or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Olympus OM-D E-M1 Mark II: 50 MP or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Olympus OM-D E-M1 Mark III: 50 MP (Hand-Held & Tripod/Still Mode) or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Olympus OM-D E-M1 X: 50 MP (Hand-Held & Tripod/Still Mode) or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*OM System OM-1: 50 MP (Hand-Held & Tripod/Still Mode) or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-G9: 40 MP or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-S1: 48 MP or 96 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-S1R: 187 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-S1H: 48 MP or 96 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-S5: 48 MP or 96 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-GH6: 100 MP or 50 MP (Tripod/Still AND Hand-Held Mode with both resolutions accessible) In Camera Stacking Results Only
*Fujifilm GFX 100: 400 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Fujifilm GFX 100S: 400 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Fujifilm GFX 50S II: 205 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Fujifilm X-H2: 160 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Fujifilm X-T5: 160 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Pentax K-3 II: 24 MP (Tripod/Still Mode) External Stacking Software Results Only
*Pentax K-1: 36 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Pentax K-70: 24 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Pentax K-3 III: 25 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Pentax K-1 II: 36 MP (Hand-Held & Tripod/Still Mode) External Stacking Software Results Only

All of the cameras have certain limitations in High-Resolution mode so you will need to research if it fits your needs because some can only shoot on a tripod, some can only use an electronic shutter (fully or Electronic First Curtain), and some can only go to certain shutter speeds (high or low speeds limits), some can use only a certain ISO limit, some will not work with strobing flash and many other limitations.

The other "limitation" is companies choose to implement the process differently:
*Sony and Pentax use their HR modes to improve image quality by improving the information data one pixel can store. Sony and Pentax move the sensor to record Red, Green and Blue information for each pixel and reduce the interpolation necessary to record the full data per pixel. I think (and this is PURE SPECULATION) that this would give FOR MACRO output the best way to counter diffraction, by recording more accurate information to reduce the information loss from diffraction.
*Fujifilm and Panasonic are going the way of more are better, gathering as much data as possible, which does improve image quality, but it is more data-intensive from the bigger resolution. I can't compare them which process would be best, one would need access to all of (each model) cameras.
*Olympus does a bit of both, they record a lot more data AND increase the information per pixel level, from my own experience of using them in the past the files require a significant amount of sharpening in post to bring out the sharpness in the image but it has also quite a bit of artefacts to the images (repetitive patterns that can be more noticeable the more you work with them).

The 3rd problem of the diffraction equation (besides sensor size and resolution) is the lens quality. The higher the glass quality in the lens the more you can stop it down before diffraction affects it which also helps with being able to keep up with high-resolution sensors and HR stacking modes. Such lenses in the macro world can get incredibly expensive and there's no way of cheating out of the laws of physics (optics). (In a similar way reviewers and pixel peepers will advise you that there's no point in putting a Sony 200-600mm f 5.6-6.3 lens on a Sony a7R IV because the lens can't resolve 61 MP and that you need a prime like Sony 400mm f 2.8 FE or 600mm f 4 FE to get more out of that sensor).
Well, the struggle for smaller sensors is greater to get the lens quality to stand up to 26 or 40 MP of Fujifilm's X-T and X-H lines and the HR modes of Panasonic G9/GH6 or Olympus's OM-D cameras. (This is not meant for bashing smaller sensor cameras, it's only meant as an observation).

There are only 2 ways to combat diffraction, at this point, and that's either by better quality glass which is improving more and more (about) every 10 years and by techniques of photography, either computational photography like HR modes or by focus bracketing and stacking, either in camera or in software.

 

[Brian]

Diffraction is always present, and is caused by the wavelike nature of light. When the physical aperture is "around" 3mm, the effects of diffraction become noticeable.
Complex, multi-group, lenses have series of converging and diverging groups. Placing the aperture in the optical path of maximum divergence allows the light to by regulated and the diameter of the aperture to be larger. This means a larger diameter opening accomplishes the same F-Stop by locating it in the optical path where the light covers the largest area. Focus stacking and other techniques allow the lens to be used wide-open, ie minimize diffraction, and build up a virtual depth of field.
Zeiss used this technique some 20+ years ago with their microscopes, I think it was called "Deep View" or something like that. The first units promised could not be delivered. They had to cancel my procurement for almost $100K.

 

[lucien]

Wow !

 

[archaeopteryx]

Thread's been sitting for a while but I had a few minutes and thought it might worth stating a few things more concisely.

• A lens's resolution is the convolution of the blur disc created by its optical aberrations and the Airy disc created by diffraction. Sharper glass has smaller blur discs and therefore transitions to diffraction limited at larger apertures (lower f/numbers) than lower resolution glass. Some recent lenses are diffraction limited by f/4 in conventional lens testing and, as it's been a few years since I last checked, there might be a few which are diffraction limited at f/2.8.
• Assuming sufficient light to avoid high ISO noise, a sensor's ability to resolve lens-diffraction limitations is determined by its pixel pitch. Much of what's written above gets at this indirectly but IMO it's simpler just to calculate the pixel pitch of sensors of interest, the Airy disc diameter for Fraunhofer diffraction at apertures of interest, and compare them.
• Most compositions are Airy disc limited in most of the frame, regardless of lens or pixel pitch. If subject motion's not an issue and the objective's to maximize information capture by the sensor, the corollary's pretty much that everything should be focus stacked. Below close up magnifications, the number of frames a stack needs is small and most any simple depth mapping stacker will do well. At close up, macro, and photomacrographic magnifications the usual approach is to blend depth map and pyramid stacks in Helicon or Zerene.
• As @Brian touched on, cameras' high resolution pixel shift modes implement a subset of well established super resolution methods for optical oversampling. They all work the same way and are all subject to diminishing returns as they're over resolving the Airy disc. Total information capture does increase. It's a good exercise for understanding to calculate Airy disc size as a function of wavelength and walk through how that interacts with the RGGB placements of pixel shift steps.

Another high resolution method used in studio photomacrography is stack and shift, where a grid of focus stacks is collected with a telecentric lens on an xy table and then stitched together. This is an excellent choice if you need things like 300 MP images of dead beetles but, in the broader context of photomacrography, it's fairly niche.

computational photography like HR modes or by focus bracketing and stacking

Richardson-Lucy deconvolution is also commonly applied (GIMP G'MIC, darktable, RawTherapee, Topaz InFocus, and Photoshop off the top of my head, likely several others as well). Since photomacrography with microscope objectives commonly settles near effective apertures of f/20 (you can get f/16 if you don't mind spending several thousand per objective) I often put those stacks through a few deconvolution iterations for Airy disc reduction.

Zeiss used this technique some 20+ years ago with their microscopes, I think it was called "Deep View" or something like that.

FWIW, Keyence calls their 50 fps stacking depth composition. That's been around in various forms for about 20 years but, other than ZEN z-stacking, I'm not finding an equivalent name for Zeiss. The bits still up about DeepView make it sound more like an aperture management system.

More recently (from 2020), there's also the optical MALS on the Visioner 1.

 

[nivaunrahne]

I haven't tested it out but I wonder how well "pixel shift" will work at minimizing defraction. The XT5 now has that, maybe I will do some playing around with it.

 

[John King]

Hello, recently I was thinking of what could help with the diffraction effect from working at f 11, f 16, f 22 and f 32 (very few lenses can close down that much). One solution is working with either manual or automatic focusing rail, which is more or less mandatory for extreme magnifications and tiny subjects but isn't always an option.
Using a high megapixel count has the benefit of giving you more "information" (however you want to look at it), like 36, 42, 45, 47, 50, 61, etc Megapixels) but these sensors push the optics of a lens into diffraction territory sooner (sometimes f 11 or even f 8) and are more affected by the degradation of sharpness. Going "down" to 24 or 20 MP isn't something some people might want either (it could mean buying a different camera) and shooting in smaller RAW files does not solve the problem because diffraction is related to the physical size of the pixel (photodiode), the smaller the "pixel" the more significant the impact. Also, smaller sensor size cameras reach diffraction quicker, f 8 for ASP-C and f 5.6 for Micro Four Thirds, and the higher the resolution you push the smaller sensors the greater the impact as well. So, while you do get more DoF with smaller sensors, your image quality will be impacted quicker and the advantages of smaller sensor size equal out no matter how you try to look at it.

One thing that could help is High Resolution stacking. This is a limited feature to only a few brands and even within the few brands is limited to specific models of cameras, the ones that I know of (and with a bit of research but you should look into it if you are interested yourself):
*Sony a7r III: 42 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Sony a7r IIIa: 42 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Sony a7r IV: 61 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Sony a7r IVa: 61 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Sony a7r V: 61 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Olympus OM-D E-M5 Mark II: 40 MP or 64 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Olympus OM-D E-M5 Mark III: 50 MP or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*OM System OM-5: 50 MP (Hand-Held & Tripod/Still Mode) or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Olympus OM-D E-M1 Mark II: 50 MP or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Olympus OM-D E-M1 Mark III: 50 MP (Hand-Held & Tripod/Still Mode) or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Olympus OM-D E-M1 X: 50 MP (Hand-Held & Tripod/Still Mode) or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*OM System OM-1: 50 MP (Hand-Held & Tripod/Still Mode) or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-G9: 40 MP or 80 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-S1: 48 MP or 96 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-S1R: 187 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-S1H: 48 MP or 96 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-S5: 48 MP or 96 MP (Tripod/Still Mode Only) In Camera Stacking Results Only
*Panasonic Lumix DC-GH6: 100 MP or 50 MP (Tripod/Still AND Hand-Held Mode with both resolutions accessible) In Camera Stacking Results Only
*Fujifilm GFX 100: 400 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Fujifilm GFX 100S: 400 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Fujifilm GFX 50S II: 205 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Fujifilm X-H2: 160 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Fujifilm X-T5: 160 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Pentax K-3 II: 24 MP (Tripod/Still Mode) External Stacking Software Results Only
*Pentax K-1: 36 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Pentax K-70: 24 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Pentax K-3 III: 25 MP (Tripod/Still Mode Only) External Stacking Software Results Only
*Pentax K-1 II: 36 MP (Hand-Held & Tripod/Still Mode) External Stacking Software Results Only

All of the cameras have certain limitations in High-Resolution mode so you will need to research if it fits your needs because some can only shoot on a tripod, some can only use an electronic shutter (fully or Electronic First Curtain), and some can only go to certain shutter speeds (high or low speeds limits), some can use only a certain ISO limit, some will not work with strobing flash and many other limitations.

The other "limitation" is companies choose to implement the process differently:
*Sony and Pentax use their HR modes to improve image quality by improving the information data one pixel can store. Sony and Pentax move the sensor to record Red, Green and Blue information for each pixel and reduce the interpolation necessary to record the full data per pixel. I think (and this is PURE SPECULATION) that this would give FOR MACRO output the best way to counter diffraction, by recording more accurate information to reduce the information loss from diffraction.
*Fujifilm and Panasonic are going the way of more are better, gathering as much data as possible, which does improve image quality, but it is more data-intensive from the bigger resolution. I can't compare them which process would be best, one would need access to all of (each model) cameras.
*Olympus does a bit of both, they record a lot more data AND increase the information per pixel level, from my own experience of using them in the past the files require a significant amount of sharpening in post to bring out the sharpness in the image but it has also quite a bit of artefacts to the images (repetitive patterns that can be more noticeable the more you work with them).

The 3rd problem of the diffraction equation (besides sensor size and resolution) is the lens quality. The higher the glass quality in the lens the more you can stop it down before diffraction affects it which also helps with being able to keep up with high-resolution sensors and HR stacking modes. Such lenses in the macro world can get incredibly expensive and there's no way of cheating out of the laws of physics (optics). (In a similar way reviewers and pixel peepers will advise you that there's no point in putting a Sony 200-600mm f 5.6-6.3 lens on a Sony a7R IV because the lens can't resolve 61 MP and that you need a prime like Sony 400mm f 2.8 FE or 600mm f 4 FE to get more out of that sensor).
Well, the struggle for smaller sensors is greater to get the lens quality to stand up to 26 or 40 MP of Fujifilm's X-T and X-H lines and the HR modes of Panasonic G9/GH6 or Olympus's OM-D cameras. (This is not meant for bashing smaller sensor cameras, it's only meant as an observation).

There are only 2 ways to combat diffraction, at this point, and that's either by better quality glass which is improving more and more (about) every 10 years and by techniques of photography, either computational photography like HR modes or by focus bracketing and stacking, either in camera or in software.

Thanks for all that research, Ovi. Appreciated.

However, diffraction effects are often greatly exaggerated, particularly by those trying to "prove" some point or another.

It depends as much on the subject as the lens and aperture as to whether it's even noticeable. e.g. with my FTs 14-54 MkII, taking repetitive shots at ever decreasing apertures, of corrosion on the edge of an originally coated aluminium outdoor table, diffraction effects become noticeable at f/7.1, and get steadily more noticeable as I stop down.

However, in more normal photographic circumstances, diffraction is not noticeable even at f/11 and f/16.

e.g. at f/11:

Join to see EXIF info for this image (if available)

Ditto with my 12-100 and 8-25 lenses, and others. It does tend to be more noticeable with my cheaper lenses (generally).

However, I have rarely seen a competently made photograph where diffraction impacted its IQ to any significant degree, when viewed at normal sizes.

e.g. a friend printed a wall size print for an office taken with his Olympus C8080WZ, an 8 MPx camera from long ago. Not perfect if one eyeballed it at one's own MFD, but absolutely fine when viewed at normal viewing distance of 2-3 metres. I've seen the same effect viewing a similar image taken with a 135 format Canon 5D (not the original, one of the later, higher MPx models).

In summary. Excellent lenses appear to show diffraction effects far less than cheaper lenses at any given aperture, and higher quality lenses appear (to me) to be considerably more important than sensor size, or pixel size, in real life.

Effects that are visible at 200% may not be visible at 400% due to scaling errors - a very much neglected subject of discussion, IMNSHO.