Going over to the dark side: beyond f/16

[Tannin]

While it is often good to stop down a bit for extra sharpness (especially with cheaper lenses), I have always tried to avoid very small apertures on the theory that diffraction will degrade image quality too much.

The diffraction-limited apertures of some common cameras are as follows:


5D: f/13.2
1D III: f/11.5
6D: f/10.5
5D II: f/10.2
5D III: f/10.1
6D II: f/9.3
1D IV & 40D: f/9.1
5D IV: f/8.6
50D: f/7.5
7D & 60D: f/6.9
5D IIIs: f/6.7
7D II & 70D: f/6.6
80D: f/6


(Sorry Nikon and Pentax users: I had easy access to the Canon specs and was too lazy to look up any others, but DLA is determined simply by the pixel pitch of your sensor, so you can be confident that any camera of any brand will have about the same DLA as a camera listed here if it has about the same size sensor and megapixel count. Most current model FX cameras will be around about f/10; FX models about f/6 or f/7.)

Now DLAs are not Do Not Exceed numbers, of course. These are simply the apertures at which diffraction begins to take effect. If you are after extra depth of field, you can readily go past these apertures, accepting a small, possibly invisible, loss of sharpness as the price you pay.

As a rule, where no other factor intrudes, I shoot top-quality lenses close to or even at wide open (because a really good lens is just as good wide-open as it is stopped down a bit), and mid-quality lenses stopped down a little My rule of thumb (not counting exceptional lenses like the 500/4) is to stop normal lenses down to about f/8, or to a full stop less than their maximum aperture. For example, if I'm doing an infinity-focus landscape with the 24-105/4, I'll go to f/5.6 or f/8 unless there is some reason not to. I'm confident that this is best practice.

But what about where you need depth of field?

My habit is to go to f/11 immediately, and to f/16 with some mild reluctance, but I pretty much never go beyond f/16 because I'm afraid of diffraction.

Can you, do you, and should you go past f/16?

If so, how far past? And under what circumstances?

[nardes]

Hi Tony

Thanks for this useful information and the underlying topic.:th3:

Should "5D IIIs: f6.7" (3rd from list bottom) read 5DS and/or 5DSr?:confused013

I have shot at F32 to show the motion of water in a waterfall, but diffraction was not an issue in this instance.:)

Cheers

Dennis

[Tannin]

Yes, they have the same pixel pitch so they should have the same DLA. Er .... but the effective pixel pitch of the 5D IIISr should be smaller because the anti-aliasing filter isn't adding blurr, right? TDP cites them both as f/6.7 but that's a really good question. I don't know the answer.

[Steve Axford]

I have a fetish for absolutely sharp images. I am known for it, so you would expect that I would take great notice of this list you have. I shoot with:
Sony A7R2
Sony A7S2
Canon 5D 3
Canon 1DS 3
Canon 7D 2
According to your list I should use no more than f7, f13, F10, f10 and f7. When I first read about diffraction I decided to test my camera/lens combinations to see what I found acceptable as there seemed to be some confusion as to what an acceptable Circle of Confusion actually was.
I came up with the following maximum values, which have served me very well and go totally against what we are told. I always shoot at the maximum fstop when focus stacking macro fungi shots.
Sony A7R2 f16
Sony A7S2 f22 (4k video)
Canon 5D 3 f18
Canon 1DS 3 f18
Canon 7D 2 f16
I do go above these values on occasions when I don't need the full resolution of the camera or dof is more critical than absolute resolution (eg time lapse). I think it is always a good idea to test these things and make up your own mind.

[Tannin]

Err .. those are not "use no more than" numbers Steve, they are "this is where the decline starts" numbers. The point at which the decline becomes more important than other factors is somewhere beyond those numbers. How much further beyond them ... well, that's the question isn't it. I think your response is a very helpful one towards answering it in a practical way.

(We are sort of opposites in this. I seldom do things that want small apertures. Birding, I mostly aim to shoot at the most-nearly wide-open aperture I can get away with. But sometimes I want to do something different (dabble with macro for example), so it would be good to have a clearer idea of what the practical limits are.)

[Steve Axford]

I also love big apertures, but not for macro. Anyway, those are my ROT's.

[The dingo]

I thought fetishes were things we did that we wouldn't tell our mother about.
the amount of diffraction that takes place is always there. It happens when light hits the diaphragm edge. What your talking about is interference that occurs when the diffraction waves overlap. That happens progressively as the edges get closer with smaller apertures.
whether it's noticeable is another matter. If your mainly viewing on a monitor you are not likely to notice any significant difference between any two consecutive f numbers or even those 2 or 3 stops different.
for narrow slit diffraction to occur you need an aperture somewhen in the magnitude of the wavelength of light. That might be noticeable with a 50mm lens set at f64 but it's unlikely to affect our image unless you print full size and view under magnification.
personally my eyesight isn't good enough to even see an image that sharp.
Besides, it's worth considering what is most important: a shite picture that is sharp or a blurred image that s full of great ideas.
i reflect on Capa's images from Normandy. Not a sharp object in sight but what impact his images have.
theres also the f64 group with those well known photographers like Adams and Cunningham who wanted that dof provided by such small apertures on the large format cameras they used.

sharpness isn't everything. It's a component of what we all do. Using aperture to control light has the counter effect of slowing the shutter speed, an feet I personally cherish when working in the street.

"We are not limited by our technical skills, only by our imagination"

[nardes]

Given that the effects of diffraction in an image should be capable of being mathematically modelled, I wonder if proprietary apps like Canon's DPP can compensate for these effects when the user applies lens corrections? From memory I have seen chromatic aberration, colour blur and peripheral illumination listed, but I wonder if other artefacts are "fixed" by their digital lens optimiser procedure.

Cheers

Dennis

EDIT:
Voila - I launched DPP V4 and lo and behold, there is a "Diffraction Correction" function that appears when the "Digital Lens Optimiser" check box is ticked. It does not appear to be selectable so I assume that the downloaded lens profile de-convolves the image based on the lens model/aperture etc. to compensate for diffraction effects.:confused013

Cheers

Dennis

- - - Updated - - -

And there is more...

Just went to the Canon Japan website and found the following useful information:

http://web.canon.jp/imaging/dlo/effect/index03.html

So, it looks like the image softening effects of diffraction at small apertures can be addressed, although I'm not sure if only the most recent Canon DSLRs can accommodate this, as the "tick box" is not accessible in DPP4 with image files from my 5D Mk III and 7D Mk II. Maybe I need to upgrade to a 5D Mk IV - LOL!:rolleyes:

Cheers

Dennis

[Steve Axford]

What a curious feature. It does seem like it does something, though there are clearly limits to how much it can do. Unfortunately, it only seems to work with the Canon RAW processor or with jpeg, so anyone who uses 3rd party converters (eg Lightroom) can't access it. The 3rd party converters are so much better in other respects that this limitation negates any advantage. I would guess that Canon have done this because they are concerned that the code is not effectively patentable, in which case why doesn't everyone do it? Perhaps some do?
I note that this feature has been around since early 2016. Why so little media comment?

[nardes]

As DPP4 has a "Launch Photoshop" function that saves the CR2 Raw file as a TIF and then opens the TIF in PS, it looks like Canon has a preference of making certain camera/lens specific editing and correction changes in DPP first, as they are tailored to camera body, lens, focal length, aperture, etc.

Then, you pass the corrected/optimised file to PS to perform less tailored and more generic changes in that environment. Although this might be more "faithful" as Canon proprietary changes are baked into the image (rather than Abode best guesses) it seems a long winded workflow.

Cheers

Dennis

[Steve Axford]

I can't see many people doing that, which tends to negate the feature for most of us. Let's hope it turns up in LR, C1 etc soon. I expect it will as it can be hard to patent mathematics.

[Tannin]

You can always use both DPP and Camera Raw, one after the other. So, if desired, you can go DPP -> TIFF -> Camera Raw -> Photoshop. One of the first things I do after installing Photoshop is set it to always open images (of any type - raw, JPG, TIFF, you name it) in Camera Raw.

There are things Camera Raw is much better at than Photoshop, and I don't just mean raw conversion. If you want to use one or more of these things, great. And if you don't, it's no hu-hu, just click "open image" to go direct to Photoshop without making any changes in CR.

[Geoff79]

Interesting thread. Probably a lot of it is over my head, but just for my simple mind... so if I have a canon 70d, which I do, as soon as I go beyond f/6.6 I will start seeing diffraction?

So this means that it is really not a great camera for land and seascape photography? Unless I shoot around f/6? I guess I'm usually shooting around f/13 for those types of shots... but that's just inviting all this diffraction, is it?

Being a simple man, I have noticed it before but didn't understand why my images had it. Good thread.

For the record, I go up to as low as f/22 for waterfall and beach photos in the middle of the day, to try and get a touch of motion in water. I guess the diffraction doesn't show too much on the busy waterfall shots, but it definitely does in a clear blue sky, hey?

Anyway, sorry, just my ramblings, from someone not really aware of stuff like this until I see it here. Thanks for the info!


Sent from my iPhone using Tapatalk

[nardes]

This is getting to be an interesting conversation, as until now, I have been applying the Adobe lens corrections and Camera Calibration profiles (X-Rite Colour Passport) in LR before jumping into PS. I am now wondering that maybe I should do all this pre-processing stuff in DPP4 and avoid the development Module in LR as I wouldn't want to double on by applying LR corrections on top of Canon proprietary corrections in DPP4.

You can save DPP corrections as a Recipe and apply them to a batch of images so that would speed up the pre-processing, much like Sync images in LR.

Cheers

Dennis

- - - Updated - - -


Interesting thread. Probably a lot of it is over my head, but just for my simple mind... so if I have a canon 70d, which I do, as soon as I go beyond f/6.6 I will start seeing diffraction?

So this means that it is really not a great camera for land and seascape photography? Unless I shoot around f/6? I guess I'm usually shooting around f/13 for those types of shots... but that's just inviting all this diffraction, is it?

Being a simple man, I have noticed it before but didn't understand why my images had it. Good thread.

For the record, I go up to as low as f/22 for waterfall and beach photos in the middle of the day, to try and get a touch of motion in water. I guess the diffraction doesn't show too much on the busy waterfall shots, but it definitely does in a clear blue sky, hey?

Anyway, sorry, just my ramblings, from someone not really aware of stuff like this until I see it here. Thanks for the info!


Sent from my iPhone using Tapatalk

Hi Geoff

As far as I understand the topic, this type of diffraction occurs at edges or hard transition zones, so any affects in clear blue skies should not be a problem as there are no hard edges, only soft tonal changes, therefore the impact will be minimal. As the example photos show above, you also have to zoom in on the detail to see the effects of diffraction, so wide scenes should be less affected that detailed crops.


Cheers

Dennis

[Steve Axford]

It would seem, on searching the net, that Canon are not the only manufacturer who does this. Sony do and probably others. Most seem to keep it under wraps, presumably to stop others copying the function.

Another thing that I have noted, and now had confirmed by internet articles, is that different lenses have different diffraction profiles. Some lenses degrade sharply at higher fstops, others degrade much more slowly. This is not something that is mentioned in reviews of macro lenses. Hopefully most macro lenses are "tuned" for small apertures and close subjects, but many other lenses may not be. f16 or f18 may work well for a macro lens, but perhaps not for a 50mm portrait lens.

[Geoff79]

Thanks for that, Dennis. Ok, yeah, I definitely notice it where, for example, the headland meets the sky. So that's the part we are talking about. Cool.

So this might be another subject then, but what is it, and what causes it when you get what I could only describe as "visualised sound waves," if you will, in a clear sky. It appears almost like a series of light brackets. I'm sure there is a common name for it, but I thought that's what we were talking about here. Apologies, as I say, I'm a bit new to all this stuff. ;)


Sent from my iPhone using Tapatalk

[Tannin]

Another thing that I have noted, and now had confirmed by internet articles, is that different lenses have different diffraction profiles. Some lenses degrade sharply at higher fstops, others degrade much more slowly.

I'm not doubting you Steve, but how is this possible? According to my understanding of the physics, is is not. Clearly there is some factor I am unaware of. Have you seen the scientific basis of this discussed anywhere?

[Steve Axford]

I'm not doubting you Steve, but how is this possible? According to my understanding of the physics, is is not. Clearly there is some factor I am unaware of. Have you seen the scientific basis of this discussed anywhere?

No, and I may not understand it if I did. All the models we see assume that the lens/sensor path is very simple and that we can ignore any variation in lens (and sensor?) design. But can we? I think the answer to that is a clear no. I remember some years back seeing an Australian video lens that could focus on a grasshopper on a blade of grass, and a crop duster aeroplane in the sky, at the same time. That shouldn't be possible, yet it clearly was. I'll do some searching.

[arthurking83]

.....

Another thing that I have noted, and now had confirmed by internet articles, is that different lenses have different diffraction profiles. Some lenses degrade sharply at higher fstops, others degrade much more slowly. This is not something that is mentioned in reviews of macro lenses. Hopefully most macro lenses are "tuned" for small apertures and close subjects, but many other lenses may not be. f16 or f18 may work well for a macro lens, but perhaps not for a 50mm portrait lens.

Not only has this been reported elsewhere(ie. by others that know enough), but I can confirm it myself.

I see diffraction in all my lenses just as every other person could with theirs when they care to bother.
(btw! I'm not one that bothers.. just my curious nature)

But, with my old Tamron 300/2.8 lens, which is not only manual, but is the old adaptall system(where the lens can mount to any camera body with the appropriate adapter), I also have it's partner 1.4x TC, and I also have a 2xTC from last century too.

With the 1.4xTC, it gives pretty good images just about from wide open(now at f/4), stop down to f/5.6, and you'd struggle to see any IQ degradation compared to the bare lens.

But the 2xTC is their really cheap, common as much variety model, and it degrades IQ at just about every f stop .. until you reach f/16.
So where the 2xTC makes the lens a 600mm f/5.6, and at any aperture from f/5.6 all the way to f/11, IQ is best described as crappy.
f/16 all the way to f/32, IQ is much more usable.
Of course the problem is that at f/16 - f/32 ISO needs to be at the very high end in anything but high noon, or shutter speeds need to be at the unusable range unless on a tripod.

If you look at any review site that measures lens MTFs(eg PhotoZone) .. any new modern fast lens usually starts to suffer diffraction at about aperture values of f/8 or so, but one thing that is common amongst macro lenses is that they all seem to suffer diffraction much later, usually at about f/11, and they hold on to pretty good MTF values all the way to f/16.
Those modern fast (eg. f/2.8 pro type zooms lenses) all start to fall off quickly at about f/8.

I think this differing diffraction levels may have to do with aperture position and or design(ie. of the blades or thickness thereof or whatever).

[nardes]

>snip

Another thing that I have noted, and now had confirmed by internet articles, is that different lenses have different diffraction profiles. Some lenses degrade sharply at higher fstops, others degrade much more slowly. This is not something that is mentioned in reviews of macro lenses. Hopefully most macro lenses are "tuned" for small apertures and close subjects, but many other lenses may not be. f16 or f18 may work well for a macro lens, but perhaps not for a 50mm portrait lens.

This lens to lens variation may not be surprising as the effects of diffraction are proportional to the diameter of the lens. A larger diameter lens will produce a tighter diffraction pattern allowing closer pin-points of light to be resolved into 2 distinct points whereas a smaller diameter lens will not be able to split the 2 point sources (Dawes Limit or Rayleigh Criteria). Note that in astronomy, aperture means the diameter of the main objective lens or mirror of the telescope and not the focal ratio or F stop.

Have a look at this nicely presented article on the resolving power of a telescope.

http://www.rocketmime.com/astronomy/Telescope/ResolvingPower.html

Cheers

Dennis

[Steve Axford]

This lens to lens variation may not be surprising as the effects of diffraction are proportional to the diameter of the lens. A larger diameter lens will produce a tighter diffraction pattern allowing closer pin-points of light to be resolved into 2 distinct points whereas a smaller diameter lens will not be able to split the 2 point sources (Dawes Limit or Rayleigh Criteria). Note that in astronomy, aperture means the diameter of the main objective lens or mirror of the telescope and not the focal ratio or F stop.

Have a look at this nicely presented article on the resolving power of a telescope.

http://www.rocketmime.com/astronomy/Telescope/ResolvingPower.html

Cheers

Dennis

So isn't the iris size the same as the objective lens size?

[Tannin]

They would be equivalent in this case, Steve, or so I understand it. A camera's aperture diaphragm, for practical purposes, makes the diameter of the lens larger and smaller.

Arthur, I'll buy that theory, at least as a starting point.

Imagine a simple lens. A light beam at the top of the glass is bent such that it passes through a central point and then continues until it strikes the bottom of the film. A ray passing through the bottom of the lens also goes through that central point and winds up at the top of the film. Let's say it's an f/4 lens.

Now, suppose an aperture diaphragm occluding 50% of the light to put us at f/5.6. It will be, essentially, a disc with a hole in the middle. If we place the disc close to the glass at the front of the lens, it will be a large one. Or we could place it close to the central point where the light rays cross over, and it could be quite small. Is this taking us in the right direction?

[ameerat42]

...Note that in astronomy, aperture means the diameter of the main objective lens or mirror of the telescope and not the focal ratio or F stop...

Cheers

Dennis...

Note that EVERYwhere "aperture" means the diameter [...being used of the main objective lens or mirror...] and not the focal ratio or F stop.

Just had to pick you up on that one, Dennis:p:D

[Tannin]

Hmmm .. I've just been reading up on what the modern approach is in tese days of high-resolution digital sensors. Apprently, it is as follows:



"Any good modern lens is corrected for maximum definition at the larger stops. Using a small stop only increases depth; beyond a certain point definition is actually impaired"


Some chap called Adams. I'm told he is quite good.

[nardes]

So isn't the iris size the same as the objective lens size?

Hi Steve

Only when wide open, but not when stopped down to e.g. F16.:)

The difference in the size of the Airy Disc is based upon the Aperture of the lens, either wide open or stopped down. But, as we can see from the screen capture below, a 400mm lens at F16 has an effective aperture of 25mm whereas a 50mm lens at F16 has an effective aperture of only 3.13mm, a significant difference.

This is one of the reasons that astronomers crave larger diameter (aperture) telescopes; they want the ability to resolve more detail.

Cheers

Dennis

[Steve Axford]

Dennis, I think that when you stop down then that is the same thing as reducing the size of the objective lens.

The reason for the difference in resolving power in astronomical and camera lenses, is that we are measuring different things. In the case of a astronomical lens we have a fixed object at a fixed distance. This object is effectively, infinitely small and the bigger the lens the more effectively we will be able to distinguish it from its neighbours. In the case of a camera lens, we are looking at a variably sized object at variable distances. The resolving power is a measure of how much detail we can see per field of view. So, we can compare a 50mm lens looking at an object at 10m, and a 500mm lens looking at the same an object at 100m (not sure if those distances are right, but you get the idea). This means that the 50mm lens can have more resolving power than the 500mm. This would make no sense in astronomy where a bigger lens always has more resolving power than the smaller (provided the lenses are roughly equal technology).

[nardes]

Note that EVERYwhere "aperture" means the diameter [...being used of the main objective lens or mirror...] and not the focal ratio or F stop.

Just had to pick you up on that one, Dennis:p:D

Hi AM

Not quite sure about this Am, as often when we refer to the shooting Aperture as e.g. F2.8, F5.6, F16 etc. for the Aperture Setting e.g. I shot this at a Shutter Speed of 1/250th sec, Aperture at F8 at ISO400, the use of "Aperture" does not describe the native wide open Diameter of the lens, but the stopped down Focal Ratio of the lens?

I never say I shot this at a lens Aperture of 72mm, or 52mm which is the lens Diameter value?:)

However, I guess we all understand the use and meaning of the term Aperture in the context of astronomical telescopes and photography lens settings so no big deal?:confused013

Cheers

Dennis

- - - Updated - - -


>snipped stuff here.

Another thing that I have noted, and now had confirmed by internet articles, is that different lenses have different diffraction profiles.

>snipped stuff here.


Dennis, I think that when you stop down then that is the same thing as reducing the size of the objective lens.

The reason for the difference in resolving power in astronomical and camera lenses, is that we are measuring different things. In the case of a astronomical lens we have a fixed object at a fixed distance. This object is effectively, infinitely small and the bigger the lens the more effectively we will be able to distinguish it from its neighbours. In the case of a camera lens, we are looking at a variably sized object at variable distances. The resolving power is a measure of how much detail we can see per field of view. So, we can compare a 50mm lens looking at an object at 10m, and a 500mm lens looking at the same an object at 100m (not sure if those distances are right, but you get the idea). This means that the 50mm lens can have more resolving power than the 500mm. This would make no sense in astronomy where a bigger lens always has more resolving power than the smaller (provided the lenses are roughly equal technology).

Hi Steve

Not sure if we are discussing the same issue in this overall topic, :) but I was responding to your post above (QUOTE: "is that different lenses have different diffraction profiles"). In this case I assumed that you meant different lens diameters and not just design, # of elements, etc. and as the diameter of a lens is one of the critical factors related to the diffraction profile of the lens, which is borne out by the formula in the attached screen capture, then yes, different lenses (as in different lens diameters) will have different diffraction profiles.

However, if your meaning of "different lenses" referred to lenses of the same diameter but different design, # of elements, etc. then there is no issue.:)

Cheers

Dennis

[Steve Axford]

I did mean lenses of different designs, but I think you may be missing the point with camera lenses. Your yellow highlighted point is true, but it is only the whole story for astophotography. With a camera, resolution is measured differently. See my previous note.

[nardes]

I'm still not sure what point I am missing Steve?:)

The relationship between the size of the Airy disk and the diameter of a particular lens is a property of the lens and is the same for visual observing as well as for astrophotograhy with the lens/telescope. It matters not whether there is an eyeball, 35mm film or a digital image sensor collecting the light, as the Airy Disk is a property of the lens opening and wavelength of light, not the sensor gathering the light.:confused013

Is this just a simple communication issue, i.e. I am talking about diffraction and you are talking about resolution?:confused013


Cheers

Dennis

[Tannin]

I thought fetishes were things we did that we wouldn't tell our mother about.

Oh I don't mind telling my mother about my fetishes. I just try to be a bit vague about how much they cost.

"Is that a new lens, Dear?"
"No Mum, it's the same one I had last week."
"That's funny, it looks about a foot longer."
"Oh, they often swell up like that. It's just the humidity."


Fun aside, that's a great point you make, Tom. But is it not wise to be aware of the technical limitations before (quite possibly) disregarding them them?

[Steve Axford]

I'm still not sure what point I am missing Steve?:)

The relationship between the size of the Airy disk and the diameter of a particular lens is a property of the lens and is the same for visual observing as well as for astrophotograhy with the lens/telescope. It matters not whether there is an eyeball, 35mm film or a digital image sensor collecting the light, as the Airy Disk is a property of the lens opening and wavelength of light, not the sensor gathering the light.:confused013

Is this just a simple communication issue, i.e. I am talking about diffraction and you are talking about resolution?:confused013


Cheers

Dennis

With astophotography we talk about the absolute amount of diffraction
With normal photography we talk about the amount of diffraction compared to the size of the image

In one case we have D (where D=diffraction)
in the other we have D/I (where I=image size)

Image size has no meaning when looking at a distant star and it can be considered a constant

[The dingo]

Oh I don't mind telling my mother about my fetishes. I just try to be a bit vague about how much they cost.

"Is that a new lens, Dear?"
"No Mum, it's the same one I had last week."
"That's funny, it looks about a foot longer."
"Oh, they often swell up like that. It's just the humidity."


Fun aside, that's a great point you make, Tom. But is it not wise to be aware of the technical limitations before (quite possibly) disregarding them them?

Wise? Who is wise? They say wisdom comes with age. I'm still waiting.

Having lived through 70 years of technology growth I recall the days when cameras were simple devices and I only needed to concern myself with pointing the camera in the right direction, which often proved to be a difficult task for me.
My mentor at the time was a man my age now. He taught me how to estimate exposure by just looking at the scene. I still do that from time to time when things are tricky and I know the light meter in the camera is confused.

I learnt a great deal about the technology while studying science at uni, especially such things relating to optics. It was interesting enough but it didn't influence the way I approached my pictures. CONTENT and CONTEXT were the prime concerns.

The biggest change for me was when autofocus came about. I no longer needed to rely on my poor eyesight. And it was fast!

The digital age and the internet have brought about a lot of changes. One of them is the number of experts there are on any one topic.

The science behind diffraction is interesting, important and relevant. How relevant to photography is often assumed with little understanding of what is happening and how it might influence a photograph.
Correcting for it, as Canon propaganda suggests, they can deal with it simply.
What they are doing is endeavouring to overcome its effects. They do this by 'sharpening'. Its not really sharpening. Its altering the contrast at high contrast points.

So. understanding the technicalities of diffraction is one thing; being obsessed about 'sharpness' is another. Anyone standing that close to a print in a gallery will be told to get their nose off the glass.

[ameerat42]

'Ts OK Dennis. I know people use terms like "an aperture of f/8" wrongly, as you point out.
The reason for replying is to alert "newbies" et al that there is a proper terminology.

I am with you 101% on your point, and I was just expanding on it.

[jim]

101% is impossible, am.

[ameerat42]

Not in the world of HYPE, in which we are immersed:(

[jim]

You might be immersed in a world of hype. I just glance idly in its general direction every now and again.

[ameerat42]

You might be immersed in a world of hype. I just glance idly in its general direction every now and again.

Ahh! - Life on Mars! [/Dreamy gaze]

[nardes]

With astophotography we talk about the absolute amount of diffraction
With normal photography we talk about the amount of diffraction compared to the size of the image

In one case we have D (where D=diffraction)
in the other we have D/I (where I=image size)

Image size has no meaning when looking at a distant star and it can be considered a constant

Sorry Steve – I still “don’t get it”.:)

I get the formula describing the size of the Airy Disc (based on the diameter of the lens and wavelength of light) which defines the diffraction profile of that lens (lens, telescope, microscope, etc.) irrespective of the sensor; eye, film or digital.

But image size (I) – what does that mean, it appears to have many interpretations:



A 10”x8” print viewed at arm’s length, or
A full screen image viewed on a 32” monitor from 25 CMS distance, or
A 28 MB, 6000x4000 pixel digital file, or
The apparent pixel dimensions of a blade of grass or insect leg in the frame, e.g. 200 pixels long by 4 pixels wide, or
The size of the sensor cells, e.g. 5.6 um in the camera?


However, I’ll stop here for now and take this off line to look at other sources of information on the topic of “Diffraction Limited Aperture” to nut this one out.:th3:

Cheers

Dennis

[Steve Axford]

No matter. If I can't explain it, then that is my problem.

[arthurking83]

What a curious feature. It does seem like it does something, though there are clearly limits to how much it can do. Unfortunately, it only seems to work with the Canon RAW processor or with jpeg, so anyone who uses 3rd party converters (eg Lightroom) can't access it. The 3rd party converters are so much better in other respects that this limitation negates any advantage. I would guess that Canon have done this because they are concerned that the code is not effectively patentable, in which case why doesn't everyone do it? Perhaps some do?
I note that this feature has been around since early 2016. Why so little media comment?

My inherent cynicism of all things magic(like that diffraction correction tool) leads me to think it's just a high pass filter or USM routine catering for specific camera bodies.
That is, for each camera body with their respective pixel densities require different amounts of sharpening to produce more contrast, but not also introducing haloing effects as the downside.

It wouldn't be too hard for Canon to get their software folks to test varying amounts of high pass filter or USM settings for each camera, and produce a tool that reads the camera data, reads the aperture used and apply a specific sharpening routine that produces zero haloing.

Other than a need to have a starting point for a sharpening routine, I doubt any overly complex mathematics have been used to create this tool.

As for using or not using manufacturer's software on raw files:
Using ACR is akin to using DPP.
Photoshop in itself can't read raw files, so you NEED to use a raw file converter first.
There are advantages and disadvantages to either method, but in effect they are still the same.
(using the common ACR workflow to describe any thirdparty raw converter)

Using ACR, you get the option to pass it directly into photoshop and in what adobe seem to think is the best format for photoshop .. which is psd(unless you specify otherwise). What you don't get tho, is access to specific camera maker raw data that reads the file as per the way the camera shot the image.
My understanding of psd is that it's a lossless compressed tiff file(I'm sure I read that somewhere).
You could just as easily use any other thirdparty raw file converter and if it allows the use of the format, you could still pass it through to photoshop as a psd file.
I don't know of any thirdparty converters that can create psd files other than PaintDotNet. Then again PDN is basically a free photoshop type software that doesn't (natively) read raw files itself! :D
I know of a few that can read the format tho.

I can't really imagine thirdparty software makers producing a similar diffraction neutralising too as such, simply due to the complexity and scale of amassing all the necessary data required for the sharpening routine ... remembering that each lens would need to be measured at each aperture setting, and matched to each camera body(ie. pixel density) .. for how many manufacturers!! :eek:

I don't know if folks prefer to use high pass or USM, they each have their own subtle differences in how they sharpen, but you could try a very mild USM routine to minimise the softening effect of diffraction.

Another point of note regarding diffraction:
It's not always a bad thing having diffraction.
Some cameras(ie. referring to very high pixel density cameras with no AA filter) can produce moire where very fine details are misinterpreted by the camera sensor used with a very sharp lens.
The two methods to eliminate moire, without resorting to software trickery ... is to either slightly defocus the lens so that the detail isn't recorded on the sensor with so much precision, or to stop the lens down past the point of diffraction(ie. same effect.. less sharp .. but still sharp enough).

[Steve Axford]

I can't really imagine thirdparty software makers producing a similar diffraction neutralising too as such, simply due to the complexity and scale of amassing all the necessary data required for the sharpening routine ... remembering that each lens would need to be measured at each aperture setting, and matched to each camera body(ie. pixel density) .. for how many manufacturers!! :eek:
.

I think they probably will, if there is any advantage in doing so. After all, it's just a table connecting camera/lens/distance/fstop . No big deal really and it only has to be done once.

[arthurking83]

I think they probably will, if there is any advantage in doing so. After all, it's just a table connecting camera/lens/distance/fstop . No big deal really and it only has to be done once.

On the basis that DxO specialises in camera and lens matching, testing and 'databasing' ... it wouldn't be unreasonable to expect them to offer such a feature first amongst the thirdparty vendors.
In fact I'm surprised that they haven't(if they haven't as yet).

[Steve Axford]

I was thinking of DxO, and there's always a battle between Phaseone and Adobe (and others no doubt). Any edge is considered worthwhile. For us mere photographers that is good.

[nardes]

Here are a couple of useful resources that I have discovered today in searching on the topic of diffraction. The "Cambridge in Colour" resource has some nice calculators for various camera/lens/pixel size combinations and spans 2 pages, so be sure to visit P2 as well.

http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm

This next resource is based on an interview with a Canon technical advisor at the Professional Engineering & Solutions Division, Canon U.S.A for when the Canon 5D Mk III was released.

http://www.arihazeghiphotography.com/AH_CW_interview/

Part way down the page we read the following excerpt:

"A new feature called Digital Lens Optimizer processes RAW images to achieve ideal optical characteristics for all types of optical aberration or diffraction, effects of a low-pass filter in front of a CMOS sensor, etc. This function improves image quality particularly in the image periphery in addition to the image center. This function is made possible because the entire design-through-manufacture process, for camera, CMOS sensor, EF lens, and DPP, is carried out entirely at Canon. Images are processed optimally using lens information in the image files (focal length, subject distance, and aperture) and lens data specially for the Digital Lens Optimizer. (However, the size of a .CR2 file will be two to three times larger after applying the Digital Lens Optimizer.) Adjustments are made for such aspects as spherical aberration, chromatic aberration, astigmatism, curvature of field, sagittal halo, chromatic aberration of magnification, axial chromatic aberration, diffraction, and the effects of a low-pass filter in front of the CMOS sensor. DPP’s Digital Lens Optimizer will be usable with any of 29 compatible lenses initially. It works with .CR2 files from EOS models released since 2006 (EOS 30D and forward)."

Cheers

Dennis

[Tannin]

On the basis that DxO specialises in camera and lens matching, testing and 'databasing' ... it wouldn't be unreasonable to expect them to offer such a feature first amongst the thirdparty vendors.

And on the basis of their demonstrated accuracy, impartiality, and reliability in other endeavours thus far, it wouldn't be unreasonable to expect them to make a complete hash of it. :(

[Steve Axford]

And on the basis of their demonstrated accuracy, impartiality, and reliability in other endeavours thus far, it wouldn't be unreasonable to expect them to make a complete hash of it. :(

I've only glanced at DxO. What are their biases?

[Tannin]

They hate Canon. Doubtless they have other oddities as well. This is the company which once (no joke) claimed that according to their 'scientific" "measurements" a Nikon D200 had lower noise than a 5D Mark 1. (Detail from memory. It may have been some other outdated crop sensor Nikon notorious for poor high-ISO performance.) I have no idea what other silly stuff they get up to these days as I stopped paying attention to their self-serving nonsense years ago.