Diffraction 5DII and minimum aperture

[bobn2]

Far too much is made of diffraction, IMO.

Especially since for any output image size and DoF the diffraction blur is constant, regardless of sensor size or pixel density. That's why I keep on saying this - once it gets into the collective consciousness, perhaps people will stop banging on about diffraction and small apertures.

It also makes me wonder about styles of photography, I do a lot of landscape, and hardly ever use smaller than f/8 on a FF, often I go wider to introduce more differential focus int the image to get more depth - yet you get posts suggesting that all landscapers are shooting at f/16 all the time, which as I've pointed out, rather negates the point of buying good lenses.
--
Bob

 

[John Sheehy]

I don't attribute it to a change in the amount of diffraction, I
attribute it to a change in the CoC requirements at the sensor plane.
The effects of diffraction on one's photography (and it is
photography that we do) are larger the higher ones requirements are
of the image.

I have a much simpler way of looking at it; if you want to make a big print, either diffraction or low pixel count can limit the value of printing large.

It's never the case that a higher pixel density is the cause for such a limit, so I don't understand your exception to some statements here. If you want a large image with satisfactory experience at close viewing, you want both higher pixel count and minimal diffraction, and the diffraction is controlled by the f-stop, not the pixel density.

--
John

 

[David J. Littleboy]

I don't attribute it to a change in the amount of diffraction, I
attribute it to a change in the CoC requirements at the sensor plane.
The effects of diffraction on one's photography (and it is
photography that we do) are larger the higher ones requirements are
of the image.

I have a much simpler way of looking at it; if you want to make a big
print, either diffraction or low pixel count can limit the value of
printing large.

True. But when you are out in the field taking a photo, you are already committed to the pixel count. So the pixel count determines how badly diffraction bites you.

--
David J. Littleboy
Tokyo, Japan

 

[Ben_Egbert]

I don't attribute it to a change in the amount of diffraction, I
attribute it to a change in the CoC requirements at the sensor plane.
The effects of diffraction on one's photography (and it is
photography that we do) are larger the higher ones requirements are
of the image.

I have a much simpler way of looking at it; if you want to make a big
print, either diffraction or low pixel count can limit the value of
printing large.

True. But when you are out in the field taking a photo, you are
already committed to the pixel count. So the pixel count determines
how badly diffraction bites you.

--
David J. Littleboy
Tokyo, Japan

I like Davids argument because he included not just the theory, but our changing expectations in the entire deal. Photography is growing, and we are becoming more critical. We are pushing the technology.

When I print, I print 16x24. I would print larger if my printer and camera supported it. When I look at a print, I want to be able to look as close as I wish. In fact, I always have done this. I see lots of professional work that is softly presented, and it looks awful to me.

But if some areas are grossly OOF it is a lot worse than a mildly soft overall image caused by diffraction. DOF trumps diffraction unless you have a specifically shallow DOF subject.

That said, a good lens will have sufficient DOF by f16 up to about 50mm.

Also like David, I use a smaller CoC, in my case I use .015 when I calculate hyper focal distances.

I suspect that 35mm format has a limit and only a larger format will take you further. I am not convinced we have reached it yet. Put some good glass on a 21mpixel sensor and prove it to yourself.

--
http://www.pbase.com/roserus/root

Ben

 

[gdanmitchell]

I also shoot a lot of landscape using FF. My typical default starting point aperture is usually f/11. I move to smaller or larger apertures depending upon the demands of the scene and the way I want to capture it.

In order to understand the strengths and weaknesses of my gear, and especially my lenses, and to get a sense of their "personalities" I do some simple tests when I get new lenses. (For the record, I am NOT testing for "good copies." I am simply trying to jump-start my learning process about the characteristics of each lens.)

It is this testing plus having made a lot of photographs and a fair number of prints that convince me that the f/8 advice is unnecessarily conservative. Some time ago I posted one example from my own tests:

http://www.gdanmitchell.com/2007/04/12/sharpness-and-aperture-selection-on-full-frame-dslrs

This is just one example from a range of tests, but f/8 and f/11 are essentially indistinguishable in these 100% crops of a scene that does test the resolution ability of the system. (A print at this resolution would be something like 60 inches wide.) I think I can see a slight bit of softening in the f/16 example, but it would be completely insignificant in a print.

Dan

Far too much is made of diffraction, IMO.

Especially since for any output image size and DoF the diffraction
blur is constant, regardless of sensor size or pixel density. That's
why I keep on saying this - once it gets into the collective
consciousness, perhaps people will stop banging on about diffraction
and small apertures.
It also makes me wonder about styles of photography, I do a lot of
landscape, and hardly ever use smaller than f/8 on a FF, often I go
wider to introduce more differential focus int the image to get more
depth - yet you get posts suggesting that all landscapers are
shooting at f/16 all the time, which as I've pointed out, rather
negates the point of buying good lenses.
--
Bob

--
---
G Dan Mitchell - SF Bay Area, California, USA
Blog & Gallery: http://www.gdanmitchell.com/
IM: gdanmitchell

Gear List: Cup, spoon, chewing gum, old shoe laces, spare change, eyeballs, bag of nuts.

 

[gdanmitchell]

For the most part, I get bored with the point by point (quote followed by response) rebuttals in these threads, so I'm not going to go there. But I do want to respond to two of your points regarding the context of my earlier posts.

My posts are precisely correct.

Yes, but in their precise mathematical correctness, they fail to
account for the fact that one has higher expectations for the 5D2
than the 5D.

My posts are totally NOT about math. They are about what happens in photographs, and they are aimed at dispelling the wrong idea that increasing photosite density has any effect on diffraction.

To say otherwise is equivalent to
saying that diffraction blur increases if you use a film with finer
grain - and that is, of course, nonsense.

But that's exactly what happens with film. With TMX100, I expect to
make much larger prints than I do from XP2-Super, and diffraction is
more of a concern. (Actually, it's not much of a concern since my MF
cameras only stop down to f/22, and f/22 is fine with 6x7 TMX100.)

You simplify. Making larger prints is, indeed, not the only reason to select a finer grain film. Another common reason was (is) simply to reduce the visibility of grain in the print. Some photographers made (make) the opposite choice - choosing a film with more grain because they like that effect for a particular photograph.

Dan

Who thinks he has pretty much made his point, but understands that a few won't be persuaded, and will try to resist the temptation to keep this going...

--
---
G Dan Mitchell - SF Bay Area, California, USA
Blog & Gallery: http://www.gdanmitchell.com/
IM: gdanmitchell

Gear List: Cup, spoon, chewing gum, old shoe laces, spare change, eyeballs, bag of nuts.

 

[rrcphoto]

ben, you're right ..

actually the limit would be that no matter what aperture you use with a given print size, diffraction is noticable - that that point, there is nothing a 35mm sensor can do.

which hasn't been reached yet - which is why canon feels FF sensors can get up to 60Mp's.

I suspect we'll see from canon (probably before anyone else) a 35mm sensor around 7600x5050 or so (38Mp)

that will yield a print size of around 24x36 as a comfortable fine art print size - however at that print size, the observer distance is around 4 feet or so back - to which the individual pixel quality isn't as important as the overall image quality and depth. F/8 still yields a sharp image according to that print size, even though the sensor is diffraction limited for pixel peepers at around f/4 to f/5.6 - to the 100% viewer and pixel inspector - that same print that would drop dead stunning at 24x36 would have a CoC of approximately three pixels in diameter - at 100% it would look like fuzz.

also there's the human perception - we'll notice DOF which will be seen by the eye as a comparison while looking at the image (sharp area, not sharp) .. but we won't notice as much if the entire print is diffraction limited within reasonable proportions.

the only time diffraction comes with is reviewers who tend to want to see per pixel quality .. and that will have to change .. sooner or later nikon and sony will come out with 15 or 16Mp cropped bodies as well .. and more and more higher Mp FF cameras will come out as well.

the "pixel" plays such a small role in the overall image quality, but it is the one thing that is hotly contested more than any other element.

 

[Skip M]

Not true at all.

There is an article at Luminous-Landscape where the images from a
large view camera are mush at F64.

BC

I'd be interested in seeing that article. Do you have a link? I couldn't find it on LL.
By the way, Ansel Adams, Edward Weston et al did ok at f64...
http://en.wikipedia.org/wiki/Group_f/64
Of course, f64 was just a name, they didn't shoot everything at f64...
--
Skip M
http://www.shadowcatcherimagery.com
http://www.pbase.com/skipm
http://skipm.smugmug.com/
'Living in the heart of a dream, in the Promised Land!'
John Stewart

 

[bobn2]

Not true at all.

There is an article at Luminous-Landscape where the images from a
large view camera are mush at F64.

BC

I'd be interested in seeing that article. Do you have a link? I
couldn't find it on LL.
By the way, Ansel Adams, Edward Weston et al did ok at f64...
http://en.wikipedia.org/wiki/Group_f/64
Of course, f64 was just a name, they didn't shoot everything at f64...

f/64 is indeed just a name. It's effect depends on the imager size and magnification you used. AA shooting f/64 on 8x10 is equivalent in both DoF and diffraction terms to f/8 on FF. If an 8x10 is 'mush' at f/64, then any FF camera is 'mush' at f/8.

--
Skip M
http://www.shadowcatcherimagery.com
http://www.pbase.com/skipm
http://skipm.smugmug.com/
'Living in the heart of a dream, in the Promised Land!'
John Stewart

--
Bob

 

[Rick Knepper]

Although I haven't used my 5D2 for a serious trip yet (March 7-13 Big
Bend NP will be my first chance), I expect diffraction to be
stronger. I'm thinking DoF blending is going to be big in my future.

It won't be. It will be exactly the same.

Could you reconcile the above statement against the findings in the link below? We can both stipulate that neither of us has this kind of noticeable blur at f8 on our 5Ds.

http://www.diglloyd.com/diglloyd/free/Diffraction/example-1DsM3.html

Also, could you reconcile your understanding against this article in the link below (the basis of my understanding)? For expediency's sake, look at the diffraction calculator. One of the inputs is Resolution. It is quite easy using the calculator to show how resolution does affect diffraction especially on larger size prints. It is also quite easy to prove your point (I think I got your point) that for most applications (and typical print sizes) it won't matter much. But, do any of us buy 21 MP for typical applications?

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

However, I do not need the calculator to see degradation going from f8 to f11 to f16 on the same 12 MP camera. I've shot over 10,000 images using the same protocol, each scene at f5.6, f8, f11 and f16 with dozens of lenses all of which produced varying degrees of the aforementioned degradation. Again, the degradation is there whether that be from diffraction, lens aberration or a combination of the two. Part of my point was: would one need f16 as often as one thinks if a better lens was involved?

Of course, when you look at, say, a 400 x 400 pixel 100% crop on the
screen you will be LOOKING MORE CLOSELY AT A SMALLER SECTION OF THE
IMAGE so it will appear larger. But in your print or other final
product it will be no different whatsoever at the apertures we are
considering here.

Let me also add that the "diffraction smaller than pixel pitch"
business is irrelevant - except in the astonishingly unlikely even
that a sub-pixel size element of your photograph lines up perfectly
with the center of single pixel or, even most unlikely, that a
perfectly horizontal or vertical line with a projected width of less
than one pixel lines up perfectly with the center of a line of
vertical or horizontal photosites. Of course that would also require
the use of a lens with zero barrel/pincushion distortion used a a
camera perfectly aligned to the scene. In other words, it ain't
gonna' happen.

In the real world sub-pixel sized subjects projected on the sensor
rarely line up in this way, and values "spill across" adjacent
photosites.

It would be accurate - but probably irrelevant - to say the the
sensor with the higher photosite density could form a more accurate
image of the diffraction blur...

Ah your disclaimer.

... but there is no "more diffraction."

I probably have been off here in the past but my understanding was that diffraction effects become "stronger" (more noticeable) as one stops down, but I will accept the description "form a more accurate image of the diffraction" as the better statement.

I have another link I will post when I find it using the diffraction equation to find diffraction limit on a 13 x 19 print shot with a 5D. The number is surprising.

 

[bobn2]

Although I haven't used my 5D2 for a serious trip yet (March 7-13 Big
Bend NP will be my first chance), I expect diffraction to be
stronger. I'm thinking DoF blending is going to be big in my future.

It won't be. It will be exactly the same.

Could you reconcile the above statement against the findings in the
link below? We can both stipulate that neither of us has this kind of
noticeable blur at f8 on our 5Ds.

http://www.diglloyd.com/diglloyd/free/Diffraction/example-1DsM3.html

Also, could you reconcile your understanding against this article in
the link below (the basis of my understanding)? For expediency's
sake, look at the diffraction calculator. One of the inputs is
Resolution. It is quite easy using the calculator to show how
resolution does affect diffraction especially on larger size prints.
It is also quite easy to prove your point (I think I got your point)
that for most applications (and typical print sizes) it won't matter
much. But, do any of us buy 21 MP for typical applications?

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

However, I do not need the calculator to see degradation going from
f8 to f11 to f16 on the same 12 MP camera. I've shot over 10,000
images using the same protocol, each scene at f5.6, f8, f11 and f16
with dozens of lenses all of which produced varying degrees of the
aforementioned degradation. Again, the degradation is there whether
that be from diffraction, lens aberration or a combination of the
two. Part of my point was: would one need f16 as often as one thinks
if a better lens was involved?

Niether of these needs 'reconciling'. If you look at the MTF figures for any lens, you will see there is a peak. In primes, somewhere between f/4 and f/5.6, zooms somewhere between f/5.6 and f/11. Smaller apertures than that, the MTF degrades, entirely due to diffraction. The only difference sensor resolution makes is how much of the degradation the camera is capable of catching. Neither of the sources you quote adds anything except precision ether in theory or practice to this simple piece of knowledge.

The real issue is plotting diffraction degradation against aperture (which is in any case constant). If you were to look at diffraction blurring vs. DoF, you will find it is exactly the same for any camera, provided the output image is the same size.
--
Bob

 

[Rick Knepper]

Although I haven't used my 5D2 for a serious trip yet (March 7-13 Big
Bend NP will be my first chance), I expect diffraction to be
stronger. I'm thinking DoF blending is going to be big in my future.

It won't be. It will be exactly the same.

Could you reconcile the above statement against the findings in the
link below? We can both stipulate that neither of us has this kind of
noticeable blur at f8 on our 5Ds.

http://www.diglloyd.com/diglloyd/free/Diffraction/example-1DsM3.html

Also, could you reconcile your understanding against this article in
the link below (the basis of my understanding)? For expediency's
sake, look at the diffraction calculator. One of the inputs is
Resolution. It is quite easy using the calculator to show how
resolution does affect diffraction especially on larger size prints.
It is also quite easy to prove your point (I think I got your point)
that for most applications (and typical print sizes) it won't matter
much. But, do any of us buy 21 MP for typical applications?

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

However, I do not need the calculator to see degradation going from
f8 to f11 to f16 on the same 12 MP camera. I've shot over 10,000
images using the same protocol, each scene at f5.6, f8, f11 and f16
with dozens of lenses all of which produced varying degrees of the
aforementioned degradation. Again, the degradation is there whether
that be from diffraction, lens aberration or a combination of the
two. Part of my point was: would one need f16 as often as one thinks
if a better lens was involved?

Niether of these needs 'reconciling'.

Of course it does.

If you look at the MTF figures
for any lens, you will see there is a peak. In primes, somewhere
between f/4 and f/5.6, zooms somewhere between f/5.6 and f/11.
Smaller apertures than that, the MTF degrades, entirely due to
diffraction.

Didn't I just essentially say this in my post? Isn't Dan saying something different?

The only difference sensor resolution makes is how much
of the degradation the camera is capable of catching. Neither of the
sources you quote adds anything except precision ether in theory or
practice to this simple piece of knowledge.

It is evident you did not read my entire post or failed to understand it because you would not be writing this as if I do not know it.

So, are you agreeing or disagreeing with Dan?

The real issue is plotting diffraction degradation against aperture
(which is in any case constant). If you were to look at diffraction
blurring vs. DoF, you will find it is exactly the same for any
camera, provided the output image is the same size.
--
Bob

Misleading. My understanding includes the output images involved. When I say my understanding involves that article in particular the diffraction calculator which includes the output image, that's what I mean.

Thanks for the reply but I believe the OP should read the Cambridge article carefully and look at the Lloyd Chambers test for the last word on the subject.

 

[bobn2]

Niether of these needs 'reconciling'.

Of course it does.

Can't see that - I've read both of those several times, I can't see anything in either of them which needs reconciling with anything Dan wrote. Precisely, he wrote that the diffraction with your 5DII would be exactly the same as the diffraction with whatever you were using before. that is true, and neither of those sources says anything different.

If you look at the MTF figures
for any lens, you will see there is a peak. In primes, somewhere
between f/4 and f/5.6, zooms somewhere between f/5.6 and f/11.
Smaller apertures than that, the MTF degrades, entirely due to
diffraction.

Didn't I just essentially say this in my post? Isn't Dan saying
something different?

No, he isn't. Dan, like both of us, knows this well.

The only difference sensor resolution makes is how much
of the degradation the camera is capable of catching. Neither of the
sources you quote adds anything except precision ether in theory or
practice to this simple piece of knowledge.

It is evident you did not read my entire post or failed to understand
it because you would not be writing this as if I do not know it.

I did read it, maybe I didn't read in it what you intended to say.

So, are you agreeing or disagreeing with Dan?

I agree with him (on this, at least).

The real issue is plotting diffraction degradation against aperture
(which is in any case constant). If you were to look at diffraction
blurring vs. DoF, you will find it is exactly the same for any
camera, provided the output image is the same size.
--
Bob

Misleading.

No, it's a clarification. There is no way past diffraction, not changing the pixel size or anything. If people would simply grasp this, these threads would disappear.

My understanding includes the output images involved.
When I say my understanding involves that article in particular the
diffraction calculator which includes the output image, that's what I
mean.

The calculator can include the output image. unfortunately, the default bases the Airy disc diameter, and therefore the diffraction 'limit' on pixel size, which is misleading. If you run the multiple tests on that calculator, setting up the same print size for different sensor and pixel sizes, and setting the aperture for equivalent DoF, you'll find that my contention is true - it's just a very long winded way of going about discovering a simple piece of knowledge.

Thanks for the reply but I believe the OP should read the Cambridge
article carefully and look at the Lloyd Chambers test for the last
word on the subject.

I think both of those articles can result in much confusion if read by someone who isn't secure in their understanding. The first thing to ask is 'why do you want to shoot at small apertures' - the general answer is for extended DoF, in which case get to know the dependence of diffraction blur on DoF, not f-number, which photographically is somewhat arbitrary.
--
Bob

 

[gdanmitchell]

Although I haven't used my 5D2 for a serious trip yet (March 7-13 Big
Bend NP will be my first chance), I expect diffraction to be
stronger. I'm thinking DoF blending is going to be big in my future.

It won't be. It will be exactly the same.

Could you reconcile the above statement against...

Since a) someone else already responded on my behalf, b) I said I wouldn't continue this bizarre debate, and c) I've already posted the answers to these questions in my earlier message...

... a very brief additional response to one part of your message about optimum apertures:

http://www.gdanmitchell.com/2007/04/12/sharpness-and-aperture-selection-on-full-frame-dslrs

Dan

--
---
G Dan Mitchell - SF Bay Area, California, USA
Blog & Gallery: http://www.gdanmitchell.com/
IM: gdanmitchell

Gear List: Cup, spoon, chewing gum, old shoe laces, spare change, eyeballs, bag of nuts.

 

[rrcphoto]

Although I haven't used my 5D2 for a serious trip yet (March 7-13 Big
Bend NP will be my first chance), I expect diffraction to be
stronger. I'm thinking DoF blending is going to be big in my future.

It won't be. It will be exactly the same.

Could you reconcile the above statement against the findings in the
link below? We can both stipulate that neither of us has this kind of
noticeable blur at f8 on our 5Ds.

http://www.diglloyd.com/diglloyd/free/Diffraction/example-1DsM3.html

his findings are based upon 100% magnification of the image.

which with respects to print size, is meaningless. both the image magnification and observer distance is flawed or poorly misprepresented in his article.

if you print a 12x18 from an 12Mp, 21Mp and a 24Mp camera all at f/16 - even though the 21, and 24 have increasing levels of diffraction at 100% magnification - on the print they will appear increasingly sharper and more detailed than the one at 12Mp which is hardly diffraction limited at 100%.

because in all cases you are keeping the observer distance the same, and the magnification of the image the same.

the noticable effect of diffraction entirely depends on the magnification, and observer distance - it's what the eye can possibly detect as being "soft" or diffused.

Diffraction limitations can be easily thought of as when the airy disc of diffraction is greater than the circle of confusion.

if you use the above meaning - then also be aware that Circle of Confusion is ENTIRELY based upon image size and observer distance and not of pixel density.

Therefore, diffraction, or the noticable effect of diffraction (which is based upon aperture) is not based upon pixel density, but simply the circle of confusion of your related image size and distance viewed by the observer.

 

[Skip M]

To answer the OP's question, my experience is that on the 5D, f/16
produces sharp images for printing at 240ppi (12x18) and f/22 is
slightly softer, maybe, but worth it if you really need the DoF.

On the 5D2, f/11 produces sharp images for printing at 240ppi (15x22)
and f/16 is slightly softer, maybe, but worth it if you really need
the DoF. You don't want to shoot at f/22 on the 5D2 if you plan at
printing at 240 ppi.

Doesn't that vary depending on lens?

Not in my experience. Lenses are more or less funky wide open, but at
the point diffraction is a potential issue, they're pretty much all
the same. (Amounts of CA and distortion vary with the lens in this
range of f stops, of course.)

--
David J. Littleboy
Tokyo, Japan

Logic, and a complete lack of understanding of the science of optics, would tell me that, since the lens is how the light gets to the sensor/film, it would have some influence on diffraction. To what degree, I would have no idea.
--
Skip M
http://www.shadowcatcherimagery.com
http://www.pbase.com/skipm
http://skipm.smugmug.com/
'Living in the heart of a dream, in the Promised Land!'
John Stewart

 

[rrcphoto]

Case: 100% magnification between 5D and 5D Mark II (or 1Ds Mark III)

Assumption: Monitor DPI is around 90 DPI - most wide screens I suspect are around this.

5D Mark II / 1Ds Mark III - Viewed Image Size = 60x40. Magification coefficient 45:1

5D - Viewed Image Size = 48x32. Magnification coefficient is: 35:1

Thus when viewed at 100% - the Circle of Confusion is actually changing, because Circle of confusion is entirely based upon the image "negative" size (ie: sensor size) and the final image size magnification and observer distance.

As you increase the image size, and maintain the same observer distance, the circle of confusion as related to the sensor or "negative" decreases.

as you stop down the aperture, the disc of diffraction increases. at the point in which the circle of confusion size intersects with the disc of diffraction, it's determined to be limited by diffraction.

so when comparing images and the effects, unless you first determine the image size, you are inducing another possible variant into the comparision - because that is changing with the increase of resolution.

digital lloyd's case - his tests are hardly real world and he doesn't indicate it as such.

the same "theory" actually applies to DOF.

DOF is actually even more so related to the image manification in exactly the same way - which is why people complain more now than ever before about back / front focus issues - and assume wrongly that QA has fallen - we've progressively changed the circle of confusion size in our testing methodology as the pixel densities increased.

 

[Skip M]

Not true at all.

There is an article at Luminous-Landscape where the images from a
large view camera are mush at F64.

BC

I'd be interested in seeing that article. Do you have a link? I
couldn't find it on LL.
By the way, Ansel Adams, Edward Weston et al did ok at f64...
http://en.wikipedia.org/wiki/Group_f/64
Of course, f64 was just a name, they didn't shoot everything at f64...

f/64 is indeed just a name. It's effect depends on the imager size
and magnification you used. AA shooting f/64 on 8x10 is equivalent in
both DoF and diffraction terms to f/8 on FF. If an 8x10 is 'mush' at
f/64, then any FF camera is 'mush' at f/8.

Exactly, and none of them are "mush" at f8.

By the way, that's what I thought, that some commonly used aperture at 35mm and f64 on 8x10 would be similar, but I figured more like f11. The DOF seems greater than f8 would produce, to me.
--
Skip M
http://www.shadowcatcherimagery.com
http://www.pbase.com/skipm
http://skipm.smugmug.com/
'Living in the heart of a dream, in the Promised Land!'
John Stewart

 

[Ben_Egbert]

Although I haven't used my 5D2 for a serious trip yet (March 7-13 Big
Bend NP will be my first chance), I expect diffraction to be
stronger. I'm thinking DoF blending is going to be big in my future.

It won't be. It will be exactly the same.

Could you reconcile the above statement against the findings in the
link below? We can both stipulate that neither of us has this kind of
noticeable blur at f8 on our 5Ds.

http://www.diglloyd.com/diglloyd/free/Diffraction/example-1DsM3.html

his findings are based upon 100% magnification of the image.

which with respects to print size, is meaningless. both the image
magnification and observer distance is flawed or poorly
misprepresented in his article.

if you print a 12x18 from an 12Mp, 21Mp and a 24Mp camera all at f/16

• even though the 21, and 24 have increasing levels of diffraction at

100% magnification - on the print they will appear increasingly
sharper and more detailed than the one at 12Mp which is hardly
diffraction limited at 100%.

because in all cases you are keeping the observer distance the same,
and the magnification of the image the same.

the noticable effect of diffraction entirely depends on the
magnification, and observer distance - it's what the eye can possibly
detect as being "soft" or diffused.

Diffraction limitations can be easily thought of as when the airy
disc of diffraction is greater than the circle of confusion.

if you use the above meaning - then also be aware that Circle of
Confusion is ENTIRELY based upon image size and observer distance and
not of pixel density.

Therefore, diffraction, or the noticable effect of diffraction (which
is based upon aperture) is not based upon pixel density, but simply
the circle of confusion of your related image size and distance
viewed by the observer.

My comments are aimed specifically at the viewing distance concept, because for me it is flawed for reason I will descirbe.

Photographers who are concerned with diffraction, printing at larger sizes, buying 21 mpixel cameras and L lenses are not always following the 1930's era viewing distance guidelines. I personally view all images no matter what size at about 10 inches, which is also the only distance where I have perfect vision and don't require eyeglasses.

It matter's little to me what other viewers think of my images, if they don't look sharp to me, I am not happy.

--
http://www.pbase.com/roserus/root

Ben

 

[Ben_Egbert]

• we've progressively changed the circle of confusion

size in our testing methodology as the pixel densities increased.

And rightfully so, the technology has improved greatly from 35 mm film days and we pay a lot more for equipment and expect a lot more. We are pushing into what was formally MF territory.

--
http://www.pbase.com/roserus/root

Ben