Q/Q2 image quality

[Jared]

Advertisement (gone after registration)

1 hour ago, jaapv said:

One thing not considered here is the diffraction limit throughout the optical system, For an (non-existent) flawless lens on an aberration-free full-frame sensor, the diffraction limit is 30 MP @ f 8.0, 60 MP @ f 5.6.

https://gmpphoto.blogspot.com/2014/11/the-pixel-race-does-it-really-make-sense.html

Which means that, if one is a landscape or street shooter and usually at something like f8.0 or even f11, there is no MP advantage in the Q2.

That’s why techniques must evolve. Hurray for focus stacking!  

Of course your general reminder is correct.  The burden placed on the lens is higher with more megapixels, and the diffraction limit becomes more relevant.  

[jaapv]

30 minutes ago, Jared said:

That’s why techniques must evolve. Hurray for focus stacking!  

Of course your general reminder is correct.  The burden placed on the lens is higher with more megapixels, and the diffraction limit becomes more relevant.  

But it also puts a realistic limit on the increase in MP number.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[Jared]

1 minute ago, jaapv said:

But it also puts a realistic limit on the increase in MP number.

It does, though we still have a way to go.  100 megapixels or even 150 could still be supported on full frame at faster apertures if the glass is good enough.  

Of course, I’ve been thinking for some time that we have already reached the point where more megapixels don’t actually help for most uses.  Lots of people have thought that—I’m no more perceptive than the next guy or gal on this topic.  Frankly, 24 megapixels is ample for almost all of my uses.  So was 18.  If I’m really honest, so was six.

The problem is that phrase “almost all of my uses.”  Like most photographers, I think, I am constantly dreaming of making that one great image that lives up to the picture in my mind’s eye, and I’d hate for that image to be lessened by lack of resolution.  Or noise.  Or imperfect focus.  Or just slightly missing the timing.  Or by a little bit of motion blur.  So, like most photographers, I keep buying cameras with “more”.  Even though I know in my mind the pictures aren’t any better, my heart keeps saying, “If only...”

[Pixeleater]

There’s no question that the Q2 is the "better" camera, but only you can decide whether these upgrades merit the extra cost in comparison to the original Q. I have the Q-P and am quite satisfied with its performance in terms of image quality and ergonomics.

[jaapv]

58 minutes ago, Jared said:

It does, though we still have a way to go.  100 megapixels or even 150 could still be supported on full frame at faster apertures if the glass is good enough.  

Of course, I’ve been thinking for some time that we have already reached the point where more megapixels don’t actually help for most uses.  Lots of people have thought that—I’m no more perceptive than the next guy or gal on this topic.  Frankly, 24 megapixels is ample for almost all of my uses.  So was 18.  If I’m really honest, so was six.

The problem is that phrase “almost all of my uses.”  Like most photographers, I think, I am constantly dreaming of making that one great image that lives up to the picture in my mind’s eye, and I’d hate for that image to be lessened by lack of resolution.  Or noise.  Or imperfect focus.  Or just slightly missing the timing.  Or by a little bit of motion blur.  So, like most photographers, I keep buying cameras with “more”.  Even though I know in my mind the pictures aren’t any better, my heart keeps saying, “If only...”

However, 24 MP on FF is, of course, the diffraction limit of most lenses (seen separately). The "traditional" range is up to f 8.0 until diffraction kicks in. On a 24 MP sensor you are very close, and I can see why Leica considered this an optimal setup, only to be forced from this position by competition pressure.

Nevertheless, a 100 MP sensor might be useful for Apo-Summicrons @ f 2, but whether it will be very relevant for real-world photography begs the question. We are into diminishing returns here. Effective high-MP photography needs large sensors, plain and simple.

 

[BXL Gotham]

Folks above. Are you saying the Q2 hits refraction at F8? So if I'm shooting F10 I am losing sharpness, is this verified? Just looking to know as it factors in one of the approaches I take to focusing (Merklinger) and I've been meaning to ask about this for a while.

[Jared]

Advertisement (gone after registration)

10 hours ago, BXL Gotham said:

Folks above. Are you saying the Q2 hits refraction at F8? So if I'm shooting F10 I am losing sharpness, is this verified? Just looking to know as it factors in one of the approaches I take to focusing (Merklinger) and I've been meaning to ask about this for a while.

This is easy enough to validate on your own—no special equipment needed.  Shoot an image of something at f/4 or so (the Q2 performs best in the center of the field at around f/4, at least mine does). Make sure your subject is in the approximate center of the field.  Then shoot the same image at f/11.  Import into the computer. Turn all sharpening and noise reduction off.  Examine the images in the middle of the field of view at 200%.  Is there a difference?  Which one is sharper?  In my experience, yes, diffraction is noticeably affecting image quality on the Q2 by f/11.

Keep in mind, diffraction always affects every image, no matter what the megapixel count of your camera and no matter the aperture.  The question is when does diffraction become the dominant factor in MTF. My simple test above should answer that for you—at least in the middle of the frame where the image quality is highest.  

Things get more complicated as you move to the edge of the field.  First, while the lens on the Q2 is extremely good, it’s not as good in the corners as it is in the middle of the field.  There’s nothing unusual in that.  The vast majority of lenses perform best in the middle of the frame.  Often, wide angle lenses like the 28mm on the Q2 get better in the corners as you stop them down.  I would say the Q2 is best in the corners at perhaps f/8 or so.  Again, your sample may differ, so try your own tests.  Keep in mind that lens assembly is never perfect, so you may find that one corner is materially better or worse than other corners.  In any event, the generally lower resolution in the corners may mean that diffraction isn’t the dominant factor at f/11.  Perhaps you would have to go to f/16 or f/22 before you see the image getting worse.  It isn’t because the effects of diffraction are any different in the corners, it’s just that other things might matter more so you don’t notice. Try it with your camera and see when the corners start to get noticeably softer.

Finally, keep in mind that these effects are relatively subtle.  In many images the increased depth of field you get with smaller apertures far outweighs the slight image degradation in the center of the field.  This will depend on the subject.  In a portrait, for example, corner performance rarely matters much at all.  In a landscape?  It could be critical.  Some photographers have turned to focus stacking to get the best of both worlds, but this won’t work for all subjects and it requires additional effort in post processing.  An even more extreme technique is the pixel shift technology incorporated into many cameras with image stabilization now.  The new Sony A7R Mark 4, for example, can produce a 240 megapixel image of a perfectly still subject that contains demonstrably more detail than the native 60 megapixels or so.  At least at certain apertures it does.

Diffraction isn’t a brick wall that you bump up against at some magic number.  It affects all images all the time.  As the aperture gets smaller, the affect gets bigger.  Eventually you get almost nothing by increasing megapixels.  We aren’t there yet (even at f/8), but Jaap’s point is a good one.  With good technique, a good lens, a tripod, and appropriate aperture it is definitely possible to get more details out of a 47 megapixel camera than out of a 24 megapixel camera.  Does it make for a better photograph?   Does your photography even allow that better technique, careful focus, appropriate aperture and perhaps even a tripod?  If you are using the Q for street photography, for example, I’d say the extra megapixels are all but irrelevant.  Portraiture?  Sports?  You’ve probably chosen the wrong camera entirely.  Travel?  Yeah, you can probably benefit, mostly because you have extra room for croppingand would otherwise be stuck at 28mm.  Landscape?  Yeah, the extra megapixels could matter.  

At some point, though, we will probably see megapixel counts stop going up because they are virtually empty in terms of the additional content.  We are already seeing that with cellular phones which have very tiny pixels.  The last three releases of iPhones, for example, have all had 12 megapixel cameras.  The improvements are in better noise control, more focal lengths, better computational photography, and faster lenses rather than increased megapixels.  Clearly, Apple is of the opinion that a chip as small as that in the iPhone has reached its limit at 12.  Any additional pixels add no more relevant detail.  Larger formats will support more pixels, of course, but even full frame will have its limit.

[jaapv]

22 hours ago, BXL Gotham said:

Folks above. Are you saying the Q2 hits refraction at F8? So if I'm shooting F10 I am losing sharpness, is this verified? Just looking to know as it factors in one of the approaches I take to focusing (Merklinger) and I've been meaning to ask about this for a while.

I linked to an article addressing that question.  Somewhere between f 5.6 and f8.0The diffraction limit of a lens/sensor imaging system is an optical/mathematical parameter that is not affected by camera brand or type, only by aperture and sensor size. The article contains a handy chart with the diffraction Megapixel numbers for all sensor sizes and resolutions. Leica lenses and dedicated sensors are quite good in this respect.

Note that these numbers are theoretical ones. Real-life factors, like a non-ideal lens, a sensor filter/microlens stack, varying physical sensel size etc. will bring the numbers down somewhat. In real life I would put the diffraction limit of the Q2 at f5.6. The Q would be f8.0. 
However, do note that there is no law forbidding to take an image beyond the limit. DOF or light conditions may well demannd it, and it is up to the photographer to balance technical considerations.

 

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

* copyright Barnack-Berek Blog, reproduced under Zitatrecht.   

https://gmpphoto.blogspot.com/2014/11/the-pixel-race-does-it-really-make-sense.html

[hdmesa]

23 hours ago, Jared said:

I agree that people inspect images at 1:1 for fun.  I do it myself.  And I agree that noise levels are generally higher when viewed at 1:1 as megapixel numbers climb.  Physics requires that.  I also explicitly mentioned that in my previous post—per pixel SNR is worse with higher megapixel counts.  While viewing images at 1:1 may be entertaining for the photographer, it is not how we tend to share images, whether online or in print form.  So I think it bears repeating that the comparison we should make is not at a particular pixel scale (unless you want to equalize it by up-sampling one or down-sampling the other), but at a given print size or view size.  That’s what is actually useful and practical.  How does the image compare in detail and in noise in the way you will actually use it?  That’s not at 100% view.

Your second point—about needing to be more careful with technique if you want to get the most out of your additional pixels—is absolutely correct and also bears repeating.  If you are just sharing on Instagram it doesn’t matter one bit, but if you are going to print large or are trying to extract every last bit of detail the lens and sensor can provide then more megapixels benefit from more careful technique and more accurate focus.

I agree, but you're coming at this from a comparison standpoint, where it's completely valid to pull back and look at apples-to-apples. But I think it's a fair assumption that this is not where most users are coming from when they ask about moving up to a much higher resolution sensor. They are looking for differentiators, and that squarely falls under 1:1 viewing, ability to crop, and large print sizes. 1:1 is a bit of a personal luxury, but cropping and large printing are very practical and common uses for moving to a higher resolution sensor. Sure, they certainly benefit from also knowing that if they don't print any larger or don't zoom in past 50% everything may look the same, but that begs the question of why upgrade to a higher resolution sensor in the first place if you don't need to crop or print larger?

Edit to add: The diffraction discussion above is another scenario important to consider when moving up to a higher resolution sensor. On a much higher resolution sensor and at increased viewing/print sizes, the areas in critical focus may appear reduced, requiring a smaller f-stop or focus stacking (IF one is printing larger or zooming in further, which in my opinion, "of course they are"). This is one topic where I never hear people throwing the equivalency argument around, and I'm not sure why – perhaps because these topics are not brought up by new users that will swallow the equivalency notion without question.

Edited September 13, 2019 by hdmesa

[bullmoon]

Always concerned about diffraction.  I remember doing some extensive macro tests with my Canon and being frustrated that decreasing the aperture beyond a certain point actually made the photos worse at close examination.  There was only so much DOF I could get and have it sharp - seemed like f16 was about it, even though the lens would do f32, I was not happy with the results.

[Jared]

2 hours ago, hdmesa said:

I agree, but you're coming at this from a comparison standpoint, where it's completely valid to pull back and look at apples-to-apples. But I think it's a fair assumption that this is not where most users are coming from when they ask about moving up to a much higher resolution sensor. They are looking for differentiators, and that squarely falls under 1:1 viewing, ability to crop, and large print sizes. 1:1 is a bit of a personal luxury, but cropping and large printing are very practical and common uses for moving to a higher resolution sensor. Sure, they certainly benefit from also knowing that if they don't print any larger or don't zoom in past 50% everything may look the same, but that begs the question of why upgrade to a higher resolution sensor in the first place if you don't need to crop or print larger?

Edit to add: The diffraction discussion above is another scenario important to consider when moving up to a higher resolution sensor. On a much higher resolution sensor and at increased viewing/print sizes, the areas in critical focus may appear reduced, requiring a smaller f-stop or focus stacking (IF one is printing larger or zooming in further, which in my opinion, "of course they are"). This is one topic where I never hear people throwing the equivalency argument around, and I'm not sure why – perhaps because these topics are not brought up by new users that will swallow the equivalency notion without question.

I'm not certain I understand your points.  Let me know if I'm accurately summarizing and responding appropriately:

1) If you are buying a camera with more megapixels, of course you want more resolution, else why buy it?  Given that, comparisons at 1:1 are perfectly valid.

If I have accurately stated your first point, then my response is that, no, such comparisons are not particularly informative.  To make those comparisons informative you need to compare at similar scales or similar resolutions.  Either up-sample the lower megapixel image or down-sample the higher if you want to use 1:1.  Simply looking at noise levels for both at native resolution 100% is not useful.  Because the scale is different, it's an apples to oranges comparison.  It tells you nothing about how the images will compare when posted and it tells you nothing about how they will compare when printed.  As long as I view the images at the same scale I can properly identify differentiators.  Looking at both at 100% zoom at their native resolutions and trying to draw conclusions would be like measuring one in inches and the other in centimeters and trying to draw conclusions without a conversion.

2) People paying money to get more megapixels almost certainly want to print larger or crop more.  This is a practical and common use for moving to higher resolution sensor.  

I certainly agree those are good reasons to move to a higher resolution sensor.  In fact, they are just about the only reasons to move to a higher resolution sensor.  I suspect, though, that a lot of people are going to higher resolution sensors just because they feel in their gut that it must be "better" not because they actually have a use for more megapixels.  This will vary from one photographer to the next, of course.  I'm not speaking for you.  If you need the extra megapixels for large prints or for cropping, go for it.  If you want them because you think your instagram posts will be better, I would recommend you do a little experimentation before spending the money.

3) On a much higher resolution sensor with increased viewing/print sizes, the depth of field is obviously going to be narrower, so focus stacking or choosing a smaller diameter aperture is going to be important.  

So far so good.  If you want to print large with a subject where depth of field is relevant, then you will need to consider focus stacking or apertures with greater depth of field.  I'd say that's true regardless of megapixel count, but whether shooting with a Q or a Q2 or a (future) S3 you'll need a smaller aperture or focus stacking to keep apparent sharpness as you zoom in more or print larger.

4) I never hear equivalency argument brought up around the topic of diffraction. 

I think it is actually embedded in there, though it is not often called out explicitly.  Equivalency basically states that, given a perfect lens and identical technologies for chips in situations where read noise is negligible, the following situations would be equivalent. That is, they would yield functionally identical results photographs in terms of angle of view, depth of field, signal to noise ratio, and resolution:

- 50mm lens shot at f/2.8 on a full frame (36 mm x 24 mm) 24 megapixel camera at a base ISO of 100

- 25mm lens shot at f/1.4 on a micro 4/3" 24 megapixel camera at a base ISO of 25

Obviously, the choices above are arbitrary.  You can use different focal lengths, apertures, formats, and megapixel counts.  It's the ratios that are important.

These two example camera/lens/aperture combinations should yield the same depth of field, required shutter speed, resolution, angle of view, and noise.  That would include diffraction.  Both systems at the listed focal ratios would have the exact same effects from diffraction.

That's about as far as equivalence goes.  It doesn't guarantee that you can actually buy cameras with the same megapixel count in all format sizes.  It doesn't guarantee that you can design, build, or afford lenses of "equivalent" aperture.  For example, I can easily find 50mm f/1.4 lenses for full frame, but I certainly can't find a 25mm f/0.7 lens for 4/3" (which would be "equivalent").  I think where most people get into fights over equivalence is when they assume it means there are not differences between formats--that you can actually purchase "equivalent" cameras and "equivalent" lenses.  That's generally not true.  It just tells you what would be equivalent if it existed.  As far as it goes, it's perfectly valid.  It's just that you can't take it all that far.

[Jared]

1 hour ago, bullmoon said:

Always concerned about diffraction.  I remember doing some extensive macro tests with my Canon and being frustrated that decreasing the aperture beyond a certain point actually made the photos worse at close examination.  There was only so much DOF I could get and have it sharp - seemed like f16 was about it, even though the lens would do f32, I was not happy with the results.

Macro photography in particular is where focus stacking can solve a huge number of problems.  Most macro subjects are stationary which makes it feasible.  If you've got a flower waving in a breeze, though, you are absolutely right that pushing your macro lens up to f/32 will inevitably degrade image quality at the point of focus.  It is sometimes the only way to get the shot which is why macro lenses allow such small apertures even though it degrades quality.  Pesky laws of physics.

[hdmesa]

1 hour ago, Jared said:

I'm not certain I understand your points.  Let me know if I'm accurately summarizing and responding appropriately:

1) If you are buying a camera with more megapixels, of course you want more resolution, else why buy it?  Given that, comparisons at 1:1 are perfectly valid.

If I have accurately stated your first point, then my response is that, no, such comparisons are not particularly informative.  To make those comparisons informative you need to compare at similar scales or similar resolutions.  Either up-sample the lower megapixel image or down-sample the higher if you want to use 1:1.  Simply looking at noise levels for both at native resolution 100% is not useful.  Because the scale is different, it's an apples to oranges comparison.  It tells you nothing about how the images will compare when posted and it tells you nothing about how they will compare when printed.  As long as I view the images at the same scale I can properly identify differentiators.  Looking at both at 100% zoom at their native resolutions and trying to draw conclusions would be like measuring one in inches and the other in centimeters and trying to draw conclusions without a conversion.

2) People paying money to get more megapixels almost certainly want to print larger or crop more.  This is a practical and common use for moving to higher resolution sensor.  

I certainly agree those are good reasons to move to a higher resolution sensor.  In fact, they are just about the only reasons to move to a higher resolution sensor.  I suspect, though, that a lot of people are going to higher resolution sensors just because they feel in their gut that it must be "better" not because they actually have a use for more megapixels.  This will vary from one photographer to the next, of course.  I'm not speaking for you.  If you need the extra megapixels for large prints or for cropping, go for it.  If you want them because you think your instagram posts will be better, I would recommend you do a little experimentation before spending the money.

3) On a much higher resolution sensor with increased viewing/print sizes, the depth of field is obviously going to be narrower, so focus stacking or choosing a smaller diameter aperture is going to be important.  

So far so good.  If you want to print large with a subject where depth of field is relevant, then you will need to consider focus stacking or apertures with greater depth of field.  I'd say that's true regardless of megapixel count, but whether shooting with a Q or a Q2 or a (future) S3 you'll need a smaller aperture or focus stacking to keep apparent sharpness as you zoom in more or print larger.

4) I never hear equivalency argument brought up around the topic of diffraction. 

I think it is actually embedded in there, though it is not often called out explicitly.  Equivalency basically states that, given a perfect lens and identical technologies for chips in situations where read noise is negligible, the following situations would be equivalent. That is, they would yield functionally identical results photographs in terms of angle of view, depth of field, signal to noise ratio, and resolution:

- 50mm lens shot at f/2.8 on a full frame (36 mm x 24 mm) 24 megapixel camera at a base ISO of 100

- 25mm lens shot at f/1.4 on a micro 4/3" 24 megapixel camera at a base ISO of 25

Obviously, the choices above are arbitrary.  You can use different focal lengths, apertures, formats, and megapixel counts.  It's the ratios that are important.

These two example camera/lens/aperture combinations should yield the same depth of field, required shutter speed, resolution, angle of view, and noise.  That would include diffraction.  Both systems at the listed focal ratios would have the exact same effects from diffraction.

That's about as far as equivalence goes.  It doesn't guarantee that you can actually buy cameras with the same megapixel count in all format sizes.  It doesn't guarantee that you can design, build, or afford lenses of "equivalent" aperture.  For example, I can easily find 50mm f/1.4 lenses for full frame, but I certainly can't find a 25mm f/0.7 lens for 4/3" (which would be "equivalent").  I think where most people get into fights over equivalence is when they assume it means there are not differences between formats--that you can actually purchase "equivalent" cameras and "equivalent" lenses.  That's generally not true.  It just tells you what would be equivalent if it existed.  As far as it goes, it's perfectly valid.  It's just that you can't take it all that far.

Thanks for the thoughtful and helpful response.

As to what I was trying to convey originally, it was simpler than where we've taken this  – Tell both sides of the story. Someone asking, "Will I have to pay more attention to my technique on a much higher resolution sensor?" isn't looking for an equal-footing comparison between their old and new cameras, they're wanting to know what differences or gotchas to watch out for. Certainly it's fair to tell them there will be no gotchas IF they make equal comparisons, but they should also know that if they like to view their images at 1:1, intend to crop, will print larger, or shoot in a style requiring sharp results from near distance to infinity, they should indeed pay attention to certain techniques to avoid disappointments.

I think we agree that a lot of casual photographers moving up in sensor resolution may not have realistic expectations based on lack of knowledge or experience. And with that in mind, I prefer to explain what to watch out for rather than dwell on what remains the same. Of course, it's most helpful to explain both concepts

Edited September 13, 2019 by hdmesa

[nicci78]

Please forget any theoretical diffraction limit. It does not occur so fast in real life. Diffraction is more a mystery than a real science.

How to know where the random light rays will bounce ? It is random ! However algorithms get better and better, pushing the limit of diffraction. 

Just look the measure made by JMS one of the best french optics tester in http://www.lemondedelaphoto.com/Leica-Q2-le-nouveau-Tri-Elmar,15348.html

the Summilux-Q is excellent from f/1.7 to f/11 Diffraction only takes its toll at f/16 Not bad for a 47MP sensor  

The 24MP Q also get diffraction effect at f/16  so it is a draw

Q2 sensor is really better in almost every way. Q one is only better at 25000 ISO only. 50000 ISO is crap for both cameras  

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Edited September 13, 2019 by nicci78

[nicci78]

By the way JMS also tested the 4 APO-Summicron-SL They are all extraordinary from f/2 to f/16 with Panasonic S1R 47MP sensor. As you can guess, excellent optics are not afraid of diffraction with high MP count. 

[Jared]

1 hour ago, hdmesa said:

Thanks for the thoughtful and helpful response.

As to what I was trying to convey originally, it was simpler than where we've taken this  – Tell both sides of the story. Someone asking, "Will I have to pay more attention to my technique on a much higher resolution sensor?" isn't looking for an equal-footing comparison between their old and new cameras, they're wanting to know what differences or gotchas to watch out for. Certainly it's fair to tell them there will be no gotchas IF they make equal comparisons, but they should also know that if they like to view their images at 1:1, intend to crop, will print larger, or shoot in a style requiring sharp results from near distance to infinity, they should indeed pay attention to certain techniques to avoid disappointments.

I think we agree that a lot of casual photographers moving up in sensor resolution may not have realistic expectations based on lack of knowledge or experience. And with that in mind, I prefer to explain what to watch out for rather than dwell on what remains the same. Of course, it's most helpful to explain both concepts

Yes, that’s completely fair and reasonable.

[Jared]

1 hour ago, nicci78 said:

Please forget any theoretical diffraction limit. It does not occur so fast in real life. Diffraction is more a mystery than a real science.

How to know where the random light rays will bounce ? It is random ! However algorithms get better and better, pushing the limit of diffraction. 

Just look the measure made by JMS one of the best french optics tester in http://www.lemondedelaphoto.com/Leica-Q2-le-nouveau-Tri-Elmar,15348.html

the Summilux-Q is excellent from f/1.7 to f/11 Diffraction only takes its toll at f/16 Not bad for a 47MP sensor  

The 24MP Q also get diffraction effect at f/16  so it is a draw

Q2 sensor is really better in almost every way. Q one is only better at 25000 ISO only. 50000 ISO is crap for both cameras  

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

I would quibble a little bit with your interpretation of the results.  First, diffraction is hardly a mystery.  It may not be possible to predict where any one photon will end up, but in aggregate ray tracing is extremely effective in predicting behavior and the results are extremely consistent with the wave nature of light.  And new algorithms don’t change anything.  Good software can simplify lens design by quickly modeling the effects of small changes in refractive index, spacing, dispersion, radius of curvature, aspherization, etc., but they don’t change our understanding of optics at all.  

As to diffraction not taking its toll till f/16... Simply not accurate.  Image quality in the center of the field may well still be “excellent” at f/11, but we have no way of knowing what the tester considers “excellent” in terms of lp/mm or MTF except by comparison to other lenses and cameras.  The effects of diffraction increase as you stop down.  Always.  The report you linked to just tells you that results are still good in the center at f/11.  I wouldn’t disagree with that.  They are still “excellent” at f/11.  That doesn’t mean they are just as good as at f/5.6 or f/8.  They aren’t.

[bullmoon]

On 9/13/2019 at 3:56 PM, nicci78 said:

By the way JMS also tested the 4 APO-Summicron-SL They are all extraordinary from f/2 to f/16 with Panasonic S1R 47MP sensor. As you can guess, excellent optics are not afraid of diffraction with high MP count. 

Maybe I misunderstand, but I don't think diffraction is dependent on optics quality.  I think it is an issue with aperture size - small hole = more diffraction; It happens because light begins to disperse or "diffract" when passing through a small opening.

[Leica Guy]

3 hours ago, bullmoon said:

Maybe I misunderstand, but I don't think diffraction is dependent on optics quality.  I think it is an issue with aperture size - small hole = more diffraction; It happens because light begins to disperse or "diffract" when passing through a small opening.

I’m not a physicist, but my understanding more than the small hole, it’s really the angle that the light has to deflect from the outer edge of the front lens element through the aperture (a sharp edge) then back up to the edge of the sensor. Perhaps there’s a real optics technical expert who can set us all straight. 

It would seem to me that diffraction must have been a huge challenge for Ansel Adams and others using box cameras at small apertures like f/64. 

Edited September 19, 2019 by Leica Guy

[Advertisement]

Hallo,

Look here: Leica Q

[Jared]

6 hours ago, Leica Guy said:

I’m not a physicist, but my understanding more than the small hole, it’s really the angle that the light has to deflect from the outer edge of the front lens element through the aperture (a sharp edge) then back up to the edge of the sensor. Perhaps there’s a real optics technical expert who can set us all straight. 

It would seem to me that diffraction must have been a huge challenge for Ansel Adams and others using box cameras at small apertures like f/64. 

Nope, it’s just the size of the hole, not the angle from the outer edge of the lens.  In fact, the outer edge of the lens is not used at all when shooting stopped down—only the central part of the lens elements illuminate the imaging plane. Light from the outer portions of the lens is simply blocked by the iris and never reaches the imaging plane. 

This also explains why larger formats are less affected by diffraction.  An f/32 aperture on an 8”x10” view camera has a physically much larger diameter opening than the same focal length with a 35mm camera, so shooting at f/11 or f/16 or even f/64 is perfectly reasonable with a larger format.