Pages: 1 [2] 3 4   Go Down

Author Topic: Understanding Lens Diffraction  (Read 8860 times)

David Sutton

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1345
    • David Sutton Photography
Re: Understanding Lens Diffraction
« Reply #20 on: June 07, 2019, 04:44:58 am »

I'd forgotten I'd done a blog post on in 2013 here.
A little dated now but I realised most sharpening software would let me go to f/22 on a full frame sensor. I don't think much has changed in that regard.
Logged

Rob C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 24074
Re: Understanding Lens Diffraction
« Reply #21 on: June 07, 2019, 07:17:20 am »

I agree 100% with Island Dave on this one.

Worrying about those things is a crippling influence that, unless you are making huge blow-ups for hospital corridors etc. mean nothing - at least to me. I have gazed at many such photographs in the last fifteen years, and all I remember are pixels at a yard away.

I used to make 60x40 inch prints for store and fashion show exhibition from my Nikons or, sometimes, 'blads, and yep, they got a little fuzzy seen up close, but that isn't the way those images were seen. They were there to attract the passer-by, the buyer from somewhere. Long live FP3/4 and TXP 120!

Rob
« Last Edit: June 07, 2019, 07:30:47 am by Rob C »
Logged

jeremyrh

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2511
Re: Understanding Lens Diffraction
« Reply #22 on: June 07, 2019, 08:03:20 am »

Well I am sorry to disagree with you Bill, but I think you are entirely wrong in this case if all you are suggesting we do is base our view of photography on such meaningless technical issues, which may indeed be absolutely correct in a science over aesthetics sort of way. Because I would argue long and hard, that photography is no longer about attaining eye watering sharpness anymore, or the technical performance of exotic overpriced lenses and camera bodies, as that ship has sailed and now even the most basic system has evolved way beyond the abilities of your average Joe photographer these days. So it is now once again (and as it should be IMHO), more about the picture and the art of photography than it is about the kit and so anyone should be able do it reasonably well with a keen eye and a reasonably priced system, that compared to the equipment of old, has gone well beyond what most of us should ever need. I mean if Ansel was alive today, I don't think you or I or anyone else for that matter, would be telling him that his pictures aren't very good, because they're not very sharp or a high enough pixel density?

Dave

Somewhat missing the point, IMO. I see a camera as, to some extent, a device for capturing data. When I have that data I can do what I want with it - make a sharp picture, make a blurry picture, whatever. If I don't have the data, I have no choice. Of course this is a personal view - many people enjoy the random, unpredictable, irreversible results of losing data when shooting with a Lomo; I don't.  Then you may ask "how much data is enough?". Impossible to answer in a simple way. For posting little pictures on Instagram, a FF sensor is overkill; for making a print for a wall, it's easy to not have enough, as evidenced by the various discussions of focus stacking.
Logged

KMRennie

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 968
Re: Understanding Lens Diffraction
« Reply #23 on: June 07, 2019, 10:47:17 am »

Without recourse to mathematics or nit picking it would seem that we have a quandary. For the time being I will ignore tilt lenses or focus stacking. If we photograph any scene other than a flat plane we will need to stop the lens down. When we do this we lose sharpness, how much and whether we can apparently get some of it back with good sharpening is not really my point. We decide what needs to acceptably sharp and their distance from the focal plane. We can use 1/3 way in, hyperfocal distance, focus on infinity and stop down until the foremost object is sharp enough or George Duovos' app and it looks promising but I don't use an iPhone and have to accept that we will not be able to produce an image that is as sharp throughout the image as is theoretically possible. Whether we have the equivalent of 46Mp or 7Mp or somewhere in between is not really the point, in one image we will not get it any sharper.

If anything is moving, clouds, foliage, water then focus stacking may be tricky and is therefore not a universal solution to the loss of sharpness. Tilt lenses altering the plane of focus, that has problems in that some scenes eg a flower filled field with trees does not have the interest points in a flat plane irrespective of how it is tilted. Shift lenses are only available in discrete focal lengths so in some occasions you will be cropping and thus not obtaining the theoretical max from the camera.

What use do we make of our images? I seldom print larger than 16 by 12 and print less than 1% of my images so the majority of my viewing is on a screen. I have never had a problem with sharpness on my prints taken with my D810 and seldom with my Olympus E-410 and that when I had stopped down to f22. I am not calculating how many pixels I need for a sharp screen image but it is not great.

So I will do as I always have and focus stack when I have an enormous difference between foreground and background otherwise focus on what I want to be sharp and stop down enough to keep the rest looking fine, this is fairly close to Dave's approach. I do find that driving the spikes of my tripod into the ground, shielding the camera from the wind, using a cable release, mirror lockup, stitching panos rather than using a lens any wider than 24mm, electronic first curtain and shielding the clean lens and filters from the sun has a far greater effect on picture quality than worrying about how many Mp are actually in the resultant image. Ken

Lastly have we all been sold into the Mp war and the need for incredibly sharp lenses when at the kind of apertures that we need to use diffraction
Logged

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Re: Understanding Lens Diffraction
« Reply #24 on: June 07, 2019, 10:56:04 am »

Aperture is you variable AA filter. The more you stop down, the stronger AA.

Except that in the case of the D850, Moire will be much less of an issue

Moire is not much of a problem with the D850 at optimal apertures (around f/5.6). I have been using the D850 for over 2 years with thousands of shots. Moire is infrequently present but can be seen in architectural shots with gratings or Venetian blinds and with some fabrics, but it can be handled with software tools such as the Moire brush in Lightroom without the need to blur the entire image by using a small aperture.

With the 36 MP D800 Nikon offered options with (D800) and without a blur filter (D800e). The smaller pixel pitch of the D850 moves the frequency at which Moire can occur to a higher frequency that is not present in most subjects and Nikon wisely chose to omit the blur filter with the D850.

bILL
Logged

Slobodan Blagojevic

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 18090
  • When everyone thinks the same, nobody thinks
    • My website
Re: Understanding Lens Diffraction
« Reply #25 on: June 07, 2019, 11:13:03 am »

... I see a camera as, to some extent, a device for capturing data. When I have that data I can do what I want with it - make a sharp picture, make a blurry picture, whatever. If I don't have the data, I have no choice...

There are two types of people when it comes to choice: maximizers and satisfiers. You are apparently a maximizer. Maximizers won't stop until they find the best among available choices. Satisfiers (like me) stop when they find good enough. Maximizers tend to spend more time deliberating and choosing, Satisfiers choose quicker and move on to do something with the chosen.

HSakols

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1239
    • Hugh Sakols Photography
Re: Understanding Lens Diffraction
« Reply #26 on: June 07, 2019, 01:24:03 pm »

I think an interesting comparison would be a landscape taken at f16 and the same landscape stacked at say f5.6. from a full frame sensor and then view the prints. How apparent is the difference between the two?  Ideally, I would want to see the results from at least a couple different focal lengths. 
Logged

John Camp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2171
Re: Understanding Lens Diffraction
« Reply #27 on: June 07, 2019, 02:20:03 pm »

Having never been much interested in the subject, as I mostly shoot shaky street in which diffraction is the least of my problems (among other things, I'm usually shooting wide open, or close to it) I don't know the math or the technical stuff about diffraction. I can say that a friend has done serious work in this area and has said that human eyes can detect very tiny differences in sharpness when comparing prints side-by-side, but has also said that m4/3 cameras and lenses can make excellent prints at fairly large sizes, because you're not often comparing those side-by-side with sharper prints. In other words, sharpness has a psychological dimension as well as a technical one, and the psychological aspect is as important, and probably more important, as the technical one. A 24mp shot can look absolutely sharp at a large size, and while a 50mp shot at the same size will be technically sharper, it won't seem that way to a viewer who is not looking at both at the same time. You can do an enormous amount of research on what "pretty" means, and nail down 150 aspects of "pretty," but a pretty girl is a pretty girl.

Again, I don't know the math, but I'm curious about one aspect of Michael's article. He says that diffraction occurs because as the lens closes down, the amount of "disturbed" light at the edges of the shutter blades increases as a percentage of the total light hitting the sensor or film. But as the shutter closes down, doesn't the amount of blade exposure/interference also decrease? Or is it that it just doesn't decrease as fast as the overall area of the shutter opening?
Logged

32BT

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3095
    • Pictures
Re: Understanding Lens Diffraction
« Reply #28 on: June 07, 2019, 02:38:42 pm »

Or is it that it just doesn't decrease as fast as the overall area of the shutter opening?

This.

The circle area decreases significantly faster than the circumference. (pi x r squared) vs (2 x pi x r).
Logged
Regards,
~ O ~
If you can stomach it: pictures

Martin Kristiansen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1527
    • Martin Kristiansen
Re: Understanding Lens Diffraction
« Reply #29 on: June 09, 2019, 04:33:29 pm »

The moment that Nyquist thingy appears my mind gets mushier than an image shot at f64. 

Here is my way of working. Depth of field I control mostly with aperture as the other controls over depth of field, namely focal length of lens and distance to the subject, have been determined by other creative or practical considerations. I decide how much or how little I require in focus and where I want that focus to be and choose my aperture accordingly. I am sure that you are correct that a larger aperture  will be sharper due to less diffraction but will the distant tree be sharper at f22 or f5,6 when I have selected to focus on an important foreground element? Yes the rock I focused on 5m away will be sharper at 5,6 than at f22 but what about the tree 100 meters away, at what aperture will it be sharper when my focus remains on the rock? At 5,6 or 22?

We are not talking about absolute sharpness with no context for goodness sakes. And yes I do know about field cameras, I bought my first one when I was 18, forty years ago. I am also aware of focus stacking, a more tedious endeavor than peeling grapes. Aperture is an exposure and depth of field control, shutter speed is an exposure and motion blur control, terror about diffraction is for astronomers.

Now and again I take a reasonable photo. No one ever said, “Nice shot, shame about the diffraction”.
Logged
Commercial photography is 10% inspiration and 90% moving furniture around.

jeremyrh

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2511
Re: Understanding Lens Diffraction
« Reply #30 on: June 10, 2019, 04:01:18 am »

terror about diffraction is for astronomers.

Difficult to understand the reluctance of people to accept that diffraction is a real phenomenon that affects photographs!  Why not just do a quick calculation for your sensor etc. and see what limits on aperture are indicated for situations that are relevant to you  (numerous apps do this) and then bear this in mind when you're shooting?
Logged

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: Understanding Lens Diffraction
« Reply #31 on: June 10, 2019, 06:29:38 am »

Difficult to understand the reluctance of people to accept that diffraction is a real phenomenon that affects photographs!  Why not just do a quick calculation for your sensor etc. and see what limits on aperture are indicated for situations that are relevant to you  (numerous apps do this) and then bear this in mind when you're shooting?

I agree. Diffraction sucks the life out of an image (loss of micro-detail), but there is some mitigation possible with proper (deconvolution) sharpening. Not all images need to be technically perfect, but when given a choice, why not go for the better result?

My rule of thumb is based on that I can see the onset of diffraction losses when the diffraction pattern diameter (at the first zero ring) exceeds 1.5x the photosite pitch. That means that, approximately, pitch x 1.108 is the Aperture number where (green wavelength) diffraction becomes visible as loss of contrast. It can still be mostly compensated for, but it's gradually down-hill from there with narrower apertures until even high contrast detail cannot be resolved anymore. One might as well have used a lower resolution sensor.

MTF will go to zero at the following (circular) aperture:
N = 1 / (cycles_per_mm x wavelength)

The cycles / mm are given by the photosite pitch, e.g. 6.4 micron pitch equals (1 / 0.0064) / 2 = 78.125 cy/mm. In that case, the limiting aperture becomes 1 / (78.125 x 0.000555) = 23.063, so at approx. f/22 for green light (555 nm).

Cheers,
Bart
« Last Edit: June 10, 2019, 06:58:34 am by Bart_van_der_Wolf »
Logged
== If you do what you did, you'll get what you got. ==

Rob C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 24074
Re: Understanding Lens Diffraction
« Reply #32 on: June 10, 2019, 06:52:41 am »

The moment that Nyquist thingy appears my mind gets mushier than an image shot at f64. 

Here is my way of working. Depth of field I control mostly with aperture as the other controls over depth of field, namely focal length of lens and distance to the subject, have been determined by other creative or practical considerations. I decide how much or how little I require in focus and where I want that focus to be and choose my aperture accordingly. I am sure that you are correct that a larger aperture  will be sharper due to less diffraction but will the distant tree be sharper at f22 or f5,6 when I have selected to focus on an important foreground element? Yes the rock I focused on 5m away will be sharper at 5,6 than at f22 but what about the tree 100 meters away, at what aperture will it be sharper when my focus remains on the rock? At 5,6 or 22?

We are not talking about absolute sharpness with no context for goodness sakes. And yes I do know about field cameras, I bought my first one when I was 18, forty years ago. I am also aware of focus stacking, a more tedious endeavor than peeling grapes. Aperture is an exposure and depth of field control, shutter speed is an exposure and motion blur control, terror about diffraction is for astronomers.

Now and again I take a reasonable photo. No one ever said, “Nice shot, shame about the diffraction”.

I agree wholeheartedly with you; don't forget, though, that we are/were full-time professional photographers where getting the shot done and dusted was what counted rather than esoteric number crunching. The challenges were all of an entirely different nature, many starting well before we got to shooting that job.

Different mindset.

Rob

Martin Kristiansen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1527
    • Martin Kristiansen
Re: Understanding Lens Diffraction
« Reply #33 on: June 10, 2019, 07:44:26 am »

Difficult to understand the reluctance of people to accept that diffraction is a real phenomenon that affects photographs!  Why not just do a quick calculation for your sensor etc. and see what limits on aperture are indicated for situations that are relevant to you  (numerous apps do this) and then bear this in mind when you're shooting?

Obviously it is a real thing. Who said it’s wasn't.

My point is an object 100 m away from where you focused is less sharp at f5,6 than it is at f22 even if there is less diffraction at f5,6. If the aim is sharpness can you not understand that a greater depth of field can be better than zero diffraction but totally blurred due to extremely shallow depth of field?

Every new lens I buy I set up and check sharpness at every aperture. I then know that below a certain point that I will get diffraction issues. However if I think a greater depth of field is more important then to hell with diffraction. A bit like my best ISO could be 200 but if I want the shot and it is getting dark and the wind is blowing the trees about which I wish to freeze then I will push my ISO to whatever is required.

It’s not a tough concept. Photography has always been about balancing technical capabilities of the medium against the message. Message trumps technical perfection every time.
Logged
Commercial photography is 10% inspiration and 90% moving furniture around.

jeremyrh

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2511
Re: Understanding Lens Diffraction
« Reply #34 on: June 10, 2019, 08:33:10 am »

I agree wholeheartedly with you; don't forget, though, that we are/were full-time professional photographers where getting the shot done and dusted was what counted rather than esoteric number crunching acquiring the optimal image.

FTFY
Logged

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Re: Understanding Lens Diffraction
« Reply #35 on: June 10, 2019, 08:40:19 am »

I agree. Diffraction sucks the life out of an image (loss of micro-detail), but there is some mitigation possible with proper (deconvolution) sharpening. Not all images need to be technically perfect, but when given a choice, why not go for the better result?

My rule of thumb is based on that I can see the onset of diffraction losses when the diffraction pattern diameter (at the first zero ring) exceeds 1.5x the photosite pitch. That means that, approximately, pitch x 1.108 is the Aperture number where (green wavelength) diffraction becomes visible as loss of contrast. It can still be mostly compensated for, but it's gradually down-hill from there with narrower apertures until even high contrast detail cannot be resolved anymore. One might as well have used a lower resolution sensor.

MTF will go to zero at the following (circular) aperture:
N = 1 / (cycles_per_mm x wavelength)

The cycles / mm are given by the photosite pitch, e.g. 6.4 micron pitch equals (1 / 0.0064) / 2 = 78.125 cy/mm. In that case, the limiting aperture becomes 1 / (78.125 x 0.000555) = 23.063, so at approx. f/22 for green light (555 nm).

Cheers,
Bart

+1

For the Nikon D850 with a pixel pitch of 4.35 microns, the Nyquist is 115 cy/mm and the limiting aperture as calculated above is f/16 with a MTF of zero. I'm not certain what an MTF of zero means. At f/16 one can still see considerable detail in a D850 image at f/16.

Cheers,

Bill
Logged

jeremyrh

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2511
Re: Understanding Lens Diffraction
« Reply #36 on: June 10, 2019, 08:43:21 am »

My point is an object 100 m away from where you focused is less sharp at f5,6 than it is at f22 even if there is less diffraction at f5,6.
It may be. It may not be.
Quote
If the aim is sharpness can you not understand that a greater depth of field can be better than zero diffraction but totally blurred due to extremely shallow depth of field?

Every new lens I buy I set up and check sharpness at every aperture. I then know that below a certain point that I will get diffraction issues. However if I think a greater depth of field is more important then to hell with diffraction.
Good plan. Sometimes if I want to get somewhere really quick I say to hell with gravity and jump over buildings :-)
Quote
It’s not a tough concept. Photography has always been about balancing technical capabilities of the medium against the message.
True, but that is another issue. The question here is how to obtain the most technically perfect image, and that is by considering the relative influences of factors which degrade the image.
Logged

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: Understanding Lens Diffraction
« Reply #37 on: June 10, 2019, 10:50:21 am »

+1

For the Nikon D850 with a pixel pitch of 4.35 microns, the Nyquist is 115 cy/mm and the limiting aperture as calculated above is f/16 with a MTF of zero. I'm not certain what an MTF of zero means. At f/16 one can still see considerable detail in a D850 image at f/16.

Hi Bill,

An MTF of zero, in this context, means that there is no discernible detail left at spatial frequencies at or above that cut-off point (Nyquist in this case) due to diffraction. Larger detail will be recorded, but with low contrast. So significant deconvolution sharpening is required to make the most of it. So it's a diffraction-limited MTF.

Here is an article about it:
http://spie.org/publications/tt52_151_diffraction_mtf?SSO=1

This would be ideal for suppression of aliasing (but with severe loss of contrast). That can be helpful as an additional security shot for subjects that give rise to aliasing, so one can create a repair layer and brush or blend away local aliasing artifacts. I use it with architecture, where a window can exhibit aliasing artifacts e.g. because there is a sunshade with a regular pattern, or bricks walls at an angle. So I take a normal shot at a wider aperture for higher resolution, and a safety shot at the diffraction cut-off aperture.

Cheers,
Bart
« Last Edit: June 10, 2019, 11:38:13 am by Bart_van_der_Wolf »
Logged
== If you do what you did, you'll get what you got. ==

Martin Kristiansen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1527
    • Martin Kristiansen
Re: Understanding Lens Diffraction
« Reply #38 on: June 10, 2019, 11:35:27 am »

I agree wholeheartedly with you; don't forget, though, that we are/were full-time professional photographers where getting the shot done and dusted was what counted rather than esoteric number crunching. The challenges were all of an entirely different nature, many starting well before we got to shooting that job.

Different mindset.

Rob

Yep. I’m try to imagine explaining to a client that the product goes badly out of focus at the back but tough because F5,6 is the sharpest aperture so that’s where we are stuck. Of course I could focus stack but I will have to put up the price three times, sorry about that.
Logged
Commercial photography is 10% inspiration and 90% moving furniture around.

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Re: Understanding Lens Diffraction
« Reply #39 on: June 10, 2019, 11:40:47 am »

Hi Bill,

An MTF of zero, in this context, means that there is no discernible detail left at spatial frequencies at or above that cut-off point (Nyquist in this case) due to diffraction. Larger detail will be recorded, but with low contrast. So significant deconvolution sharpening is required to make the most of it. So it's a diffraction-limited MTF.

Here is an article about it:
http://spie.org/publications/tt52_151_diffraction_mtf?SSO=1

This would be ideal for suppression of aliasing (but with severe loss of contrast). That can be helpful as an additional security shot for subjects that give rise to aliasing, so one can create a repair layer and brush or blend away local aliasing artifacts. I use it with architecture, where a window can exhibit aliasing artifacts e.g. because there is a sunshade with a regular pattern, or bricks walls at an angle. So I take a normal shot at a wider aperture for higher resolution, and a safety shot at the diffraction cut-off aperture.

Cheers,
Bart

Bart,

Thanks for the clarification and your method of taking an additional security shot at a smaller aperture.

BTW, did you see my previous post in this thread using your sinusoidal Siemens star. Your comments would be appreciated.

https://forum.luminous-landscape.com/index.php?topic=130733.msg1111937#msg1111937

Thanks,

Bill
Logged
Pages: 1 [2] 3 4   Go Up