Luminous Landscape Forum
Equipment & Techniques => Medium Format / Film / Digital Backs – and Large Sensor Photography => Topic started by: hdomke on December 04, 2011, 07:23:58 am
-
What is the best way to test a lens for diffraction?
Is it true that medium format lenses can go to much higher f-stops than 35mm DSLR lenses before diffraction is an issue?
I am asking this because I would like to avoid having to use focus stacking (http://en.wikipedia.org/wiki/Focus_stacking) in my landscape images.
In the Luminous Landscape video series “Camera to Print & Screen (http://www.luminous-landscape.com/videos/tutorials/camera_to_print_and_screen.shtml)” Michael Reichmann says: “With a pocket camera you may start to hit diffraction at f/8, with a DSLRs it is usually about f/16, but on a large or medium format camera you might be able to go to f/32 before you hit diffraction.”
Jeff Schewe: “Maybe, but that is something you have to test with that lens and sensor combination.”
Note: This is an approximate quote from about 4 minutes in to chapter 3 of “Camera to Print & Screen”
-
Are you a pixel peeper? If not, your aperture limit is proportional to the sensor size, the larger the sensor, the more you can stop down. So are you using a 44x33mm MF sensor or a a 53x40mm MF sensor? That will have something to do with it. I am quite happy to use f/16 with my 645D. I can get excellent results at f/22 as well.
If you are shooting with a tech/view camera, you can also use tilt.
-
Is it true that medium format lenses can go to much higher f-stops than 35mm DSLR lenses before diffraction is an issue?
Yes it is, but it is also true that medium format sensors need to go to higher f-stops than smaller format sensors to achieve the same DOF. In the end, there is no practical advantage/disadvantage for any format size regarding diffraction and DOF issues.
-
Hi,
It depends a lot on your expectations, your definition of an "issue" and what back you have.
Diffraction is usually defined by: r = 1.22*Lambda*N, where r is the radius of the Airy circle, Lambda is the wavelength of light and N is the f-number, like 16 for f/16.
So, for green light around 0.5 microns and f/16 we would get 9.76 microns. But this is the radius, the diameter would be twice that, namely 19.5 microns.
Recall that the Airy ring is a circle, so it's area would be r^2 * pi, that is 299 microns.
If pixel size is 9 microns, typical of many MFDBs around 20 MP the Airy ring would cover about 3.5 pixels. O
n the IQ180 with 5.17 micron pixel pitch it would cover 11 pixels. Both sensors would loose some sharpness, but the 9 micron pixel sensor may loose little and the 5.17 micron sensor a lot.
So if you are very careful about having maximum sharpness and in utilizing your lenses and have a high resolving sensor like P45, P65+, IQ180, Pentax 645D or a Leica S2 you would consider diffraction. If you don't really care about maximum achievable sharpness it may be less of an issue.
This page illustrates diffraction on a 4.7 micron sensor: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/49-dof-in-digital-pictures?start=1 . That would be pretty close to what you would see on an IQ180.
That said, diffraction is relatively benign to sharpening, as demonstrated here: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/49-dof-in-digital-pictures?start=2
This is a good link for using the hyperfocal distance correctly: http://optechsdigital.com/Alpa_and_Hyperfocal.html the article also demonstrates diffraction effects.
These two recent threads here on LuLa forums discuss this issue:
http://www.luminous-landscape.com/forum/index.php?topic=59862.0
http://www.luminous-landscape.com/forum/index.php?topic=59930.msg483360
Best regards
Erik
What is the best way to test a lens for diffraction?
Is it true that medium format lenses can go to much higher f-stops than 35mm DSLR lenses before diffraction is an issue?
I am asking this because I would like to avoid having to use focus stacking (http://en.wikipedia.org/wiki/Focus_stacking) in my landscape images.
In the Luminous Landscape video series “Camera to Print & Screen (http://www.luminous-landscape.com/videos/tutorials/camera_to_print_and_screen.shtml)” Michael Reichmann says: “With a pocket camera you may start to hit diffraction at f/8, with a DSLRs it is usually about f/16, but on a large or medium format camera you might be able to go to f/32 before you hit diffraction.”
Jeff Schewe: “Maybe, but that is something you have to test with that lens and sensor combination.”
Note: This is an approximate quote from about 4 minutes in to chapter 3 of “Camera to Print & Screen”
-
And naturally, the size of the Airy disk to the number of pixels is not really that important. What is important is the size of the Airy disk to the permissible circle of confusion for the format--the eye has a limit to resolve.
-
Hi,
Yes but that limit is quite flexible, it depends much on how large you want to print, but also on subject. Sometimes a small detail is important, like a penguin in an arctic landscape. It's generally agreed that the eye resolves 1 minute of an arc. Let's assume a viewing distance of 1 m. One minute of arc is 1/60 of a degree, at one meter this would be 0.29 mm. Two tell two points apart you would need a point in between. So you need three points spaced at 0.1 mm. So if you print at 70x100 cm and look at the image at 1 meter you would need 10000 pixels along the longer side to match the resolution of the unaided eye. On a 49 mm sensor that corresponds to 4.9 microns.
On the other hand I'd agree that the requirement can be much relaxed.
My opinion is that in many cases it's better to have the essential subject as sharp as possible and the less important parts acceptably sharp, than having everything a bit unsharp.
Another interesting option is that sharpening can be selectively, smart sharpen in Photoshop can do deconvolution for out of focus images that works quite well.
This article compares three different methods to achieve extended DoF, stopping down to f/22, Scheimpflug and focus stacking, with real world examples. http://echophoto.dnsalias.net/ekr/index.php/photoarticles/29-handling-the-dof-trap
Best regards
Erik
And naturally, the size of the Airy disk to the number of pixels is not really that important. What is important is the size of the Airy disk to the permissible circle of confusion for the format--the eye has a limit to resolve.
-
Erik, we seem to be bumping up against each other recently. Everything is based on standard viewing distance. The permissible circle of confusion is not something set in stone. And the permissible circle of confusion is not set at the eye's ability to resolve to point sources. Sharpness and resolution are not the same thing.
What MFD system are you shooting with where you are thinking you need to work at such tight tolerances? I just hung an exhibition of 5x4 foot portraits taken with my 645D. There is nothing to suggest that MFD has tighter tolerances than before digital photography--no one would say shooting with Tech Pan has less depth of field and is more susceptible to diffraction than TMax 400 just because it has finer grain and more resolving power. Nor with film was there any question you could use smaller apertures with larger formats. Digital has not changed how a human would perceive an image. The idea that an image has to be optimized to 100% monitor view really has no basis in any real condition an image will be viewed at, especially with the pixel numbers on MFD sensors.
Yes, technically you are right diffraction affects an image--after all, an image is created from diffraction. But the problem is relative one, not an absolute one based on pixel pitch.
-
Hi,
The example given was intended to illustrate what the eye can actually resolve.
An MFDB has tighter tolerance than film for five reasons.
- The first is the that the sensor is a flat device, while film has a curvature that varies between exposures.
- The second is that digital is said to have higher resolution than film. Joseph Holmes talks about a factor five.
- The third factor is simply that new optical systems are built for digital, sharper lenses, tighter tolerances.
- The fourth factor is that the digital image is a first generation product. With film there was always an additional optical step potentially degrading the image.
- The fifth is that we print large. Making large prints in the dark room was not easy.
Another factor is that there are people buying 80 MPixel backs, and specially calculated digital lenses with extreme performance. To utilize all that expensive technology the workmanship also needs to be on top.
You may check this article: http://optechsdigital.com/Alpa_and_Hyperfocal.html that describes a way to achieving optimal focus. Interestingly he also arrives at f/16 as minimal optimum aperture, using his P65+ plus on Alpha. It is a good article worth reading.
My take on the issue is that I'd rather have optimum sharpness on the main subject and have a bit extra sharpening on the out of focus areas than reduce overall sharpness.
The other point is that high end backs have more to loose than low end backs. With a P25 the loss at f/22 may be minimal, with a P65+ it is a lot.
Joseph Holmes articles are clearly worth reading, they are about making most of MF digital:
http://www.josephholmes.com/news-sharpmediumformat.html
http://www.josephholmes.com/news-medformatprecision.html
Best regards
Erik
Erik, we seem to be bumping up against each other recently. Everything is based on standard viewing distance. The permissible circle of confusion is not something set in stone. And the permissible circle of confusion is not set at the eye's ability to resolve to point sources. Sharpness and resolution are not the same thing.
What MFD system are you shooting with where you are thinking you need to work at such tight tolerances? I just hung an exhibition of 5x4 foot portraits taken with my 645D. There is nothing to suggest that MFD has tighter tolerances than before digital photography--no one would say shooting with Tech Pan has less depth of field and is more susceptible to diffraction than TMax 400 just because it has finer grain and more resolving power. Nor with film was there any question you could use smaller apertures with larger formats. Digital has not changed how a human would perceive an image. The idea that an image has to be optimized to 100% monitor view really has no basis in any real condition an image will be viewed at, especially with the pixel numbers on MFD sensors.
Yes, technically you are right diffraction affects an image--after all, an image is created from diffraction. But the problem is relative one, not an absolute one based on pixel pitch.
-
Erik, I shoot MFD. I have shoot film. The human eye has not evolved in my lifetime. Perhaps I have not printed large enough, but I think when you hit 12 feet, you really are at a point that larger is not going to be any different. My experience is that diffraction is very much overstated.
Personally, I don't know why Mr. Holmes finds this so difficult.
What MFD system do you use?
-
Hi,
No, I'm not shooting MFD, I think I have clearly stated this at least on of the discussion we both are involved in. Nor do I print large, 70x100 cm is my maximum, for practical reasons. My normal print size is A2 (around 18"x26").
I have been using Pentax 67 with film and I was considering to buy a Pentax 645D, but decided against it, basically because I don't print large.
The camera I shoot is the Sony Alpha 900, it has a pixel size similar to your Pentax 645D. The Alpha 900 has an OLP filter which the Pentax 645D does not have.
I also bought a DVD with images from a shootout made by Michael Reichmann, Bill Atkinsson and Charlie Cramer. I made actual prints from those images. The main reason I have so much interest in this is that I'm quite interested in optics.
I'm not entirely alone stating that diffraction starts to be an issue past f/16. I don't mean that you should never stop down beyond f/16 just that you start to loose a lot.
Another question is what can be regained with sharpening. That applies to both defocus and diffraction. I'd suggest that diffraction is more benign to deconvolution based sharpening than defocus. But adding sharpening to the discussion just makes the issue more complex.
Best regards
Erik
Erik, I shoot MFD. I have shoot film. The human eye has not evolved in my lifetime. Perhaps I have not printed large enough, but I think when you hit 12 feet, you really are at a point that larger is not going to be any different. My experience is that diffraction is very much overstated.
Personally, I don't know why Mr. Holmes finds this so difficult.
What MFD system do you use?
-
Erik, sorry if I missed you were not using MFD
All I can say is I use MFD and I print large with it. I also know about the concepts in photography that we have been discussing, which also support what I see. You can chalk me up to a lost cause if you are going to try to convince me otherwise.
And you will also not convince me that small print size means a larger format film or digital camera will have no impact on the quality of the image. Sampling theory is not enough. An 8x10 will look different depending on the format, whether film or digital. I have seen it and so have others. Make an 8x10 with a 6x7 and a 35mm negative and the images will not be the same. Again, print size is not the defining characteristic. Even the human visual system is not defined by a single characteristic. Resolution and detection are two different things where detection can actually perceive even finer detail than resolving power will imply. And resolving power, both in optics and the visual system, are not a fixed quality, but will change with object contrast--a lens will only resolve to its maximum ability, only if you also present it with a high-contrast target.
I have no doubt you have a genuine interest in this subject--the internet is probably the worst place to have conversations about topics people are passionate about. You certainly have experience. But this is not a simple optical problem. Photography is subjective and very relative because it all goes back to a viewer and their perception of an image. But I think we may be having a problem with trees and the forest...
-
... My experience is that diffraction is very much overstated...
So is gravity, in my experience.
If all the physicists would just stop the hysteria .... ::) ::) ::)
-
And you will also not convince me that small print size means a larger format film or digital camera will have no impact on the quality of the image. Sampling theory is not enough. An 8x10 will look different depending on the format, whether film or digital. I have seen it and so have others. Make an 8x10 with a 6x7 and a 35mm negative and the images will not be the same.
Yes, and that's the main reason I used both 6x6 and 135. The other reason for using 6x6, in my case, was that much of my fashion work required that the client get contact sheets... using 135, the solution often meant my selecting a few better shots and making blow-ups of them instead, but how can you do that quickly without cutting your own standards and creating a poor impression, regardless of how much better the selected, finished prints might be? Time costs money, and spending your own isn't so clever in many cases.
But regardless of size, tonality was always won by larger format!
Rob C
-
I was curious about this and shot a set up with a ruler and a $2 bill with both my CF 39 Multishot and an Aptus 12 through the whole range of apertures. I still have all the files so one day when I have time, I'll post a thread with the crops, but the summary is that I felt I could shoot to f/18 with the CF 39 in single shot, f/25 in multishot (yes its much different), and f/16 with the Aptus 12. I did this test mostly for my graphite hands art series which has very fine detail in the fingerprints but because of the focus distance needed as much DOF as possible. Previously I shot with a phase p20 which I thought I could stop down to f/20. Note that these are my own personal values, where I felt the trade off in more DOF vs loss of sharpness was balanced for larger prints. Everyone will have their own set of criteria. I'm sure were I printing smaller or with subject that had less fine detail, I might think I could stop down a bit more.
-
Hi,
Just a short comment, my initial posting which was third on this list began with:
"It depends a lot on your expectations, your definition of an "issue" and what back you have."
I also have a small anecdote. Once upon the time my best friend and I were shooting my friends Hasselblad SWC and my Minolta XE with MC Rokkor 24/2.8 in parallell. My friend was shooting Tri-X at 200 ASA (that was before ISO) developed in D-76 and I shot Plux-X, also at 200 ASA developed in HC-110. These were the films we used in those days.
We printed both images in his darkroom, with his enlarger using his APO-Rodagon. When we developed the smallish 18x24 cm prints (that was than) he was pretty sure that my picture was the one he shot with the Hasselblad.
I did shot with Pentax 67 on film for a long time, but found that I worked in a different way with the 67 than with 135. One of the issues I found that it was essentially impossible to achieve the depth of field I was used to with 135. So I essentially gave up on depth of field.
When I started scanning I found out that depth of field was more limited than I initially thought.
Anyway I was perfectly satisfied with the Pentax 67, but than I went digital and never looked back.
Best regards
Erik
Erik, sorry if I missed you were not using MFD
All I can say is I use MFD and I print large with it. I also know about the concepts in photography that we have been discussing, which also support what I see. You can chalk me up to a lost cause if you are going to try to convince me otherwise.
And you will also not convince me that small print size means a larger format film or digital camera will have no impact on the quality of the image. Sampling theory is not enough. An 8x10 will look different depending on the format, whether film or digital. I have seen it and so have others. Make an 8x10 with a 6x7 and a 35mm negative and the images will not be the same. Again, print size is not the defining characteristic. Even the human visual system is not defined by a single characteristic. Resolution and detection are two different things where detection can actually perceive even finer detail than resolving power will imply. And resolving power, both in optics and the visual system, are not a fixed quality, but will change with object contrast--a lens will only resolve to its maximum ability, only if you also present it with a high-contrast target.
I have no doubt you have a genuine interest in this subject--the internet is probably the worst place to have conversations about topics people are passionate about. You certainly have experience. But this is not a simple optical problem. Photography is subjective and very relative because it all goes back to a viewer and their perception of an image. But I think we may be having a problem with trees and the forest...
-
I certainly wouldn't shy away from using f22 on a 22MP back when necessary.
Yes definitely, these are/were the best backs for small apertures. I'd say if someone was doing primarily table top / product work or macro, then the 22mp multishot backs are perhaps the best they will find. With multishot one can stop down several clicks more and still get a crisp detailed shot.
-
There is quite an interesting discussion on the Shneider web site that discusses the technical reasons that forced them to redesign their older film lenses to create the newer digital versions.
Basically, digital was quite clearly showing flaws that were hidden in the noise of the film base, among other things. I believe it discusses the MTFs required for both 35mm andIlarger digital sensors. They used a hypothetical 24MP 35mm full frame sensor shortly before the Canon 1Ds came out (among other sizes) Interesting read.
FWIW, the digital lenses were optimized for a much wider aperture - between 5.6 and 8.0 - than the older lenses. You should be able to find the curves and discussion in a second paper there. (too hard to search now on my phone.)
OK, one older paper:
http://www.schneiderkreuznach.com/knowhow/digfoto_e.htm#1.%20What%20is%20"Picture%20Sharpness"?
More general discussion that precedes the above:
http://www.schneiderkreuznach.com/knowhow/opt_quali_e.htm
Best,
Michael
-
Theory is all very well but why not just do a test at different apertures with YOUR lens and back, and decide for yourself where the acceptable limit is?
-
Theory is all very well but why not just do a test at different apertures with YOUR lens and back, and decide for yourself where the acceptable limit is?
+1
-
Theory is all very well but why not just do a test at different apertures with YOUR lens and back, and decide for yourself where the acceptable limit is?
With no theoretical background, it is very hard to interpret the results. Perhaps a given image looks "sharp" to me when viewed @0.75 meters on a 27" computer monitor using default Lightroom development when shooting my 85mm on a crop-camera @ f/6.3 and 1/200s and "base" ISO using a nice stand and focused on a reasonably shallow object at 10 meters. Would it make sense for me to find a soap-box and proclaim loudly that f/6.3 should be "good enough for anyone, anywhere"?
Are my results "good" and "as expected"? Do my results indicate that there is a flaw in my methods or my equipment? Are my expectations too low? Too high? Can I extrapolate my findings to all possible combinations of camera settings, motifs and light conditions, or do I need to accumulate field-experience with every single one before making up my mind?
I agree that good, relevant, hands-on empiry is always a good thing, and that arm-chair photography seldom results in nice images. I do believe that theoretical analysis and practical experience should be combined for optimal results.
-h
-
Theory is all very well but why not just do a test at different apertures with YOUR lens and back, and decide for yourself where the acceptable limit is?
Hi Graham,
I agree that experience based on empirical observations is useful. However, it may not be immediately clear how to translate that experience to situations that have not been tested yet. That's where the theory can help to predict behavior well in advance, and take appropriate measures in advance.
By using the theory one can also simulate all sorts of scenarios that may be difficult to test, until confronted with them at the decisive moment. I think it's better to be prepared than sorry ...
Cheers,
Bart
-
Hi,
You can just shoot a series of images at different apertures and compare them, for instance using Lightroom´s 'Compare View'. This will demonstrate the deterioration of the image, than you need to decide how much deterioration you are willing to accept.
Best regards
Erik
With no theoretical background, it is very hard to interpret the results. Perhaps a given image looks "sharp" to me when viewed @0.75 meters on a 27" computer monitor using default Lightroom development when shooting my 85mm on a crop-camera @ f/6.3 and 1/200s and "base" ISO using a nice stand and focused on a reasonably shallow object at 10 meters. Would it make sense for me to find a soap-box and proclaim loudly that f/6.3 should be "good enough for anyone, anywhere"?
Are my results "good" and "as expected"? Do my results indicate that there is a flaw in my methods or my equipment? Are my expectations too low? Too high? Can I extrapolate my findings to all possible combinations of camera settings, motifs and light conditions, or do I need to accumulate field-experience with every single one before making up my mind?
I agree that good, relevant, hands-on empiry is always a good thing, and that arm-chair photography seldom results in nice images. I do believe that theoretical analysis and practical experience should be combined for optimal results.
-h
-
Hi,
You can just shoot a series of images at different apertures and compare them, for instance using Lightroom´s 'Compare View'. This will demonstrate the deterioration
Good advice. I like theory now because anything physical hurts! :P
You might also print a portion of each image at the size you will most likely print at. Some things that look terrible on the screen look fine on paper. And visa versa.
If your target size is 24 inch x 30 inch, or whatever, size the image in Photoshop, then print out an 8x10 sample of each of the images - from the center, corners, etc.
Or better, I run cheap dye inks that are about $20 a liter in a calibrated Epson 7600 ($500 printer, $100 for Claria match OCP dye inks.)
I run my "proofs" off at 24"x30" (4x5) to be able to really see the image - my final output is almost always a print. (Fine art stuff now.). It costs less than $.30 for that size print on 170 weight proofing paper. Less than $.05 for the ink.
Great way to learn!
When I settled on default papers about 5 years ago, I ran the same image on 8.5x11 on more than 40 papers. Then at 24x30 on at least 10. Settled on 1 standard proof paper, 1 standard print, and 3 premium (normal, high end, b&w.). Sorry - wandering here! ???
Cheers, Michael
-
Hi,
You can just shoot a series of images at different apertures and compare them, for instance using Lightroom´s 'Compare View'. This will demonstrate the deterioration of the image, than you need to decide how much deterioration you are willing to accept.
Best regards
Erik
And while testing, dial in some varying sharpness settings, and perhaps even take them to photoshop for further sharpening and prints.
As you have mentioned, current sharpening can offset the visual artifacts of diffraction quite nicely in many circumstances. I shoot at f/22 all the time and have printed some large prints that look terrific.
-
And while testing, dial in some varying sharpness settings, and perhaps even take them to photoshop for further sharpening and prints.
As you have mentioned, current sharpening can offset the visual artifacts of diffraction quite nicely in many circumstances. I shoot at f/22 all the time and have printed some large prints that look terrific.
Wayne,
The prints may look terrific, but you have effectively reduced the resolution of your IQ180 to perhaps that of a 6 MP dSLR. Still, you can take some very nice shots with a 6 MP camera.
Nathan Myhrvold (http://www.luminous-landscape.com/essays/Equivalent-Lenses.shtml) puts it succinctly:
"Now, I don’t think anybody would be very excited about turning their EOS 1Ds Mark II, or Canon 5D or other full frame camera into a 2 megapixel camera. It sounds pretty drastic, but that is exactly what you do when you stop down to f/22 – the diffraction limit imposes this condition. If you shoot with a full frame 24 x 36 sensor at f/22 you are throwing away a lot of resolution. There is no getting around this – it is fundamental in the physics of light."
For some subjects, excessive detail may detract from the impact of the image. Can you imagine Monet's water lilies presented with photographic reality.
Regards,
Bill
-
Theory is all very well but why not just do a test at different apertures with YOUR lens and back, and decide for yourself where the acceptable limit is?
That's exactly what I did
-
Hi,
You can just shoot a series of images at different apertures and compare them, for instance using Lightroom´s 'Compare View'. This will demonstrate the deterioration of the image, than you need to decide how much deterioration you are willing to accept.
Best regards
Erik
My point was that there are many variables in-between camera aperture setting, and me being satisfied with the end-results. If I am going to ignore every piece of theoretical understanding I may have, I would strictly have to investigate every single one of them. My life is too short for that, therefore I try to combine theory with practice. One example may be that I assume that the results at f/6.3 will be an intermediate of the results at f/5.6 and f/7.1. Or that the deterioration due to diffraction at small apertures will be similar in two lenses. Or that the images shown on my computer display tells me something about how the image will look on paper.
I do agree that shooting at different apertures to get a qualitative feeling for how it affects images is a good thing.
-h
-
Bill,
Wayne has a good point. Don't forget that we discuss diffraction or parts of the image being out of focus. With deconvolution sharpening, even with unknown PSF, we can improve much on diffraction and less on out of focus. Diffraction is more like a broad peak, with most of the energy at the center. Out of focus is more like an uniform disk with aberrations added depending on construction of the lens.
We had a lengthy discussion here on LuLa forums regarding deconvolution that you, Wayne ;-), me and many other were involved in: http://www.luminous-landscape.com/forum/index.php?topic=45038.0
This is just one sample from that discussion:
http://www.luminous-landscape.com/forum/index.php?topic=45038.msg378744#msg378744
Best regards
Erik
Wayne,
The prints may look terrific, but you have effectively reduced the resolution of your IQ180 to perhaps that of a 6 MP dSLR. Still, you can take some very nice shots with a 6 MP camera.
Nathan Myhrvold (http://www.luminous-landscape.com/essays/Equivalent-Lenses.shtml) puts it succinctly:
"Now, I don’t think anybody would be very excited about turning their EOS 1Ds Mark II, or Canon 5D or other full frame camera into a 2 megapixel camera. It sounds pretty drastic, but that is exactly what you do when you stop down to f/22 – the diffraction limit imposes this condition. If you shoot with a full frame 24 x 36 sensor at f/22 you are throwing away a lot of resolution. There is no getting around this – it is fundamental in the physics of light."
For some subjects, excessive detail may detract from the impact of the image. Can you imagine Monet's water lilies presented with photographic reality.
Regards,
Bill
-
Wayne,
The prints may look terrific, but you have effectively reduced the resolution of your IQ180 to perhaps that of a 6 MP dSLR. Still, you can take some very nice shots with a 6 MP camera.
Nathan Myhrvold (http://www.luminous-landscape.com/essays/Equivalent-Lenses.shtml) puts it succinctly:
"Now, I don’t think anybody would be very excited about turning their EOS 1Ds Mark II, or Canon 5D or other full frame camera into a 2 megapixel camera. It sounds pretty drastic, but that is exactly what you do when you stop down to f/22 – the diffraction limit imposes this condition. If you shoot with a full frame 24 x 36 sensor at f/22 you are throwing away a lot of resolution. There is no getting around this – it is fundamental in the physics of light."
For some subjects, excessive detail may detract from the impact of the image. Can you imagine Monet's water lilies presented with photographic reality.
Regards,
Bill
Well, I've shot with nearly every digital camera since my Kodak DCS 520 and 560, and have many images taken with decent sensors such as the original 1Ds. I'm not a scientist, but I do know none of these printed large (90" pano for example) can come close to what I see with the phase backs, even at f/22. they're just absolute mush at that size and the phase files look terrific ... even up close.
-
My approach to this is a purely practical one. For almost all the Zeiss 'C' lenses which I use, the best MTF is at f8, according to the factory data.
Now, I find I can see almost no loss of IQ at f11, but I do gain a bit of useful DOF over f8. So I tend to use f11 as my working aperture pretty much most of the time. If the light is really bright and I need the DOF I will use f16 just now and then, but you can certainly see a fall-off in the centre at f22 so I avoid it. That's for the normal run of lenses - on my 120mm S-Planar f22 is useable but f32 and f45 are pretty hopeless, really. The oddball lens is the 250mm Sonnar, which has its best MTF wide-open at f5.6. The DOF is pretty skinny, though. That's it - having discovered those things some while ago I no longer worry about it.
John
-
why not just do a test at different apertures with YOUR lens and back
I agree. Any suggestions on a practical way to do that?
-
Hi,
Find a good subject, like a dollar bill and photograph it with different apertures using the best technique you have.
My test, below was done with a 100 mm lens at 3.0 meters. I also varied "defocus" by moving the camera in 3cm steps.
http://echophoto.dnsalias.net/ekr/index.php/photoarticles/49-dof-in-digital-pictures?start=1
Last page shows effect of deconvolution sharpening with Lightroom and Topaz InFocus.
In general, the more you have, the more do you loose.
Best regards
Erik
I agree. Any suggestions on a practical way to do that?
-
I agree. Any suggestions on a practical way to do that?
take some pictures at different apertures and print them large. Not difficult at all really.
-
Wayne,
The prints may look terrific, but you have effectively reduced the resolution of your IQ180 to perhaps that of a 6 MP dSLR. Still, you can take some very nice shots with a 6 MP camera.
Nathan Myhrvold (http://www.luminous-landscape.com/essays/Equivalent-Lenses.shtml) puts it succinctly:
"Now, I don’t think anybody would be very excited about turning their EOS 1Ds Mark II, or Canon 5D or other full frame camera into a 2 megapixel camera. It sounds pretty drastic, but that is exactly what you do when you stop down to f/22 – the diffraction limit imposes this condition. If you shoot with a full frame 24 x 36 sensor at f/22 you are throwing away a lot of resolution. There is no getting around this – it is fundamental in the physics of light."
For some subjects, excessive detail may detract from the impact of the image. Can you imagine Monet's water lilies presented with photographic reality.
Regards,
Bill
That is the silliest thing I have read in a long time. Like Wayne, I have shoot MFD at f/22 and that claim is so wrong on so many levels. So you are saying we are seeing great images, but actually they are really not good, we are just not seeing it? So if you cannot see the loss of quality, then why the stress on quality that cannot be observed?
I have a feeling the folks stressing absolute technical perfection really do not shoot and print their work. Only the working photographers seem not to be bothered by small working apertures. And if resolving power is really that important to people, I would like to note that my backgrounds and foreground have more resolution at small apertures than large ones--that is what DoF does...
This is one of the funniest conversation I have seen on LL in a while.
-
And above all do relevant tests. Pictures of dollar bills or brick walls aren't going to be of much help to a landscape photographer.
But forgers and masons may benefit...
-
Hi,
You need to have fine detail down to pixel level, and it needs to be reproducible. A dollar bill shot at 300 cm with a 100 mm lens will have enough detail.
BR
Erik
And above all do relevant tests. Pictures of dollar bills or brick walls aren't going to be of much help to a landscape photographer.
-
That is the silliest thing I have read in a long time. Like Wayne, I have shoot MFD at f/22 and that claim is so wrong on so many levels. So you are saying we are seeing great images, but actually they are really not good, we are just not seeing it? So if you cannot see the loss of quality, then why the stress on quality that cannot be observed?
I have a feeling the folks stressing absolute technical perfection really do not shoot and print their work. Only the working photographers seem not to be bothered by small working apertures. And if resolving power is really that important to people, I would like to note that my backgrounds and foreground have more resolution at small apertures than large ones--that is what DoF does...
This is one of the funniest conversation I have seen on LL in a while.
It may appear funny to you, but it is based on the laws of physics. Nathan Myhrvold (http://en.wikipedia.org/wiki/Nathan_Myhrvold) is no crank, but a genius with a PhD in physics form Princeton University. He is not only an accomplished scientist but an accomplished photographer and gourmet cook--a true renaissance man. That is why I chose to quote him. Of course, one can stop down further with a MFDB than a full frame dSLR, but a few calculations are instructive.
The pixel pitch of the IQ-180 is 5.17 μ, the resolution is 7816 x 10380 pixels, and the sensor dimension is 40.4 x 53.7 mm. The Airy disc diameter for light at 530 nm is 28.5 μ. The sensor becomes diffraction limited when the Airy disc diameter is 1.4 to 2 times the pixel pitch. Diffraction limited resolution of a lens at f/22 is 70 lp/mm at the Rayleigh limit (about 9% MTF) and 33 lp/mm at 50% MTF (figures from Roger Clark). Rayleigh resolution in an astronomical telescope can separate binary stars in this high contrast setting, but is not very useful for terrestrial photography, where an MTF of 50% is more reasonable.
One can use these figures to calculate the resolution of the IQ-180. The Rayleigh resolution at f/22 is 2840 x 3375 lp; since it takes 2 pixels to resolve 1 lp, that corresponds to 2702 x 3592 pixels or 43 MP. For a MTF of 50% the figures are 2702 x 3592 pixels or 9.7 MP.. For the Canon 1DsMIII that Dr. Myhrvold quoted in his post, the Rayleigh resolution at f/22 is 1687 x 2351 lp 17 MP. The MTF 50 resolution is 803 x 1204 lp or 3.9 MP, somewhat better than his prediction of 2MP.
One needs to correlate these figures to the perception of sharpness. According to David Pogue's (http://www.nytimes.com/2007/02/08/technology/08pogue.html?pagewanted=all) tests, megapixels are not as important as often believed, so it is reasonable to get good results at the 9.7 MP you would get with the IQ-180 at f/22.
Regards,
Bill
-
Hi,
I don't agree at all. Neither a lens or a sensor will care about if you photograph a Dollar bill, or details on a redfern or twigs on a dead tree 200 m away.
But I want to have my test reproducible.
Best regards
Erik
Great if you're photographing dollar bills, less so if your photographing landscapes.
The OP wants to know what impact diffraction will have on his landscapes. Photographing dollar bills will tell him sod all.
-
My approach to this is a purely practical one. For almost all the Zeiss 'C' lenses which I use, the best MTF is at f8, according to the factory data.
Now, I find I can see almost no loss of IQ at f11, but I do gain a bit of useful DOF over f8. So I tend to use f11 as my working aperture pretty much most of the time. If the light is really bright and I need the DOF I will use f16 just now and then, but you can certainly see a fall-off in the centre at f22 so I avoid it. That's for the normal run of lenses - on my 120mm S-Planar f22 is useable but f32 and f45 are pretty hopeless, really. The oddball lens is the 250mm Sonnar, which has its best MTF wide-open at f5.6. The DOF is pretty skinny, though. That's it - having discovered those things some while ago I no longer worry about it.
John
Most sensible take on this matter yet!
Rob C
-
Really, you know, there is no big deal about any of this stuff.
With almost all the stock MF lenses, the widest aperture will be crap - it is just there for focusing - and the smallest aperture will also be crap. (We are not talking Leica here, and amazing wide-open performance, neither are we discussing LF lenses and the capability to use f64). So if the biggest hole and the smallest hole should be avoided, that leaves the ones in between. It's pretty safe to assume that the hole in the middle of the range (around f8) will be the best to use. So use f8, but if you need a bit more DOF then f11 will be just fine too. On the other hand, if you would like a softer background for portaits, open her up to f5.6.
That's pretty much it, really. What else are you going to do?
John
-
...and I and my clients would rather my images were reproducible.
Photographing dollar bills will tell me how diffraction impacts my photographs of dollar bills and frankly I could care less.
The extreme "hands-on" take would be that testing camera settings when photographing pine trees say nothing about the results you will get when photographing birch trees. I dont think anyone are that fundamentalist.
The extreme "hands-off" take would be that everything worth knowing about photography can be studied by reading a book on physics, optics or the golden ratio. I dont think anyone are that fundamentalist.
Most of us try to patch together a coherent understanding/method based on _both_ personal experience and what we may read or calculate.
-h
"Experience keeps a dear school, but fools will learn in no other." Benjamin Franklin
-
That is the silliest thing I have read in a long time...
I think that it is actually a slippery slope to compare diffraction artifacts to pixel density artifacts. I would be hesitant to claim that this or that f-number corresponds to this and that number of megapixels, or that there is a hard limit beyond which the benefit of adding more megapixels are exactly zero. Nature and physics seems to generally "dislike" hard limits or rectangular functions.
But if a 180MP image contains no more details than a 6MP camera could have, so what if it looks good? I believe that people have had exhibitions and national geographic pages using 6MP cameras, the limiting factor for most of us is between the ears, or luck or travelling budget, or ...
-h
-
...placing too much importance on tests, reading books on physics, optics and particularly the golden ratio.
Or walking around with misconceptions stemming from doing a lot of hands-on testing but not understanding the physics behind it, or the issues with non-blind subjective testing.
People who reject science are particularly easily fooled by people who sells snake-oil.
-h
-
Hy,
Physics and chemistry are obviously important, without that our planet would not support 7 billion (US) people, nor would we have cameras, computers and forums like this. I don't know if Neanderthal art was very profitable.
It's OK to be an artist but do have some respect for the folks that develop the tools you have. No lens, camera or sensor could have been created ever without a massive amount of science. OK, there have been a few lenses crafted by hand, without good knowledge of optics. But since the Petzval lens (daddy of all Tessars) all lenses are based on science.
This is a good read: http://www.lensrentals.com/blog/2011/08/lens-geneology-part-1
Best regards
Erik
Or walking around with misconceptions stemming from doing a lot of hands-on testing but not understanding the physics behind it, or the issues with non-blind subjective testing.
People who reject science are particularly easily fooled by people who sells snake-oil.
-h
-
Really, you know, there is no big deal about any of this stuff.
With almost all the stock MF lenses, the widest aperture will be crap - it is just there for focusing - and the smallest aperture will also be crap. (We are not talking Leica here, and amazing wide-open performance, neither are we discussing LF lenses and the capability to use f64). So if the biggest hole and the smallest hole should be avoided, that leaves the ones in between. It's pretty safe to assume that the hole in the middle of the range (around f8) will be the best to use. So use f8, but if you need a bit more DOF then f11 will be just fine too. On the other hand, if you would like a softer background for portaits, open her up to f5.6.
That's pretty much it, really. What else are you going to do?
John
John, you are right, there is no big deal about this stuff. I would rather have a perfect image rather than perfect resolving power.
The documentary work I do in low light handheld with MFD, the widest aperture is great. I find that some of the limitation like vignetting can add to the image and some of the "bad" aberrations are still better than handshake. Many of my medium-format lenses are actually very nice wide open. My smaller format lenses are far worse--although their max aperture is also larger.
And since the science has been brought up, the funny thing is that it does not say shoot only at f/11. The aperture range is actually determined for the format and the lenses ability. Except for a few lenses where maximum aperture is really just for focusing, the optical designers are making the apertures within the useful range.
-
My test, below was done with a 100 mm lens at 3.0 meters. I also varied "defocus" by moving the camera in 3cm steps.
http://echophoto.dnsalias.net/ekr/index.php/photoarticles/49-dof-in-digital-pictures?start=1
Last page shows effect of deconvolution sharpening with Lightroom and Topaz InFocus.
In general, the more you have, the more do you loose.
Erik,
An excellent demonstration! I have tried to use authoritative references to back up my calculations, but nothing beats actual data in a well done test. So far in this thread you are the only one to present actual images. The subjective impression of the IQ of an image is the most important factor in the end, but it is good to back up the impression with quantitative data such as MTF. You could add Bart van der Wolf's resolution chart to the setup to check resolution and aliasing. Thus far, none of our MFDB colleagues have taken up the challenge.
The results of the Topaz InFocus deconvolution restoration look good and have less artifact than the SmartSharpen method. I did buy InFocus when Bart recommended it and it was on sale, but I haven't used it much. It takes a bit of work to get the parameters right, and I will have to crank it up and work with it a bit more.
Regards,
Bill
-
Hmm, I don't know you from Adam, but I'll hazard a guess that you make your living as a scientist rather than an image maker?
Some of us have both Art and Science backgrounds. Me? I test my stuff so I know what I'm going to get and what I can and can't do with it. Saves a lot of time. And I have degrees in both fine art and physics and worked in both professions. What's wrong with that? The people who don't have the time/energy/experience to test all benefit from those that do. You must have some interest in this otherwise you wouldn't read...
-
Pentax 645D, D FA 55mm, ISO 200, f/22. The scene. 100% crop (and I am unsure if it is from the plane of focus). A plot of a small section of the image which is also inset in the plot. The image is resolving features 12 micros or a less on the image plane.
-
I don't agree at all. Neither a lens or a sensor will care about if you photograph a Dollar bill, or details on a redfern or twigs on a dead tree 200 m away.
But I want to have my test reproducible.
Macro vs. Infinity lens performance are not the same. A lens may well be much sharper at infinity than it's closest focus distance (or vice versa). The sharper the lens the less diffraction can be tolerated if the goal is to get the sharpest image the system can produce.
Also in macro photography the effective aperture regarding diffraction are not the same as the marked aperture of the lens.
e.g. here is an extreme macro example where diffraction kicks in at a lens marking of f/2.8 and is severe by f/5.6.
http://www.captureintegration.com/2009/08/25/extreme-macro/
Yet another vote for real-world testing in situations as close to your actual intended use as possible.
Doug Peterson (e-mail Me) (doug@captureintegration.com)
__________________
Head of Technical Services, Capture Integration
Phase One Partner of the Year
Leaf, Leica, Cambo, Arca Swiss, Canon, Apple, Profoto, Broncolor, Eizo & More
National: 877.217.9870 | Cell: 740.707.2183
Newsletter (http://"http://www.captureintegration.com/our-company/newsletters/") | RSS Feed (http://"http://www.captureintegration.com/2008/08/11/rss-feeds/")
Buy Capture One 6 at 10% off (http://"http://www.captureintegration.com/phase-one/buy-capture-one/")
-
When I shoot Landscapes with my Mamiya RB67, I want everything in focus from front to rear. So I work with the Hyperfocal distance, figure out what aperture I need, and then stop down one more stop for good measure. Is this right? Is this wrong?
-
Pentax 645D, D FA 55mm, ISO 200, f/22. The scene. 100% crop (and I am unsure if it is from the plane of focus). A plot of a small section of the image which is also inset in the plot. The image is resolving features 12 micros or a less on the image plane.
Interesting - to me that image is too soft for my purposes. I wonder if you have tried working this image using the detail slider in Lightroom or ACR or one of the other deconvolution programs?
-
Hi,
The recommendation was to shoot it at 3 meter distance with a 100 mm lens, so it's hardly macro. The reason I suggested the Dollar bill is that it has a lot of fine detail and is quite challenging in many ways.
Of course, my posting was a one liner, but it refers to this posting: http://www.luminous-landscape.com/forum/index.php?topic=59978.msg484272#msg484272
Your point is very valid. Thirty times focal length is still a bit on the short side for testing lenses.
Best regards
Erik
Macro vs. Infinity lens performance are not the same. A lens may well be much sharper at infinity than it's closest focus distance (or vice versa). The sharper the lens the less diffraction can be tolerated if the goal is to get the sharpest image the system can produce.
Also in macro photography the effective aperture regarding diffraction are not the same as the marked aperture of the lens.
e.g. here is an extreme macro example where diffraction kicks in at a lens marking of f/2.8 and is severe by f/5.6.
http://www.captureintegration.com/2009/08/25/extreme-macro/
Yet another vote for real-world testing in situations as close to your actual intended use as possible.
Doug Peterson (e-mail Me) (doug@captureintegration.com)
__________________
Head of Technical Services, Capture Integration
Phase One Partner of the Year
Leaf, Leica, Cambo, Arca Swiss, Canon, Apple, Profoto, Broncolor, Eizo & More
National: 877.217.9870 | Cell: 740.707.2183
Newsletter (http://"http://www.captureintegration.com/our-company/newsletters/") | RSS Feed (http://"http://www.captureintegration.com/2008/08/11/rss-feeds/")
Buy Capture One 6 at 10% off (http://"http://www.captureintegration.com/phase-one/buy-capture-one/")
-
Alan,
What is right or wrong can only you decide, if you are happy, you are happy.
Using the hyperfocal distance is not without problems. The question is which hyperfocal distance you use?! Hyperfocal distance and DoF essentially are about the acceptable amount of unsharpness. So hyperfocal distance calculation is essentially based on an acceptable amount of unsharpeness criterion, normally called CoC (Circle of Confusion). When lens is at hyperfocal distance, everything from infinity to half hyperfocal distance will be at least as sharp as the criterion.
The problem is that your eyes may be more demanding than the criterion.
Here is a very good article about adapting hyperfocal focusing for high resolution work: http://optechsdigital.com/Alpa_and_Hyperfocal.html
When I shot 67 on film I normally used the f/8 markers for depth of field when shooting at f/16.
Best regards
Erik
When I shoot Landscapes with my Mamiya RB67, I want everything in focus from front to rear. So I work with the Hyperfocal distance, figure out what aperture I need, and then stop down one more stop for good measure. Is this right? Is this wrong?
-
When I shoot Landscapes with my Mamiya RB67, I want everything in focus from front to rear. So I work with the Hyperfocal distance, figure out what aperture I need, and then stop down one more stop for good measure. Is this right? Is this wrong?
If it works, it's right.
What else are you reasonably going to do?
Rob C
-
Hmm, I don't know you from Adam, but I'll hazard a guess that you make your living as a scientist rather than an image maker?
Neither, actually. In what way does my occupation make my arguments more or less valid?
Both "hands-on" and "hands-off" approaches can be done in stupid ways, and easily ridiculed. Do you want to ridicule one but not the other?
-h
-
Hi,
Doug's posting referred to oneliner. The original posting clearly indicated that the Dollar bill was shot at three meters with a 100 mm lens.
Best regards
Erik
Eric, the above is the opening line of my original post to this thread.
Doug's post highlights the importance of relevant testing.
I've nothing against testing as such, hell, I do my own, but I make damn sure it is relevant to my own work!
-
The recommendation was to shoot it at 3 meter distance with a 100 mm lens, so it's hardly macro. The reason I suggested the Dollar bill is that it has a lot of fine detail and is quite challenging in many ways.
Of course, my posting was a one liner, but it refers to this posting: http://www.luminous-landscape.com/forum/index.php?topic=59978.msg484272#msg484272
Your point is very valid. Thirty times focal length is still a bit on the short side for testing lenses.
My apologies for missing that detail.
I guess my general issue with "lab" tests is that they only tell you the answer to questions you know to ask. The "known unknowns" as Cheney would say. Often the "unknown unknowns" are often the ones that bite you in the butt.
At Capture Integration we get a lot of odd-ball testing requests from photographers who are going to use the camera exclusively for landscape. Generally I try to convince them going out and shooting a landscape is the best way to determine how a camera will perform when shooting a landscape :-). Ideally with the different equipment you are considering and alongside someone who knows where the pitfalls/misuse/mistakes/limits of each system are (e.g. don't stop down that canon lens to f/22).
-
Interesting - to me that image is too soft for my purposes. I wonder if you have tried working this image using the detail slider in Lightroom or ACR or one of the other deconvolution programs?
You mean soft at 100%? 100% is not a real world viewing condition. So you have made a 3 foot print? Besides, I this is a low contrast scene I shoot as a test, the test I mention in the thread, to see what the result would be at f/22 and a bunch of other apertures to show the effect of diffraction. Result: diffraction is overstated.
-
It may appear funny to you, but it is based on the laws of physics. Nathan Myhrvold (http://en.wikipedia.org/wiki/Nathan_Myhrvold) is no crank, but a genius with a PhD in physics form Princeton University. He is not only an accomplished scientist but an accomplished photographer and gourmet cook--a true renaissance man. That is why I chose to quote him. Of course, one can stop down further with a MFDB than a full frame dSLR, but a few calculations are instructive.
The pixel pitch of the IQ-180 is 5.17 μ, the resolution is 7816 x 10380 pixels, and the sensor dimension is 40.4 x 53.7 mm. The Airy disc diameter for light at 530 nm is 28.5 μ. The sensor becomes diffraction limited when the Airy disc diameter is 1.4 to 2 times the pixel pitch. Diffraction limited resolution of a lens at f/22 is 70 lp/mm at the Rayleigh limit (about 9% MTF) and 33 lp/mm at 50% MTF (figures from Roger Clark). Rayleigh resolution in an astronomical telescope can separate binary stars in this high contrast setting, but is not very useful for terrestrial photography, where an MTF of 50% is more reasonable.
One can use these figures to calculate the resolution of the IQ-180. The Rayleigh resolution at f/22 is 2840 x 3375 lp; since it takes 2 pixels to resolve 1 lp, that corresponds to 2702 x 3592 pixels or 43 MP. For a MTF of 50% the figures are 2702 x 3592 pixels or 9.7 MP.. For the Canon 1DsMIII that Dr. Myhrvold quoted in his post, the Rayleigh resolution at f/22 is 1687 x 2351 lp 17 MP. The MTF 50 resolution is 803 x 1204 lp or 3.9 MP, somewhat better than his prediction of 2MP.
One needs to correlate these figures to the perception of sharpness. According to David Pogue's (http://www.nytimes.com/2007/02/08/technology/08pogue.html?pagewanted=all) tests, megapixels are not as important as often believed, so it is reasonable to get good results at the 9.7 MP you would get with the IQ-180 at f/22.
Regards,
Bill
Nathan may be a genius, but optics and imaging are not his fields. It would not be the first time someone making statement about something outside their field have made errors.
But anyway. Perhaps you can add to the discussion by showing examples of this megapixel equivalency hypothesis. Numbers are not results, especially if the underlying model has errors.
-
Nathan may be a genius, but optics and imaging are not his fields. It would not be the first time someone making statement about something outside their field have made errors.
But anyway. Perhaps you can add to the discussion by showing examples of this megapixel equivalency hypothesis. Numbers are not results, especially if the underlying model has errors.
Perhaps you should read his article and then point out where he made errors. Your own results seem to verify the hypothesis (vida infra).
Pentax 645D, D FA 55mm, ISO 200, f/22. The scene. 100% crop (and I am unsure if it is from the plane of focus). A plot of a small section of the image which is also inset in the plot. The image is resolving features 12 micros or a less on the image plane.
Your test shot does show light and dark bands likely representing resolved areas of the bark or other tree structures. Aliasing and demosaicing artifacts as well as excessive sharpening can produce false detail, but diffraction at f/22 should have eliminated most aliasing and I assume we are seeing true resolution.
Resolution is measured in lp/mm or cycles/mm--that represents peak to peak. Measuring peak to peak, I get 25 μ which would be 40 lp/mm. You seem to have measured peak to trough. From my previous calculations, the resolution of a diffraction limited lens at f/22 is 70 lp/mm at the Rayleigh limit and 33 lp/mm at 50% MTF. Your resolution is somewhere between these two values and is consistent with theory. The Nyquist frequency of your sensor is 84 lp/mm and you are achieving considerably less in your shot. A statement regarding resolution is incomplete unless the contrast achieved for that resolution is stated. Resolution at 50% contrast (MTF 50) correlates best with perceived sharpness.
BTW, what software did you use for your measurement? Photoshop extended? Anyway, nice work.
Regards,
Bill
-
The software is ImageJ, free for anyone who wants to use it. Google and name and you should find it easily
This is not a test target and so you need to dig a little further. The plot of the image is far more complex, which is why just using resolving power to make some sort of grid to define "equivalent MP" does not really work--you are not going to find convenient line pairs and line pairs in and of themselves do not describe an image (it is just a quantification of system that allows comparison, but does not have enough detail to express image formation or perception). Since the 645D has a pixel pitch of 6um, the image does actually provide clear separation of detail between 1-2 pixels or 6um-12um. You can see that in the image and if you analyze the plot, it is there as well--as in other areas of the image. You will also see it is defining edges very well. Just taking peaks on the plot will not work.
This is also a low contrast target which would lower the resolving power of the system. Which suggests that the Rayleigh criterion/limit, which you use in your post, is far too crude to model real systems. Actually, the Rayleigh limit is really just a useful learning tool, but it is out of date.
-
I've tested all of my lenses with my MFD back at distances that are relevant to my subject matter and my output. I recommend that anyone who is concerned about diffraction does the same.
Well spoken. 'Tis the only way to do it. And once you have done it, internalise the results and stop worrying about it.
I did some testing recently off a tripod using a subject with mixed distances from about 30 feet to infinity, to check a couple of Zeiss 60mm lenses which I was servicing. Folks, don't you worry too much about f22. In this case it was f4 which was scary - I was horrified how the image fell apart when those lenses were wide-open. The corners with these old Distagons are not that great at the best of times, but . . .
John
-
I've tested all of my lenses with my MFD back at distances that are relevant to my subject matter and my output.
That's just crazy talk!
-
The software is ImageJ, free for anyone who wants to use it. Google and name and you should find it easily
Thanks. I do use ImageJ extensively, but did not recognize the output in your graph.
This is not a test target and so you need to dig a little further. The plot of the image is far more complex, which is why just using resolving power to make some sort of grid to define "equivalent MP" does not really work--you are not going to find convenient line pairs and line pairs in and of themselves do not describe an image (it is just a quantification of system that allows comparison, but does not have enough detail to express image formation or perception). Since the 645D has a pixel pitch of 6um, the image does actually provide clear separation of detail between 1-2 pixels or 6um-12um. You can see that in the image and if you analyze the plot, it is there as well--as in other areas of the image. You will also see it is defining edges very well. Just taking peaks on the plot will not work.
Your test method seems quite convoluted and I doubt that it is validated. Why don't you use a validated method? Some choices are:
ISO 12233 Chart (http://www.graphics.cornell.edu/~westin/misc/res-chart.html): allows visual determination of resolution by inspection and the slanted edge can be used with Imatest and other software. However, one must calibrate for the magnification factor if using the visual method. With the slanted edge target, the magnification is not critical.
Koren Resolution Chart (http://www.normankoren.com/Tutorials/MTF5.html): Can be used visually and with ImageJ, where one measures peak to peak. The magnification factor must be accounted for.
Van Der Wolf modified Siemens star (http://www.openphotographyforums.com/forums/showthread.php?t=13217): Target can be used at any distance and the slanted edge can be used with Imatest.
This is also a low contrast target which would lower the resolving power of the system. Which suggests that the Rayleigh criterion/limit, which you use in your post, is far too crude to model real systems. Actually, the Rayleigh limit is really just a useful learning tool, but it is out of date.
The Rayleigh criterion is hardly out of date and, along with the Dawes criterion, is a fundamental concept in optics. The old USAF charts measure resolution near the Rayleigh limit and this is the result you get when using this method. As I stated, a MTF of 9% is too low to be used in terrestrial photography. The Dawes limit is where MTF is zero. A MTF of 50% corresponds best with perceived sharpness. What MTF did you get with your method. Resolution without reference to contrast is not that helpful.
Regards,
Bill
-
Depends from scene to scene. Lighting to lighting. Lens to lens.
I see it a more on my P65 with V Series lenses more so than the H Series which I don't really see it.
It's always cleaned up with some sharpening though and hasn't ever really been a problem.
It starts to appear at f16.
It's never been an issue when I used the V with a P45. Probably more exaggerated with the IQ180.
-
Your test method seems quite convoluted and I doubt that it is validated. Why don't you use a validated method? Regards,
Bill
They are my tests for myself--you had challenged us to show results. Also, test charts do not provide the information I need. These test give very useful data. And the final part of the test is always real output.
Bill, lets see some of your tests on diffraction.
-
They are my tests for myself--you had challenged us to show results. Also, test charts do not provide the information I need. These test give very useful data. And the final part of the test is always real output.
Bill, lets see some of your tests on diffraction.
Certainly. I used Bart's target with my Nikon D3 with the 60 mm f/2.8 AFS Micro-Nikkor. I rendered the images with ACR 6.5 with the default settings, which include a slight amount of capture sharpening.
Here is the result for f4.0 with Bart's method and with Imatest. The blur circle is 98 pixlels giving a resolution of 55 cy/mm or 0.47 cy/pixel, near the Nyquist limit of 0.5 cy/px. Bart's chart shows aliasing particularly well, but I think that the resolution results are for a relatively low MTF, perhaps near Rayleigh. Maybe Bart can comment if he is following this thread. Bart's chart also allows visual inspection of the test images; the low contrast at f/32 is evident at a quick glance, more so than the reduced resolution. The Imatest results are self explanatory. The resolution per picture height is not valid since I used cropped images, but the other figures are valid. Imatest allows calculation of the MTF at various resolutions. The MTF at 50% contrast is considered most important for image quality.The effects of diffraction would be more marked with the D3x or other higher resolution camera. At larger apertures, the lens out resolves the sensor with the D3.
(http://bjanes.smugmug.com/Photography/DiffractionResolution/i-KMW989q/0/O/f4.png)
(http://bjanes.smugmug.com/Photography/DiffractionResolution/i-j2zcdQW/0/O/0002YR5601cpp.png)
And for f/32. The blur circle is 109.5 mm, corresponding to a resolution of 41 cy/mm or 0.42 cy/pixel. Note that the aliasing is completely eliminated.
Also note that Imatest shows the diffraction limit for f/32, which is closely approximated by the curve.
(http://bjanes.smugmug.com/Photography/DiffractionResolution/i-LJs9hc3/0/O/f32.png)
(http://bjanes.smugmug.com/Photography/DiffractionResolution/i-47jcrNL/0/O/0008YR5401cpp.png)
It would be most interesting if you would shoot Bart's chart at optimal aperture and the minimum aperture and post the results.
Regards,
Bill
-
... f/22 ... The image is resolving features 12 micros or a less on the image plane.
This fits a pattern I have noticed in various field observations: the diffraction effects coming in on a length scale in microns of about half the f-stop. Fitting roughly with the wavelength of visible light being about half a micron. This seems to fit fairly well with the theoretical figures based on 50% MTF, or am I remembering that incorrectly?
Of course the effect is incremental, not a brick wall, and some more fastidious observers draw the line one stop earlier (like f/8 with 5.6 micron pixel spacing, and some willing to go about a stop further, especially with high pixel count MF images expected to be printed at high PPI.
-
Certainly. I used Bart's target with my Nikon D3 with the 60 mm f/2.8 AFS Micro-Nikkor. I rendered the images with ACR 6.5 with the default settings, which include a slight amount of capture sharpening.
Here is the result for f4.0 with Bart's method and with Imatest. The blur circle is 98 pixlels giving a resolution of 55 cy/mm or 0.47 cy/pixel, near the Nyquist limit of 0.5 cy/px. Bart's chart shows aliasing particularly well, but I think that the resolution results are for a relatively low MTF, perhaps near Rayleigh. Maybe Bart can comment if he is following this thread.
Hi Bill,
Yes, the limiting resolution (which approaches Nyquist) is at a relatively low MTF. It is closer to the human visual resolution limit, at approx. 10% response on the MTF, than to the 50% MTF metric (which gives a more overall perception of sharpness and contrast, and is closer to the maximum constrast sensitivity of the human eye). The Rayleigh criterion is more relevant for separating 2 point light sources, such as stars, and works out to something like a 20% contrast at that limit for a diffraction pattern.
To place that in perspective, a subject with 10:1 contrast at the 10% MTF point will exhibit a 1% resulting contrast, and that's at the perception threshold of detection (without sharpening) for human vision. The unaltered test chart, when printed on glossy paper will have something like a 100:1 contrast, and therefore plenty to produce aliasing even at very low MTF responses.
And for f/32. The blur circle is 109.5 mm, corresponding to a resolution of 41 cy/mm or 0.42 cy/pixel. Note that the aliasing is completely eliminated. Also note that Imatest shows the diffraction limit for f/32, which is closely approximated by the curve.
Sharpening will raise the MTF curve somewhat above the diffraction limited MTF, as long as there is sufficient signal left for sharpening. Low subject contrast combined with diffraction will result in loss of resolution, unrestorable even for deconvolution sharpening.
Cheers,
Bart
-
Yes, the limiting resolution (which approaches Nyquist) is at a relatively low MTF. It is closer to the human visual resolution limit, at approx. 10% response on the MTF, than to the 50% MTF metric (which gives a more overall perception of sharpness and contrast, and is closer to the maximum constrast sensitivity of the human eye). The Rayleigh criterion is more relevant for separating 2 point light sources, such as stars, and works out to something like a 20% contrast at that limit for a diffraction pattern.
To place that in perspective, a subject with 10:1 contrast at the 10% MTF point will exhibit a 1% resulting contrast, and that's at the perception threshold of detection (without sharpening) for human vision. The unaltered test chart, when printed on glossy paper will have something like a 100:1 contrast, and therefore plenty to produce aliasing even at very low MTF responses.
I repeated my experiment at f/22, but omitted capture sharpening. The resolution calculated from your target was 57 lp/mm and the MTF 10 calculated by Imatest was 59 lp/mm, confirming the approx. 10% figure you quoted. It is interesting to note that the MTF10 at f/22 is at Nyquist, which would imply that deconvolution sharpening could restore some of the lost detail imposed by shooting at this small aperture.
(http://bjanes.smugmug.com/Photography/DiffractionResolution/i-j7djc2n/0/O/0007noShYR5501cpp.png)
(http://bjanes.smugmug.com/Photography/DiffractionResolution/i-q97qCXn/0/O/f22NoShB.png)
I am somewhat confused by your statement that MTF at Rayleigh is around 20%, since I had always thought that it was aroung 10% as stated by Roger Clark (http://www.clarkvision.com/articles/scandetail/index.html#diffraction). However, on checking in Wikipedia (http://en.wikipedia.org/wiki/Angular_resolution), they do mention a figure of 20%. How does one resolve these conflicting values?
Regards,
Bill
-
I repeated my experiment at f/22, but omitted capture sharpening. The resolution calculated from your target was 57 lp/mm and the MTF 10 calculated by Imatest was 59 lp/mm, confirming the approx. 10% figure you quoted. It is interesting to note that the MTF10 at f/22 is at Nyquist, which would imply that deconvolution sharpening could restore some of the lost detail imposed by shooting at this small aperture.
Hi Bill,
That's correct. Unfortunately we'll not be able to recover all detail all the way up to Nyquist in your example because the actual image contrast gets too low (in case of an adequate Low-pass filter). Image contrast multiplied by MTF response approaches zero. This is also where a low glare lens helps to maintain contrast in the microdetail. I expect the image contrast of a MF system without low-pass filter to reach Nyquist without much attenuation other than that from an area sample (and produce aliasing).
I am somewhat confused by your statement that MTF at Rayleigh is around 20%, since I had always thought that it was aroung 10% as stated by Roger Clark (http://www.clarkvision.com/articles/scandetail/index.html#diffraction). However, on checking in Wikipedia (http://en.wikipedia.org/wiki/Angular_resolution), they do mention a figure of 20%. How does one resolve these conflicting values?
Frankly, I traditionally assumed Roger's 9% number to be correct, but when checking Wikipedia found a higher contrast being quoted. To verify, I made a simulation in ImageJ of two (f/16, 564 nm) diffraction patterns I already had, offset by the radius to the first zero, and I find something like a bit over 26% contrast (depending on sensel size (!) and alignment with the sensel grid). This simulation is without the effect of a low-pas filter.
Cheers,
Bart
P.S. The sensel's aperture size (or microlens) will quickly reduce the above oversampled (1 micron) rendering of the neighboring diffraction patterns (Rayleigh criterion) into a complete blur with zero contrast. I'm preparing a demonstration of it, but it takes the computer a while to calculate the base images.
-
When I shot 67 on film I normally used the f/8 markers for depth of field when shooting at f/16.
Eric: Basically you're doing the same thing I described I do except you're using a two stop margin rather than my one stop. I checked that link you provided on Hyper-focal. It appears a method for calculating for digital cameras and lenses rather than for anything for my film RB67. But thanks for the link. It'll be useful with my DSLR (as soon as I get one!) ;)
-
Alain,
A few observations:
1) When I started shooting 67 I discovered that I had issues with sharpness, and realized that those were caused by the "optimistic" DoF scales. I didn't buy MF to get same sharpness as from 135, but to have better images.
2) I sort of revisited the issue of stopping down. The philosophy I have right now is essentially to focus on what is important and stop down as much as needed. But I try to avoid going to f/22. I found out that of my around 58000 images about 300 are shot at f/22, many of those are tests. But after reconsidering my images I'd say that stopping down to f/22 is more OK than I expected. It's still a waste of resolution, but resolution may less important than what I (we) may believe.
3) Sharpening matters a lot. Cannot replace the real stuff but matters a lot.
I hope you are going to enjoy digital once you have it. For many of us digital is a revelation but for some it's just a steep learning curve.
Best regards
Erik
Eric: Basically you're doing the same thing I described I do except you're using a two stop margin rather than my one stop. I checked that link you provided on Hyper-focal. It appears a method for calculating for digital cameras and lenses rather than for anything for my film RB67. But thanks for the link. It'll be useful with my DSLR (as soon as I get one!) ;)