Loading…
www.cambridgeincolour.com
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Diffraction...quick test at f 6.7 ..f8...f19/ subject: Great Blue Heron
- Thread starter ANAYV
- Start date
briantilley
Veteran Member
From a quick scan, I can't see where that article answers BasilG's question.
I have seen you weigh in on technical questions with answers that appear to make sense but are not quite right several times. This is one of these cases.
The suggestion might originate from an erroneous interpretation of the point spread function.
Diffracted rays contribute to all areas of the image.
I dont know.
I guess DXO is adding that correction, or is it only lens abberation and distortion corrections?
Me either. But as I was shooting this very cooperative sbject, I adjusted Aperture , thinking I will see what is sharpest from this lens. Came home and looking at the images, I see similar details and f/19 still had feather details, but my bad test had ISO raised , adding noise and NR into the equasion.
Thats are goal ehen we shoot.Getting the fastest exposure with the least amount of noise is stressful enough.
Some noise not an issue as it can be removed in post, but I am still learning this balance between noise and detail retention.
ANAYV
Z 50 has that option too, but as you said only for jpeg, and RAW gives me much better results than jpegs with the cameras I shoot. So that setting not applies to my workflow.
I think that setting adds some sharpening, not really offsetting the effects of diffraction, but I see a small gradual degrading of details when stopping down, so diffraction most likely not an issue for me and my gear.
f/19 has enough details for me, but will never shoot at that small aperture anyway.
But another better controlled test will show me more of the limits
ANAYV
Thats what I seem to gather..so far.
Indeed.
Yes, diffraction applies to any lens that can be stopped down enough, but some folk buy faster glass to avoid diffraction, say at f/5.6...so they use a f/4 lens, wishing it was a f/2.8
Both will have more resolution than my 20MP Z 50Despite "diffraction fears" - you should get more image resolution at f16 from 45 MP than from 24 MP
No fears, just wondering what I read, and when cannot see it being a problem with my gear..hence I'm puzzled.
Especially since I should have loads more softer diffraction caused images...shooting with a 1/2.3" type pinhead sensor...so my fastest f/ 6.3 is well into diffraction territory, or so I read.
ANAYV
You're right, and so am I, because diffraction (generally a ripple effect) has 50% more impact on the center of an image than the outer 1/3; depending on the subject, lens, and sensors size.
I try to take a more practical/realistic approach to photography. Will most viewers notice the lighter diffraction across the image? I don't fear or avoid stopping down when I need to. I'm pretty confident a Z lens with diffraction will still provide sharper results than an f-mount lens with it. I'm here to learn and I keep an open mind. If I see an image or example that changes my mind, I will accept any changes/admissions I may need to make.
I'm usually stopping down for reasons that override any concerns with diffraction. I'm sure there are instances where it can be an issue.
I try to take a more practical/realistic approach to photography. Will most viewers notice the lighter diffraction across the image? I don't fear or avoid stopping down when I need to. I'm pretty confident a Z lens with diffraction will still provide sharper results than an f-mount lens with it. I'm here to learn and I keep an open mind. If I see an image or example that changes my mind, I will accept any changes/admissions I may need to make.
I'm usually stopping down for reasons that override any concerns with diffraction. I'm sure there are instances where it can be an issue.
robert1955
Veteran Member
- Messages
- 7,425
- Solutions
- 10
- Reaction score
- 4,965
Seems to me you still haven't explained only repeated a claimYou're right, and so am I, because diffraction (generally a ripple effect) has 50% more impact on the center of an image than the outer 1/3; depending on the subject, lens, and sensors size.
I try to take a more practical/realistic approach to photography. Will most viewers notice the lighter diffraction across the image? I don't fear or avoid stopping down when I need to. I'm pretty confident a Z lens with diffraction will still provide sharper results than an f-mount lens with it. I'm here to learn and I keep an open mind. If I see an image or example that changes my mind, I will accept any changes/admissions I may need to make.
I'm usually stopping down for reasons that override any concerns with diffraction. I'm sure there are instances where it can be an issue.
briantilley
Veteran Member
Where is your evidemnce for that claim?
Unless you have other sources, I agree with BasilG that you may be misunderstanding something.
anotherMike
Forum Pro
Diffraction (the law of physics) doesn't give a crap about where in the frame something is.
What I suspect you might be seeing is related to this: Diffraction impacts higher frequency (in MTF lingo, which I would equate to very fine structures) data first and most, then mid frequency somewhat, then low frequency not as much. Consider that most lenses have the highest resolution (at higher frequencies aka finer structures) centrally, you might *notice* it more there, but there is no special behavior to diffraction relative to frame position. I've noticed the impact of diffraction, on a scene with complex/fine detail across and deep, taken with a very high end lens with very little resolution drop off in the edges, as I move across the commonly used landscape apertures, and that's anywhere in the frame that I notice it, and more importantly, I notice it "in whole" as a subjective loss of "bite" or realism, albeit slight in some manner, when I look at the image (and print) as one unit as opposed to looking at sections of it.
Now, getting back to some things earlier. If the image does not contain that much higher frequency detail - and there are a plethora of reasons why it might not, from imprecise focus, being handheld instead of tripod, the subject matter not containing such data, then you're not going to notice (visually) the diffraction as much. In the OP samples, which are downsized heavily and the focus point isn't exactly the same, not much to base any discussion from - the test isn't valid, nothing more to talk about.
Note I showed, objectively the impact of diffraction with regard to resolution frequency in another post from a while ago with tests using mtfMapper, but beware, it's a long post.
What I suspect you might be seeing is related to this: Diffraction impacts higher frequency (in MTF lingo, which I would equate to very fine structures) data first and most, then mid frequency somewhat, then low frequency not as much. Consider that most lenses have the highest resolution (at higher frequencies aka finer structures) centrally, you might *notice* it more there, but there is no special behavior to diffraction relative to frame position. I've noticed the impact of diffraction, on a scene with complex/fine detail across and deep, taken with a very high end lens with very little resolution drop off in the edges, as I move across the commonly used landscape apertures, and that's anywhere in the frame that I notice it, and more importantly, I notice it "in whole" as a subjective loss of "bite" or realism, albeit slight in some manner, when I look at the image (and print) as one unit as opposed to looking at sections of it.
Now, getting back to some things earlier. If the image does not contain that much higher frequency detail - and there are a plethora of reasons why it might not, from imprecise focus, being handheld instead of tripod, the subject matter not containing such data, then you're not going to notice (visually) the diffraction as much. In the OP samples, which are downsized heavily and the focus point isn't exactly the same, not much to base any discussion from - the test isn't valid, nothing more to talk about.
Note I showed, objectively the impact of diffraction with regard to resolution frequency in another post from a while ago with tests using mtfMapper, but beware, it's a long post.
Loading…
www.dpreview.com
I'm really not digging my head in the sand on this issue. Every visualization and diagram I have found researching visualizations based on the Airy Disc and the circle of confusion. I haven't seen proof that it doesn't impact the center of the image more than the rest either.
Regardless, I'm still under the overall opinion/impression that the benefits of stopping down outweigh any practical loss in detail. Especially given how much sharper the z-mount lenses are and how much larger sensors are these days. There are just too many other variables in photography that are more likely to impact the image than diffraction.
Loading…
alikgriffin.com
Thank you for that additional information and post. "Consider that most lenses have the highest resolution (at higher frequencies aka finer structures) centrally, you might *notice* it more there, but there is no special behavior to diffraction relative to frame position."
Where I might notice it more is all that matters to me, and I'm way more critical than most viewers will be. I have a 45mp sensor, much sharper glass, with much less fall off than I had just 5 years ago, camera correction, and software correction. Some of my "best photos" are far from perfect captures, and I'm fine with that.
Where I might notice it more is all that matters to me, and I'm way more critical than most viewers will be. I have a 45mp sensor, much sharper glass, with much less fall off than I had just 5 years ago, camera correction, and software correction. Some of my "best photos" are far from perfect captures, and I'm fine with that.
anotherMike
Forum Pro
I also always like to point out that diffraction seems to be another one of these "either/or" binary arguments when the reality is more fluid and less binary in real life.
IF your technique, scene, etc aka "situation" allows for the best data to be transmitted to the sensor, THEN diffraction matters more than if those items aren't allowing good data, so to speak. Diffraction matters, but like you alluded to in your other reply to Brian, if you need the DOF, well, you stop down, take the (diffraction) hit, and that's that. But ultimately if you're striving for high quality capture, you should be aware of diffraction - it's really why I choose really good lenses that deliver to the corners/edges at apertures that are pretty far away from where diffraction visually impacts D850/Z7 files. It's why one might spend the money for the Voigtlander 35 Apo Lanthar since it can be shot at freaking F/4 and get amazing quality across the frame if the DOF needs are met. To a degree no other 35mm lens you can natively mount on a Z7 can be. Which is why it's a 5 star landscape lens that meets the hype.
So in a way, it's not always a "OMG, Diffraction is so horrible" or a "Diffraction is over-rated" thing; depends on the situation and scenario for sure.
IF your technique, scene, etc aka "situation" allows for the best data to be transmitted to the sensor, THEN diffraction matters more than if those items aren't allowing good data, so to speak. Diffraction matters, but like you alluded to in your other reply to Brian, if you need the DOF, well, you stop down, take the (diffraction) hit, and that's that. But ultimately if you're striving for high quality capture, you should be aware of diffraction - it's really why I choose really good lenses that deliver to the corners/edges at apertures that are pretty far away from where diffraction visually impacts D850/Z7 files. It's why one might spend the money for the Voigtlander 35 Apo Lanthar since it can be shot at freaking F/4 and get amazing quality across the frame if the DOF needs are met. To a degree no other 35mm lens you can natively mount on a Z7 can be. Which is why it's a 5 star landscape lens that meets the hype.
So in a way, it's not always a "OMG, Diffraction is so horrible" or a "Diffraction is over-rated" thing; depends on the situation and scenario for sure.
These ripples are extremely small (a few pixels at worst), so even if your logic were correct, only a few pixels at the very edge of the sensor would be spared. However, diffraction is not a sensor effect, it's an optical effect that is independent of the sensor, so the ripples from rays hitting just outside the sensor will also affect the pixels at the very border of the sensor.
I think you might misunderstand those dome/ripple graphics in the article you linked to. Like this:
This is the result if you image a single point source (the text says a laser, which is one example of an approximation of a point source). However, if you capture an actual image, which is a collection of many, many point sources, there is not a big dome in the middle of the sensor and then small ripples towards the edges. This doming/ripple effects happen at every position of the sensor (think - every pixel, or every light ray). Like this (this is the top left point source and the bottom left diffraction pattern applied to all "rays" hitting that "sensor"):
Last edited:
This is a flying spaghetti monster argument.
That is something I'd agree with in many cases.
Last edited:
Boudewijn van der Drift
Senior Member
Ah, nice article, thanks. I was trying to simplify.The above is not correct.
For non-macro distances the size of the Airy disk at the film/sensor plane due to diffraction is entirely down to the F/number and has nothing to do with the focal length.
For macro distances the particulars of the lens design, specifically pupil magnification, is another factor in Air disk size.
https://en.wikipedia.org/wiki/Airy_disk#Cameras
For the OP:
Diffraction is proportional to F-number so it doesn't just suddenly rear its head. You just slowly lose resolving power the smaller the aperture. You can make up for a bit of that with sharpening in post processing. If you are downsizing the image it may end up being hardly noticeable. Yes, by F/8 one can "detect" the effects of diffraction with a high resolution FF sensor or a 20MP APS-C sensor and a very high quality lens. Whether it is of any consequence in the final processed image viewed at output size rather than 100% pixel peeping is another matter.
Here's a more controlled test from a long time ago with a 16MP m43 sensor shot at F/2.8 and F/16:
Clearly everything is much softer at F/16. Notice for the stars on black that you can also see the effect of wavelength - the red star is more blurred than the blue star since diffraction is also proportional to wavelength. The comparison between F/2.8 and F/16 is of course dramatic. The comparison between say F/16 and F/11 would be more subtle.
So let me simplify some more: A sunstar in the picture is a diffraction phenomenom as well, IMHO.
Some of the imageforming rays that are supposed to form a sharp image of the sun pass in close proximity to the diafragm, and are bent. And the wikipedia article probably includes formulae of how much bending as a function of passing distance.
Result is a sharp (be it overexposed) picture of the sun from the central part of the lens, plus diminishing size/exposure "star rays" where the light is bending more and more outward, the closer the imageforming rays were to the inner edge of the diafragm.
There are other factors, mainly how many diafragm blades and their shape. The ideal being a round diafragm at small apertures. Or a haxagonal one if you are after sunstars.
Yes. Settings for sharpening were the same...but noise levels are more in last shot ( ISO 4500)Third shot looks like it's been sharpened in PP a lot more. Are you sure you are putting all of these through the exact same PP pipeline?I read at times about diffraction and how it can lessen details as aperture is stopped down.
I also read that the smaller the sensor, the more this can become an issue.
Some even say at f8 details are already effected by diffraction...yet I have seen some very detailed images at f8 with a pinhead superzoom camera at 3,000mm.
Honestly ..I really don't see the impact when stopping down.
I shoot 90% wildlife...maybe diffraction shows up more with other type scenes?
Here's a quick non lab test.
I was shooting a very cooperative Great Blue Heron.
Took some shots at different apertures:
The last shot, at f/19, forced ISO up and in removing noise, some of the fine hair details most likely were smeared a bit.
Basically this is a flawed test...I need to do a much better one.. ( Keep ISO same but change shutter speed)..use tripod and timer release.
.but I still dont see a lot of softening when I stop down aperture....at least until f13 or smaller
ANAYV
I agree.......this is a flawed test.....but It didnt start out as one.
I wanted to see where sharpness was and fell off on this Sigma lens...and since subject was cooperative..,I took a bunch of shots...stopping down aperture....but I now see the flaw.
Better If I use totally still subject...tripod ...timer...keep ISO same ...and change shutter speed
Not that I would ( unless I buy the TC1401 1.4x TC for this lens...and I read it can also be used with some Nikon lenses).
Yes...but also DOF increased...a tradeoff
ANAYV
Keyboard shortcuts
- f
- Forum