Diffraction in Photography - Pixel Pitch, Sensor Format and More
ฝัง
- เผยแพร่เมื่อ 3 มิ.ย. 2024
- How the optical phenomenon of diffraction really works and how it impacts our everyday photography.
.
Video Content:
0:00 Introduction
1:19 Diffraction Explained
2:45 Diffraction and Resolving Power
3:45 Diffraction in Practice
6:22 Diffraction and Pixel Density
8:10 Diffraction and Sensor Format
10:25 Diffraction - A New Perspective
11:02 Recommendations
12:37 Conclusion!
.
Corrections
12:40 Large f-numbers instead of small f-numbers
.
Article: thomaseisl.photography/blog/d...
.
Instagram: / thomaseisl.photography
Website: www.thomaseisl.photography/
Facebook: / thomaseisl.photography
.
🎥 OM System OM-1
.
#diffraction #pixelpitch #sensorformat #sharpness #resolution #photographyeducation #ThomasEisl #ThomasEislPhotography
Please consider supporting the channel, donate a cup of coffee ☕
ko-fi.com/thomaseislphotography
Done!
Done
Incredible explanation, the quality and clarity of information you've made available to the hobbyist and budding pro is astounding. This is a perfect warm up and side conversation to Hyugen Optics Mirror Lenses Part 2 segment which is a deep dive into explaining the physics at play here.
Thank you so much! That means a lot, really!
I love the comment: 'The only ones worried about diffraction are those that haven't actually ever seen it.' Wonderful! and true.
Hehe, thank you!
Thank you for a sound and thorough explanation of the several things which influence image quality. We both know how widely misunderstood these are. I was brought up with astronomical optics, where resolution and diffraction are prominent, so it was basic knowledge when I started serious photography. But in the following half-century I have had to explain these effects and their interactions and their importance so many times to others! I only wish I had your straightforward top to bottom presentation at hand to refer them to.
Thanks again. I’ll bookmark this and post it where it can be useful, as you did.
Absolutely brilliantly explained Thomas... I look forward to hearing your next topic every time.... The perfect refresher course in the physics of lenses and the laws of photography...
Thank you very much Gerry!
Thank you for your clear explanations. Rule for me, don't bother about defraction, bother about depth of field, not printing anything larger than A3 (almost ever)...
What I really like is the easy to understand sciense instead of the shortcuts and myths in many explanations.
Enjoy your coffee!
I agree - don't worry about it as you cannot really avoid it anyways!
Thank you very much for your kind feedback and the tasty ☕!
Even if you printed larger than A3, most viewers would stand further back anyway and not try to pixel peep.
Great video, I actually learned something.
I wouldn't say resolution is "irrelevant" though; but it doesn't cause diffraction, and higher resolutions don't increase diffraction.
When thinking only about diffraction limits, resolution is irrelevant (but it is not useful to increase it far beyond the DLA, either).
When thinking about resolving power of the whole sensor+lens system, it can be limited by various things including depth of field, motion blur, pixel noise, lens artifacts, diffraction and resolution. Increasing one can reduce the other. Resolution is a relevant factor in that whole, and bigger is not always better because it impacts the other things.
Best discussion of diffraction I have seen so far 👍. Just subscribed.
Awesome, thank you!
Only a pro can argue that is more important a big depth of field than blurried out of focus. Great!
Thank you very much Matteo!
Well, I've never had "too much of it" really haha
Spot on once again! For ultra macro (life size or greater), I use f16 (m43 where I usually limit to f8) as DOF is usually more important. If cropping a macro greatly (or making a large print), then there is a clash of physics, and other techniques like focus stacking are needed. Landscapes with a required DOF but superb sharpness are a challenge to balance diffraction, DOF and picking the best point of focus.
A good future topic for your superb style would be "How sharp is sharp enough"... I find that MTF values of say 90lpmm vs 65lpmm are misleading when described as excellent vs "just" good. In practice, only staring at side by side comparisons shows a difference. For a particular lens, f8 with 65lpmm is to all intents, as good as 90lpmm at f2.8. Many opinions will say to avoid the f8 setting as (diffraction is usually blamed) resolution is not good enough.
I believe most modern lenses are sharp enough that priority in shooting should be the character your lens provides... quality of bokeh and "amount" of out of focus background (or DOF of subject), characteristic vignetting and drop off of sharpness from centre to edge. Lastly, "sharpness" once good enough for the type of subject has been achieved. (Improvement of some aberrations like colour fringing by a little stopping down can be advantageous, or left to software removal later).
Basically, explore the character of your lens, not just record breaking resolution... don't stop an expensive fast lens down just because a test reveals 65lpmm wide open vs 90lpmm stopped down, when a shallow DOF is the best artistic expression, and conversely, don't hesitate to use f11 (for fear of diffraction producing "only" 65lpmm) when a deep DOF is necessary for the appropriate expression in your photo.
I am sure your delivery would make sense of my jumbled up thoughts on "sharpness".
Thank you very much for your excellent comment! I don't find your thoughts jumbled at all, your points are very valid!
Indeed, I intend on making a "how to really pick a lens" video - I think this will cover exactly what you mention here.
A lens is - as you have stated - so much more than LPM and test charts, and I think you are completely right that this deserves attention!
Thank you very much for your inspiring thoughts and taking the time to write this comment!
@ThomasEisl.Photography thank you. I did leave out size, cost, mechanical features etc. I'm sure I will enjoy your take on lenses... I have been fascinated by them for 45 years, since I was 15 years old. Information was hard to find then, quality differences were great, plus it was expensive shooting film. Lenses were relatively, usually more expensive too, and a fast, expensive lens was soft but necessary for slow film. Now, fast lenses are sharp but wanted for blurry backgrounds... so many different things to think about... but I think sharpness has now been "solved", by and large, as a lens characteristic. Some genres, like product pack shots, architecture, birds and many landscapes, often depend on sharpness to succeed, but output size and viewing conditions are part of the equation.
@@bpcs63 Brilliant analysis! Thanks for sharing that - I will reflect on your very valuable points. Our thoughts on the topic are very similar!
Great comment, I totally agree.
If I could only watch one TH-cam channel, this would be the one I'd choose. Thank you so much.
Thank you very much - can't thank you enough for such a positive feedback!
Same
your ability to test and relay the scientific results is top notch! thank you so much!
Thank you very much for your kind words!
Super! Perfectly explained! Danke!
Thank you very much!
Very clear. Thanks for this video!
Thank you very much!
I have seen a lot of your presentation. Simple and succinct. Basic physics and optics which we have forgotten are brushed nicely. Thanks a lot.
You are most welcome, that is just great to read! Thanks for watching!
Excellent presentation of the math behind depth of field and diffraction. Thanks. This gives the rationale behind the old rule of thumb that most lenses are best stopped down about two stops.
Many thanks!
This was well done. One thing that helps is being able to shoot on M43 and FF cameras. I see no difference, both systems run into problems with diffraction in the same way. However, on my large format camera, things seem different. For some reason, f/32 looks amazing!! I can even shoot at f/48 and I rarely shoot wider than f/22. But we are no longer talking about pixels on LF!!
Thank you very much!
The great thing about Large Format is that you practically do not have to enlarge the image - as the negative is just so big. That means that all issues from diffraction to lens defects are almost "invisible" at normal print sizes.
You cannot beat large format with any digital camera!
@@ThomasEisl.Photography Okay that makes sense. Ou are saying the diffraction IS there, it's just that you can't see it very much because we are not enlarging the image enough to see it (normally).
I also thought that LF lenses were designed to operate at f/22 to f/48 in normal operation and they are therefore optimized for this. Maybe it's both....with the insane amount of depth of field in LF, anything close to f/11 has near paper thin depth (of course depending on subject distance). Anyway, thanks!!
Yes, you are perfectly right! LF lenses were indeed designed for smaller apertures, but that has more to do with aberrations being less at these apertures.
Thanks for this great conversation!
Good stuff Thomas, funny that you mentioned a ghost hunt, I was a paranormal investigator for 13 years. Made me chuckle. Cheers!
Great to hear - so my joke was approved by an expert! 😅
Thanks for the comment, and thanks for watching Randy!
Again, very good. Going to watch all your clips! Related to this are “experts” who claim film grade lenses perform poorer on new high rez sensors because of lens resolution. They claim the new sensor produces an image that is worse than a lower rez sensor and link it to diffraction somehow. Again, probably because they pixel peep and hence look at the image at higher magnification
Completely agree - the problem is the increased magnification. One thing that might be an issue - which has nothing to do with diffraction of course - is that the light rays are not focused perpendicularly to the sensor plane, especially in the corners. Film does not care, but camera sensors do depending on their micro lens design.
Thanks again for your kind words and your excellent contribution!
Another deep lesson about as camera works and consequentially as the photographer can use his camera to achieve the desired effect. Bravo! Sergio from Milan
Thank you very muhc Sergio! I very much appreciate your words and thank you so much for watching!
Really interesting! Thank you
Thank you very much Stefano!
Brilliantly explained - so clear and so well reasoned. I learn something new every time. Thank you Thomas!
Thank you very much for the kind words! So great to read that!
Another brilliant video! Surely the result of a great deal of work in order to construct such a theoretically sound and practically focused explanation. Thank you! I have begun telling everyone I know to watch your channel and follow you.
Wow! I'm out of words, thank you again!
That is really awesome - I'm super thankful for that!
incredible video Thanks
Thank you very much!
Outstanding information!
Thank you very much!
Evidence based explanation brilliantly presented. Danke!
Thank you!
Thank you
Most welcome!
Exzellent!
Thank you - vielen Dank!
Thank you for making this video 😎👍
My pleasure!
Finally a proper explanation, thank you. Higher f-stops are a feature not a flaw. If you are truly diffraction-averse then use a high quality lens and focus stack at the optimum aperture and then anti-diffraction/deconvolution during post processing. Or use AI to generate the perfect photo for you -- no expensive lens needed.
Thank you!
I like the professorial lecture style. The quality is exceptionally high, which is why I bother commenting on potential glitches. Please debunk any glitches in my comments as well.
Viktor, I very much appreciate critical feedback - ONLY (!) by scientific method and critical approach we can come to valid conclusions.
Thank you very much for your open-minded efforts and kind comments!
Profesor dr Thomas Eisl! 👏👏👏👏👏👌👌👌👌👌
Haha thank you very much!
Thanks 🙏🏻❤️
You’re welcome 😊
Thanks!
Thank you very much for your generous support!
one of the best explanation video i have ever seen. Subscribed.
Welcome aboard! Thank you very much!
Amazing quality 💓
Thank you 🙌
Definitely the best explanation of diffraction I've seen. Well done 👍
Wow, thanks!
Great stuff
Thank you very much Martin!
Excellent presentation Thomas. I really appreciate all of your presentations. You are not sloppy with your choice of words. Something quite rare these days. As another viewer stated, the clarity of your work, examples and recommendations is superb. Thank you. As an Olympus user, we are fortunate to see your support of the M43 format.
Many thanks for your exceptional feedback, that means a lot to me.
As a "TH-camr", I'm fortunate to see YOUR support of my channel with comments like these. I'm very thankful, really!
Best wishes from Vienna!
Vielen Dank!
Ich danke fürs Ansehen!
Thanks
Thank you very much for your support!
Thomas, here's a funny conclusion from this video and the depth of field/aperture equivalency video. The depth of field that is the namesake of Ansel Adams & Co.'s famous Group f/64 can be achieved on micro four thirds at just f/4.5 - with the same level of diffraction!
Haha 😆
Wow! 👏👏👏👏👏👏👏
Thanks!
Your explanation helps me to understand the sweet spot on my lenses. Most of my lenses look the sharpest between F4 to F8, depending on the lens.
I did have one time I was shooting an N Scale layout with my macro lens. However, the limited depth of field was creating problems with my shots. I tried everything, but stopping the lens down to F22 gave the best results. I'm sure I lost some sharpness due to diffraction, but the gain in depth of field was worth it.
Absolutely - don't worry about diffraction, if you need the DoF, there is no other choice!
Thanks for taking the time to share your experiences, that is very valuable!
MAte, I love all your videos.
I appreciate that! Thank!
Thanks Thomas! I love the clear and succinct ways that you describe things. I loved the statement "without diffraction, there is no photography". I literally laugh out loud when I hear people say things like "above f5.6 diffraction sets in". Aside from being technically incorrect, it places a self imposed limit on how people create images. Diffraction is ALWAYS present. Too many times people repeat the same kind of things because they've heard it so many times...they believe it must be true! I simply don't worry about diffraction. I often use f11 when taking single shot macro images. Nobody has ever commented "wow, looks like a lot of diffraction there...". As you mentioned, it's better to have the DoF in your image...along with some diffraction than it is to have an out of focus area with no diffraction. Sharpening in post will likely be much more successful on the softness of diffraction than it will on trying to correct for shallow depth of field. Keep up the good work :-)
First, thank you very much for your kind feedback!
Second, thanks for your very relevant and detailed comment!
Third, I completely agree with what you stated, especially the sharpening part is highly relevant, I have not thought of that yet but you are of course completely spot on!
Thank you very much for taking the time to contribute!
As always Thomas a very enjoyable and informative video. As a seasoned professional it’s good 👍 to view your presentation/s from a person who actually knows his subject. Congratulations 🎊🍾 Best, Geoff
Many thanks!
Thanks for watching, although you know this stuff pretty well already!
@@ThomasEisl.Photography I do 😊 however, you are brilliant at presenting it.
Thank you very much!
Nice video!
A few days ago I took a series of test shots with my Canon 750D APS-C camera with the 60 mm F=1:2.8 CANON EF-S macro lens.
I used 2.8/4/5.6/8/11/16/22/32
Magnification was ca 1:1 and the subject was the flat rear side of my watch.
I cut out the same 1600 x 100 pix near the center from each image and used ImageMagick (a tool) to add all 8 images in a single image.
The loss of sharpness is clearly seen. At 16 there is a bit of degradation and at 32 the image is way too soft.
You did not mention that the aperture as marked on a lens relates to the infinity setting. For macro, the effect gets even worse.
Yes, you are absolutely right!
Thanks!
Thomas, that was excellent. I hung on your every word... and this on a subject that, for all daily purposes, i really couldn't care less about. 'Impeccable' is the only word that comes to mind to describe that video - and that given that i paid very close attention. I'm very happy for you (work that out, if you can!). Kind regards from the swamp, where my love affair with my 'new' E-1 continues unabated. PS: nice shutter sounds... should i recognize those?
Many thanks for the great feedback - this is a bit of a dry subject, so I'm happy to hear that I was able to deliver!
I think I used the shutter sounds of the OM-1! One of my favorite shutter sounds by the way.
Very usefull, thanks a lot!
NB As always I'm very nervous looking at your hands and lenses on the table )
Cheers
Haha, yes, they are actually right at the edge. If they hit the floor, I have an excuse for a system change lol
Thank you!
Very interesting and insightful.
Just a thought: how about using focus bracketing for extra DOF while maintaining a large(r) aperture, as a way to limit diffraction?
Yes, absolutely!
That is why in-camera focus stacking like in the OM-1 is really a neat feature.
If you can live with a certain amount of diffraction, you can, however, stop down your lens quite extensively. Diffraction is not as bad as it sounds.
This was interesting again, thanks Thomas. IMO we tend to forget about the whole image by obsessing about details. Modern lenses are that great that the inferior kit lens become a myth. So we had to find a different topic, diffraction it was. Easy rules, like NEVER shoot a smaller aperture than F whatever. I personally just do not see the effects of diffraction or I'm not aware of them. Maybe as a film shooter I took what I got away with or my standards are so low, IDK. But it's my firm belief that it's about the whole image first.
You are absolutely correct, completely agree!
Nice. How would it relate to circle of confusion? Thanks.
Circle of confusion is used in the context of depth of field - the bigger the circle of confusion, the more we perceive a part of the image as "out of focus". It is best to examine both phenomena separately - because if you stop down, you get smaller CoCs but more diffraction. So overall perceived sharpness goes up, while resolving power goes down.
Hope this helps!
Interesting stuff, the Airy disc science especially.
Since you've been doing a lot of videos about the Olympus OM-D cameras lately, I found myself a fairly cheap second-hand OM-D E-M5 Mark 1 with a couple of kit lenses and compact flash included, on ebay. Only has about 3,000 actuations, so it has barely been used. I'm curious to see what all the fuss is about MFT cameras, so it should give me some insight. Should arrive sometime next week. Will let you know what I think in the near-future although it likely won't see much outdoor use until a bit later in the year as I don't go out during colder season for health reasons. Thanks for the inspiration.
Wow, Marcus!
Now I really hope that you will be happy with your new OM-D! The M5 is a great camera, really. I love to use mine. It is a legend that started the whole OM-D lineup. And yours is really brand new, as you've said!
I am very sorry to hear about the health issues, I really hope that things will get better soon!
Maybe you can at least try the camera indoors and get to know your new tool. One can do great photography indoors as well.
Hope to hear from you soon and all the best, Marcus!
I'm sorry to hear that!
The good news is that we are probably about to have a great warm summer!
(Note: as your comment contained sensitive information, I decided to hide it from the comment section to protect your privacy)
@@ThomasEisl.Photography I really hope we do. Last year wasn't too bad, but I wish it'd last longer - summers here are usually short and sweet with the odd "Indian summer" later on. Personally, I'd prefer long and moderate summers as I'm not into tanning or all that, I just want an excuse to get out and about more, the lack of exercise after all these years of isolation is making me lazy and lacking in motivation.
Thanks for the privacy, I appreciate it.
@@MarcusBritish yes, those super humid and hot summers are not my thing either - but we don't get much choice unfortunately haha 😆
1:07 (subtitles:) Diffraction and the Ghost Ship have much in common [Chip = Ship]
Ah, these titles are auto - generated 😀
Your explanations are the clearest and best I have found. I still struggle that pixel size will not matter. The airy disk even at the same size should effect more pixels when they are smaller. Keeping sensor size the same a 12mp 35mm sensor will have bigger pixels than a 60mp sensor and the airy disk will effect more pixels.
Thank you very much Sebastian!
I completely agree - it is hard to wrap the head around that resolution does not matter when it comes to diffraction.
The fun thing is that the 12 MP camera will be affected the same way, it is just not so apparent as the camera is not able to record "the defect" in high resolution.
Another way to look at it: There is no single airy disk anywhere, but just a lot of light rays projecting these disks. So no matter how big the pixels are, some rays will always "hit the space" in between the pixels and so on.
Again, thanks for your very kind words!
@@ThomasEisl.Photography I read another article that said basically what you are saying, and I think I have a better understanding now. Diffraction does not change with smaller pixels, but smaller pixels are able to better record detail even if that detail is diffraction so it can have a visual impact if you are pixel peeping. At the end of the day, you should still be better off with the high resolving sensor. Thank you. I think this is as best as I have understood these concepts in my life.
That is great to read!
And you are right, more resolution does not harm, except your hard drive 😅
@@ThomasEisl.Photography I don't want to waste too much of your time, but do you think the claims that lenses are not able to resolve the new high MP camera's to be credible? With a fast lens such as f/1.4 I can see that being true, but then again, I doubt the aberration control was good enough to resolve a lower MP sensor as well. However, stopped down to f/4-f/8 I would think these lenses should not have any issue?
I've recently had the chance to talk with Peter Karbe chief of optics at Leica about this.
Here is the thing: Older lenses were not built for insanely high magnifications - zooming in on a higher MP camera is nothing other than enlarging the image. So it is less about the MP, like with the airy disk, but more about the problem that we preview images unreasonably large. This means, yes, you can definitely use old lenses on high MP cameras. If you enlarge the resulting images like crazy, you will see the limits of the lens at some point. this is also why modern mirrorless lenses are these insanely large abominations (😉) - they are overly corrected to allow these huge enlargements (that no one really needs). So you can use your old lenses wide open, but also stopped down, as long as you do not enlarge the images more than you did in the past, there won't be any (!) difference.
Thomas, your videos are exceptional. You are a very trusted source of information. I have a question. Would your please explain the difference between Field of View and Angle of View. It seems that camera manufacturers indicate they are the same when giving specifications. Intuitively, I think there must be some difference. Thank you in advance
Many thanks!
So there is indeed a difference, but both terms are mixed up all the time - so in practice, the AOV is usually the metric that is used, describing the angle that a given imaging system can image. I recommend this article (note they use the term AFOV for what is usually known as AOV):
www.edmundoptics.com/knowledge-center/application-notes/imaging/understanding-focal-length-and-field-of-view/.
@@ThomasEisl.Photography Thank you, Thomas, for the article. It is very helpful. As I understand it, for a prime lens the AOV is an attribute of the lens and is constant; the FOV can vary by changing the working distance (also by a change in focal length of sensor size). While this sounds a little strange, it is very obvious when you actually try it with a lens. Since I do not have an OM-1, the articles about the OM Workspace and the more technical side of the MFT system are very interesting. I hope they continue. Again, thank you very much.
10:19 You did pull a modest slight of hand here by mixing linear and area ratios. The ratio of the discs is 4 in linear dimensions and the ratio of the sensors is 2 in linear dimensions so there is still a factor of 2 in play.
Hey Colin! Great to hear from you again.
Here is the good thing - I did not as I calculated the area size for the airy disk 📀
If you compare it with the table earlier in the video, you will see that it is exactly the area and not the diameter (which is in the table) of the green light airy disk.
Thanks for getting in touch!
@@ThomasEisl.Photography Many thanks for the reply! I must sit down and go through all this for myself - it will no doubt be therapeutic for my aging brain to do some physical optics as I haven't had much use for it since the 1980s
No worries!
I was actually that quick with the response as I made this mistake when doing my original calculations. Gladly, I found out.
In any case, I very much appreciate your feedback and that you really watched the video - should there be a mistake, please let me know! That would be great!
@@ThomasEisl.Photography Many thanks again. I think I will also try to nail down the concept of depth of field. It was always something with which I was never entirely happy. Of course, once you have a word like "acceptably" in a definition it's problematic!
Good recommendations! What good is a bit lower diffraction when whole parts of your picture are not sharp because of a too narrow depth of field? Well, for e.g. product photography or landscapes. When it comes to portraits, you can get a narrow depth of field plus lower diffraction when choosing a lower f- number. As always it all depends on what you want to photograph...
Absolutely agree! Thanks for sharing your thoughts!
Thank you Thomas. It would be nice if you could address this additional issue: Comparing 80 MP and 20 MP images in M43 is comparing different pixel sizes in the image, but the SAME pixel size on the sensor, not 4 times smaller.
Also, your recommendations for safe upper limits seemed conservative to me. My meaning is that f5.6 is conventional wisdom for M43, and I was expecting something different. In any case, I now feel safer playing with settings beyond that limit. :-)
Hey Roderick!
Yes, the upper limits are SUPER conservative indeed. I was trying to provide numbers that will work with ultra high end and low end lenses. That was of course only a rule of thumb.
Yes, it is the same sensor size - I was trying to emphasize the point that resolution does not "produce more diffraction" but our "zooming in" does!
And yes, try shooting f/16 on MFT from time to time, thanks to clever JPEG algorithms in camera, it is actually quite good!
Thanks for your comment, much appreciated!
@@ThomasEisl.Photography Thanks. I should have understood that the first time around but my mind is a bit tired this end of the week. It was stuck back at “the sensor just records “. :D
Very lucid explanations. And very, very interesting. To be honest, I have never actually seen diffraction. On the other hand, I am convinced that every camera I have ever possessed has diffracted and color-fringed in extensio. Apart from that, my motive was probably never absolutely in focus or even otherwise sharp because most of the time I was running around hand held. Nevertheless. Many of my pictures were absolutely gorgeous. But I do admit that a similar number of pictures were not. But this has to do with my brain and never had anything to do with my cameras. In other words, over the past fifty years affordable cameras have always outperformed their owners.
Although I have a technical background, I do not worry too much about the characteristics of my camera. I know very well that the designers have had to make a hundred compromises. Then the manufacturer messes everything up. The lenses are not quite bang on the optical axis, the glass and the coatings are slightly off spec, and last not least, the zoom mechanism never puts everything back twice into the same place. Then I realize that the sensor has in reality only about a half of the nominal resolution - basically four color pixels make one luminance pixel and to do this magic without anybody noticing it, the software is full of tricks (and bugs), and at the end of line, just for good measure, the jpeg dumps two stops of dynamic range.
But all these shortcomings are either quite subtle or do not play a really relevant role.
For example, the Sony RX100M3, a few years ago the best of its class - i.e. compact 1", did not even have its optical system in line with its sensor. The sensor is completely off-axis. If you look at the raw file - without the appropriate distortion compensation algorithm in place - you fall off your chair and roll around on the floor in a fit of hysteric laughter! Once they have got the picture back into a non-distorted rectangular shape, about an eighth of the recorded information at the bottom edge of the sensor has to be thrown over-board. And even so they still managed to achieve top marks and outperform their competitors.
To sum it all up - today's middle-range cameras are on such a high level that there is not much left to complain about.
The batteries could last a bit longer. 😁
Very well said!
I completely agree!
I use f/6.3 on my OM-D as the smallest f-stop before diffraction sets in. Its given me the possibility to achieve super sharp images that are indistinguishable from the ones of my Nikon full frame. Especially for landscapes where I use no smaller than f/11 on a full frame when detail is critical. One could argue that the full frame 24 megapixel is superior to the 16 megapixel four-thirds, but that's only true if you explode your images above 1.5 meters wide and view them up close... and the optical quality of the lens is superior, which is seldom the case with Zuiko lenses. No lens I own (OM-System, Canon L, Nikkor and Zeiss) has a superior resolution of my Zuiko 12-40 mm 1:2.8 even wide open. The Nikkor 180 mm 2.8 AI comes very close indeed, but that's not why I bought this magnificent lens. Every lens I own, was bought with a certain goal in mind and that's seldom resolution although it's a nice surprise when that's also true.
Absolutely agree!
M43 is extremely capable and often vastly underestimated!
Why is the recommended F stop (timeslot 11:41 min) the same foor APS-C and Full Frame?
Thanks for asking - two quick clarifications:
1) Those are rule of thumb numbers and I did not want to provide 1/3 f stops - not really relevant in practice.
2), those are not the ideal f stops for best resolution and performance, but safe upper limits for general purpose - you should definitely test your lenses for your own tolerances as described in video.
Hope this helps!
Hi Thomas. I watched the video twice. I'm left confused how sensors with smaller pixel size/pitch aren't differently(worse) affected than larger pixels/pitch by diffraction. if i am on the right track...there are two parts to an Airy disc and the concentric circles of the Airy disc are the undesirable parts. the center portion (the light ray) of the Airy disc being the transmitted 'data' that is of importance. if the rings of the airy disc are substantial larger than the pixels of the sensor then some pixels will receive little actual light ray data and more of the just the rings. that would be bad, no? less information for the processor to use for interpolation.
I know exactly what you mean - I've been on the same train of thought for a long time.
Here is the thing - in reality, there is never a single light ray hitting one pixel. They are instead hit by numerous always overlapping airy disk patterns (which are light rays).
When you keep distance, you won't be able to see the unsharpness.
But with a high MP camera, you are using a microscope and zoom in way more than on a low MP camera. So you are just able to magnify the defects more.
In reality, even a low MP sensor is also hit by the same amount of overlapping airy disks - it is just as sharp. But you cannot magnify it enough to actually see it.
It is like a billboard, if you are not able to get close enough, you can never see the poor image quality. If you can, you will see all the defects.
But even if you cannot get close enough, the defects are still there.
The pixel airy disk coverage thing is a very common misconception and you will hear it all the time - I heard it many times as well.
Was I able to help?
@@ThomasEisl.Photography Thanks Thomas. I did not take into account that the pixels are being constantly bombarded by Airy discs/light rays.
@@ThomasEisl.Photography I too have been pondering this part of the topic, with you and Fuzzy. Your further insights above, have allowed the light to reach my cerebrum! Thank you for another brilliant video, Thomas!
Thank you so much!
So it seems that if your required DOF requires too small apertures (let's say F16 or more on a 45MP camera), then focus stacking is a better option to get maximal sharpness?
A technique used by most landscape photographers. Almost all of them will stay "under" f/16 for all of their shots. They will focus stack to account for the depth of field on photos where everything is in focus.
Absolutely agree - focus stacking enables incredibly sharp photographs.
In most cases, accepting diffraction is also perfectly fine.
I will have to watch a second time ... The explanation is top, but my brain is struggling ... :-)
Most welcome Rudy!
Thank you very much!
I shoot 1x macro in F/13, there is loss of details due to diffraction, but most current AI tools can "repair" them, F/8 dof is too small for my insects :p
Absolutely - that's exactly the point I was trying to make: stop down as much as you need and don't worry about it too much! Thanks for sharing!
There should be mandatory training videos required before signing up to DPreview et al. 😅 but i suppose if everyone had good quality info there would be less to debate 😂.
Haha, agreed 😂
I thought the loss of contrast around the arrows at f/16 and beyond was already very very visible, even in a 1080p export. f/22 was substantially worse as the texture on the keycap was lost. Even at f/8 you can see the texture gets just a tiny tiny bit soft. If you tried this test on something with lots of fine detail (e.g. money) , the results will be more visible.
Regarding the test about pixel density, you're testing extrapolated pixels from pixel shift and not real physical pixels. That will introduce artifacts/inconsistencies depending on how software combines and then debayers them. It's better to test with actual high res vs low res sensors. E.g. GH5S vs regular MFT or A7S vs A7R bodies. One cannot trust proprietary or specialized software that will introduce variances in such testing. E.g. Some DoPs use a variety of debayering softwares on their movie RAW files because each one gives a different result and they are used in different situations. It would be also better if you could show exact frames instead of left/right comparisons as they're different things and viewers cannot draw any conclusion from different scenes.
Yes, technically even at f/4 the lens cannot resolve as much detail - well it can, as aberations are the more prevalent issue.
Regarding the example - that is correct. I thought about different variations but by setting it up like this I was actually able to keep the most parameters the same. As I already knew that it was mathematically and optically correct, this was a compromise I was willing to take.
Thanks for your feedback!
Very informative video. Everything is well justified except for the "rule of thumb table" at 11:38. If Thomas Eisl hadn't come out in another video as a FF shooter (and recently dual FF/MFT shooter), I really would have believed that he was an APS-C shooter. Because it goes along with the common APS-C delusion that "APS-C is basically like FF."
To me, format-obsessed APS-C shooters are like younger brothers who hang out with the big boys and thus feel entitled to make fun of the "little kids" barely younger than them (non format-obsessed APS-C-ers just don't care and use this format because it works for them). As a reminder, the "conversion factors FF --> APS-C and APS-C --> MFT are
In the Nikon world: 1.5 and 1.33
In the Canon world: 1.6 and 1.25
In other words, there is a bigger jump from FF to APS-C than from APS-C to MFT.
So, in my opinion, the table should have read f/7.1 (or f/6.4 if one wants to be very kind) for APS-C.
Thanks for your kind words and your feedback!
Of course, the value should not be a full stop - however, the table is only a rule of thumb so I decided to provide an easy to remember full stop. That is the only reason - there are no format wars here. Also, the MF should also not be a full stop if calculated precisely - but again, the rounding is done to give everyone an easy number to remember.
"Without diffraction there would be no single photograph", you said. Please explain briefly why. Is the logic that diffraction is a property of light and without diffraction no light and without light no photograph? Or is it because diffraction is doing a "useful job" that is essential to create a photograph -- the diffraction creates the photograph? However, if you instead had said (and here is my version): "Without refraction there would be no single photograph", it would have been crystal clear because refraction is what a lens is doing to create a focused projection to create a photograph.
Hello Viktor!
Another very sharp observation by you - thanks for contributing.
I agree. The reason why I said "without diffraction no photograph" was that I wanted to illustrate that every focused beam always creates an Airy disk (which is the basis for diffraction) and put the "fear of diffraction" a bit into perspective. That was the intention.
Thanks for the contribution, as I stated earlier today, I greatly appreciate that!
"Nothing can fix an out of focus image"
I know you mean that in the conventional sense, but for your own interest, check out light field cameras. I think it's a shame those never became mainstream.
Yes agreed!
Pixel pitch effecting diffraction is a myth I've never heard. It used to be said that increased pixel pitch could effect dynamic range*, especially on older cameras with slower processors. I don't know if this is true with modern high megapixel cameras, but complaints continue to emerge about poorer dynamic range, banding and artefacts at high ISOs, compared to cameras with fewer megapixels.
When you say "sensor size is irrelevant when it comes to diffraction", it could lead someone to imagine that this pertains to every setting. I think you are saying a m4/3 lens set at f4 will offer the same diffraction as a full frame lens at f8, not that diffraction in both formats will be the same at the same aperture.
*Fewer, larger pixels offering better dynamic range than more smaller pixels.
Maybe it is something that is more prevalent in German photo circles. I've heard it quite a bit, but I am super happy that you don't as it is a rather strange train of thought!
Dynamic range and what gives more and less DR is really difficult - as you've stated, there are so many factors at play. One thing is for sure - almost every digital camera has more than enough of it.
Thanks for your feedback!
(Deleted my comment about Airy disk, which was incorrect).
Small F Number: 2 or 64?
At 2:30 in the video and in the table of Airy disk, 2 is small (but represents a large opening).
whereas in the conclusion 64 is small. ("Small f numbers were always used in photography history for artistic expression, think about group f 64"). The way the terminology is practised (not only here but many places), it invites confusion and mistakes. F number is a very bad misnomer. It literally states: focal length number.
I suggest: always think of the number as the denominator in a fraction: aperture_opening=focal_length/Number. Never write F-Number, but do as Ken Rockwell, write f/Number. Regarding large or small never describe the Number, but do describe the f/Number.
Whoop, you found a slip of tongue!
Small f number means of course small number = wide opening. In the end, I accidentally said small although I meant large! Thanks for pointing that out.
f/64 is equivalent to f/8 on a full frame sensor
I'm sorry, could you clarify that?
@@ThomasEisl.Photography Hi Thomas. The group f/64 used a bigger sensor compared to full frame or m4/3. Thus, the equivalence in terms of depth of field is about f/8 on a full frame.
Sorry, obviously it was not a sensor. They used film or plates
Yes!