-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
FFF should be 600 MP ...
- Thread starter xpatUSA
- Start date
- Messages
- 4,193
- Solutions
- 1
- Reaction score
- 4,056
There is no noise penalty, at the image level, to more resolution.
Do you have a source?The 200MP images those phones produce are created with seveal images atcked onto each other to alleviate the noise issue. I doubt any modern camera is doing this.
And besides, I don't think I'd ever need or want 200MP images.
And it would have the slowest readout speed ever seen in a camera, would have enormous files requiring a NASA computer to edit them, the worst part still being the fact that image quality wouldn't be that great : most lenses can't resolve a sensor like that, so you'd be throwing most of your resolution in the bin, and then there is the issue of noise which is horrendously bad on those photosites.Imagine a full-frame sensor with photosites as small as those? It would probably be 2 gigapixels or something crazy like that.
No thank you, I'm perfectly fine with the cameras we alreayd have today.
My understanding is that we are *already* diffrection limited on a lot of lenses on 61MP full frame sensors.I'd like to see a camera like that. Even with the 65mm f2 at 4 the sensor would capture all the detail the lens can produce.
About 200 MP/layer Scott
https://www.dpreview.com/forums/post/67790814
--
What you got is not what you saw.
Last edited:
- Messages
- 4,193
- Solutions
- 1
- Reaction score
- 4,056
I guess it depends on how one defines “problem”. Beyond say, f/11 on a modern FF, the image will become blurrier due to diffraction. This limits the ability to stop down to get more in sharp focus. The depth of field increases, but so does the blurring due to diffraction.
Makes me wonder why this was published long ago for some normal lenses:
Courtesy of Bob Atkins https://bobatkins.com/photography/technical/mtf/mtf2.html
"Ideal MTF" is the diffraction-limited line for all lenses. 'S' is a normalized frequency so that a lenses f-numbers can be compared in the same graph. Bottom right would not be my favorite lens, even "at normal distances" ... top left looks good ... at MTF50, f/11 still has some steam left before "hitting" diffraction.
--
What you got is not what you saw.
Last edited:
The key to what I said Ted . . . is exactly what I said . . . "Imagine a full-frame sensor with photosites as small as those?"
That tiny 200 MP sensor in the Samsung S23 Ultra (or is it the S24 Ultra?) has such small pixels that if you made a full-frame sensor with such small pixels it would have billions of pixels, no doubt. In the Samsung phone it is already 200 MP, so a full-frame sensor with the same size pixels would not have just 200 MP, obviously.
- Messages
- 4,193
- Solutions
- 1
- Reaction score
- 4,056
When I said "about 2um", maybe I was mislead by Notably, the HP9 excels in low-light conditions, addressing a common challenge for traditional telephoto cameras. Its Tetra²pixel technology merges 16 pixels (4×4) into a large, 12MP 2.24μm-sized sensor.
If it could shoot and store 0.56um pixel pitch FF images then 1543 MP.
There's this thing called "Moore's Law" . . . which states that processor speed will double every 18 months (but I think they've adjusted it to every two years now). In a few years the processors will be so fast and powerful that processing gigapixel images will be just as fast as processing 50 MP images today, and Sony's A1 can process 50 MP images quite quickly. In fact they've been processing 24 MP images plenty fast for many years now (i.e. in the 24 MP Sony A65 that I have the images process really really fast - so fast that the camera can shoot 10 fps, and right after I shoot a series at 10 fps if I want to review them it only takes a couple of seconds for me to start reviewing the shots I just took). I got my Sony A65 used long before I bought my Sigma SD1 Merrill, and that was years and years ago. I'm sure that today's processors can handle 100 MP photos (and probably even 200 MP photo) quite quickly. I haven't heard of anyone complaining that their Samsung S23 Ultra takes forever to process those 200 MP photo, and I'm sure there is a whole lot of processing going on with those photos.And? I don't want a 200MP RAW file if that means I get tons of noise. The 200MP images those phones produce are created with seveal images atcked onto each other to alleviate the noise issue. I doubt any modern camera is doing this.
And besides, I don't think I'd ever need or want 200MP images.
And it would have the slowest readout speed ever seen in a camera, would have enormous files requiring a NASA computer to edit them, the worst part still being the fact that image quality wouldn't be that great : most lenses can't resolve a sensor like that, so you'd be throwing most of your resolution in the bin, and then there is the issue of noise which is horrendously bad on those photosites.Imagine a full-frame sensor with photosites as small as those? It would probably be 2 gigapixels or something crazy like that.
No thank you, I'm perfectly fine with the cameras we alreayd have today.
No, we're not diffraction limited on lenses at 61 MP (maybe with some crappy old lenses, but for good lenses - no way). I have a sixty year old lens made in Russia, which I mount on my Fuji GFX100 sometimes, and it produces excellent results at f8. Go ahead and take a look for yourself:My understanding is that we are *already* diffrection limited on a lot of lenses on 61MP full frame sensors.I'd like to see a camera like that. Even with the 65mm f2 at 4 the sensor would capture all the detail the lens can produce.
GFX100
Sample photos (mostly OOC jpegs) from my new (to me) Fuji GFX100 with various lenses - first images are with a Helios 44-2 58mm f2 on an adapter (mostly at f8 and f11).
www.bigprintphotos.com
All the samples you find there can be downloaded, and you will see lots of photos from that little old lens. I have included crops, so you don't have to download anything in order to see the amazing resolution of that lens. I have added samples from other lenses near the end of that gallery, but for the first ten or fifteen shots they are all from that really old Russian lens.
Keep in mind that this is a 100 MP camera that lens was mounted on. In the future, with newer lenses like the 65mm f2 DG DN from Sigma, you could shoot at f4, and get amazing resolution images with a crazy-high-resolution sensor. 61 MP is nowhere near enough to capture all that the lens has to offer at f4. Sure, we're diffraction limited if we set the lens to f8, but most lenses can produce a better image if you set it to f5.6, and that newer Sigma lens can produce an even better image at f4. Newer lenses in the near future will be even better than that Sigma lens, which is only a C lens - not an Art lens. Already there are amazing lenses on the market that we will need far better sensors to fully take advantage of. Shooting at f4 with the best lenses available for mirrorless systems today would require 100 MP sensors to get close to capturing all the detail the lenses have to offer. 200 MP would probably required to capture every nuance . . . and that's with lenses available today. When we finally can buy an L mount camera with a 200 MP sensor I'm sure there will be even better lenses out there, which will offer even better image quality at f2.8, and then we'll need a 400 MP sensor to capture all the detail those lenses have to offer.
Anyway, I like the idea of a 200 MP per layer full-frame Foveon sensor Ted.
I still hope for the 20 MP per layer full-frame X3 sensor though. I think that would be a good start, and then they could double that to 40 MP per layer maybe a couple of years later. With an 80 MP equivalence (i.e. in my opinion such a sensor would be equivalent to a Bayer pattern CFA sensor that is 80 MP), I would really like to have such a camera, but I'd buy the 20 MP per layer camera first, and then hope they would make a camera real soon with a higher resolution sensor. If they could "quattrify" the 20 MP per layer sensor that would be spectacular!
All I know is that I stepped up from 12 MP to 36 MP over the years in the full-frame world, and all I get is more detail in my photos. The 61 MP cameras (Sony A7r IV, Sigma fp L, etc.) seem to get even more detail. We now have even better lenses that we can buy, such as the Sigma 65mm f2 DG DN C and Sigma's amazing 105mm f2.8 DG DN Art, as well as other lenses from the likes of Panasonic and Leica, which offer crazy-high image quality at f5.6, and even more resolution at f4. Image quality has indeed improved significantly since the days of 12 MP full-frame (Canon 5 D and Nikon D3 and D700). It will get even better in the years to come.Makes me wonder why this was published long ago for some normal lenses:
Courtesy of Bob Atkins https://bobatkins.com/photography/technical/mtf/mtf2.html
"Ideal MTF" is the diffraction-limited line for all lenses. 'S' is a normalized frequency so that a lenses f-numbers can be compared in the same graph. Bottom right would not be my favorite lens, even "at normal distances" ... top left looks good ... at MTF50, f/11 still has some steam left before "hitting" diffraction.
Diffraction does not seem to be a wall we cannot pass. Sure, it will become more and more of a problem with ultra-high resolution photography in the future, and already is starting to become a slight issue today, but until we have 200 MP full-frame cameras I don't think diffraction is or will be a major, limiting problem. Even then we can get amazing landscape photos with a 500 MP full-frame camera if we just use a high-quality lens that produces truly amazing image quality at f2.8, and we are already starting to see such lenses released. I believe my 135mm f1.8 Art is such a lens, though I'm sure some lenses in the future will be even better, offering a nice step up in resolution from f4 to f2.8 if you have a sensor that can take advantage of that.
--
Scott Barton Kennelly
Big Print Photos - Photography by Scott Barton Kennelly
Landscape and nature photography, photographic prints by Scott Barton Kennelly, an Australian photographer living in south Florida.
www.bigprintphotos.com
I'm wondering if we have two different interpretations of "diffraction-limited"?There's this thing called "Moore's Law" . . . which states that processor speed will double every 18 months (but I think they've adjusted it to every two years now). In a few years the processors will be so fast and powerful that processing gigapixel images will be just as fast as processing 50 MP images today, and Sony's A1 can process 50 MP images quite quickly. In fact they've been processing 24 MP images plenty fast for many years now (i.e. in the 24 MP Sony A65 that I have the images process really really fast - so fast that the camera can shoot 10 fps, and right after I shoot a series at 10 fps if I want to review them it only takes a couple of seconds for me to start reviewing the shots I just took). I got my Sony A65 used long before I bought my Sigma SD1 Merrill, and that was years and years ago. I'm sure that today's processors can handle 100 MP photos (and probably even 200 MP photo) quite quickly. I haven't heard of anyone complaining that their Samsung S23 Ultra takes forever to process those 200 MP photo, and I'm sure there is a whole lot of processing going on with those photos.And? I don't want a 200MP RAW file if that means I get tons of noise. The 200MP images those phones produce are created with seveal images atcked onto each other to alleviate the noise issue. I doubt any modern camera is doing this.
And besides, I don't think I'd ever need or want 200MP images.
And it would have the slowest readout speed ever seen in a camera, would have enormous files requiring a NASA computer to edit them, the worst part still being the fact that image quality wouldn't be that great : most lenses can't resolve a sensor like that, so you'd be throwing most of your resolution in the bin, and then there is the issue of noise which is horrendously bad on those photosites.Imagine a full-frame sensor with photosites as small as those? It would probably be 2 gigapixels or something crazy like that.
No thank you, I'm perfectly fine with the cameras we alreayd have today.
No, we're not diffraction-limited on lenses at 61 MP (maybe with some crappy old lenses, but for good lenses - no way). I have a sixty year old lens made in Russia, which I mount on my Fuji GFX100 sometimes, and it produces excellent results at [f/8].My understanding is that we are *already* [diffraction-limited] on a lot of lenses on 61MP full frame sensors.I'd like to see a camera like that. Even with the 65mm f2 at 4 the sensor would capture all the detail the lens can produce.
DavidMillier
Forum Pro
Scott
Just a note that "poor" lenses don't cause or suffer especially from diffraction effects, they have limited resolution because they suffer from aberrations. Diffraction exists because of the physics of light. If we had perfect lenses with no aberrations they would be limited only by diffraction. "Crappy" lenses don't suffer from diffraction any worse than fine lenses.
Stopping down often improves aberrations, and here most lenses get sharper. However, diffraction gets worse the more you stop down. With most practical lenses it is a trade off - the photographer chooses an aperture that has the right balance between suppressing aberrations while avoiding excessive diffraction softening. High end modern lenses are getting good at being aberration free at wider apertures, providing a bit more space from the diffraction softening apertures.
Just a note that "poor" lenses don't cause or suffer especially from diffraction effects, they have limited resolution because they suffer from aberrations. Diffraction exists because of the physics of light. If we had perfect lenses with no aberrations they would be limited only by diffraction. "Crappy" lenses don't suffer from diffraction any worse than fine lenses.
Stopping down often improves aberrations, and here most lenses get sharper. However, diffraction gets worse the more you stop down. With most practical lenses it is a trade off - the photographer chooses an aperture that has the right balance between suppressing aberrations while avoiding excessive diffraction softening. High end modern lenses are getting good at being aberration free at wider apertures, providing a bit more space from the diffraction softening apertures.
I wish you hadn't said that, Scott! It is misleading and totally incorrect. Go tell it to Rayleigh, Abbe, Airy, et al. The degree of diffraction is a physical limit (i.e. a wall) that has nothing to do with image sensors at all!All I know is that I stepped up from 12 MP to 36 MP over the years in the full-frame world, and all I get is more detail in my photos. The 61 MP cameras (Sony A7r IV, Sigma fp L, etc.) seem to get even more detail. We now have even better lenses that we can buy, such as the Sigma 65mm f2 DG DN C and Sigma's amazing 105mm f2.8 DG DN Art, as well as other lenses from the likes of Panasonic and Leica, which offer crazy-high image quality at f5.6, and even more resolution at f4. Image quality has indeed improved significantly since the days of 12 MP full-frame (Canon 5 D and Nikon D3 and D700). It will get even better in the years to come.Makes me wonder why this was published long ago for some normal lenses:
Courtesy of Bob Atkins https://bobatkins.com/photography/technical/mtf/mtf2.html
"Ideal MTF" is the diffraction-limited line for all lenses. 'S' is a normalized frequency so that a lenses f-numbers can be compared in the same graph. Bottom right would not be my favorite lens, even "at normal distances" ... top left looks good ... at MTF50, f/11 still has some steam left before "hitting" diffraction.
Diffraction does not seem to be a wall we cannot pass. <speculation clipped>
--
What you got is not what you saw.
Last edited:
shouldn't you be in precious metals by now? surely you don't expect the markets to stay intact in the next few years...I'm old, invalid, can't hardly operate a camera and the Jeep and the stock market is eating my breakfast. Doesn't take much to get me going ...Working on it - Jim Kasson wrote an article, though the memory isn’t totally sharp.I knew that - should have written "is said to set in".Diffraction doesn’t “set in”. <>I see. I am reminded of lens tests where diffraction "sets in" above a certain f-number irrespective of the sensor resolution.Yes, I am familiar with this “kidding”. The smiling face was a major clue.He was kidding.Beyond this you are not going to get any more detail, due to diffraction.Why stop there?
On the other hand, some might find it interesting that the 600 MP range is the threshold above which more sensor resolution won’t improve image quality.
That would work for me. Got a number for that threshold "due to diffraction"?
Why are you so snippy, by the way?
regardless, I hope you are able to find and retain inner peace regardless of the external issues.
I hold physical gold and silver.shouldn't you be in precious metals by now?I'm old, invalid, can't hardly operate a camera and the Jeep and the stock market is eating my breakfast. Doesn't take much to get me going ...Working on it - Jim Kasson wrote an article, though the memory isn’t totally sharp.I knew that - should have written "is said to set in".Diffraction doesn’t “set in”. <>I see. I am reminded of lens tests where diffraction "sets in" above a certain f-number irrespective of the sensor resolution.Yes, I am familiar with this “kidding”. The smiling face was a major clue.He was kidding.Beyond this you are not going to get any more detail, due to diffraction.Why stop there?
On the other hand, some might find it interesting that the 600 MP range is the threshold above which more sensor resolution won’t improve image quality.
That would work for me. Got a number for that threshold "due to diffraction"?
Why are you so snippy, by the way?
Ouch ...
Thank you!
TN Args
Forum Pro
Hi Ted,I wish you hadn't said that, Scott! It is misleading and totally incorrect. Go tell it to Rayleigh, Abbe, Airy, et al. The degree of diffraction is a physical limit (i.e. a wall) that has nothing to do with image sensors at all!All I know is that I stepped up from 12 MP to 36 MP over the years in the full-frame world, and all I get is more detail in my photos. The 61 MP cameras (Sony A7r IV, Sigma fp L, etc.) seem to get even more detail. We now have even better lenses that we can buy, such as the Sigma 65mm f2 DG DN C and Sigma's amazing 105mm f2.8 DG DN Art, as well as other lenses from the likes of Panasonic and Leica, which offer crazy-high image quality at f5.6, and even more resolution at f4. Image quality has indeed improved significantly since the days of 12 MP full-frame (Canon 5 D and Nikon D3 and D700). It will get even better in the years to come.Makes me wonder why this was published long ago for some normal lenses:
Courtesy of Bob Atkins https://bobatkins.com/photography/technical/mtf/mtf2.html
"Ideal MTF" is the diffraction-limited line for all lenses. 'S' is a normalized frequency so that a lenses f-numbers can be compared in the same graph. Bottom right would not be my favorite lens, even "at normal distances" ... top left looks good ... at MTF50, f/11 still has some steam left before "hitting" diffraction.
Diffraction does not seem to be a wall we cannot pass. <speculation clipped>
I think Scott is referring to this phenomenon:-
Diffraction-limited to 3400 lw/ph at ƒ/11 on a 24 MP Nikon D3X full frame camera
Diffraction-limited to 4007 lw/ph at ƒ/11on a 45 MP Nikon D850 full frame camera
So, the same lens, on a higher resolution camera, achieves more resolution than it could on a lower-resolution camera, even at apertures like ƒ/11 where the resolution is clearly diffraction-limited.
Sources: http://www.photozone.de/nikon_ff/631-nikkorafs5018ff?start=1 and http://www.photozone.de/nikon_ff/1110-nikkorafs5018fx?start=1
cheers
Interesting references, thank you. I did note that the numbers are lw/ph, not lp/ph, and are for MTF50.Hi Ted,I wish you hadn't said that, Scott! It is misleading and totally incorrect. Go tell it to Rayleigh, Abbe, Airy, et al. The degree of diffraction is a physical limit (i.e. a wall) that has nothing to do with image sensors at all!All I know is that I stepped up from 12 MP to 36 MP over the years in the full-frame world, and all I get is more detail in my photos. The 61 MP cameras (Sony A7r IV, Sigma fp L, etc.) seem to get even more detail. We now have even better lenses that we can buy, such as the Sigma 65mm f2 DG DN C and Sigma's amazing 105mm f2.8 DG DN Art, as well as other lenses from the likes of Panasonic and Leica, which offer crazy-high image quality at f5.6, and even more resolution at f4. Image quality has indeed improved significantly since the days of 12 MP full-frame (Canon 5 D and Nikon D3 and D700). It will get even better in the years to come.Makes me wonder why this was published long ago for some normal lenses:
Courtesy of Bob Atkins https://bobatkins.com/photography/technical/mtf/mtf2.html
"Ideal MTF" is the diffraction-limited line for all lenses. 'S' is a normalized frequency so that a lenses f-numbers can be compared in the same graph. Bottom right would not be my favorite lens, even "at normal distances" ... top left looks good ... at MTF50, f/11 still has some steam left before "hitting" diffraction.
Diffraction does not seem to be a wall we cannot pass. <speculation clipped>
I think Scott is referring to this phenomenon:-
Diffraction-limited to 3400 lw/ph at ƒ/11 on a 24 MP Nikon D3X full frame camera
Diffraction-limited to 4007 lw/ph at ƒ/11on a 45 MP Nikon D850 full frame camera
So, the same lens, on a higher resolution camera, achieves more resolution than it could on a lower-resolution camera, even at apertures like ƒ/11 where the resolution is clearly diffraction-limited.
Sources: http://www.photozone.de/nikon_ff/631-nikkorafs5018ff?start=1 and http://www.photozone.de/nikon_ff/1110-nikkorafs5018fx?start=1
cheers
I took Scott's post to mean that a lens or something could go beyond the diffraction-limit "wall".
For a perfect lens at f/11, green light, refractive index=1, MTF50 occurs at about 3150 lw/ph (full frame). That's the "wall" so maybe their test was not done with pure green light ...
http://www.mif.vu.lt/atpazinimas/dip/FIP/fip-Sampling.html
web.ncf.ca/jim/misc/resolution/mtf/
--
What you got is not what you saw.
Last edited:
Sorry to say that I am really struggling to accept the phenomenon.Interesting references, thank you. I did note that the numbers are lw/ph, not lp/ph, and are for MTF50.Hi Ted,I wish you hadn't said that, Scott! It is misleading and totally incorrect. Go tell it to Rayleigh, Abbe, Airy, et al. The degree of diffraction is a physical limit (i.e. a wall) that has nothing to do with image sensors at all!Diffraction does not seem to be a wall we cannot pass. <speculation clipped>Makes me wonder why this was published long ago for some normal lenses:
Courtesy of Bob Atkins https://bobatkins.com/photography/technical/mtf/mtf2.html
"Ideal MTF" is the diffraction-limited line for all lenses. 'S' is a normalized frequency so that a lenses f-numbers can be compared in the same graph. Bottom right would not be my favorite lens, even "at normal distances" ... top left looks good ... at MTF50, f/11 still has some steam left before "hitting" diffraction.
I think Scott is referring to this phenomenon:-
Diffraction-limited to 3400 lw/ph at ƒ/11 on a 24 MP Nikon D3X full frame camera
Diffraction-limited to 4007 lw/ph at ƒ/11on a 45 MP Nikon D850 full frame camera
So, the same lens, on a higher resolution camera, achieves more resolution than it could on a lower-resolution camera, even at apertures like ƒ/11 where the resolution is clearly diffraction-limited.
Sources: http://www.photozone.de/nikon_ff/631-nikkorafs5018ff?start=1 and http://www.photozone.de/nikon_ff/1110-nikkorafs5018fx?start=1
cheers
I took Scott's post to mean that a lens or something could go beyond the diffraction-limit "wall".
The examples don't jibe with for example Atkins' lens diagram which truly shows difraction-limited behavior for both a perfect and a reasonable lens:
Between "Perfect" f/4 and "Perfect" f/8 the latter's MTF50 frequency is about half that of the former.
But in the photozone.de examples, the percentage difference between f/4 and f/8 is quite small.
I calculate the "zero MTF" limit for f/11 due to diffraction as 164 lp/mm or 328 lw/mm.
https://chatgpt.com/share/9017880b-34f7-4c0c-9c40-ebc04c31df6f
For a 24mm high sensor, that's 7872 lw/ph and MTF50 occurs at 40% of that 3149 lw/ph, so where their 4007 lw/ph comes from, I have no idea.
Summat's up or summat ...
Last edited:
"I felt a great disturbance in storage, as if millions of disk drives suddenly cried out in terror and were suddenly filled."
Similar threads
Keyboard shortcuts
- f
- Forum