Posts: 1,597
Threads: 22
Joined: Apr 2010
Reputation:
8
Quote:This is incorrect.
As we can see from the charts here at PZ and pretty much everywhere else, diffraction hits well before these aperture values.
A good example is the best MFT lens tested here in the zone, the Oly 75mm f1.8: http://www.opticallimits.com/m43/830-oly75f18?start=1
From f5.6 onwards, the resolution is hit by diffraction already.
Keep in mind that the higher the sensor resolution, the sooner it will be hit by diffraction.
Please read my entire post again.
Diffraction is always at work, because it is a function of the fact that light bends at edges. In the end it is a matter of determining, in combination with lens aperture and sensor limits, what works best. Diffraction only does not limit a lens at F/1 or bigger apertures; limits at larger apertures are always lens limits, i.e., resolution limits caused by lens flaws etc.
This whole thing about "diffraction hits at..." is purely based on the way diffraction influences sytem resolution after a lens hits it highest resolution, as it can only go down thereafter.
The problem lies with the fact that people tend to use a shortcut way of saying how diffraction influences the final result. it would be better to say that a particular lens reaches it highest resolution, in combination with diffraction effects, at such and such an aperture, and thereafter it cannot resolve any higher lp/mm. However, that is a rather long and roundabout way to do so.
Also, the saying that a particular lens is diffraction limited at a specific aperture, or a bunch of apertures, only means that such a lens cannot resolve higher lp/mm only becuase it is so good that it runs into the diffraction limits, which is a physics constraint due to the properties of light, and in such cases not because of lens constraints.
And it is logical that the higher the resolution, the sooner diffraction limit are hit. My tables displayed above clearly indicate that. It si a natural phenomenon. Saying that a lens or system hits a sensor diffraction limit only means that teh diffraction limit for a specific aperture is reached, and based on the criteria you use for determining resolution, that si a fixed aperture, based on sensor resolution or nyquist frequency. It does not mean that a system is not good, or that it cannot do better than another system with a smaller "sensor diffraction limit".
All of the above can be calculated fairly easily, BTW. If you are interested, I'd suggest to look up Norman Koren's pages here on the internet. he explains things quite well.
HTH, kind regards, Wim
Gear: Canon EOS R with 3 primes and 2 zooms, 4 EF-R adapters, Canon EOS 5 (analog), 9 Canon EF primes, a lone Canon EF zoom, 2 extenders, 2 converters, tubes; Olympus OM-D 1 Mk II & Pen F with 12 primes, 6 zooms, and 3 Metabones EF-MFT adapters ....
Posts: 1,252
Threads: 112
Joined: Apr 2010
Reputation:
18
06-19-2016, 05:38 PM
(This post was last modified: 06-19-2016, 05:41 PM by thxbb12.)
Quote:Please read my entire post again.
Diffraction is always at work, because it is a function of the fact that light bends at edges. In the end it is a matter of determining, in combination with lens aperture and sensor limits, what works best. Diffraction only does not limit a lens at F/1 or bigger apertures; limits at larger apertures are always lens limits, i.e., resolution limits caused by lens flaws etc.
This whole thing about "diffraction hits at..." is purely based on the way diffraction influences sytem resolution after a lens hits it highest resolution, as it can only go down thereafter.
The problem lies with the fact that people tend to use a shortcut way of saying how diffraction influences the final result. it would be better to say that a particular lens reaches it highest resolution, in combination with diffraction effects, at such and such an aperture, and thereafter it cannot resolve any higher lp/mm. However, that is a rather long and roundabout way to do so.
Also, the saying that a particular lens is diffraction limited at a specific aperture, or a bunch of apertures, only means that such a lens cannot resolve higher lp/mm only becuase it is so good that it runs into the diffraction limits, which is a physics constraint due to the properties of light, and in such cases not because of lens constraints.
And it is logical that the higher the resolution, the sooner diffraction limit are hit. My tables displayed above clearly indicate that. It si a natural phenomenon. Saying that a lens or system hits a sensor diffraction limit only means that teh diffraction limit for a specific aperture is reached, and based on the criteria you use for determining resolution, that si a fixed aperture, based on sensor resolution or nyquist frequency. It does not mean that a system is not good, or that it cannot do better than another system with a smaller "sensor diffraction limit".
All of the above can be calculated fairly easily, BTW. If you are interested, I'd suggest to look up Norman Koren's pages here on the internet. he explains things quite well.
HTH, kind regards, Wim
Totally agree on what you say above. However, I still don't understand what you mean by this statement:
Quote:For MFT sensors at 16 MP and 20 MP these sensor diffraction limits are F/12 and F/11 respectively, but to get the most out of this you generally need to stop down a full aperture less, so F/9 and F/11.
F9 and F11 are well into diffraction territory with current (16 and 20MP) MFT sensors.
Posts: 2,724
Threads: 595
Joined: Apr 2010
Reputation:
15
Folks, I just came from a long day where I enjoyed shooting, I am now working the RAW files and happy to have very nice keepers.
This too long for me to read today, but seems interesting.
All I know is diffraction rises with aperture. The higher the sensor density the more it is visible.
Is there anything else I missed ?
Hope you had all a good weekend.
Posts: 1,597
Threads: 22
Joined: Apr 2010
Reputation:
8
Quote:Totally agree on what you say above. However, I still don't understand what you mean by this statement:
F9 and F11 are well into diffraction territory with current (16 and 20MP) MFT sensors.
The problem really is the word diffraction territory - there si no such thing.
A sensor is "diffraction limited" (essentially the wrong word, which is why I put it in quote marks), when the aperure used produces a resolution which by definition is less than the resolution of the sensor. Diffraction limits for a sensor in principle occur at teh Nyquist frequency, and if you compare diffraction limits for different apertures, you'll find that when using teh Rayleigh criterion, this limit occurs at around F/12 and F/11, and ideally, because of other factors, one should use a stop larger to get optimal results, i.e., F/9 and F/8 (made a typo in my earlier post; I'll correct it), or, even larger apertures, obviously.
And what do you mean that it is more visible the the higher the sensor density is?
You will see sharper jumps in resolution going from aperture to aperture with higher density sensors, than you do with lower density ones, but thatis not so much related to diffraction as to sensor resolution itself.
If you click on the pictures I attached, you will see that the differences between resolution figures at larger apertures between the different sensors is greater than at smaller ones. This means that the drop off in system resolution is steeper once you go beyond teh apertures where a lens is best. That is all there is to it.
When I get a chance, I'll draw some resolution graphs to show what I mean.
HTH, kind regards, Wim
Gear: Canon EOS R with 3 primes and 2 zooms, 4 EF-R adapters, Canon EOS 5 (analog), 9 Canon EF primes, a lone Canon EF zoom, 2 extenders, 2 converters, tubes; Olympus OM-D 1 Mk II & Pen F with 12 primes, 6 zooms, and 3 Metabones EF-MFT adapters ....
Posts: 1,597
Threads: 22
Joined: Apr 2010
Reputation:
8
Quote:Folks, I just came from a long day where I enjoyed shooting, I am now working the RAW files and happy to have very nice keepers.
This too long for me to read today, but seems interesting.
All I know is diffraction rises with aperture. The higher the sensor density the more it is visible.
Is there anything else I missed ?
Hope you had all a good weekend.
Hi toni-a,
It is only more visible because the graphs fall off more steeply with higher density sensors than with lower density sensors.
In addition, you may find that one hits maximum system resolution off-center with lenses sooner too, because lenses tend to have lower resolution off-axis then on-axis. The higher the resolution of the sensor (or film for that matter) the clearer this becomes.
HTH, kind regards, Wim
Gear: Canon EOS R with 3 primes and 2 zooms, 4 EF-R adapters, Canon EOS 5 (analog), 9 Canon EF primes, a lone Canon EF zoom, 2 extenders, 2 converters, tubes; Olympus OM-D 1 Mk II & Pen F with 12 primes, 6 zooms, and 3 Metabones EF-MFT adapters ....
Posts: 8,004
Threads: 1,851
Joined: Apr 2010
Reputation:
49
From luminous landscape:
https://luminous-landscape.com/wp-conten...TABLA3.jpg
Thus if we are using the fairly strict criterias, it's all downhill for MFT from f/5.6 onward (a bit earlier than that for longer wavelengths (red) actually).
Now is that a real problem ? Well, why should you stop down to f/8 (aka "f/16") on MFT anyway (other than to reduce vignetting on a slow lens) ?
Ref:
https://luminous-landscape.com/do-sensor...ve-lenses/
Posts: 8,004
Threads: 1,851
Joined: Apr 2010
Reputation:
49
BTW, these are the official MTFs for the lens:
http://panasonic.jp/dc/lens/leica_dg_var...mage02.gif
(took a while to find them)
While such MTFs only show half of the truth, they are usually not totally off either.
In this case, it seems unlikely that there's a massive drop towards the edges especially considering the fact that this is a tele lens.
Posts: 1,597
Threads: 22
Joined: Apr 2010
Reputation:
8
06-20-2016, 12:09 AM
(This post was last modified: 06-20-2016, 12:29 AM by wim.)
Quote:From luminous landscape:
https://luminous-landscape.com/wp-conten...TABLA3.jpg
Thus if we are using the fairly strict criterias, it's all downhill for MFT from f/5.6 onward (a bit earlier than that for longer wavelengths (red) actually).
Now is that a real problem ? Well, why should you stop down to f/8 (aka "f/16") on MFT anyway (other than to reduce vignetting on a slow lens) ?
Ref:
https://luminous-landscape.com/do-sensor...ve-lenses/
No, it is never a problem, to be very honest.
Even a bad lens profits from a sensor with a higher pixel density, up to a point anyway, basically because the system will be able to image the picture it projects closer to its own resolution limit, whether we get to or over the diffraction limit at a certain aperture. Note that the only real limit for sensors with regard to diffraction limit is the Nyquist frequency, the limit for lenses is determined by quality of the glass, and aperture, and the latter is what determines the size of the Airy disks, in theory anyway.
As to using smaller apertures, as long as we stay at or above resolution a good amateur could get from comparative film, i.e., 30 lp/mm, or for a professional about double that, we are fine. And that equates to about 6 MP and 12 MP respectively. If somehow that was not the case, our pictures would be unprintable, or unviewable - and this is where the whole discussion on "sensor diffraction limits", and "sensors outresolving lenses" is completely flawed. The same could be said about film, and somehow we never really did - but then, that was mostly before the internet, and before pxelpeeping ![Wink Wink](https://forum.opticallimits.com/images/smilies/wink.png) .
Even with smaller apertures, or maybe not such good lenses or sensors, at lower resolution, in the end it is about getting a picture, or not getting a picture.
Kind regards, Wim
P.S.: For sake of reference, I used 1600 lp/mm as a reference point for diffraction, as Norman Koren indicates. In the LuLa article they use 1490, but that is also for a slightly longer wavelength ![Smile Smile](https://forum.opticallimits.com/images/smilies/smile.png) .
Gear: Canon EOS R with 3 primes and 2 zooms, 4 EF-R adapters, Canon EOS 5 (analog), 9 Canon EF primes, a lone Canon EF zoom, 2 extenders, 2 converters, tubes; Olympus OM-D 1 Mk II & Pen F with 12 primes, 6 zooms, and 3 Metabones EF-MFT adapters ....
Posts: 1,597
Threads: 22
Joined: Apr 2010
Reputation:
8
Quote:BTW, these are the official MTFs for the lens:
http://panasonic.jp/dc/lens/leica_dg_var...mage02.gif
(took a while to find them)
While such MTFs only show half of the truth, they are usually not totally off either.
In this case, it seems unlikely that there's a massive drop towards the edges especially considering the fact that this is a tele lens.
Actually, with telelenses there often is less of a drop-off than with wideangle lenses. What a lot of people do not seem to realize is that the laws which control non-mechanical vignetting, also control resolution to a DEGREE. Essentially the spread of light towards the edges caused by the AoV, not only cause vignetting by spreading the same light over a larger area, but also spreads resolution over that same area, with as result that one could go. e.g, from 100 lp/mm in the center, to say 50 lp/mm the corners just by that effect alone.
Because the AoV in telelenses is so low, especially in long ones, vignetting and the loss of resolution towards the edges is in principle less. This apart from the fact that telelenses are easier to correct for certain flaws than wideangle lenses. Vignetting in telelenses often is caused more by mechanical vignetting in the lens system rather than by optical vignetting.
What we currently see with high MP sensors like the latest FF cameras and with MFT cameras, is that we are seeing more or less actual resolution again of a lens over the entire width of the lens, rather than its limits caused by maximum sensor resolution and the artificial impact of analyzing systems (as compard to film).
Sorry, just thinking aloud here: maybe it could be worthwhile, or at least interesting, to calculate what the maximum edge resolution of a lens is based on AoV in relation to on-axis center resolution, and extrapolate from there how well a lens performs over the whole image. That would actually mean that it would be possible to establish a good quality scale for lenses and their measuring points at all the points used currently in lens tests, and hence make qualification of the results easier, like for the 5DsR ![Wink Wink](https://forum.opticallimits.com/images/smilies/wink.png) .
Hmm. I'll see if i can find anything on this subject.
Kind regards, Wim
Gear: Canon EOS R with 3 primes and 2 zooms, 4 EF-R adapters, Canon EOS 5 (analog), 9 Canon EF primes, a lone Canon EF zoom, 2 extenders, 2 converters, tubes; Olympus OM-D 1 Mk II & Pen F with 12 primes, 6 zooms, and 3 Metabones EF-MFT adapters ....
Posts: 1,340
Threads: 55
Joined: Apr 2010
Reputation:
0
Just catching up on the thread. I think perhaps a better way to think about diffraction is there is a point where it starts to visibly "soften" the image, and there is a later point where it is the final limiting factor. Photographers tend to refer to the former, whereas I think Wim is referring to the latter.
There are times where it is still advantageous to work well into diffraction limited region. Anyone tried planet imaging? My scope would be, in photographic terms, 2000mm f/10. With barlows (think of them as functioning like teleconverters) I also tried 4000mm f/20, and 6000mm f/30. Of the three, the best results were obtained around f/20, as I was into the noise of my sensor at f/30, and the subject was too small at f/10. The sensor had just over 4 micron pixel pitch if you want to work out an effective MP to compare with photographic sensors, but I think that was in the rough ball park as 18MP APS-C from memory.
<a class="bbc_url" href="http://snowporing.deviantart.com/">dA</a> Canon 7D2, 7D, 5D2, 600D, 450D, 300D IR modified, 1D, EF-S 10-18, 15-85, EF 35/2, 85/1.8, 135/2, 70-300L, 100-400L, MP-E65, Zeiss 2/50, Sigma 150 macro, 120-300/2.8, Samyang 8mm fisheye, Olympus E-P1, Panasonic 20/1.7, Sony HX9V, Fuji X100.
|