A (very rare) recommendation for a DPR thread, on fundamental image quality.

[Bart_van_der_Wolf]

Hi folks,

It's a rare occasion that I'll recommend a read of a thread on an other (especially the DPR) forum. However, this thread does currently offer a nice view/development (so far) on a subject, diffraction, that has had my attention for quite a while already.

Diffraction is predominantly caused by our aperture settings. When light is forced down a small physical restraint/aperture, the light will 'diffract' around the edges of the aperture blades (especially noticable if the blades do not provide a perfectly circular edge).

For reasons that will be disclosed in future threads, I've (a while ago already) produced a chart that shows theoretical limitations for several camera sensor pitch dimensions:

The indications from this chart, translate quite well to some of the findings of the referenced article (within that article's margins of visual judgement accuracy), as they do to my personal observations on various sensor/lens/sensor-array combinations.

My basic premise is that when the diffraction blur diameter (first minimum (= zero) of the 'Airy disc' pattern) for the (most important for visual acuity) green wavelengths exceeds 1.5x the sensel pitch, there will be a visually significant impact on resolution. One could quibble if the lower boundary is at 1.5x or 2x the diffraction diameter, but that also depends on the Anti-aliasing filter used in the specific sensor designs at hand.

Bart

 

[Emil Martinec]

Thanks for the chart, Bart.

I was spurred by a related DPR thread

http://forums.dpreview.com/forums/read.asp?forum=1032&message=26504942

to do a few calculations. I haven't yet done the calculation for a round aperture, as our lenses have (approximately, sue to the aperture blades); the math is much easier for a square one. For a square aperture, MTF as a function of spatial frequency linearly decreases with aperture until the size of the diffraction spot equals the spatial wavelength, at which point MTF is zero thereafter. I expect the result for a round aperture will be similar.

So if you want maximum fine detail in the focal plane, even in the "green zone" of your chart, it's best to keep the aperture at the minimum value needed for DOF. I haven't yet done the calculation for the tradeoff between diffraction and DOF for the best MTF near the focal plane. Maybe later.

 

[Bart_van_der_Wolf]

Thanks for the chart, Bart.

I was spurred by a related DPR thread

http://forums.dpreview.com/forums/read.asp?forum=1032&message=26504942

to do a few calculations. I haven't yet done the calculation for a round aperture, as our lenses have (approximately, sue to the aperture blades); the math is much easier for a square one. For a square aperture, MTF as a function of spatial frequency linearly decreases with aperture until the size of the diffraction spot equals the spatial wavelength, at which point MTF is zero thereafter. I expect the result for a round aperture will be similar.

I haven't gone as far as calculating the MTF, my chart is just based on my empirical judgement that visual degradation seems to set in at approx. 1.5x the sensel pitch. This is also often supported by an observed declining MTF score at higher spatial frequencies when using narrower apertures.

An MTF degradation can be used to calculate a lowest limit of usefulness, and certainly a Sinc function is easier to use than an Airy pattern on an assumed square (sensel) aperture. I don't think we have to draw the line at a zero response, because at 10% MTF we're already pretty much at the practical limiting visual resolution (for average contrast subjects). The ISO also mentions a good correlation between 10% MTF and limiting visual resolution.

So if you want maximum fine detail in the focal plane, even in the "green zone" of your chart, it's best to keep the aperture at the minimum value needed for DOF. I haven't yet done the calculation for the tradeoff between diffraction and DOF for the best MTF near the focal plane. Maybe later.

I could also run a simulation in Mathematica (which I use to calculate a 2-D diffraction pattern PSF kernel), but it would still give a theoretical limit, because of the unknown properties of the AA-filter and residual optical aberrations. That's why I use the rule of thumb of 'sensel widths' which is useful enough to predict the visual on-set of diffraction degradation. Any narrower aperture will degrade the per pixel resolution, and reduce the chance of successful deconvolution to restore the losses.

Bart

 

[Emil Martinec]

I haven't gone as far as calculating the MTF, my chart is just based on my empirical judgement that visual degradation seems to set in at approx. 1.5x the sensel pitch. This is also often supported by an observed declining MTF score at higher spatial frequencies when using narrower apertures.

And that judgment should be pretty accurate. Many demosaicing algorithms use the green channel as an ersatz luminance channel for the purposes of interpolation, and the green pixels are spaced at sqrt2~1.4 times the pixel pitch, so when the diffraction spot exceeds this scale then diagonally nearby green pixels are blurred together.

An MTF degradation can be used to calculate a lowest limit of usefulness, and certainly a Sinc function is easier to use than an Airy pattern on an assumed square (sensel) aperture. I don't think we have to draw the line at a zero response, because at 10% MTF we're already pretty much at the practical limiting visual resolution (for average contrast subjects). The ISO also mentions a good correlation between 10% MTF and limiting visual resolution.

Right. I wasn't using a square sensel aperture, I had in mind a square lens aperture (usual "spherical cow" approximation we physicists use to simplify calculations). Agreed we don't have to draw the line at zero, but I was wanting to calculate how much the finite aperture degrades MTF for any aperture, even above the diffraction limit. I was surprised to find a linear relation, basically for an intensity pattern
I = I_0 cos^2 kx
and diffraction pattern sin^2ax/ax, the output of averaging the intensity pattern against the diffraction pattern is

I_diff = I_0 (1-a/k)
(for a<k), and
I_diff = 0
(for a>k)

I was merely expressing my surprise that the result was linear in the parameter _a_ which is proportional to the f-number. So MTF is decreasing directly with aperture even above the point where pixel-level resolution is diffraction limited (granted, not much for small _k_ but the amount of degradation doubles every time we stop down two stops).

I could also run a simulation in Mathematica (which I use to calculate a 2-D diffraction pattern PSF kernel), but it would still give a theoretical limit, because of the unknown properties of the AA-filter and residual optical aberrations. That's why I use the rule of thumb of 'sensel widths' which is useful enough to predict the visual on-set of diffraction degradation. Any narrower aperture will degrade the per pixel resolution, and reduce the chance of successful deconvolution to restore the losses.

Bart

Yes, doing the full integral against an Airy pattern is probably overkill. But it represents an upper bound on MTF. The MTF due to the AA filter simply multiplies that due to diffraction. Optical aberrations and diffraction effects might not cleanly factorize in this way, I'm not sure, since they arise from the same part of the optical pathway (the lens), in any case can only decrease overall MTF. In any case, these effects are present independent of diffraction and the influence of diffraction is I think reasonably modelled by the simple 1d calculation.

 

[Bart_van_der_Wolf]

Yes, doing the full integral against an Airy pattern is probably overkill. But it represents an upper bound on MTF.

Agreed, there's not much going to 'contribute' to the image at, or beyond, zero modulation.

The MTF due to the AA filter simply multiplies that due to diffraction. Optical aberrations and diffraction effects might not cleanly factorize in this way, I'm not sure, since they arise from the same part of the optical pathway (the lens), ...

FWIW, I do wonder, when given the point-spread function (PSF) kernel of the total optical chain, if we divide out the diffraction PSF kernel (for a given wavelength pass-band), wouldn't we be left with the lens aberration + AA-filter + sensel geometry effects for the specific aperture? And since the AA-filter and the sensel geometry are constant, and the lens aberrations are aperture dependent, wouldn't it be possible to derive a good model for the different PSFs involved? With a good model it wouldn't be too difficult to synthesize a combined PSF which would allow decent deconvolution (without the need for empirical camera/lens/aperture specific PSF determinations).

In any case, these effects are present independent of diffraction and the influence of diffraction is I think reasonably modelled by the simple 1d calculation.

While I agree it's adequate for general conclusions, it doesn't provide enough guidance for corrective measures, IMHO.

Bart

 

[Asher Kelman]

I think this is a thread that is worthy to keep in mind when getting enthusiastic about the increased resolution possible with higher MP counts.

As the sensel pitch decreases the effects of diffraction become more increasingly important at smaller apertures.

This would argue for going to a larger camera format when one seeks to use a higher MP count for the image plane.

Asher

 

[Asher Kelman]

Thanks so much Bart and Emil for you impressive contributions to our understandings of how some limitations of different sensors, apertures and lenses might be overcome by choice or post processing mathematics.

I feel that there is so much more here to learn from, benefit and leverage.

A. What practical lessons are there in terms of our

1. Choice of sensor size and it’s resolution
2. Aperture
3. Deconvolution methods
4. Sharpening
5. Other

B. Then, what have we actually achieved that improves our work for its particular purpose?

C. What shortcuts get us to ~80% of that full benefit?

D. Can that shortfall be recognized as we stare at pictures for a minute or two, (or pass by), in a gallery or a book?

Asher

 

[Doug Kerr]

Hi, Asher,

I think this is a thread that is worthy to keep in mind when getting enthusiastic about the increased resolution possible with higher MP counts.

As the sensel pitch decreases the effects of diffraction become more increasingly important at smaller apertures.

This would argue for going to a larger camera format when one seeks to use a higher MP count for the image plane.

To review what brings that about:

For any wavelength we care to work at, and assuming focus at infinity, the "diameter" of the Airy disk (the blur figure created by diffraction) is proportional to the f-number. This holds as we change focal length for whatever reason.

It is not unreasonable to "score" the impact of diffraction on degrading the resolution potential of a camera in terms of the "diameter" of the Airy disk compared to the pixel pitch. (Indeed, earelier in this thread we have a more sophisticated look into this in terms of MTF, but here at the kitchen table . . .)

Thus, if we wish to hold to a certain impact of diffraction (in the sense I just stated) as we increase the f-number at which we wish to work, we must increase the pixel pitch. And if we wish to hold to a certain pixel count (to still attain a certain "ideal" resolution in terms of pixels per image height), we must go to a larger format size.

We can of course consider the "tradeoffs" among the various properties in different ways. For example, if we wish to consider a certain operating f-number, but wish to increase the pixel count, then to attain the same impact of diffraction on the resolution potential of the camera, we must still hold to the same pixel pitch. and accordingly must increase the format size.

Best regards,

Doug

 

[Jerome Marot]

This is quite an old thread. I think that we discussed this subject in more recent threads. In a nutshell, the conclusions where:

• for reasonable resolutions (e.g. 24MP or so) and sensor sizes (1" to 24x36), diffractions starts to be just noticeable between f/5.6 and f/11
• the combination of diffraction and depth of field implies that there is an upper limit in resolution if we want to image tridimensional subjects at finite distances
• if we increase sensor size, we need to close the aperture to get usable depth of field in practice
• therefore that limit is somewhat independent of sensor size
• for typical photographic practice, that limit is around 100-300MP
• this is also roughly what we get with large format film at practical apertures for that "sensor" size
• this is also roughly what we need for a print at human size (about 1.5m large) to appear sharp with average human vision when examined close.

The limit does not concern subjects far away (they don't need depth of field), like landscape or astronomy. For tridimensional subjects, there are programatic ways to circumvent the limit (e.g. focus stacking or 3D imagers).

 

[Asher Kelman]

Jerome,

I personally very much appreciate your work in summarizing discussion findings.

I practice, my Sony 7D IIR seemed to achieve the impact of an 80 MP Phase One MF. I wonder whether the “impression” might relate to your description of 3D feeling of images near to us?

Still, when getting to 10” away from the print, one can recognize that hair is better define with 80 MP as opposed to an
Bout 30MP.

I am not at all certain how I can distinguish between

• the more realistic effect of the 3D sense of a near object from focal length, aperture and sensor/film size
• similar effects by changing the S-curve in the histogram
• altering/standardization of the distribution of colors in different make sensors.

I have upgraded to the Fuji 50 MP MF camera and believe in am more likely to get more impressive detail-rich dimensional images of studio subjects.

Down the road I will try to do direct controlled comparison with the Sony A7RII but for now, I am acting on faith and impressed with the result.

Part of my euphoria could be self-delusion since the new camera was expensive and I am paying for it monthly!

I could do a side by side test using one great lens, like the Canon 70-200 2.8L IS on both.

Any ideas would be appreciated.

Asher

 

[Jerome Marot]

I see. When comparing high megapixel cameras, one has to realise that mpix numbers are linear, while our perception increases like a logarithmic series. That means: there is no noticeable difference between 43 and 50 mpix, one needs to double the resolution for the increase to matter. There should be a noticeable difference between 43 and 80 mpix.

When one print big enough, of course... You would need at least an A0 print to see the difference (about 80x120cm or 33"x47").

As to the difference in rendering between different cameras, sensor size has less to do with it than lens design (and also some technical choices in post-processing). We have had a thread about sensor sizes, the part I wrote about lenses are here:

https://openphotographyforums.com/forums/threads/size-matters-but-just-how.17642/#post-146256

https://openphotographyforums.com/forums/threads/size-matters-but-just-how.17642/page-2#post-146350

Fuji designed glorious lenses for their GFX series. Some Sony lenses are also excellent, but you have to be a bot more choosy. What lenses do you have for your A7RII?

 

[Asher Kelman]

Jerome,

I am not Catholic, so I have never gone to confession and my friends who confessed to their wife got kicked out of the house!

But I do feel one can trick oneself in buying better cameras than justified by practical use of the actual prints. That is why I confessed my disquiet in acquiring the GFX and its 32 to 64 mm zoom lens, reportedly stellar!

The fact was that my youngest son “needed” my A7RII as his camera, as the Canon 5DII was, of course passé to his pals!

So, I leveraged that to justify debt, (of over $7,000), to acquire the GFX. My print size is Ao, (the size you suggested), or 10% larger.

I have yet to print anything from the GFX, but now that my eldest son has the A7RIII, I feel very motivated to do some comparisons in the studio. One can use:

• the best lens for that camera or
• adapters and one common lens.

I have G lenses available for the Sony cameras or we can use a good Canon or Sigma art lens in common.

I would love to discover the Fuji does better!

The test would be

• “Visual impact to commanding dimensional existence in space” or something like that
• “Detail-rich color rendering” of a model draped in a beautiful fabric.

I do note that they now offer Fuji GFX version at 100 MP! But I doubt I deserve it! If I could readily see an improvement in the mental impact, of course I would be tempted sorely!

My birthday is soon!

Asher

 

[Doug Kerr]

Hi, Asher,

I feel very motivated to do some comparisons in the studio. One can use:

• the best lens for that camera or
• adapters and one common lens.

I have G lenses available for the Sony cameras or we can use a good Canon or Sigma art lens in common.

I would love to discover the Fuji does better!

The test would be

• “Visual impact to commanding dimensional existence in space” or something like that
• “Detail-rich color rendering” of a model draped in a beautiful fabric.

I for one look forward to the results of this initiative.

I do note that they now offer Fuji GFX version at 100 MP! But I doubt I deserve it!

Well, you certainly deserve it!

My birthday is soon!

Le mien aussi!

Best regards,

Doug

 

[Doug Kerr]

Hi, Jerome,

I see. When comparing high megapixel cameras, one has to realise that mpix numbers are linear, while our perception increases like a logarithmic series. That means: there is no noticeable difference between 43 and 50 mpix, one needs to double the resolution for the increase to matter. There should be a noticeable difference between 43 and 80 mpix.

And note that we have to quadruple the pixel count to double the (prospective) resolution!

Doug

 

[Asher Kelman]

Hi, Jerome,



And note that we have to quadruple the pixel count to double the (prospective) resolution!

Doug

Yes, Doug, that’s true. But for my studio, to get the best Ao or larger print with immediate impact, it’s not the hair that wins gears immediately. Yes, it can be recognized, but that’s later!

But then Jerome directed me to think not just in terms of detail, “character” and color rich resolution, but perhaps more so the 3D “presences” of “fairly near” objects. That’s what’s needed in portraits, persons and pears!

Asher

 

[Doug Kerr]

Hi, Asher,

But then Jerome directed me to think not just in terms of detail, “character” and color rich resolution, but perhaps more so the 3D “presences” of “fairly near” objects. That’s what’s needed in portraits, persons and pears!

Yes, I think that is (to me) a very important new outlook. I'm grateful to Jerome for articulating it.

Best regards,

Doug

 

[Jerome Marot]

The fact was that my youngest son “needed” my A7RII as his camera, as the Canon 5DII was, of course passé to his pals!

So, I leveraged that to justify debt, (of over $7,000), to acquire the GFX. My print size is Ao, (the size you suggested), or 10% larger.

There is no need to justify yourself, especially on a photography forum. I have far too many cameras and lenses myself, but then I know people who collect cars...

I have yet to print anything from the GFX, but now that my eldest son has the A7RIII, I feel very motivated to do some comparisons in the studio. One can use:

• the best lens for that camera or
• adapters and one common lens.

My idea would rather be to compare camera systems as they are used in practice, that is with the lenses that one is likely to use: the lenses of that system. And using a lens designed for a 24x36 sensor on a MF sensor is a source of frustration.

I have G lenses available for the Sony cameras

Which exact lenses?

I do note that they now offer Fuji GFX version at 100 MP!

Not yet and probably not in time for your birthday.

 

[Asher Kelman]

Jerome,

You asked which G lenses?

I have written to my son but here’s what I know so far:

For sure!

Yes, got this from me! Not a G lens but stellar, just the same!

Yes, he has this one, its superb!!

Also the Sony 24-105 F4

But we would need a Canon Mount lens for testing on both Sony and GFX bodies!


Asher

​

 

[Jerome Marot]

Note of these lenses is a portrait lens. Still: try to compare your fuji zoom with the 24 and 50mm and see how it goes.

Using a Canon lens for comparison is not what I would do.

 

[Asher Kelman]

Jerome,

The 24 mm would be a close match for the 32 mm end of the Fuji zoom, as there is a ~ 0.8 factor to be added to get to 25.6mm equivalent.

What about using a manual Leica 50mm Summicron for both?

Asher

 

[Jerome Marot]

If you use the same lens on the two cameras, the angle of view will be different and the resulting pictures will not be comparable.

 

[Asher Kelman]

Thanks, that’s a good point.

So we need the same quality!

....and the same angle! The 55mm Sony and the 32-64 mm set to the same angle would be a start.

Asher