Sharpness

[Abitconfused]

Sharpness in a photographic work: Excelling at the display of fine detail. The ability to highly resolve minute elements of an image. Tending toward greater contrast, distinct color separation, and the discernment of the edges of the elements of an image. Producing an image of high dynamic range, striking resolution and superior micro-contrast.

I recently read an article where the author criticized DXO methodology for determining sharpness. While I applaud an investigative mind, I thought it odd that the author went to a dictionary to define the term sharpness. If you can think, write, record and communicate, you are a scribe. Define your own terms.

A sharp image also has an almost indescribable element of visual appeal. It resonates.

I recently returned two lenses that produced dull images. Photoshop could not save them. Oddly, many many people gave these lenses four and five star ratings. What are they thinking? All my lenses, such as the Sigma 50mm Art, that are highly rated by DXO are truly sharp and produce images of enviable crispness.

Art is also a drive toward perfection. Resist bad lenses! Return them!

[Shadowman]

Sharpness in a photographic work: Excelling at the display of fine detail. The ability to highly resolve minute elements of an image. Tending toward greater contrast, distinct color separation, and the discernment of the edges of the elements of an image. Producing an image of high dynamic range, striking resolution and superior micro-contrast.

I recently read an article where the author criticized DXO methodology for determining sharpness. While I applaud an investigative mind, I thought it odd that the author went to a dictionary to define the term sharpness. If you can think, write, record and communicate, you are a scribe. Define your own terms.

I'm only speculating but the author was probably going for a universal definition, was it possible to connect with the author, erhaps you could challenge him on using an academic rather than aesthetic definition?

[/QUOTE]A sharp image also has an almost indescribable element of visual appeal. It resonates.

I recently returned two lenses that produced dull images. Photoshop could not save them. Oddly, many many people gave these lenses four and five star ratings. What are they thinking? All my lenses, such as the Sigma 50mm Art, that are highly rated by DXO are truly sharp and produce images of enviable crispness.

Art is also a drive toward perfection. Resist bad lenses! Return them![/QUOTE]

How were you using the lens and how do others use the lens, I've been tempting by the "Art" description of the lens but doubt the glass could offer more than I'm able to achieve with a few of my already owned lenses, I've read descriptions of creamy focus, sparkling bokeh or some other term but with the right lighting I can achieve similar results, problem is that special lighting is sometimes fleeting and I don't always have that special lens in my bag. Hope the quotes came out as intended.

[rpcrowe]

IMO: the "sharpness" of any image is dependent on other factors besides the acuity of the lens that you are using. I am not talking about the shots of a test target but, rather of the final image product...

A few, but not all of these factors are the inherent contrast of your subject. As an example (but certainly not the only example) we have recently experienced some rain in my area of Southern California. The precipitation has cleared the atmosphere and a landscape image shot would "look" a lot sharper than an image shot a few days before when there was some smog or other particulates in the atmosphere.

Lighting also will give an impression of sharpness, flat lighting usually does not look as sharp as an image with contrasty lighting.

The subject chosen will also give an impression of sharpness or lack of sharpness.

Of course, the method you use to shoot your image will alter the impression of sharpness. This has been proven with inexpensive kit lenses which can produce very sharp looking images when the camera/lens is tripod mounted and the image is shot around f/8 to f/11...

This is definitely not to say that lenses are all equal. Nothing could be further from the truth. Often the price of a lens can indicate the relative sharpness except when you are using a very wide aperture lens wide open. Then, a less expensive lens stopped down might give the impression of producing sharper images...

[xpatUSA]

Sharpness in a photographic work: Excelling at the display of fine detail. The ability to highly resolve minute elements of an image. Tending toward greater contrast, distinct color separation, and the discernment of the edges of the elements of an image. Producing an image of high dynamic range, striking resolution and superior micro-contrast.

Sharpness and Contrast are different. An image can have perfectly sharp edges but low micro-contrast; vice-versa; both; or neither:

https://medium.com/ice-cream-geometr...t-f279c6a3eb31

As to sharpness, it is easily measured and quantified by the spread in pixels and fractions thereof of an edge, using such apps as Imatest or QuickMTF.

http://www.falklumo.com/lumolabs/art...ess/index.html

Somewhat less easily measured with vague terminology describing shrubbery or rolling hills ...

[Manfred M]

Could you post the link to the article in question? It would be an interesting read.

In general, when looking at lens sharpness has been the resolution of the lens, measured in lines / mm has been the historical norm in doing so. DxO Mark seems to have tried to update the old standby definition and I suspect that is the reason for the push-back. If their methodology becomes the new "gold standard" in understanding lens sharpness, then I will agree with their methodology, but as long as they are the only ones making this measurement, I would also tend to push back.

That being said, I have yet to see a modern lens that is not sharp enough to work with but if one is pixel peeping, then there could be an argument made. In virtually every case I have seen complaints about lack of sharpness, the problem was not the lens, but rather the technique used by the photographer. Focus issues, inappropriate camera settings and motion blur tended to be the root cause, not a lens design or manufacturing defect. Unless the less is tested on a heavy duty tripod, in a vibration and wind free environment (i.e. a lab setting), I find it difficult to accept that the lens is not sharp. Sigma, with their Art series is marketing sharp lenses, but frankly sharpness is only one characteristic we need to look at when evaluating a lens. The reason so many people are snapping up old lenses is because of some of the characteristics outside of sharpness that make for a great image; modern lenses are sharp but produce boring results much of the time.

[DanK]

DXO says this:

This new unit of measure replaces the MTF scores that we have been publishing up until now. Perceptual Mpix weights MTF measurements with the sensitivity to detail of the human visual system, thus providing a score for the sharpness perceived by our eyes

However, I haven't yet found anything that describes how this "weighting" is done. If someone has a link, please post it.

I tend to agree with Manfred, although I think his statement about sharpness is stronger than I would have made. I think there are cases where a lack of sharpness from the lens does matter, particularly if one is printing large or cropping severely. However, I agree that it often doesn't matter. And sharpness isn't going to matter much if at all when people downsize their images to 1000 pixels or so to view them on a computer screen.

[xpatUSA]

DXO says: This new unit of measure replaces the MTF scores that we have been publishing up until now. Perceptual Mpix weights MTF measurements with the sensitivity to detail of the human visual system, thus providing a score for the sharpness perceived by our eyes

However, I haven't yet found anything that describes how this "weighting" is done. If someone has a link, please post it.

Sounds like 'SQF' to me, Dan, quite an old metric and certainly not unique to the much-revered DxO.

http://www.bobatkins.com/photography.../mtf/sqf1.html

http://www.imatest.com/docs/sqf/

The weighting is to do with human visual acuity.

I'm not a great fan of DxO's one-number-says-it-all metrics. I'd rather look at an MTF vs. lp/mm curve and make my own mind up. It's not like the human contrast-sensitivity function is a big secret and I'll bet that DxO does not even account for lighting, color, ethnicity etc.

I would have put DxO's Mpix page up but the link here is broken:

https://www.dxomark.com/About/Glossa...-S-T/Sharpness
.

[dje]

Yes I've also seen reference to the SQF concept in relation to the DXOMark weighting function. But I haven't seen any detail published by DXO themselves. They call this weighted figure acutance in their test result charts I think. And of course their results are for a camera system, ie body plus lens.

The other factor with a lens that is often overlooked is it's basic contrast (low frequency contrast). This is affected by veiling flare and is dependent on the lens elements and inner body coatings. Generally I've found primes to be better in this regard.

Dave

[Abitconfused]

I am sorry, I seem to have lost the link. I believe it came to my email along with info from about ten other photography sites that I subscribe to, these do not constantly pester with advertisments but truly offer interesting articles.

[Manfred M]

This is affected by veiling flare and is dependent on the lens elements and inner body coatings. Generally I've found primes to be better in this regard.

That is hardly surprising as one of the key contributors to this issue is the number of air / glass interfaces. Fixed focal length lenses tend to have fewer lens elements and hence fewer air / glass interfaces.

I seem to remember reading that more complex optical designs of lenses were not feasible until thin-film anti-reflective coatings hit the scene on a larger scale. Flare negated the advantages of having additional lens elements. Even though the process was discovered in 1935 by the Carl Zeiss Company, it did not have a significant impact on lens design until after WWII.

[Abitconfused]

I know that German binoculars were much better than anywhere else in the world during the war. I suspect tank sighting optics were similarly exceptional.

[xpatUSA]

I seem to remember reading that more complex optical designs of lenses were not feasible until thin-film anti-reflective coatings hit the scene on a larger scale. Flare negated the advantages of having additional lens elements. Even though the process was discovered in 1935 by the Carl Zeiss Company, it did not have a significant impact on lens design until after WWII.

Quite so, Manfred. After years of owning a pair of Russian-made binoculars (Zenit?, Zenith?) I splurged on a oldish used pair of Carl Zeiss Jena 'Octarem' 10x50mm. The difference was almost literally night and day as regards clarity and sharpness ...

[rpcrowe]

Zeiss Optics in WW-II German Tanks
https://www.quora.com/Just-how-super...sights-in-WWII

[Abitconfused]

In passing, I must give some credit to the author who went to a dictionary to define sharpness. It is difficult to succinctly define sharpness as a general concept but I will try without looking at a dictionary. Sharpness is the ability to render matter into less contiguous components.

[DanK]

Sharpness is the ability to render matter into less contiguous components.

I don't think so. I think it is the extent to which one can differentiate things that are contiguous. For example, a black letter on white paper has a border at which black is contiguous with white. If an image is sufficiently unsharp, you won't perceive that border, but it's still there: the white and black are contiguous whether we can perceive it or not.

[Abitconfused]

DanK, I think you are correct!

[GrahamS]

Sharpness and resolution - two totally different things......

[DanK]

Sharpness and resolution - two totally different things......

Not quite, if I understand the terms. Resolution contributes to sharpness. Correct me if I am wrong:

Resolution is a measure of the amount of detail that can be resolved by the equipment. this is usually measured in lines/mm.

Accutance refers to edge contrast. I just went searching for a technical definition and finally found a clear one in Wikipedia:

Acutance is related to the amplitude of the derivative of brightness with respect to space.

In other words, for people unfamiliar with calculus, if you plotted distance on the horizontal axis and brightness on the vertical, the steeper the slope of the curve, the higher the accutance.

Finally, if I understand correctly, sharpness has no single technical definition; it refers to a subjective impression that is affected by accutance, resolution, and other factors. For example, the sharpness tutorial on this site points out that the impression of sharpness is also affected by image noise and viewing distance.

[george013]

Not quite, if I understand the terms. Resolution contributes to sharpness. Correct me if I am wrong:

Resolution is a measure of the amount of detail that can be resolved by the equipment. this is usually measured in lines/mm.

Accutance refers to edge contrast. I just went searching for a technical definition and finally found a clear one in Wikipedia:



In other words, for people unfamiliar with calculus, if you plotted distance on the horizontal axis and brightness on the vertical, the steeper the slope of the curve, the higher the accutance.

Finally, if I understand correctly, sharpness has no single technical definition; it refers to a subjective impression that is affected by accutance, resolution, and other factors. For example, the sharpness tutorial on this site points out that the impression of sharpness is also affected by image noise and viewing distance.

See how DxO defines sharpness with Imatest. http://www.imatest.com/docs/sharpness/
My problem with this test is that it calculates sharpness with pixels. One lens will give different numbers of sharpness with different camera's. Or with different pixel sizes. In this way pixel density or resolution limits the sharpness of the lens.

George

[xpatUSA]

See how DxO defines sharpness with Imatest. http://www.imatest.com/docs/sharpness/
My problem with this test is that it calculates sharpness with pixels.

Pixels is a common unit of measure for edge spread response, George. QuickMTF is the same; easily converted into other units - as a matter of fact QuickMTF does it for you ...

One lens will give different numbers of sharpness with different camera's. Or with different pixel sizes.

That's life in the big city of resolution, sorry, George.

In this way, pixel density or resolution limits the sharpness of the lens.

Nope. The sensor could be replaced by film or even a piece of toilet paper; the sharpness of the lens will not be affected one little bit!

This might help:

http://www.falklumo.com/lumolabs/art...ess/index.html