Erwin: Sharpness in the digital domain

Leica lens used in the test.

 

http://imx.nl/photosite/leica/sharpness_1/sharpnessstart.html

 

Enjoy.

That's extremely interesting. I'm sure there will be endless discussion of this, but the basic result isn't too surprising. Your enlarger affected the results when you were capturing on film; you should expect your RAW processor to do you the same favor.

"Is also evident that the choice of raw processor and the choice of sharpness settings has a

major impact on the resulting quality of the image."

 

Shocking...

Jonathan - do you have any idea of how really clueless this Erwin dude is? It is like reading what a person who cant swim has to say about surfing - really really pathetic.

Read the article, Peter. If you don't understand it, just say so, but don't complain that it doesn't make sense to others.

Billy boy - what's your point?

Well I don't mind saying so. I did not understand it.

 

Bill or Jonathan. I would appreciate if one of you could explain it to me please.

 

Thanks

Trevor, I posted the link because I think people here might be interested in what Erwin has to say. I'm neither his translator nor his exegete and, to be honest, I've only skimmed this article myself.

 

With those disclaimers in mind, what I think Erwin is saying is that, in digital photography, the sharpness of the lens may be less important than the post-processing used. While this might be an obvious point to people who are well-versed in digital -- I can't be sure as I am not one of those people -- it might come as a surprise to film photographers since, in film photography, the quality of the lens is often a major variable in the technical quality of the final image.

 

This is apparently the first installment of a three part series. I am anticipating where he's going with this, perhaps incorrectly. But that is what I have gathered from a quick reading. I didn't pay much attention to which post-processing program produced which effect since, at present, I don't use any of them.

 

Hope this helps.

Thankyou

Le me add an observation about the implications of what Erwin seems to be saying.

 

Generally, in any system, including imaging chains, it's helpful to know how much of the variance in outcome (here, image quality as measured by resolution) is attributable to each of the variables in the system.

 

If the variables have different weights in digital imaging than they do in silver halide imaging, that would be good to know, since it might influence how we, as photographers, attempt to improve the technical quality of our images.

 

It might also have implications for a company (Leica) that has built its reputation largely on the quality of its lenses. If lens quality makes a lesser contribution to variance in digital image quality than in silver halide image quality, then Leica will face a new challenge, since its primary strength, lensmaking, will no longer serve to differentiate it effectively in the marketplace.

All of the RAW processors I've seen employ some degree of edge sharpening, which affects the MTF curve. The default in Adobe Camera RAW is just enough to counter the anti-aliasing filter in the camera (most cameras), but can be increased, decreased or turned off if desired. I prefer to sharpen in Photoshop, where the process is completely parametric (hence reproducible).

 

Sharpness and resolution are functions of both the medium and the lens, and are always less than with either element measured individually. In statistical terms, it is estimated by the root sum of squares of uncertainties. Z-transforms (digital Fourier transforms) give a more accurate analysis.

What he's saying is quite interesting, that just like in film photography, the developer chosen can also influence the image sharpness in a major way.

 

Of course, the difference between film and digital is that you can only process film once, but you can process the raw files many times with different developers. Which means if you don't like the look from Capture One, you can use Bibble or Lightroom or whatever.

Could Capture One results be extended to Capture NX? It`s interesting, many people find Nikon software the best for converting their own RAW files...

While all programs did a competent job, I noticed the fringing (outside edge of circle) introduced was quite noticeable. I would also say it is unacceptable except what we see here are lab results. This would be so much more relevent if Puts used an image of something besides a test pattern.

It's a semi meaningless test. It tells you something but he doesn't provide enough parameters to really explain his findings.

 

 

For instance Lightroom 1.1 has several "capture sharpening" presets for different general types of subject matter and you can additionally tune those sharpening effect presets nearly infinitely to suit both the camera + lens + aperture combination and the specific subject matter. Tuning that capture sharpening will in all likely hood go a long way towards eliminating the moire produced by overly aggressive sharpening.

I think Erwin is saying that the post processing in Digital is just as likely to affect the final product as chemistry/time/emulsion does with film based photography. I just wish he would make his language a bit easier for non-phd's like me to fully grasp.

<i>I think Erwin is saying that the post processing in Digital is just as likely to affect the final product as chemistry/time/emulsion does with film based photography</i><p>And, as Brad implies above, this is news? This is worth an article? You can find stuff going back to the release of the first independent (non-mfr) RAW converter showing this. It's not news, it's history. How do you make a living re-writing what people wrote five years ago, it seems like a good racket.

>>> How do you make a living re-writing what people wrote five years ago, it seems like a

good racket.

 

 

Throw in some some MTF charts, add a pinch of old-world tweed, and title it "Optical

Analysis." Some will find it tasty...

My heart races in anticipation of his next report:

 

<blockquote><i>The Impact of Sensor Dust on Image Quality: A Quantitative Analysis</i>

</blockquote>

<i>How do you make a living re-writing what people wrote five years ago, it seems like a good racket.</i>

<p>

While one may agree/disagree with what Erwin Puts writes (or anyone related to photography write, be it a magazine, or an internet publication), I find it wrong to slight them. The article in question here is free to read. So, there isn't any monetary gain there.

<p>

If someone writes about the grain size of a particular emulsion would that make them reinventing the wheel?

<p>

I believe the slighting is misdirected.

There are far more extensive, interesting and valuable investigations of the subject. This article is like an abstract for a much more useful article.

Jeez, chill out ,Viv. I'm sure Erwin can defend his own sandcastle. If it helps i'm sure he is a very nice man.

 

But you have to admit it's all a lot of old hat cloaked in Erwin talk.<div></div>

"I'm sure Erwin can defend his own sandcastle."

 

It isn't about Erwin Puts, Al (how does that sound as a shortened name? :) ).

< This article is like an abstract...>

 

True, but Erwin says it's the first installment in a three-part series. In fairness, perhaps we should suspend judgment until all three parts have been posted.

Al (how does that sound as a shortened name? :) ).

 

Cool,Viv, i'm happy for you to call me Al...some of my friends call me Al. You are making me laugh. My good mate Trevor Hare, i sometimes call him Trev. Sort of a friendly niceness.

 

Are you having a crusty day or something Vivek Iyer Esq. You should post some more of your UV work, it will cheer you up. And we could all enjoy.