Summarit- talk to me

[rosuna]

Advertisement (gone after registration)

Good is image quality, and it is a dependent variable of many explicative variables (and a error term). Resolution and contrast (axis, field, corners... at different apertures and distances), flare, vignetting, focus precision, color rendition and tonal gradation, residual aberrations (and its effects, including bokeh), handling, build quality...

 

The new Summarit will not be a "cheaper" class of lenses at all. They will be performers as "good" as Elmarits, Summicrons and Summiluxes in the same sense in which the Heliar 75mm is supposed to be as good as the APO-Summicron (this is, not an inferior lens in terms of most of the explicative variables or "weights" in any IQ index).

[sean_reid]

Hi Ruben,

 

Until and unless all photographers decide that they all want the same kind of drawing from their lenses, we won't ever be able to accurately say that lens A is "better" than lens B. Its done all the time but its a mistaken concept - too simplistic, too many assumptions. It is, in part, at the root of what dissappoints me about most lens reviews that I read.

 

We can define technical goals and describe lenses accordingly but a lens that performs very well (in conventional technical terms) may well be a poor choice for a specific photographer. Contrast is a matter of taste and purpose, there's no such thing as "good" or "bad" bokeh, flare can be essential to certain pictures, tonal rendering is a matter of taste and purpose, etc. Most writing about lenses doesn't tell me what I most need to know and saying that lens "A" is "as good as", etc. exhausts the exploration before its even begun (ir concludes an exploration that is very incomplete).

 

Perhaps you already agree with some of this. One of my jobs is to act as a gadfly when needed.

 

Best,

 

Sean

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[roubaix]

Sean,

 

Looking forward to your reviews.

 

Peter

[sean_reid]

Me too, thanks.

 

Cheers,

 

Sean

[rosuna]

An old lens can be an interesting working tool for a photographer, but optical progress consists in more control of aberrations and other problems (flare, vignetting) that deteriorate an accurate reproduction of the subject (image quality, or quality of reproduction). From a technical point of view there are "better" and "worse" lenses. When a lens is better in at least one vector component (or variable determining IQ) and not worse in any of the remaining components, then the lens is better (from a technical point of view). When a lens is better in one component/variable and worse in other(s), you cannot say it is better or worse as a whole. For instance, a lens can show more vignetting just because the lens designer used it for compensating aberrations. It was just a choice. The problem is even more complex when you consider economic variables into the equation, because it makes relative (to cost, and production technology!) any evaluation of a lens (from a technical point of view).

 

The whole subject is very complex. You are right pointing to lenses as photographic tools for creative photographers, but the technical (pure scientific) and technical/economic points of view are also relevant, because they aren't subjective and, therefore, are very useful as guides and points of reference. A landscape photographer can find the pre-ASPH Summilux 35mm lens a wonderful tool, but that means nothing to other photographer (even other landscape photographer!). Particular and subjective points of view make comparisons and evaluations meaningless ("it is good for me", "it renders this subject (under these conditions) beautifully"). Only objective variables make any comparison meaningful (but limited in scope).

 

I expect the Summarit lenses to be technically as good as the other Leica lenses (aberration control, image reproduction) and, given the lower price, I expect the technical/economic (value for money) will be very high. Of course, these lenses can be inadequate for a particular photographer ("too contrasty", "too small in size", "too 'clinic' rendition", or "too classic rendition", etc, etc.), but due to the same reasons than any other Leica lens with similar fingerprint (not because they are "cheaper" or "worse" or "second rate" in any sense).

[pklein]

I have no evidence to back this up, but, I think and hope that the new Summarits will be designed to be relatively free from focus shift. In other words, no focus worries on current or (ahem) future digital bodies.

 

Even my 50/2 Summicron shows a bit of focus shift. Perhaps f/2.5 is slow enough to bring the phenomenon down to the photographically insignificant level, while still being fast enough for many purposes.

[ho_co]

Advertisement (gone after registration)

A personal tidbit: A videographer/photographer I know (daughter of a close acquaintance) prefers to use the cheapest, 'consumer-grade,' plastic-mount Nikkors in her portrait business. For her use, they are the 'best' choice.

 

We all know the Summarits will be Leica lenses. They will be worth the price and will perform at or above the level that will make most of us happy.

 

Going beyond that is wasted excess because we don't have them to play with.

 

From E Puts' pre-evaluation, I don't expect them to elicit from him the kind of exuberance the current 50/1.4 or the 75/2 have. That is, they won't "pull out all the stops" in his phrase.

 

Think of the current 50/2 R. No aspherics, no floating elements. Good images from a six-element double-Gauss derivative--with 5 plane surfaces! Unbelievably simple and at the same time sophisticated design, and affordable as well.

 

I have no evidence to back this up, but, I think and hope that the new Summarits will be designed to be relatively free from focus shift.

Judging from the technical discussion in the latest LFI, I think we have implied evidence to support your guess!

 

 

--HC

[sean_reid]

An old lens can be an interesting working tool for a photographer, but optical progress consists in more control of aberrations and other problems (flare, vignetting) that deteriorate an accurate reproduction of the subject (image quality, or quality of reproduction).

 

So...for example, would you suggest that higher lens contrast delivers a "more accurate reproduction of the subject" even when it serves to obliterate detail in the shadows or highlights? What about a lens that shows no barrel distortion but more "stretching" of objects near the edges of the frame. Technically, it distorts less (simply allowing the mathematical rules of perspective to play out as three dimensions become two) but one could argue that it provides a less accurate reproduction of the subject. Do we define accuracy mathematically? Does the eye play a role? Do we follow Erwin's suggestion to ignore what our eyes tell us (in a visual medium, no less)? Do we all agree as to when a subject has been "accurately" reproduced? Given that a photograph is not and never can be "the thing itself", have we all agreed as to which transformations are considered accurate and which are not? So often in these "objective" lens assessments, there's an assumption of consensus where none exists. It's a veneer of objectivity that is often mistaken for objectivity itself.

 

To be sure, we can talk about optical/technical goals (for better or worse) and a given lenses' accomplishment with respect to those. But what deserves ongoing critical thought is the hierarchy and assumptions implied by the technical goals. They tend, too often, to be accepted tacitly.

 

Also, as you are an economist, I imagine you maintain a healthy skepticism about "objective" data. For example, resolution charts are photographed at a given distance, as chosen by the tester. Lens' resolutions can be slightly different at different focus distances. Are those resolution results then to be treated as "objective" descriptions of resolution performance, in general, for that lens at a given aperture? As soon as the information they imply is generalized as a description of the lens, is it still "objective" information? People, probably not yourself, tend to take comfort in the supposed "objective" nature of information that is described in numbers. Its easy to forget the subjective assumptions that had to be made before the numbers could be specified. It's a sleight of hand and some people don't even know to look for a trick.

 

You disagreed once with Mark Twain's observation that there are three kinds of liars - liars, damn liars and statistics. But I think he was on to something.

 

I think many, probably not yourself, forget that, at best, we might see data that denies the null hypothesis. Otherwise, there are a lot of illusions of certainty where none exists.

 

Its not just a question of "objective" vs. "subjective" evaluation, as Erwin Puts mistakenly suggested. The important differences lie more in the assumptions that the evaluations proceed from. Most discussions of these differences in approach are not rigorous enough, they tend to fall easily into prefabricated categories, dichotomies, etc.

 

Time for me to reread Thomas Kuhn's "The Structure of Scientific Revolutions". He, I think, was clear on much of this.

 

Cheers,

 

Sean

[sean_reid]

From E Puts' pre-evaluation, I don't expect them to elicit from him the kind of exuberance the current 50/1.4 or the 75/2 have. That is, they won't "pull out all the stops" in his phrase.

 

--HC

 

An interesting "pre-evaluation" BTW, as Leica tells me that no reviewer has had access to test copies of the Summarit lenses, pre-production or otherwise.

 

Cheers,

 

Sean

[rosuna]

Hello Sean.

 

So...for example, would you suggest that higher lens contrast delivers a "more accurate reproduction of the subject" even when it serves to obliterate detail in the shadows or highlights?

 

Of course, yes. That loss of detail is due to the limited "dynamic range" of the capture medium, not the lens. Higher lens contrast is a purer light transmiter. No losses in light transmition, no light deviation from bright zones to dark zones. That's all. From a technical point of view, it is a better lens. Another history is a camera system as a whole, considering the film or sensor. Lenses are long term investments, and capture mediums evolve very fast.

 

Do we follow Erwin's suggestion to ignore what our eyes tell us (in a visual medium, no less)? Do we all agree as to when a subject has been "accurately" reproduced?

 

Well, I think Puts is misunderstood here. Lens comparisons from a technical point of view are meaningful, have sense, provide information what the photographer can take account of. It is not all the history, of course. Erwin Puts doesn't suggest to ignore what our eyes tell us. He says it is very subjective. For instance, he says lens A has higher resolution than lens B, but in normal working conditions (hand held shooting, etc.) we will not see it. There are many examples of this kind of comments in Puts' reviews.

 

 

Also, as you are an economist, I imagine you maintain a healthy skepticism about "objective" data. For example, resolution charts are photographed at a given distance, as chosen by the tester. Lens' resolutions can be slightly different at different focus distances. Are those resolution results then to be treated as "objective" descriptions of resolution performance, in general, for that lens at a given aperture? As soon as the information they imply is generalized as a description of the lens, is it still "objective" information? People, probably not yourself, tend to take comfort in the supposed "objective" nature of information that is described in numbers. Its easy to forget the subjective assumptions that had to be made before the numbers could be specified. It's a sleight of hand and some people don't even know to look for a trick.

 

You are right, but this is not a problem of objectivity vs subjectivity. This is a mere statistical problem. There are infinite possible combinations of distances, angles, light conditions, etc. Therefore, you must to take a sample, a representative sample. In statistcs you take a random sample, if possible at all, but in optics you take a typical sample under controlled conditions (close distance, medium distances, infinite; bracketing in focusing; different lens copies,). It is "subjective" only in the following sense: you can take other test parameters, in theory. For instance, you can use 5 distance positions instead of 3, etc. This isn't all the possible information in all the possible working conditions, but it provides a set of relevant and secure data (in the same sense, even MTF graphs don't provide all the information about a lens). The key point here is comparability. If the testing procedure is the same for different lenses, you can make a comparison among them.

 

You disagreed once with Mark Twain's observation that there are three kinds of liars - liars, damn liars and statistics. But I think he was on to something.

 

Statistics aren't lies. Sometimes marketing guys and politicians are liars. The lie is possible because the liar doesn't give all the information and people doesn't know much about statistics. It is a simple problem of asymmetric information.

 

I think many, probably not yourself, forget that, at best, we might see data that denies the null hypothesis. Otherwise, there are a lot of illusions of certainty where none exists.

 

The null hypothesis is rejected with a confidence margin. Similarly, a zero probability doesn't mean something is impossible. But you are mixing different things. In optics you are in the laboratory, and you have control on almost all the relevant variables. You can have a sample data has big as you want, but there are diminishing returns in this gathering of information process. Hypotheses and uncertainty has more to do with subjects for which we haven't enough information or representative samples. It is not the same to test the consumers' decisions about retirement with a (representative?) sample of 2,000 families than testing a particular and tangible car, computer or lens.

 

Its not just a question of "objective" vs. "subjective" evaluation, as Erwin Puts mistakenly suggested. The important differences lie more in the assumptions that the evaluations proceed from. Most discussions of these differences in approach are not rigorous enough, they tend to fall easily into prefabricated categories, dichotomies, etc.

 

I think we can say a lot of things about lenses rigurously. On the other hand, there is a danger in the other approach, when we try to say something general about a lens from a particular case, like resolution comparisons from a particular photograph (repeated samples of resolution cards under different conditions is a much more solid procedure). Even if we accept a generalization from that case (and we cannot), it can be relevant for you and not for me. Generalization is the basis of any science, and the old problem is how we can generalize accurately. It is easy if your subject is under control, and much more difficult if you only can "see" it partially (samples problem). Resolution effectively measured can be uninteresting for a particular photographer too, but the the data is objective, and can be relevant and important for other photographers. It is comparable data, useful or not. Descriptions of the fingerprint of a lens based on many different photographs under different working conditions can be interesting too, although more subjective, and this requires close familiarity with a lens (Michael Reichmann does this, working with a camera or lens for a while, but even then, he is careful in his conclusions).

 

Time for me to reread Thomas Kuhn's "The Structure of Scientific Revolutions". He, I think, was clear on much of this.

 

I recommend you Ortega y Gasset's "The Idea of Principle in Leibnitz and the Evolution of Deductive Theory" and "About Galileo" (included in "Scheme of Crisis").

 

Best,

 

Rubén

[rosuna]

From E Puts' pre-evaluation, I don't expect them to elicit from him the kind of exuberance the current 50/1.4 or the 75/2 have. That is, they won't "pull out all the stops" in his phrase.

 

I expect comments similar to those for the Zeiss ZM lenses.

[hankg]

Its not just a question of "objective" vs. "subjective" evaluation, as Erwin Puts mistakenly suggested. The important differences lie more in the assumptions that the evaluations proceed from. Most discussions of these differences in approach are not rigorous enough, they tend to fall easily into prefabricated categories, dichotomies, etc.

 

 

Which lens is best will totally depend on how you construct the test. If you compare a macro lens to a lens designed to perform at it's best at infinity what distance should the resolution target be shot at? Higher contrast in a lens will return better results when shooting a test target but what if you constructed the test so that loss of detail at the ends of the scale in high contrast scenes was penalized? Do you give a fault like chromatic abberation or barrel distortion equal weight in valuing a lens as corner or center resolution? How should you score values like size and price? Depends on the shooting you do. An architectural photographer will have very different criteria for what's best then say an event photographer.

 

Lens design is as much art as science. It involves balancing various compromises to satisfy the priorities you choose. Like a chef deciding how to balance a sauce. Assuming 2 different very well prepared dishes with very different tastes, no 2 diners will have the same preferences for what tastes 'best'.

 

For some photographers the Nocti is a ridiculous, expensive, oversized beast for others it's their dream lens. Who's right? they both are. Engineers can be impressed at the technology developed to achieve specific design goals like flatness of field and corner to corner resolution but those technical details don't really tell you about the character of the lens in the photographers hands.

 

The Zeiss 120 Makro for the Hasselblad was regularly trashed on the internet mostly by people who did not own the lens based on some MTF chart published online. However myself and many other photographers loved the lens for specific tasks. It's pentagonal bokeh sucked and at infinity it was bested by other lenses like the excellent 100 but using it for what it was designed for -it had no peer.

[ArtZ]

Ok for the technical/artistic considerations discussed here but I see things like this:

 

1.- Leica is a victim of its success. Because of the build quality of Leica Lenses, many people buy second hand lenses which is not good for Leica finances. Leica would like selling more new lenses and here comes the Summarit: Many people would prefer to buy a brand new lens than a second hand lens for the same price (or even less).

 

2.- CV and Zeiss have very nice lenses 2, 3 or 4 times cheaper than the Leica ones. But to buy these lenses means to get them coded, get the right IR filter and, sometimes, to get adapters (i.e. for the IR filter). This, obviously, increases the price of the CV and Zeiss lenses. (I always thought the 6-bit coding and the IR filter was done intentionally with the idea that customers will buy in this way more Leica lenses… but that’s another story). So why getting in trouble to code my CV or Zeiss lens and to get the right filter if I can get an equivalent Leica lens for almost the same money?

 

I do believe that many weekend photographers (and travelers) will buy M8 with two or three Summarit lenses. They will be very proud to show off their “LEICA” equipment… and this will be good for Leica finances!

 

I must say “Chapeau !” (Bravo!) to the Leica Marketing department. I simply wish that Quality department would be half good as the Marketing department

.

[sean_reid]

Which lens is best will totally depend on how you construct the test.

 

 

...and upon how one defines "best".

 

Cheers,

 

Sean

[Peter Branch]

So...for example, would you suggest that higher lens contrast delivers a "more accurate reproduction of the subject" even when it serves to obliterate detail in the shadows or highlights?

 

Sean, I think that there is a flaw in this argument. What you are implying is that if a lens maintains more of the contrast of the original subject, (something that can be evaluated very precisely), but by so doing produces an image with a contrast which exceeds the dynamic range of the sensor with which it is being used then that lens is less good than one which is less able to maintain the subject contrast but as a consequence produces images whose contrast does not exceed the capabilities of the sensor. This cannot be right. The lens is technically and objectively better. What we are considering here however is not “Lens” performance but the performance of a sub-set of the components in the image chain. It may be that this lower contrast lens and limited dynamic range sensor used together give a “better” result in the particular image chain in which they are being used but we should not kid ourselves that nothing has been lost or distorted.

 

It is important to distinguish between the overall,i.e. low frequency, image contrast range which a lens can produce and what most people are actually concerned about which is the change of intensity over a limited range and a very small dimension which is the realm of MTF curves.

 

I have long maintained that when it comes to evaluating lenses lessons can, and should, be learned from information theory where “error” is measured and not “goodness”, as in MTF curves. It is customary in many spheres to evaluate performance in terms such as parts per million, (“ppm”), of error. If this were adopted we would soon see that some lenses suffer something like a 2% drop in contrast at low frequency. This would normally be expressed a 98% goodness on an MTF chart but is in fact 20,000ppm error. A better lens may exhibit only a 0.5 % drop in contrast. This is either 99.5% good or 5.000ppm error. Such apparently small differences are actually very easily seen and are often what is being argued over. They are, in my view, better expressed as a x4 improvement, 20k -> 5kppm, rather than as a negligible change in the MTF figures.

[sean_reid]

Hi Ruben,

 

You wrote:

 

"Quote:

Of course, yes. That loss of detail is due to the limited "dynamic range" of the capture medium, not the lens. Higher lens contrast is a purer light transmiter. No losses in light transmition, no light deviation from bright zones to dark zones. That's all. From a technical point of view, it is a better lens. Another history is a camera system as a whole, considering the film or sensor. Lenses are long term investments, and capture mediums evolve very fast."

 

Of course it means the lens is purer transmitter of light. But while high lens contrast is a technical accomplishment (as I've written many times) it doesn't necessarily make a lens "better" in actual use. If one defines the problem in terms of how lenses work with mediums that we know to exist, it makes sense to consider a lens a relation to its role in a system. Collectors might be interested in the academic accomplishments of a lens, working photographers need to care about how it will work with the actual cameras they use.

 

 

"Quote:

Also, as you are an economist, I imagine you maintain a healthy skepticism about "objective" data. For example, resolution charts are photographed at a given distance, as chosen by the tester. Lens' resolutions can be slightly different at different focus distances. Are those resolution results then to be treated as "objective" descriptions of resolution performance, in general, for that lens at a given aperture? As soon as the information they imply is generalized as a description of the lens, is it still "objective" information? People, probably not yourself, tend to take comfort in the supposed "objective" nature of information that is described in numbers. Its easy to forget the subjective assumptions that had to be made before the numbers could be specified. It's a sleight of hand and some people don't even know to look for a trick.

You are right, but this is not a problem of objectivity vs subjectivity. This is a mere statistical problem. There are infinite possible combinations of distances, angles, light conditions, etc. Therefore, you must to take a sample, a representative sample. In statistcs you take a random sample, if possible at all, but in optics you take a typical sample under controlled conditions (close distance, medium distances, infinite; bracketing in focusing; different lens copies,). It is "subjective" only in the following sense: you can take other test parameters, in theory. For instance, you can use 5 distance positions instead of 3, etc. This isn't all the possible information in all the possible working conditions, but it provides a set of relevant and secure data (in the same sense, even MTF graphs don't provide all the information about a lens). The key point here is comparability. If the testing procedure is the same for different lenses, you can make a comparison among them."

 

You're missing an important component.There are assumptions that govern how statistics are created, as you well know. Those assumptions are not objective and people often forget that subjectivity in the seeming objectivity of numbers. I think Twain's point about statistics stems from this.

 

"Quote:

Its not just a question of "objective" vs. "subjective" evaluation, as Erwin Puts mistakenly suggested. The important differences lie more in the assumptions that the evaluations proceed from. Most discussions of these differences in approach are not rigorous enough, they tend to fall easily into prefabricated categories, dichotomies, etc.

I think we can say a lot of things about lenses rigurously. On the other hand, there is a danger in the other approach, when we try to say something general about a lens from a particular case, like resolution comparisons from a particular photograph (repeated samples of resolution cards under different conditions is a much more solid procedure)."

 

Solid, perhaps but really most useful to photographers whose subjects are resolution cards. I don't know of many. The problem with that approach (aside from the assumptions that must precede the testing) is that it tends to make mountains out of molehills. It emphasizes differences that often have little relevance to photographers working with a normal range of subjects.

 

Quote:

Time for me to reread Thomas Kuhn's "The Structure of Scientific Revolutions". He, I think, was clear on much of this.

I recommend you Ortega y Gasset's "The Idea of Principle in Leibnitz and the Evolution of Deductive Theory" and "About Galileo" (included in "Scheme of Crisis").

 

Can I get these in English? If yes, I'll read them over the winter. Have you read Kuhn? Paradigm shift (leaving aside the cliche that phrase has become in some cases) is relevant here.

 

Cheers,

 

Sean

[sean_reid]

"Sean, I think that there is a flaw in this argument. What you are implying is that if a lens maintains more of the contrast of the original subject, (something that can be evaluated very precisely), but by so doing produces an image with a contrast which exceeds the dynamic range of the sensor with which it is being used then that lens is less good than one which is less able to maintain the subject contrast but as a consequence produces images whose contrast does not exceed the capabilities of the sensor."

 

Hi Peter,

 

No, I never suggested one was better than the other, universally. Reread my post.

 

First off, I should say that I don't define lenses as better or worse, as a rule. I try to describe them and let the photographer see how his or her needs fit with a given lens. With respect to high lens contrast... it can be desirable or not depending upon many factors (as I've been writing for many years).

 

The distinction is this. If one is looking academically at the performance of a lens, in isolation, then high contrast indicates a technical accomplishment (and if you reread my lens reviews, you'll see that I've made that point for years). So, a collector can certainly display the lens and be proud of its technical accomplishment.

 

But when people start talking about "better" or "worse" lenses then I believe we have to consider them as part of real systems where real pictures are made. And, in that context (actual use with results in print or on screen) high lens contrast can be either a strength or a weakness. The fact that Zeiss' current lenses tend, on the whole, to deliver higher contrast than Leica's lenses do not necessarily make them "better" in a general sense.

 

Do you understand the distinction I'm making? Many people talk about "better" and "worse" lenses as if we had all come to a consensus on how we want lenses to draw. We haven't. It is also, by the same token, vague and not very useful to say that lens "A" is "as good as"generation X" of lens "B". Yet, people make comments like that all the time.

 

There is, very often, too much attention given to technical ideals (in isolation) and too little to how a lens allows us to draw specific kinds of pictures.

 

If one defines specific technical goals, he or she can say that a given lens does better or worse with respect to them. But one cannot generalize from there to talking about better and worse lenses. It all comes down to how the individual photographers wants the lens to draw and this is a point that I have been trying to emphasize for several years now.

 

Even if, and when, sensors develop to the point that they can handle extreme contrast ranges, photographers will still vary with respect to how much contrast which they want to see from their lenses. Even then, high lens contrast will remain a technical accomplishment, not a universal mark of a lens being "better" or "worse". And contrast is just one of many aspect we could look at.

 

When I test the new Summarits, my guiding question will not be "are they as good as XYZ lenses" which I find to be a dead-end approach. Rather, I'm going to try to describe how they draw in relation to other lenses.

 

 

Cheers,

 

Sean

[pklein]

An interesting "pre-evaluation" BTW, as Leica tells me that no reviewer has had access to test copies of the Summarit lenses, pre-production or otherwise.

 

Hmmm. . .

 

My learned profession I'll never disgrace

By taking a fee with a grin on my face

When I haven't been there to attend to the case

Said I to myself, said I

 

(W.S. Gilbert, Iolanthe)

 

[sean_reid]

Hmmm. . .

 

My learned profession I'll never disgrace

By taking a fee with a grin on my face

When I haven't been there to attend to the case

Said I to myself, said I

 

(W.S. Gilbert, Iolanthe)

 

 

Well chosen. Tis a mysterious thing, that.

 

Cheers,

 

Sean

[Advertisement]

Hallo,

Look here: Leica M

[pklein]

Sean's comments are right on.

 

I've always found it interesting that there are people who have sold all their "classic" Leica lenses to buy new aspherics. There are other people who would never buy an aspheric lens, finding them too sterile and clinical. Each are fanatical about their opinions.

 

I have sympathy with both camps at times, which is why I have a few too many lenses.

 

Terms like "roundness" and "glow" are imprecise, but they do mean something to the person using them. They fit right in with Sean's "how the lens draws." It is a lot easier to quantify MTF and the precise reproduction of microstructures than it is to quantify some of the things that really matter to photographers.

 

For example, I could probably "prove" that Monet, Manet, Pizzaro, Seurat, etc. were inferior painters, using the same arguments used about lenses. They didn't paint as precisely as their predecessors, they made fuzzy dabs and dots of paint. I could provide you with precise measurements of the size and width of their brushstrokes compared to earlier painters, compare the numbers, and show convincingly that based on hard, objective numbers, they just didn't paint very well. And of course, I'd be dead wrong.

 

Now granted, the Impressionist painters did what they did deliberately, and an older lens does what it does because that was what available technology allowed at the time. But since viewing the final print is what counts, pixel peeping may miss the point. It may tell you why the grass is more detailed with a current lens. It doesn't tell you why the old DR Summicron draws a more pleasing landscape than a modern Summicron with "better" MTF.

 

Sometimes the "measure everything" crowd gets it wrong, because they measure what is easy to measure, and ignore other things that matter. This phenomenon is not limited to photographers. Those of us who work for large organizations see it all the time. The search for numbers to put on spreadsheets and charts sometimes (often?) makes people lose track of what they are really after. And like Internet discussions, "winning" the argument becomes more important than finding more truthful but often ambiguous answers.

 

--Peter