[rawr]

I think one of the factors considered by Hirakawa was the way resolution tests conducted against 2-D, totally flat plane test charts did not reflect the eye's perception of real-world 3-D subjects. He has said that some of his lens designs - like the 77mm lens - were specifically tuned to add 'spice' to the depiction of solid objects.


[source unknown - interview article with Hirakawa - translator unknown]

It certainly seems to work with some lenses.

[Rondec]

Resolution and sharpness are very much over rated. Obviously, images come down to light, subject and composition -- probably in roughly that order. If your focus is off if there is no sharpness, then that can certainly spoil an otherwise acceptable photo, but usually is is a pretty distant factor as to whether or not an image succeeds. Often little differences in color balance make a difference as to perception of viewers as well. Many people like their photos warmer and with (to my eye) overly saturated colors.

Be that as it may, beauty is in the eye of the beholder.

[nomadkng]

And I would take exception to the suggestion that lighting must be studio controlled and perfectly identical. As are all my tests, these are real world tests. But, my answer is, fine, do some of your own tests, prove me wrong.

The only thing more frustrating than someone with imperfect data, is someone with no data. Long experience has taught me, the fact that someone can nit pick doesn't mean they know a thing.

I never once suggested lighting must be studio controlled. That is a straw man argument. Secondly, it is not my responsibility to provide data, it is YOUR responsibility to provide valid data when presenting a claim. You did not provide test subjects identical images from which they could make subjective choices. Therefore, you introduced several variables beyond the scope of your testing hypothesis that could effect the results. That is faulty methodology and I need not produce contradictory data as part of the review process. You may very well have a valid hypothesis, but the burden of proof lies with the claimant. You espouse your results as empirical, and given your response to my questioning, irrefutable. But It is one thing to present a theory and ask us to consider your theory, it is another to attack one who might question your conclusions.

If you wish only praise and validation, perhaps an open forum is not the place to present ideas. With your background in education, I though perhaps you were introducing a mental exercise to challenge us to think of things differently than perhaps we do. However, based on your response, it appears you are not open for debate and discussion, but rather wish to bludgeon others into accepting your philosophy as writ.

I take even greater exception to THAT approach, than I do your faulty testing methodologies or your derisive comments.

[normhead]

Too funny.... any one can claim someone else used faulty methodology. It's one of the oldest scams on the books. Used by the tobacco industry, the sugar industry and lots of other scammers. Interesting that's the company you chose to keep.

[nomadkng]

Too funny.... any one can claim someone else used faulty methodology. It's one of the oldest scams on the books. Used by the tobacco industry, the sugar industry and lots of other scammers. Interesting that's the company you chose to keep.

I don't have to claim anything. I cited two concrete examples of your faulty methodology. If you do not wish to accept, or if you chose to dismiss my citations, that is one thing. But you have yet to admit to the discrepancies, nor have you addressed my hypothesis that these discrepancies might have had a effect on the pool results. Instead you continue to try and belittle me, or undermine my credibility. Now you equate me with scammers. Are you really suffering from such a high level of cognitive dissonance that all who disagree or challenge your beliefs must be relegated to the trash heap?

[normhead]

You cited two concrete examples of faulty methodology, but you have no information that would lead us to believe those short comings in any way altered the results. that is just pure speculation on your part.

SO, ya that's what I'm saying. you're countering empirical examples with criticism of methodology, which is just speculation.

Are you really suffering from such a high level of cognitive dissonance that all who disagree or challenge your beliefs must be relegated to the trash heap?

Too funny, you stand before us with nothing but your beliefs, and challenge me for interpreting some physical evidence I've presented. SO really, who is drawing line in the sand based on their beliefs? You have presented exactly nothing.

Now you equate me with scammers. Are you really suffering from such a high level of cognitive dissonance that all who disagree or challenge your beliefs must be relegated to the trash heap?

If you think you are the only one capable of criticism you're wrong. You feel the right to point out errors in my methodology which admittedly isn't that tight, but fine as a demonstration of concept, but when I come back and point out your criticism is pretty much baseless and irrelevant, and point out how destructive that kind of criticism can be, thousands died because the tobacco industry got away with it for years, now, I'm relegating you to the trash heap?

No, I'm pointing out the danger in the logic you're using.

It has always amazed me how people who use this type of argument, always fold their hands and assume, it's some kind of "last word". There is no last word. Just more research to clarify your last research.

Every thing you do, if you're an honest researcher, leads to more questions than it answers. This test raised some questions. It could be done better. But you start from what you've got and work from there, you don't dismiss what you have until you have something better. What you posted was not "something better".

Interesting.

---------- Post added 03-18-16 at 11:34 AM ----------

I think one of the factors considered by Hirakawa was the way resolution tests conducted against 2-D, totally flat plane test charts did not reflect the eye's perception of real-world 3-D subjects. He has said that some of his lens designs - like the 77mm lens - were specifically tuned to add 'spice' to the depiction of solid objects.


[source unknown - interview article with Hirakawa - translator unknown]

It certainly seems to work with some lenses.

That's really cool... I haven't seen that before. Thanks for sharing.

I hope everyone interested in this topic reads it.

[Nicolas06]

What I see on the prefered picture is that it has slightly warm tone with shadow and it look like mid sunny/cloudy light with the light there but some cloud. Like after a storm. A great light.

The next one look like exactly the same kind of day but when a cloud go in front of the sun and you get suddenly a quite different feel. You were warm before, now you are cold, the things looked beautifully now look quite average...

And this is exactly what I find in the least popular photo: the white balance give a blueish cast on my screen, shadows are mostly gone and the subject look more flat because of that.

So for me it was the lighting that did the trick, not the lens designer.

Do we really have to proove that lighting play a big role in photos? I don't think so, this is common knowledge and the reason people go to the pain of setting studio shots or wake up early to get that landscape shot... Photography is not about cameras or lenses. It is about subjects, composition and light. If you want to deny it, you would be alone in that quest. Not like it is a novelty concept.

Now this doesn't mean lens don't play a role like the 77 and so on... But unfortunately for Pentax, this is really difficult to prove and measure...

---

Last edited by Nicolas06; 03-19-2016 at 03:47 AM.

[Nicolas06]

Yeah, sorry for the wording. Constrast, I correct it in post, as for sharpness I prefer to have a sharp lens to beging with. That's what I do. I don't know what others do according the their lenses, curious to know as well.

A beautiful lady in front of any lens with good lighting and perspective, and I don't think about anything such as if it is a FA or DA or DA* lens, smc or HD coated, and even I forget the brand and format of the camera .

I am with you there. With current technology, we can easily play on color rendering, contrast/micro constrast even if from personnal experience, i get a much easier job from my FA77 as an example. And nothing say that we could not add fake detail in post to make a low rez picture look good. That what our brain does with our own vision after all.

While it is clear we have tool that work not so great for resolution enhencement (or that us both didn't manage) and tools that work much better for contrast/micro contrast, this doesn't mean theses tool are perfect and achieve the same as a lens that would do it out of the box.

Post processing is very powerful and enhance image quite a lot and you can do it regardless of the lens, subject, light... But you can only work from what you have or you must have great skill to add fake things to compensate. At that point we are more into artistic and less into phography.

Futhermore, while it might be possible with the right tool and experience, we may also fail to similutate the interresting part of some lenses rendering.

But yes ! Subject, lighting and composition trump all the rest.

[robgski]

There might be value in trying the same test using B&W photos as well.

[northcoastgreg]

Contrast: you can easily improve in post processing. Sharpness, you can improve a little , but can never recover the lost details (information not present in the image file data).

But what if the contrast you get out of a lens is qualitatively superior from what you can get via post-processing? After all, lens contrast, which is a kind of local contrast involving fine details and structures of an image, is hardly the same thing as the "global" contrast you get in post. Move that contrast slider to far to the right and you end up hosing detail in shadows and hi lights, whereas lens contrast accentuates those details while at the same time allowing the image to have that crispness and pop that some of us crave.

Now Lightroom provides what they all the "Clarity" slider, which is supposed to imitate something akin to lens contrast. The clarity slider improves mid-tone contrast, thus preserving detail in the hi-lights and shadows while at the same time adding contrast to the image. However, that still isn't the same thing as lens contrast, and, as anyone familiar the clarity slider probably knows, adding copious amounts usually harms more than it helps. The clarity slider, in that respect, is very much like the sharpness and saturation sliders. Adding some sharpness, saturation (actually, vibrance is much better), and clarity in Lightroom improves almost any image; but adding a lot of these elements is usually to the detriment of the image. This being the case, wouldn't it be desirable to start off from a better position — i.e., to have an image with better local contrast (or "microcontrast"), better color, better rendering, (and yes, better resolution as well, if you're planning on making large prints) as there are limitations as to how much you can improve those things in post?

[Nicolas06]

But what if the contrast you get out of a lens is qualitatively superior from what you can get via post-processing? After all, lens contrast, which is a kind of local contrast involving fine details and structures of an image, is hardly the same thing as the "global" contrast you get in post. Move that contrast slider to far to the right and you end up hosing detail in shadows and hi lights, whereas lens contrast accentuates those details while at the same time allowing the image to have that crispness and pop that some of us crave.

Now Lightroom provides what they all the "Clarity" slider, which is supposed to imitate something akin to lens contrast. The clarity slider improves mid-tone contrast, thus preserving detail in the hi-lights and shadows while at the same time adding contrast to the image. However, that still isn't the same thing as lens contrast, and, as anyone familiar the clarity slider probably knows, adding copious amounts usually harms more than it helps. The clarity slider, in that respect, is very much like the sharpness and saturation sliders. Adding some sharpness, saturation (actually, vibrance is much better), and clarity in Lightroom improves almost any image; but adding a lot of these elements is usually to the detriment of the image. This being the case, wouldn't it be desirable to start off from a better position — i.e., to have an image with better local contrast (or "microcontrast"), better color, better rendering, (and yes, better resolution as well, if you're planning on making large prints) as there are limitations as to how much you can improve those things in post?

This is always better sure.

Beside, there setting behind global contrast. You have clarity and DxO has clearview but there also the different tones cursors, micro contrast, fine micro constrast available by tones too. You can also work dirrectly on the curves.

But yes, if your lens has less contrast, say half of it (MTF50 equiv) you basically loose 1 stop of dynamic range. if your lens has 1/4 of it (MTF25) then you loose 2 stops. That like saying you have 1" sensor picture quality on your APSC ! Or APSC quality on your MF...

Combine that with a 400 or 800 iso shot where you loose respectively 2 or 3 stops and you soon find yourself with a picture were pushing the slider will make it look overcooked and noisy with no other option than to try to salvage it with some noise reduction.

[biz-engineer]

A lot of the lens color rendering can be aligned via metering of the white balance. There was a discussion about lens color rendering one or two years ago, it was about how Tamron rendering was compared to Sigma and Pentax etc.... So, I did an experiment: I mounted my Tamron lenses and Pentax lenses, manually metered the white balance with the same white balance calibration target. I was shocked ! All images looked the same, except by pixel peeping to look for differences in sharpness, it wasn't possible to distinguish. Basically, white balance calibration align RGB components whatever spectrum hit the front of the RGB filter matrix that's on top of the photodiode array sensor. On the sensor, you have only three primary colors coded: RGB with each color filter having its own sensitivity, every color being synthesized from various abouts of RGB components, even if the lens is super wide band (all colors are transmitted with a transmittance of 100%), the extra color variants transmitted trough the lens are filtered out by the RGB matrix. Feel free to do the experiment for yourself: calibrate the white balance on the same target with each lens, you'll be shocked.

[Nicolas06]

A lot of the lens color rendering can be aligned via metering of the white balance. There was a discussion about lens color rendering one or two years ago, it was about how Tamron rendering was compared to Sigma and Pentax etc.... So, I did an experiment: I mounted my Tamron lenses and Pentax lenses, manually metered the white balance with the same white balance calibration target. I was shocked ! All images looked the same, except by pixel peeping to look for differences in sharpness, it wasn't possible to distinguish. Basically, white balance calibration align RGB components whatever spectrum hit the front of the RGB filter matrix that's on top of the photodiode array sensor. On the sensor, you have only three primary colors coded: RGB with each color filter having its own sensitivity, every color being synthesized from various abouts of RGB components, even if the lens is super wide band (all colors are transmitted with a transmittance of 100%), the extra color variants transmitted trough the lens are filtered out by the RGB matrix. Feel free to do the experiment for yourself: calibrate the white balance on the same target with each lens, you'll be shocked.

Well it make sense that if you calibrate your lenses and camera the color rendering will be roughly the same. White balance may be good enough for color difference.

For contrast difference, this would not work. DA lenses for example have more punchy agressive colors than the more suble, softer rendering of FA lenses (that is also warmer).

I guess you can fix that, at least partially with the contrast slider...

Now that still leave other aspects of the rendering like the in focus/out of focus transitions, the ability to make things look "3D" and pop. Constrast/micro contrast do help, but that not the only factor.

Also I have checked that the more you puch sliders of contrast/micro contrast, the worse you make the bokeh look. To achieve the kind of result a better lens would manage naturally, you'd have to do it only on some zones of the picture. That's quite possible, but the time needed for it start to be annoying.

This may be worth it for somebody with time and no money, but if you actually are paid for that activity, I guess you want tool that get it right out of the box.

[biz-engineer]

For contrast difference, this would not work.

That's right.

Now that still leave other aspects of the rendering like the in focus/out of focus transitions, the ability to make things look "3D" and pop. Constrast/micro contrast do help, but that not the only factor.

Regarding 3D pop, I've observed a number of photographs from others, with and without "pop", I paid close attention to photographs from Brenizer, and I noticed that the famous "3D pop" or "pixie dust" is actually achieved by the perception of volume around the subject, bojeh being only a secondary contributor to the effect. For example, the 300mm@f4 is sharper wide open than a fa77@f1.8 and the 300@f4 lens creates more background bokeh than the fa77@f1.8, but the fa77 includes more background elements. For instance, place a model between two lines of trees, focus on the model... and you've go the 3D pop effect, because the eyes are guided by the background to the subject, so that it looks like everything revolve around the subject. Basically, it is the composition and the blur that create this 3D pop, the lens plays a secondary role.

---------- Post added 20-03-16 at 07:37 ----------

This may be worth it for somebody with time and no money, but if you actually are paid for that activity, I guess you want tool that get it right out of the box.

For sure, you save a lot of time is you get great image quality straight out of the box. There is a point of diminishing return where you just want to have fun instead of having deal with hassles of unsuitable gear (also true for auto-focus...you can improve your technique up to a certain point as long as it does not kill the fun). Time and money, that's exactly the point: if you have a lot of time and no money, you can play around with cheap stuff to make it deliver decent results, but if you have money and no time, then it's also worth getting the gear that just deliver what you like. For wildlife, I've seen super images from retired people... but they are there in the national park every day of the year, while I'm in the office.

[reeftool]

There are a lot of components to a really good photo that need to line up to make it "good". Composition is always number one. No matter how sharp a lens is or how great a 100% crop of a a section of the shot looks, a badly composed shot is just going to look exactly what it is, bad. The light you are shooting in is important. Good focus is important. Using the proper aperture to obtain the right bokeh is important. Get a good composition and get the technical stuff right and a shot from a K10D and the kit lens will look just a s good as one from a K3 and a Limited.

What you are viewing on makes a big difference. Shots that don't look so great up close on a 24 inch hi-res monitor can look great on a 4x6 or 5x7 print that you give somebody. Those of us who do events like parties, wedding or even sports, and shoot over 1000 shots on an afternoon, see this often. The people you give these shots to don't see what we do. Most people don't pixel peep. That's a madness that inflict photographers. I have ignored shots that my wife absolutely loves. She sees the image. She doesn't zoom in on the computer or look at a print with a magnifying glass. She wants something to hang on the wall.

The motorcycle racers I photograph download my shots on their phones and post them on Facebook. Many proudly use shots as their cover photos that I wouldn't post here in the forum because to me, they just aren't as sharp or crisp as they could be but none of that can be seen on a phone or a tablet.