A plea for common sense and a return to THINKING...
Before I begin, I would like to emphasize that while I have a preference for modern optics that are well corrected, this post of mine is in no way meant to be an "attack" on those who truly prefer the subjective rendering of lenses different than what I choose. As anyone who knows me knows, I am a believer in matching tool for task, and for some tasks, one lens will be the best match relative to another lens choice, even if one lens choice "tests" better. Even within the context of modern lenses that test well, there are different design goals, and differing performance personalities. No lens is perfect, and thus there will always be tradeoffs involved. However, at the same time, there is a distinct possibility that what some may perceive to be negative aspects of image quality in images they see and then call out the lens (whether specific lens, lens brand, or era the lens was designed in) as the reason for this image quality valuation, when the reality is likely that something else entirely is the primary cause. A lack of education on how things interrelate is usually the case here, along with some peoples desire for the simple answer and some peoples choice to believe conspiracy theories and whack job explanations.
The other thing I wish to note that is that nothing is more potentially misleading than an improperly done test. Besides having been involved with photography for over 35 years, I've been involved in both hardware and software testing for over 20 years. And this is what this whole thing comes down to. Somebody didn't like a quality in their images and decided to blame it on the element count and vintage of their lenses, and then proceeded to do an abysmal job of testing and attempted an arrogant presentation of why they were right with a serious lack of understanding of things photographic.
The only part of this thread where I'm going to get "personal" is right here: I have absolutely no use for such people. To me, the writer of that blog and the "angry photographer" rank amongst the most useless people in the entire photographic community due to the combination of their arrogance and gross incompetence. Those sorts of people are dangerous, because it causes otherwise sane people to go deep into the proverbial rabbit hole when instead they should be spending their time in other avenues that will, in the end, provide vastly better solutions than chasing some contrarian arcane theory about more lens elements reducing depth and other nonsense.
So that ends the personal attack. You can either continue reading or continue chasing the false truths. It's really your choice. But on the way out, if you think I'm being arrogant and unkind, I challenge you (or anyone reading this who is sticking around) with this: For the next 30 days, don't spend a solitary minute on dpreview or any other photography forum or blog. Spend the time you would have on this instead on educating yourself on post processing, partiularly on the role of contrast, and spend some time learning about visual perception; how we as humans perceive things. Spend some time googling forums or blogs or articles from reputable sources and learn some basics about the challenges faced by an optical lens designer. Spend some time practicing what you've learned. If you take that 30 day challenge, I'm confident you'll come back with an entirely different view, and if you are open minded, you may have changed your preconceived notions about what constitutes a good lens. Note that I don't care how good you think you are at post. I've been at it since 1998 and have been paid to do it for others. Peers consider me extremely competent, yet I don't spend a day thinking about how I could improve. Nobody quickly gets to a mastery level in this stuff and stays there without awareness and work, and that includes 100% of people reading this thread and me too.
So let's talk about the perception of depth, what makes some images "flat" and others "lifelike". Once we get past the way the human visual processing system works (we see a red rose set against a bed of green flowers as "more red" than the same exact red rose set against a bed of, say, orange tulips, and so on), it comes down to contrast. Always has, always will. The thing about contrast is that it's always tied to resolution. The two are married. So when the blogger thinks microcontrast isn't related to resolution, he just lost his entire argument, before he even stepped out the door, because he's just wrong. It's not a debatable issue. If you were to walk into any lens design team, whether it be in the 40's, 50's, 70's or modern era, and proposed to the designers what that blogger proposes, you'd be laughed out of the room. You'd have a better argument convincing others that the sun is made of swiss cheese and emanates gummy bears for energy. In short, you'd be considered insane. But in the modern day of the internet where a blog and/or a youtube channel somehow give one a false sense of credibility, pseudoscience and poor testing result in madness. We were better off in the magazine era, because no magazine editor would have *ever* allowed such nonsense to be published. But in the land of the internet, bull excrement can flow, and it flowed plenty in that bloggers thread.
So let's talk about lens design and the number of elements for a bit. The need for multiple elements is to better correct aberrations. If you feel the goal of a lens is to accurately portray what is in front of it with a minimum of disturbance, the better you correct the lens, the more accurate the portrayal. The problem of course comes in when you realize that you can't be perfect, so you do have to make choices. A bit of deliberate undercorrection of spherical aberration likely will help out your bokeh, but it also might take away a bit of contrast and resolution. But if the subject matter benefits from this design choice (remember tool-to-task matching), then it makes sense. In the contrary it may not. The designers from the "older" era didn't have the computing power we do today. Dr. Nasse, in a Zeiss whitepaper, mentions that in the old days it took nearly 2 minutes to do the calculations for a single ray trace through a single element. It's no wonder the older designers tended to work with more "simple" designs. Stories of Nikon having "math girls" - a team of students proficient in math - doing the calculations amaze me. But it's true - I'm not an optical designer, but everything I've read indicate that once you start adding more elements, the complexity of the calculations goes up seriously high, seriously quick. So I'm pretty sure that if a designer from the "golden era" of lens design were to have avaible both modern coatings (which reduce light loss per element down to inconsequential amounts) and most important, the computing power (and software) we have today, you'd have seen lens designs with more lens elements, simply because they could have better corrected their lenses. There simply isn't enough light loss between a 6/6 and a 13/8 lens design to suddenly have this massive failure to capture "depth". You'd have to have one insanely poor lens design to have so much contrast loss to suddenly go from "3D" to "flat". In fact, most early lenses had more veiling flare (which causes contrast loss) compared to modern era lenses. Now - yes - in some designs where OOF transitions and bokeh have been placed as higher priorities in the design process, you will see some differences in results. But we're back to tool for task - the user of a Nikon 58/1.4G is likely photographing completely different subject matter than a landscape or studio shooter (working stopped down) shooting a Sigma 50/1.4 Art or Zeiss Otus. But to globally assign all "simple" lenses from an older era with the inherent ability to show depth? Nope, not going to work out that way.
Now, I know what's coming from the blog kool-aid drinkers: "but look at his examples". I have. They are not done properly. Take for example the comparison of the Zeiss 28 Otus against the older Nikon. To me the first thing is that the lighting on the subject is different - the subject moved. So there is a seriously large difference in contrast between one shot and the other. Secondly, the Zeiss shot is slightly out of focus. Since it is a fact that contrast and resolution are married, we can no longer make a determination about microcontrast from the examples right there, and when you include that the lighting is different, you have to throw the comparison out. And then there's his B&W conversion, or his attempt to normalize exposure and WB. Well, if you truly are going to talk about contrast differences in lenses, after you've shot the same subject in the exact same light, you would have also had to somehow vary the lighting so that the exposure is identical between the shots. If you change exposure in many a raw converter, you may changing the mid tone contrast along with it. Converting to B&W invokes a default B&W conversion in the converter. How do we know if it is the best conversion (in terms of how much do we use from each channel) for the subject? The problem here is that the tests were quick cowboy jobs - borrow a Zeiss Otus at an event, snap a few shots, and suddenly you're an expert at the lens. If you don't take the time to properly test, everything you get in the results is open to question, and very likely is wrong. Another example: A shot of a childrens toy, shot with an old 35 and a new 24-70 IIRC. He pointed out the difference in "roundness" on the noise in the toy. Yet the exposure isn't the same, nor is the color, so who knows. I'm pretty sure if one had taken the shots with equivalent exposure and could somehow normalize the color differences, then and only then could you start to talk about whether one lens was flat. And even if it was, who is to say it's the number of elements?
With that last thought in mind, a while ago, before I read any of this guys blog or had any discussions about "more elements meant flatter images", on at least two occasions, I took a set of lenses out for evaluation. Back in 2012, it was a Sigma 35/1.4 Art I had just picked up for evaluation, my Nikon 35/1.4G that I owned at the time, my Nikon 24-70/2.8, and an old 35/2 AIS that came highly recommended by a forum participant. Note this was 2012; the blogger had not yet written his article. I test across multiple days, in multiple scenarios, always with the goal to try and root out test error and show outliers. When it was all said and done, the Sigma 35/1.4 Art stood out to me as being the best of the bunch, not only for sharpness, but also for a better sense of conveying a realistic sense of depth in the images. The lens that came in last place? The 35/2 AIS. While very sharp in the center, overall its presentation was a bit cooler in cast than I like, and from a "3D" point of view, while certainly good, it wasn't as good as the Sigma art or even the Nikon zoom. I did something similar when I tested the Sigma 24-35/2 Art, a zoom with a ton of fancy glass, the Sigma 35/1.4 Art, and the same 35/2 AIS. Similar results. Sharpness wise, the Sigma prime won, "depth/3d" wise the 24-35/2 zoom won (!!!!) and once again, the 35/2 AIS, the simplest design, the older lens, while extremely sharp in the center, simply lost the battle. And these are the same results as everything I've ever seen from the older glass. I've shot over 55 Nikkors - basically everything Nikon has made since the '70's, except for the Noct, the 50/1.2, and few of the exotics. I own four 55mm (old) AIS micro nikkors, one from each vintage of the F/3.5 version. Stupidly sharp, some of them, as in, crazy sharp. But none of them show any sense of dimension or 3D that makes me think the older lens is better. When I used to have the 60/2.8G AFS around, it schooled the old Nikkors in rendering. The Sigma 50/1.4 Art is in a completely different league, and I most definitely include the presentation of depth in that evaluation. I've done blind print comparisons with a group of photographers and not one of them chose the print that came from the older lens. Not one.
So that gets us to the question: Why does the blogger think lens element count reduces depth. My best guess is that something else in the photographic chain isn't being managed properly. While I do not doubt that there are subtle rendering differences between lenses, they aren't so incredibly great in magnitude that they are the reason for bad images. Take the usual "attack" on Sigma arts: "Too Clinical" they scream. Now think about this: If I were to take a shot in the outdoors with on camera flash - a pretty unfavorable light source, and the image started out as harsh, which lens is going to make the image look worse - an older lens that has veiling flare and lower contrast (and thus hides the defects in the lighting) or a modern lens that has better contrast, less veiling flare, and is sharper (and thus does NOT hide the reality of the poor lighting). The answer is obvious. I'm not saying everyone has to like Sigma arts or Zeiss Otus, but if you believe that the lens should be honest, then if you don't like the results, it's likely somewhere else you must look. Exempting bokeh quality of course, which is a different discussion. So the next question then would be: Have the haters of the modern lens actually sat a model down in front of proper studio lighting and worked a couple of thousand frames with lens candidate A and then switched over to lens candidate B to make their judgement? I tend to doubt it. I know I have - when I compared the 35/1.8G FX to the Sigma 35/1.4 Art and the Nikon 24-70/2.8 (at 35), I did just that - shot with each for 1/3 of all the shots where I needed 35mm, and then repeated the same thing on different sessions. Then and only then did I arrive at a conclusion. So if you don't like an aspect of your images, you have to get real and honest with yourself and ask if it's the lens? Or something else. If you don't know how to post process (which includes raw conversion) correctly, and don't understand how a very subtle contrast move in the curves dialog in photoshop can cause an image to go from "three dimensional and rounded" to blocky (over correction) or flat (not enough contrast), then how in the hell can you blame the lens?
Obviously, again, if you are unhappy with the OOF transitions and bokeh, then you can start talking about a lens and it's qualitiy in that context. But saying a modern multi element lens is flat? No. Something else you're doing is wrong. Sure - it's far easier to read this guys blog and have the lights come on and you suddenly think "Eureka - EVERYONE ELSE IS WRONG AND I AM NOW RIGHT BECAUSE I HAVE READ THE ALMIGHTY SPEAK' but if you really think, if you use some common sense, if you look at the consensus view on things, you'll find out that what is in that blog is all nonsense.
Take the time to learn what matters: Light properly, or make sure the subject is in the good light, or is illuminated by good light. Understand the nuances of raw conversion and post processing. Become an expert in subtle and fine granularity manipulations of contrast to achieve your goal. Once you've become pretty proficient at those, then you an start exploring the subtle aspects of lens performance. I don't care if you choose the same lenses I like - I really don't - but I truly hope that you've chosen the lenses you use for VALID reasons and not gone chasing nonsense such as that which has caused me to likely waste over an hour of my time writing this post. Strive to think more, to look at the conensus view, to realize that all of us, no matter what leve of mastery we are at, can always improve. Ignore the outliers, the conspiracy theory guys and angry youtubers and spend the time on education. Walk by those who can't test properly. You'll be far better off for it.
I will not be partaking in any further discussion of this. If you made it this far and still think the blogger was right, frankly, there is no hope for you, and there is no point in any further conversation between you and I. It is, actually, that fundamentally basic. I'm out....
References and suggested reading:
1) Zeiss PDF on the Planar designs by Dr. Nasse
http://lenspire.zeiss.com/en/wp-con...2015/09/en_CLB_40_Nasse_Lens_Names_Planar.pdf
2) Zeiss PDF on the Distagon design (and others). Also by Dr. Nasse.
http://lenspire.zeiss.com/en/wp-con...2015/09/en_CLB41_Nasse_LensNames_Distagon.pdf
3) Zeiss PDF on the Tessar design by Dr. Nasse
http://www.zeiss.com/content/dam/Photography/new/pdf/en/cln_archiv/cln39_en_web_special_tessar.pdf
4) Thread in Fred Miranda forums. Pay attention to those from (optics student) Brandon Dube
http://www.fredmiranda.com/forum/topic/1395836/0
