Interview re 58mm f/1.4G

<p>This is an interesting article: http://nikkor.com/technology/02.html</p>

<p>Also, check out mid-article where one of the designers recommends the much-maligned 18-200!</p>

<p>Regarding future advances: "I'm sure that we will forever continue our efforts to increase sharpness at the point of focus. We do not, however, feel that this alone is sufficient."</p>

<p>Thom Hogan did some commentary on this interview, although I didn't actually read the interview itself when I read his post. As he points out, the "we did a 58mm lens because we decided we wanted to do a 58mm lens rather than because anybody actually asked for one" approach does speak to some of Nikon's strange sense of priorities - not that I have anything against the 58mm, but if they'd made a decent replacement for the 135mm f/2, I'd have bought one by now. I don't really feel the need for the 58mm with the Sigma Art out there for less money. (Assuming I can get the thing to focus accurately, obviously.)<br />

<br />

I approve of designing a lens for good bokeh (that's why I own several of the lenses I do), and some of Nikon's primes are a little lacking in that area, but it's not exactly a new thing - hence the DC lenses. To my mind, some other fast primes have shown that it's possible to maintain good bokeh along with sharpness - you could argue that this is important, because if you sharpen digitally you tend to make the bokeh worse, and blurring to improve bokeh also blurs the sharp areas of the image (insert disclaimer about advanced image editing techniques and this being a simplification). I'd like Nikon to look more into controlling LoCA in small primes - Canon seem to have taken a step with their weird organic goo. On sharpness, I always thought MTF was over-simplistic, and I'm not sure why there aren't more graphs of contrast vs resolution out there.<br />

<br />

Recommending the 18-200 is either a heavy hint that they have warehouses full of the things that they can't ship, or that the Nikon lens designers never shoot DX (which explains the DX lens line-up). At least the 24-120 recommendation wasn't for the old one!</p>

<p>Andrew, I see the 58mm as a kind of prestige project as opposed to a meat-and-potatoes product. All I can say is that I have seen some very lovely images taken with this new 58G. As for your take on the 18-200 recommendation - why so cynical?? :)</p>

<p>It was clearly a prestige project, but it seems to reveal a poor grasp of what attributes would confer prestige. <br>

Also, why not do this with the 50f/1.4? Why the goofy 58mm focal length? 58mm only invites comparison with the 58mm f/1.2, and highlights the fact that F mount is too small to handle both electronic contact and f/1.2?</p>

 

<p>It is a fair question as to why the 58mm FL. I did not see any discussion about that in the interview. I like Mr. Sato's description of a "3D hi-fi lens." What does that mean? I don't know. But it sounds good. Also, the images I have seen from this lens do have a special character I think.</p>

<p>I have the old 58mm f/1.4 from like 1960 it was Pre AI and I had John White do a conversion for me. When I bought the lens I used it mostly on a D300 and then on my Kodak Pro14NX. I really like the focal length on both DX and FX and I like the way this old lens draws an image. Honestly it is not razor sharp wide open but it is no worse than any other f/1.4 lens that I have used with the exception of the 35mm f/1.4 Sigma which is very sharp wide open.</p>

<p>It can be a challenge to nail the focus but I find with practice and making sure that you have set the diopter correctly I can get a high percentage of keepers. I have used it on a friends D800 and it still holds up well with the high resolution of that camera</p>

It is 58mm because to achieve the desired characteristics was easier at 58mm than 50mm without resorting to more

complex designs that cause a tripling of the weight. I love the fact that the 58 AF-S produces beautiful results yet helps

keep the total bag weight from increasing too high. When I was in London I had the 20/1.8, 58/1.4 105/2 and 300/4 PF

and it was much more pleasant than past trips with heavy lenses yet the results look pleasing to my eye. The 300 PF

probably the least of the set in my view but I used it a lot because I hadn't had that focal length before in a travel-friendly package.

<p>Michael - I have seen more than a few praises for the old 5.8cm Nikkor. It is reported to have a nice character, and it sounds like you concur. I'll have to keep my eyes open for a good deal on one of these.</p>

<p>Very nice read. And, I am glad Nikon designers are addressing the overall lens reproduction capabilities.</p>

<p>It's pretty clear that the aftermarket lens suppliers have a single goal in lens design, plane of focus sharpness and resolving power. This is driven to a large extent by consumers with high resolution cameras shooting test images and declaring "this is the sharpest lens I have ever owned!" Case in point, the Sigma 35/1.4 Art lens. Widely declared by the forum elite as vastly superior to the Nikkor 35/1.4G. Yet, I find that Nikon images much more pleasing than the Sigma's.</p>

<p>My 2¢</p>

<p><em>it seems to reveal a poor grasp of what attributes would confer prestige.</em><br>

<em><br /></em>I don't look at it as a "prestige" item, but simply a project to design a fast normal lens that produces images that the designers thought were pleasing and advance their conception of how a fast normal lens should render images (the idea is that the transition to out of focus areas should be smooth rather than "jump" at the viewer), and because most high end lens buyers are thought to want sharpness above all, this lens was never about selling in high volume (though it will probably sell >10x as many copies than all versions of the Noct Nikkor combined, so what is considered specialty item is relative). To me it is one of the most exciting lenses in recent years because I use wide apertures a lot in my documentary and event photography and I love the natural results that this lens gives. The sharpness is adequate (not so great at short distances wide open, but excellent at mid to long distances) and since these are documentary photos they're rarely printed large in any case, so the issue of slight softness at close distances wide open is hardly of much consequence (the 60mm AF-S Micro is the lens to use for those shots). I totally understand why someone would not share my view on this lens preferring one's fingers to be cut if touching the print but that is the very reason why there is such variety of products in the normal lens range of various manufacturers, so that people with different expectations and preferences can be satisfied. And while the people whose photographs I make (or whose events I photograph) might not know why they like it, I get a lot of complements for images made with this lens. They can see it too without having a photographer's background. The 105 DC is another lens which has a truly great rendering of people subjects and again it is not too heavy. Interestingly I had severe difficulties autofocusing the 105 DC on the D800 (I don't recall any such problems with 35mm film cameras) requiring up to a 20 point difference in autofocus fine tune at long vs short distances. With the D810 this problem is absent, and I use -12 or -13 depending on whether the subject is static (-12) or approaching (-13 if I recall correctly) and get excellent accuracy. This highlights the reason why there was never any question whether I would switch to the newer body once I heard reports of its autofocus performance with fast primes. With the 58/1.4 the focus is ok, not the best but overall no serious AF issues with that lens in my use. The aberrations must contribute to the AF performance, so a truly sharp lens with a very precisely defined in focus sharp area is likely to autofocus decisively, this is certainly true of the 300mm PF and 70-200/2.8 II, but a lens that is designed to project the smoothest transition from in to out of focus is likely to have less precise AF behavior, or this is my perception about it.</p>

<p>Returning to what is important in a lens, the reality is that Sato and his colleagues actually understand the way photographs are used, which is to communicate, remember, highlight, and transform something perhaps even mundane into art. Most images that are actually <em>used </em>for some purpose are not printed large, but shown on a magazine page, viewed on a computer screen (full image, not zoomed into a partial view), or printed small in an album. The exception is images shown in banners, posters, galleries, and as wall decoration where large prints are commonly used, but here's the gist: to see the proper meaning of the image you look at it as a whole and take its message in from that view; not 10cm away with a magnifying glass! Thus the finest details rarely convey any significant part of the message of the image since the viewing distance is such that your visual field includes the whole image at once. I totally understand that people want to have the sharpest lenses because it allows some flexibility to crop, and this is important for some subjects such as birds (especially when working with a lens that doesn't quite have a long enough focal length or is a prime lens so the framing isn't easy to adjust) or landscapes (where the rich texture conveys a feeling of realism and the prints can often be large), but the AF-S 58/1.4 isn't primarily a lens meant for those applications (it <em>can</em> be used for special effects in landscape but I wouldn't think it was designed for that purpose) but for photographs of people with a more dynamic and perpective and sense of depth than with a longer focal length lens (also useful in tight quarters), and secondarily for the original application of the Noct Nikkor which is for night urban scenes. Incidentally the softness of the 58/1.4 AF-S is present really only in the near range (in my opinion this is not a great flaw in a portrait lens and it is easy to avoid if you're not making close-ups) and at distance it is very sharp even at f/1.4 and improves upon the Noct Nikkor in that there is less vignetting (and some reports say less coma as well though I've never had the chance to compare). I think those who think that a lens's purpose is to render images that can be zoomed in by 50x on a computer screen and oohed and aahed at the detail are the ones lacking perspective quite frankly, and Sato and his colleagues are looking at the big picture of photography, where the image content is presented to an audience and the medium and viewing distance limit the importance of the finest detail. Of course, Nikon and other manufacturers also produce lenses which have the highest possible rendering of detail, to satisfy those customers who want these characteristics above all. I have a few such lenses but apart from the 200/2 they don't really get all that much use. Somehow the 200mm is able to be extremely sharp and deliver very pleasing images ... but the price is paid in monetary terms and also in the size and weight of the lens, and it is one of the lenses that Nikon sells only in small quantities, even though it produces very fine results. High quality doesn't automatically result in high volume of sales; practicality and cost are very important as well. This is why I doubt the new 24-70/2.8 VR's commercial success. It is extremely fine ... but also very expensive and very large for its focal length and aperture.</p>

<p>I'm <em>not</em> a big fan of some of the other lenses mentioned, e.g. while the 18-200 is a thrill to shoot with, it is really mostly interested to a consumer and I didn't particularly care for the results at the long end. But I am picky about the cleanliness and clarity of an image and the superzooms just don't deliver images that I would prefer. Also the 24-120/4 is not exactly a favorite lens of mine but I do think it is much better than previous extended focal range wide to tele lenses and had similar range. I judge lenses on a case by case basis and the most important factors are that they produce results that look pleasing to my eye when applied to situations which I want to photograph, according to my subjective criteria (lens choice contributes to the look and personality of my images).</p>

<p><em>highlights the fact that F mount is too small to handle both electronic contact and f/1.2?</em></p>

<p>The 58/1.4 has less vignetting and more even performance across the frame than the 58/1.2 (Noct) which is probably why they selected to manufacture an f/1.4 lens instead of one of the f/1.2 designs that they worked on (some of which they patented). Also the front element is smaller and makes the lens less heavy. Finally in an autofocus lens f/1.2 focusing would be quite difficult to achieve in a consistent fashion, f/1.4 is a bit more realistic.</p>

<p>Another advantage of not pursuing focal plane sharpness as the overriding objective is that based on recent history, reasonable degree of focal plane sharpness would be overtaken in just a few years. Lens considered outstandingly sharp in 2008 is now considered below average compare to their peers.<br>

However, there seems to be less pressure to ratchet up the general standard of performance in other areas.<br>

So a lens considered outstanding because of its focal plane sharpness is not likely to last long in its tenure. However, a lens considered outstanding in other areas would last longer in its tenure as being an outstanding lens.</p>

 

<p>Good point Chuck, the 135/2.0 DC is a case in point. Not the sharpest tool, but gosh, it produces lovely images time and time again.</p>

<p>Excellent point, Ilkka; and I hope these Nikon designers do get some leeway to come up with more lenses as the 58mm f/1.4 (I wish I could afford one, but no such luck). Not make the MTF chart and DxO Marks the leading design philosophy, but how a lens renders and how a resulting image looks. Hopefully we'll see more of it in future lenses.<br>

And sure, to keep the company financially sane, up and running with a marketshare large enough to warrant what they're doing, they do need 18-200VR's and lenses like it. Nothing wrong with that, there is a much bigger market for those, even if most of the people posting here aren't in that particular part of their marketshare.</p>

<blockquote>All I can say is that I have seen some very lovely images taken with this new 58G. As for your take on the 18-200 recommendation - why so cynical?? :)</blockquote>

 

<p>I've seen nice photos taken with the 58mm too. I don't dispute its merits, and I never liked the price/performance trade-offs of the Nikkor f/1.4 50mms (or 85mms!) particularly in the bokeh. But, like Ilkka, I've seen lenses that can combine sharpness with nice bokeh, and the impressions I've had of the 58mm are that it doesn't compare all that well to the Sigma Art or the Otus. But I don't own a 58mm, so all I'm doing is summarising "the internet". I really would have thought that it would be in Nikon's interests to prioritise other focal lengths, though.</p>

 

<blockquote>and highlights the fact that F mount is too small to handle both electronic contact and f/1.2</blockquote>

 

<p>I've not tried to work this out (despite f/1.2 lenses coming up a few times), but is that "fact" definitive? I appreciate that a wider rear element makes it harder to fit things in, but I wasn't aware that the AF contacts are any more central than, say, the aperture lever. I'll have to look at a lens to be sure of that. Not that I'm requesting that Nikon come up with some f/1.2 AF lenses any time soon, just checking the validity of information.</p>

 

<blockquote>I like Mr. Sato's description of a "3D hi-fi lens." What does that mean? I don't know. But it sounds good.</blockquote>

 

<p>I once saw a printer described as a "multimedia printer". No, it didn't have a CD-ROM drive and speakers. But yes, "it sounds good" is exactly what I'd take from "hi-fi lens". At least we're not talking an "HD DSLR" (ooh, two megapixels...)</p>

 

<blockquote>And, I am glad Nikon designers are addressing the overall lens reproduction capabilities.</blockquote>

 

<p>And they weren't before? The 1001 nights stories about the DC lenses, the f/1.4 lenses etc. all talk about bokeh. I can't imagine someone would have looked at the 85mm f/1.8 AF-D and created the 85mm f/1.4 AF-D if they were aiming for sharpness, especially off-axis. It's true that Nikon seem to have been trying to increase sharpness more recently, partly because MTF charts get compared and partly because you can see the mush better if you use an old fast prime at wide apertures.</p>

 

<blockquote>It's pretty clear that the aftermarket lens suppliers have a single goal in lens design, plane of focus sharpness and resolving power.</blockquote>

 

<p>That's far from clear to me. I certainly don't think "sharpness" was a concern for the SLR Magic Hyperprime 50mm f/0.95 (not Nikon mount, obviously). The 50mm pre-Art Sigma f/1.4 seemed to have more to do with bokeh than not. Sigma and Zeiss both claim to be doing things about LoCA, as - with the 135mm - does Samyang. Yes, there are some sharp lenses with nervous bokeh - you could argue that for several Zeiss lenses, but you can also argue it for the Nikkor 85mm f/1.8 AF-D.</p>

 

<blockquote>This is driven to a large extent by consumers with high resolution cameras shooting test images and declaring "this is the sharpest lens I have ever owned!" Case in point, the Sigma 35/1.4 Art lens. Widely declared by the forum elite as vastly superior to the Nikkor 35/1.4G.</blockquote>

 

<p>Bokeh is rarely my highest priority in a 35mm lens, though I admit 35mm is borderline. I <i>do</i> care at longer focal lengths, and at that point it's usually easier to have sharp corners. I'm not a fan of wide lenses that are only sharp in a tiny area in the middle. But then I mostly use 35mm lenses in the dark, not for creative subject differentiation. I can't really claim to be a 35mm specialist, though.<br />

<br />

I completely agree with Ilkka that other aspects of rendering are more important to me than resolution - bokeh and LoCA are visible from a far longer distance than microcontrast. I would assert that some new designs are showing that sharpness and good bokeh can be combined. I include the 85mm f/1.8 AF-S in that, for the record (although its LoCA bugs me). The 200mm f/2, while not perfect, is an extremely good combination of optical features if you don't care about weight and cost. The 135mm DC manual actually had a warning about autofocus accuracy, at least with the DC set to non-neutral; mine was horrendously inaccurate even on a D700. I didn't mind that it wasn't terribly sharp (at wider apertures) - again, it was the LoCA that killed me. A bride with a blurry crowd behind is good; a bride with green hair, not so much. So large-level detail was the problem, not fine detail - the problem with issues on edges is that hair and clothing have a lot of edges. I appreciate that others seem to have got on better with theirs.</p>

 

<blockquote>However, there seems to be less pressure to ratchet up the general standard of performance in other areas.</blockquote>

 

<p>Not from me!</p>

 

<blockquote>Excellent point, Ilkka; and I hope these Nikon designers do get some leeway to come up with more lenses as the 58mm f/1.4 (I wish I could afford one, but no such luck). Not make the MTF chart and DxO Marks the leading design philosophy, but how a lens renders and how a resulting image looks. Hopefully we'll see more of it in future lenses.</blockquote>

 

<p>I want Nikon to have leeway to focus on aspects other than sharpness, especially for focal lengths where subject isolation typically matters (longer ones). I'm not sure that's he same thing as "I fancied doing a 58mm as a personal project". Bean counters and MTF bad. What customers actually need, good. Where's my replacement 135DC?</p>

 

<blockquote>And sure, to keep the company financially sane, up and running with a marketshare large enough to warrant what they're doing, they do need 18-200VR's and lenses like it.</blockquote>

 

<p>Well, they would if the 18-200 was substantially better. Instead, we have an ancient 18-200 design that can't keep up, and nine different current 18-xxx zooms (not including the 16-xxx and 17-xxx zooms).</p>

 

<blockquote>Nothing wrong with that, there is a much bigger market for those, even if most of the people posting here aren't in that particular part of their marketshare.</blockquote>

 

<p>I feel Nikon should sell one 18-200 (for people who value convenience over cropping), one 18-xxx moderate DX street sweeper that's actually optically decent (like the 24-120), and then actually make some fast DX primes for people who want a small system. Not that I care myself - I only have an FX body - but there's a lot of very consistent complaining, and I don't buy that market research really said "we need lots of basically identical lenses and we want to let people keep saying that we have a less complete DX lens line-up than Fuji". But I don't really claim to be qualified to run a megacorporation.</p>

<blockquote>

<p>Widely declared by the forum elite as vastly superior to the Nikkor 35/1.4G.</p>

</blockquote>

<p>this comment suggests that claims of optical superiority are purely anecdotal. the fact is that every single lens review i have seen scored higher marks for the Sigma 35ART, making the Nikkor seem vastly overpriced. Maybe it has other aspects than sharpness going for it, but "unique character" doesnt register on MTF charts. </p>

<blockquote>

<p>Maybe it has other aspects than sharpness going for it, but "unique character" doesnt register on MTF charts.</p>

</blockquote>

<p>This is my take away from the 58G interview. "Special character" does not allow for easy internet comparison. Sharpness is possible to measure and sharp lenses make waves with many photographers. My take on Dan's comment: the Nikkor may not have 500hp, but that doesn't necessarily make it inferior. Aka "sharpness is overrated." My favorite lens is my Sigma 30mm, which is not my sharpest lens by a long shot. </p>

<blockquote>

<p>this comment suggests that claims of optical superiority are purely anecdotal.</p>

</blockquote>

<p>No, it suggests that the internet forum elite generically seems to have a different preference than Dan; and nobody is right (or wrong). The point isn't "optical superiority", the point is how well you feel a lens does the job you want it to do. Somehow people always want to declare a winner, "the best", the one we should all consider the king of the hill - and I agree with Dan that this is a nasty (internet) habit. We do not all have the same priorities, so what is best for one, might be a lousy choice for another. Good thing there is choice, and again, I think good thing Nikon is seeking to provide that choice more consciously by respecting different design priorities.</p>

<p>Andrew, sure listening to the customers is a wise route; but "What customers actually need, good. Where's my replacement 135DC?" - without knowing the actual sales of 135mm lenses versus the 58mm lenses, you're making the same error as the rogue engineer that goes for 58mm. We don't know to which extend the bean counters have had a say - we only have this interview, we don't know what really happened indoors.</p>

<p>Personally I have had no problems with the 135 f/2 DC. I even use it to shoot equestrian events with. Works great on a D4 and I have not had any problems with it on borrowed D800's</p>

<p>D4 and 135 f/ 2 at f/3.2<br>

<p>Tight crop from the same shot</p>

<p>

<p>The 58 does take creamy-nice photographs. There's a 58 picture thread on the FM/nikon forum, lots of good stuff in there. But it only gets 3.5 stars on photozone. Mediocre grades from the other testers as well. There's more to it than plane of focus test results. <br /> Sato-san from the Nikon team said it pretty well:</p>

<blockquote>

<p>Photography is the process of compressing three-dimensional subjects into two-dimensional images, so we must consider what is preferable or what makes for a pleasant image; in other words, the optimal method of evaluating the act of compression.</p>

</blockquote>

<p>Even allowing for some loss in translation, that description of the 58mm f/1.4 Nikkor's design philosophy comes across as complete new-age BS.</p>

<p>Bokeh can be completely described by existing and established optical design rules and measurement. It's well-accepted that the character of OOF points is governed by the amount and direction of residual spherical aberration, and there really isn't any conflict between developing a lens with high resolution and good bokeh. Except that SA effectively extends the plane of focus into a region-of-focus and hence increases apparent depth-of-field, at the expense of image micro contrast. It's also hard to separate LoCA from residual SA, such that there's no colour "halo" around the OOF points close to the plane of paraxial focus.</p>

<p>This region of "extended" focus between the plane of paraxial focus and the plane of best contrast was discussed at least as far back as 1943 by Arthur Cox in his book Photographic Optics. So looking at this aspect of lens design is nothing new at all. And Cox uses plain and scientific language, not stupid expressions like 3D Hi-Fi.</p>

<p>Anyhow. It seems to me that deliberately designing a lens with less than optimal resolving power and then selling it for a high price against the likes of Sigma, is just asking for poor reviews. As seems to have happened. Plus I don't read any slamming reviews of the Zeiss Otus lens, which has absolutely outstanding resolution, along with bokeh that no-one seems to object to. But then any lens is only as good as the photographer who's wielding it.</p>

<p> </p>

<p>Why design a lens for less than optimal resolving power, unless through it one gains some other attribute? If less than optimal resolving power is really desired, why not use firmware or manual focus to achieve it so that optimal resolving power can be recovered if desired?</p>

 

<p>The point is that the Sigma 50/1.4 Art and the Zeiss Otus 55mm f/1.4 weight almost 3x that of the 58/1.4 Nikkor, and thus ignoring completely the weight their designers could implement more complex optical designs than the Nikkor where the lens weight was kept similar to that of classical fast 50mm lenses for the 35mm format. Using more complex optical designs it may be possible to achieve better optical characteristics than in a simple, lightweight lens, but for me at least if my kit increased to 3x in weight I would consider that a significant drawback. Optical design is all about balance of characteristics, much as all engineering. Similarly Volkswagen run into trouble with US environmental regulations for nox emissions in their small diesel cars. Larger cars are easier to implement with urea-based systems that emit much less nox because there is more space available within the chassis of the car without compromising the applications characteristics of the car (interior space, acceleration etc), but small cars don't have enough space for such systems (and a good sized fuel tank) and if they do implement them, some other aspects will have been compromised (interior volume, fuel tank size, possibly have to use a smaller engine), not to mention the additional cost of such systems which affect the buyer of the smaller car more than that of a large car (the relative price increase is smaller in the larger car).</p>

<p>You're welcome to suggest a fast normal lens with better out of focus transitions, global contrast, flare resistance, and resolution than the 58/1.4 AF-S Nikkor with similar or lower weight and show that it achieves a better result. I would be happy to consider it on its merits.</p>

<p>The point Rodeo seems to be making is smoothness of out of focus highlights is independent of the lens' resolving power. So 58/1.4 could have achieved resolving power at least matching those of any lens at its price or lower without sacrificing the nominal objective of better out of focus transition.<br>

The fact that it didn't achieve that level of resolving power suggest resolving power was deliberately reduced.<br>

So the question is for what real reason was it reduced, if the real reason could not have been to improve out of focus highlight.</p>

<p> </p>