reframing the world one picture at a time
KLCC Convention Center. Fuji FinePix F10
One from the archives – 2005, to be precise.
I’m in the middle of retouching a huge batch of product images for a client – 36MP also means 3x more retouching time, unfortunately – so today will be pretty quiet. I’m still planning to get the reader critiques up later today (thank you to all of you who submitted – there are far more than I expected, so I’m afraid I won’t be able to do everybody). MT
This post is a quick update to my D800 autofocus issues, as well as commentary on the specific performance on some of the more popular lenses people have been asking about. I don’t have time to post crops, but I think most of you would trust that I know what I’m doing.
I spent the morning at Nikon. Ostensibly, to collect my replacement D800, and a PC-E 85/2.8 Micro. However, it turns out the replacement D800 exhibits the SAME autofocus issue – namely, with wide angle lenses, the center and right side AF points yield in focus images, the left side bank is way out. This is especially obvious with the 24/1.4 G. We also tried their NPS loaner demo units and their D800E sample. The results were mostly the same – all of the D800s showed identical results. The D800E was a bit better, but still noticeably soft on one side. It gets worse: I’ve had a number of emails from people with cameras in the same serial number block – below 1000 – and the 24/1.4, who are finding the same thing. Apparently it is a serious issue, because my NPS rep told me that HQ has asked for updates and is looking into it on the production line.
Perhaps it was my 24/1.4 sample that was the issue – nope, because it works fine on a D3x, D4 and D700; we tried another 24/1.4 which showed consistent results – that rules out lens problems. I think we can also rule out sensor alignment problems as I don’t see any odd shifts in the focal plane when focusing using live view.
The upshot is that it will take them two days to diagnose the problem, and possibly longer to fix. Since it’s only an issue with wide angles, and not an issue with anything above about 50mm, I elected to keep this body for the time being – I’m only using it in the studio with the 60/2.8 G Micro, and now the 85/2.8 PCE. It looks like I will be reviewing both E and non-E after all – I’ve elected to take a D800E for the replacement unit.
There were more surprises in store, though – specifically, with lenses.
Summary of Nikkors tested so far on the D800:
AFS 14-24/2.8 G: Not good at 14mm; obvious corner sharpness issues. Displayed AF issues at 24mm. Center is sharp. T stop is probably closer to f4 than f2.8. Average to good performer.
AFS 24-70/2.8 G: Sharp everywhere in the range, at every aperture. No AF issues, even at 24mm. Excellent performer.
AFS 70-200/2.8 G VR II: Sharp everywhere in the range, at every aperture. 85mm setting better than the 85/1.4 G at f2.8, and comparable to the 85/1.8 G at f2.8 (yes, you read that right. The 85/1.8 G is better than the 85/1.4 G.). No AF issues either. Excellent performer.
AFS 24/1.4 G: Sharp everywhere except extreme corners at every aperture if you live view – remains an optically amazing lens, but now even more fiddly to use thanks to the AF issues. Three copies all displayed left-side softness on the D800, but not on other bodies. Cautiously, I’d say excellent performer, to be confirmed once I have a properly working body.
AFS 28-300/3.5-5.6 G VR II: Sharp everywhere if you close down the aperture on stop. Microcontrast not great, but serviceable. Overall global contrast is good. Color a bit odd. Good to very good performer. No AF issues, probably covered by depth of field and small apertures.
AFS 35/1.4 G: No good wide open. Center is okay, both sides are not good – even after AF fine tune. Not recommended. The 24-70 performs much better at 2.8 than the 35/1.4 does at the same aperture. Slight AF issue noticed, same as 24/1.4.
AFS 85/1.4 G: Inconsistent. Wide open displays LCA and LoCA at edges. Nowhere near as good as it was on the D700/ D3/ D3s. Stopped down to f2.8, it improves, but only to about the same level as the 70-200/2.8 II wide open. Note T stop is pretty high for this lens though – probably 2/3 stop more than the 70-200/2.8 II, and half a stop more than the 85/1.8 G for the same aperture. Good to very good stopped down. Honestly, I’m not liking this lens very much anymore.
AFS 85/1.8 G: Incredible. Sharp everywhere at every aperture, no LCA or LoCA. Bokeh is neutral, not quite as good as the 85/1.4 G. Surprising considering this lens has no ED glass, Nano coating or aspherical elements. It’s honestly an optical masterpiece, and very, very cheap. If you need an 85mm and don’t have the 85/1.4 G already, I’d suggest buying one of these. Performance at wide open at f1.8 is better than the 85/1.4 G at 2.8; it matches or slightly exceeds even the 60/2.8 G Micro at the same distances. You’re probably wonder what’s the catch: two things; T stop and build quality. T stop is half a stop down on the 85/1.4 G for the same aperture, and it’s light and plasticky. Still weather sealed, though. Excellent plus performance, no visible AF issues.
PC-E 85/2.8 Micro: This is the only lens of the group tested that could best the new 85/1.8 G, and by the slightest of margins (or maybe both lenses out resolved even the D800E sensor and we’d need something even higher density to see the difference). Global contrast is a little lower than the 85/1.8 G, but micro contrast has more bite and structure to it – reminds me of the Zeiss macros. Excellent plus performance. (I took this one home, after relieving my credit card of some of its available balance. Look out for a full review in the future once I get a chance to shoot it in the studio.)
PC-E 24/3.5: A truly excellent piece of glass. Matches the performance of the 85/2.8, but at 24mm. Shame about the small aperture, though. Handily focuses to about 20cm – which is about 3cm from the front element of the lens. I’d say sharpness performance of this and the 24/1.4 G at f4 are about the same, however the micro contrast structure of this lens is almost Zeiss-like in detail. Excellent plus performance again. And whoever said it won’t mount is wrong – it mounts and offers full movement just fine, but you must zero all of the movements before trying to mount it, and there are certain orientations that work better than others (big knobs vs small knobs near the prism etc.)
AFS 60/2.8 G Micro: Although this was my reference standard on the 12MP FX bodies, it’s performance clearly isn’t up to the D800’s demands: I’m seeing plenty of longitudinal CA (especially in the bokeh) that wasn’t there, or almost negligible, on the D700. It’s sharp already at f2, but not critically bitingly sharp til f4-5.6; your working aperture range is somewhat limited because diffraction kicks in noticeably by f16, and it’s unusably soft by f25. I’d say f22 is probably best reserved for emergencies. This is the main reason I got the 85/2.8 PCE: lack of depth of field control. I’d put it in the good-to-excellent range.
A word on the D800E: I didn’t have a lot of time with it, but from what I can see, there is a slight but noticeable difference in fine micro contrast, as well as sharpness and resolving power. It seems to offset diffraction to some extent. However, file sizes will be even bigger, and lens demands even higher. Recommendation: use with caution, requires controlled circumstances to get the most out of it (tripod or studio lights, low ISO).
Conclusion: If you plan on getting the most out of your D800/D800E, you’re going to have to rethink your lens lineup. What worked brilliantly for me on the D700 – as in I felt I couldn’t get any more image quality out – isn’t working on the D800. And there are a lot of surprises here; not all of them good – the 85/1.4 G and 24/1.4 G are good examples of this. It seems that one has to now choose for a lens set optimized for studio work (or slightly brighter light conditions) – 24-70, 70-200, 85/1.8G, 85/2.8 PCE – with the compromises that brings for available light work, especially now that you’re going to require more shutter speed to handhold and the sensor loses out a stop to the D700 at the pixel level – or run two sets of lenses. This obviously isn’t ideal, or cheap. I feel the latter route is likely the way I’ll have to go – probably with the 85/2.8 PCE for the majority of my studio work, and a Zeiss 21/2.8 or 24/3.5 PCE for architecture and interiors.
Am I happy with feeling like a bit of an expensive guinea pig? Not one single bit. I think this latest push in resolution has brought up manufacturing tolerance and QC issues that were never previously noticeable. But at least a) it works under a known range of conditions, and more importantly NPS here deserves credit for doing their best to rectify the situation, and at least provide me with a working solution in the intermediate period (D3x on extended loan for high-res WA work).
It’s not ideal, but when you get everything right, the D800 is capable of delivering pretty darn amazing image quality. The trouble is, once you’ve seen it, you really don’t want to give it up – even if it is a colossal pain to achieve. Of course, none of this will be news to seasoned medium or large format shooters – but for anybody expecting to go from a DX consumer body, or even 12MP FX, to D800 and get pixel-level crispness across the frame, there’s going to be something of a steep learning curve to climb. MT
Check back for more updates once my D800E replacement body (finally) arrives at the end of the month.
____________
Visit our Teaching Store to up your photographic game – including Photoshop Workflow DVDs and customized Email School of Photography; or go mobile with the Photography Compendium for iPad. You can also get your gear from B&H and Amazon. Prices are the same as normal, however a small portion of your purchase value is referred back to me. Thanks!
Don’t forget to like us on Facebook and join the reader Flickr group!
Images and content copyright Ming Thein | mingthein.com 2012 onwards. All rights reserved
A little while ago, I was on a company trip to Melaka/ Malacca – happened to attend Friday prayers at one of the more pleasant mosques I’ve been to, on the outskirts of the historical city of Melaka/ Malacca. The open, airy architecture made the most of its location by the sea, oriented well to allow cool breezes to flow through and keeping the internal temperature pleasant. I got some images before and after the main prayer session.
Series shot with a Sony NEX-5 and 18-55 kit lens.
If you’d like to have your image torn to shreds in a very public (but objective) manner by yours truly – with the ultimate objective of sacrificing yourself for the greater educational good – I’m taking submissions again for the reader image critique. There are a couple of past examples here and here. Please send an image no more than 800 pixels wide on the long side to mingthein(at)gmail.com. Thanks! MT
The two images below demonstrate the difference between choosing your perspective before you shoot and then composing around it, against choosing your angle of view then just picking whichever focal length happens to match where you are standing:
2008. (Yurakucho, Tokyo; Nikon D3, 24-70)
Being self-critical, I was trying to ensure that all parts of the frame were filled; shooting with 24mm on full frame was a new experience for me, and the one mistake I continually made was not putting the subjects close enough to the foreground to be identifiable. Whilst a visually appealing image with nice lighting and geometry, there’s no real focus; the man should be in the gap closer to the traffic light to balance out that space. The most important thing to note here is that I wanted to get the whole scene in the frame to give it context – but failed to put a subject in.
Late 2011. (Vienna U-Bahn, Leica M9-P, 50/1.4 ASPH)
Note how the diagonals converge onto the center of the image, precisely where it’s focused; there’s symmetry in the arrangement of the people to the left and right of the pillar, which balances the composition. I was shooting with a normal lens this time, and still managed to get some perspective in the frame. The red frame further focuses your eye onto the subject, which (in a large print) would be clear – the sign. The image is a metaphor for choices – left, right, back, forward – everybody is different, so everybody makes the choice that suits them best, which is why there are people on both sides of the platform. I’m pleased with this image now, but I have no doubt that I’ll find something to improve on if I look at it another five years down the road.
Metaphors and stories in an image aren’t so easy to capture, especially when you only have a very short amount of time to get the shot – most of the time it’s instinctive, if you happen to capture anything of deeper meaning at all. If you don’t, try again and practice until it does actually become intuitive. I’ll be the first to admit that I don’t manage this consistently, and it’s something I need to work on. MT
Fog. La Tania, Les Trois Vallees, France.
As ever, there’s a moral to this story. Look at the above image.
Suppose I tagged it ‘shot with the first preproduction Leica M10’ – there would be soaring traffic, minute discussion, questions over grain and image quality, people wondering why I didn’t upload a full size image, others gushing over the lens sharpness…etc.
Now suppose I left it as is, i.e. with caption only and no camera info – it would be seen as an image only, and merits judged accordingly – commensurate to the subject, composition and technique. (I like the image very much, but then again I clearly suffer from personal bias.)
Consider a third scenario. The truth: the image was shot in 2005, with a Nikon D2H – a camera that was already perceived as being under-specced noisy technology at the time of its release, with a mere 4MP and ISO 1600 that had to be used with extreme caution. Does it make it any less of an image? I should think not; if anything, the fact that it was possible to capture this tough scene – it was dark, extremely foggy and low in contrast – the camera focused cleanly on the trees at f2, with one of Nikon’s sharp but frequently miscalibrated DC lenses (in the days before AF fine tune) – should say something. The tonal range was also pretty challenging; frankly, at the time I remember being amazed that there was anything there at all, other than a white mess. If I reshot this today with a D800, would it be better? Better technically, yes. Better compositionally, I doubt it. The camera doesn’t influence that part of the image making process.
Forget what other people think: what would you think of the image in each one of those three scenarios?
Moral of the story: it really doesn’t matter what camera you use. Get over the gear lust: cameras are tools; some work better than others for a given purpose – know your tools, select the right ones, and that should be the end of equipment masturbation. Ultimately, it’s the sack of meat behind the viewfinder that makes the difference, not the metal. MT
Print announcement
Starting today, all of the current and past POTD images will be available as a limited edition (maximum 20) 13×19″ numbered and signed fine art print at US$300 delivered by courier anywhere in the world; please contact me for details.
The editing suite. Maitres du Temps Chapter 2, Nikon D800, 60/2.8 G Micro
At 100% view on a 15″ MacBook Pro, (1440x900px) note the little red box in the navigation pane at top right: this is how much (how little?) of the whole image you can see at once. Let’s just say that retouching takes a lot longer when you’ve got this many pixels, and the product has to look perfectly flawless. I of course have to remove my reflection from the watch crown – you could probably tell what lens I’m using, since the 60/2.8 G has a fairly distinctive front (small element, big vanity plate, no markings). MT
There’s a magical whimsy about hot air ballooning. It’s always appealed to me because of the silence, grace and intimate connection with your environment that you can’t get from other forms of flying; unfortunately there’s also the issue of complete lack of control over navigation, and the fact that your sole protection from the elements is a wicker basket, and the only thing keeping you from crashing is a thin bag full of gas. But still, I do enjoy watching them fly and very occasionally going up in one myself. MT
Images from a variety of events.
Fire. Nikon D700, Zeiss ZF.2 2/28 Distagon
Unhappy at being left behind. Leica M8, 35/2 ASPH
Rising. Leica M8, 35/2 ASPH
Watching the show. Nikon D700, 24/1.4
Evening glow. Nikon D3100, 85/1.4 G
High priest saluting the power of fire. Nikon D700, 24/1.4
Steve, head chef. Nikon D700, 85/1.4 G
Although I’m not a portrait shooter primarily, there are occasions on which one is required to rise to the occasion – usually for something related to another job. A long time ago, I did shoot fashion. If you have professional models or outgoing, confident people, it’s pretty easy to get a decent headshot. But it’s difficult to capture their personality because they’re always in ‘professional’ mode around a camera – you can’t show them for who they really are, because they’re not comfortable enough to show it around you.
The flip side of the coin is when you have a person who clearly has a very distinct character, but suddenly gets uncomfortable and shy in front of the camera. Steve is a great guy, and very, very passionate about what he does – but completely blocks up and gets stiff in front of the lens. The trick I always use is to fire away a lot of frames until they get used to you; then go in for the stealthy kill: this shot (for a client) was not one of the hundred or so ‘posed’ frames we did; that was my intention all along. It was captured during a coffee break while he was talking to the client’s marketing director. I had the camera by my side on the chair, which I raised stealth-ninja fashion and got off a double tap to the head: bullseye. We landed up using this image for the final selection. Moral of the story: anticipate, and be ready at all times. MT
For those of you with legacy system lenses, compact system cameras seem to make sense, yes? You can use all of your old lenses on newer bodies via adaptors and you’ll not only get to use your favorite optics, but you’ll save money to boot. And the really short back flange distances of some of the lenses mean that just about everything from an SLR or RF mount will fit with the right adaptor.
Well, not quite: firstly, all the fields of view will be different unless you’re going between APSC and APSC. Then there’s also the issue of size: why buy a compact mirrorless systems with the intention of reducing weight when you then go stick an enormous lens on the front?
Bottom line: there are many disadvantages to using legacy glass on mirrorless systems; more so than advantages. I knew that going in, and only use my micro four thirds system with dedicated lenses; but I have got a whole drawer full of adaptors, mostly purchased out of sheer curiosity.
If you are hell bent on making unholy pairings, then read on.
Quick aside: ‘Good’ means lenses work well, and deliver excellent or outstanding optical results; better than the system’s native lenses. ‘OK’ means that they work about as well. ‘Bad’ means that you shouldn’t bother; most of the time this is due to non-telecentric wide angles not playing nice at the edges due to very short back flange distances and a lack of offset micro lenses on the sensor to counter sharp angles of incidence between image rays and sensor. This manifests as purple fringing, chromatic aberration, vignetting, and corner softness. It could also mean the lens just doesn’t have enough resolution to deal with a very high density sensor – for instance the 24MP Sony NEX-7. There are exceptions to this rule, and where I’ve found them, I’ve noted them. The Ricoh GXR M-module is an exception because it does have offset micro lenses and was specifically designed for RF glass.
Sony NEX
Good: Legacy Sony/ Minolta telephotos and normals; new Sony midrange and high end zooms; Leica M telephotos
OK: New Sony wides; Leica M normals
Bad: Older wides; Leica M wides; C-mount/ CCTV lenses (won’t even cover the image circle!)
Micro Four Thirds
Good: SLR high end telephotos, SLR normals, RF telephotos. Exception: Zeiss ZF/ZF.2 glass, Leica 35/1.4 ASPH FLE.
OK: SLR midrange telephotos; some of the slower RF normals
Bad: Leica M wides (pay attention to your adaptor: cheaper ones probably won’t be planar, and land up causing obvious astigmatism. They may not even focus to infinity, or minimum distance!). The Leica Noctilux 0.95 does not do well on micro four thirds; it displays a lot of CA and blooming. I suspect that it is only optimized with the focal plane precisely at 28mm from the flange. Most C-mount/ CCTV lenses – these lack the resolving power and flatness of field.
Nikon 1
Good: Any of the new Nikon lenses, say post-2009; almost all of these are telecentric and of sufficiently high resolving power. Any of the SLR high end telephotos, RF telephotos
OK: RF normals, all other cheap telephotos; older Nikon MF glass; RF normals
Bad: Cheap C-mount/ CCTV lenses – these lack resolving power; RF wides
Ricoh GXR
Good: Any Leica M
Behind the scenes videos
If you enjoyed the content, please consider 
-ing or subscribing - thanks for your support!
