Camera Labs

This review is now finally edited, and can be found here.

If you have any question, comment or suggestion regarding my review head over here.
Und wenn es irgendwelche Fragen, Kommentare und Anregungen zu meinem Review gibt, hier gibt's die Möglichkeit auch für unsere deutschsprachigen Forums-Besucher!

===facts from the catalog===



Now, while waiting for Nikon to deliver on their latest and hopefully greatest 85/1.4G lens, let's have a look at the technical data and make some comparison to similar lenses.

Size: 87x84mm = medium size 83x80mm is the size of the 85/1.4D so not much to be seen here.
Weight: 595 g = Only 45 grams more than the old version.
Optics: 10 elements in 9 groups vs. 9 elements in 8 groups on the 85/1.4D. This is much better than the zooms that tend to have something like 21/16 for the Nikon 70-200/2.8 . That bodes well for contrast and flare-resistance - helped by that mysterious Nano-coating that Nikon is using like snake-oil to improve lens-performance. Looking at the cross-section you see that there're no aspherical or ED-element at all:

Closest focus distance/max.magnification: 0.85m / 1:8.3 This is too bad as I normally need somewhere at 1:3-1:5. So no brownie points for this performance. When you use a Canon 500D close-up filter (as some would call it) you could reach something around and beyond 1:5.8. So perhaps all is not lost.
Filter-thread: 77mm = standard with pro-lenses
IS: No = a pity The Nikon 70-200/2.8 zoom has VRII. So you can only hope that with a 2 stops larger aperture you could crank up the shutter speed to where shake is less likely.
AF: AF-S with SWM (silent wave motor), so does work on D60/3000/5000-bodies , manual-focus override by turning the focus ring
Covers full frame/FX or smaller = very good
Comes with a flexible lens pouch, not a nice soft-case
Price: around 1400 EUR new (incl. 19% VAT) = expensive The old Nikon 85/1.4 is around 1100 EUR now, the Zeiss (manual focus only!) at 1100 EUR, the Sigma at 900 EUR and there's also the Samyang as El Cheapo.
The lens-caps are standard Nikon's The others still play catch-up
Distance information is relayed to the camera, so the Nikon body can do all the advanced exposure-related stuff with this lens But this is true for all the alternatives too.
Aperture ring = no, just like all Nikon G-lenses
Lens-shade included and revertible for transport.
Sealing: seems so, but I'm not sure. The 85/1.4D has none.

Motivation:
Large aperture lenses are for isolating your subject. I need this as a nature shooter and even as a landscape shooter I love to have not everything in equal sharpness.

Alternatives:
- The old Nikkor AF 85/1.4D
- The Sigma AF 85mm 1.4 EX DG HSM
- The Zeiss ZF Planar T* 85/1.4
- Samyang

Even a first inspection on the LCD-monitor of the camera revealed that this lens must be sharp, even wide open. So I went forward with the usual tests on Siemens-star targets to get a better view on what's going on here.

I present center results (first row) followed by border (almost corner) results on a D300 (!). So we're looking at the lens only within the limited DX image circle but with a pixel-pitch that is roughly equivalent to a D3x.
Processing was done in CaptureNX2 at standard settings, with white-balance adjusted to a neutral white and some exposure compensation to make the brightness match. CA-removal is ON (in this case also loCA-removal. I'll come to this later). Focus was achieved as usual live-view based with a little optimization by hand and it was done separate for the center and the corners because the lens has slight field curvature. But I'd doubt that this is field-relevant.

These are 100% crops!

Center (D300):


Border (D300):


The results in the center are excellent, with only little softness even wide open (remember: you're looking at a f/1.4 lens). In the corners you see only a little drop off in sharpness below f/2.8. But this is incredibly good for a f/1.4 lens.

I'd say that excellent results that I'd have never expected from such a large aperture lens. Would be interesting to see how this lens performs on an FX body.

After the mysterious remark that I used loCA-removal to develop the test-images, let's see what's behind this.
loCA or longitudinal CA (sometimes also called bokeh-CA or axial color) is the effect that some out-of focus subjects in the foreground get a reddish tint and out-of-focus subjects in the background a greenish one.
First of all, here's a link to both test images without loCA removal:
Center, corner.
Now you see the reddish cast you may ask why I had used a little back-focus in these shots - because if I had focused precisely the cast would not occur. You're right: If I had pulled the focus just a little bit closer, I could have eliminated the occurrence of loCA in the test-shots altogether. But interestingly in this case I would have lost a little sharpness. So, as I was after the sharpest focus I could get I accepted the loCA and eliminated it in post-processing.

A photo of the micro-focus test-chart is soon to follow so that you can make your own mind as to how prominent loCA is with this lens. And later on I'll try to catch the effect in real live photos...

Here's a shot showing the AF-S 85/1.4G (second from the right) in comparison to other primes and one zoom. All lenses are FX:


Primes w Hood 28444 by Thomas, on Flickr


Primes wo Hood 870_02-7 by Thomas, on Flickr

from left to right: AF 35/2.0D, AF-S 35/1.4G, AF-S 50/1.4G, AF-S 85/1.4G, AF-S 28-300 VR.

Well, I managed to get a shot today that made my jaw drop in awe: Shot some trees at f/1.4 against the glaring sky.


Trees 28432 by Thomas, on Flickr

All systems/processing was set to "standard" with the only exception that I dialed in loCA-removal in CaptureNX2 at 100% as there was some nasty red/green "ghosting" around the plane of sharpest focus (which was to be expected*). But there was no contrast-enhancing measure, no nothing. Now if you click through the image to the large fine original JPEG you'll see what I mean: clean, well defined twigs with lots of details and a histogram that covers almost the complete range of tones.
And let me remind you again: This was shot at f/1.4!

In my book this test constitutes one of the real "nasties" and passing with flying colors at f/1.4 makes me think that this is quite an extraordinary lens.
---
* See the original with loCA here.

Now here's the test of axial color (longitudinal CA) and it shows some pretty nasty coloration.

(almost)100% crop shot at f/1.4:


Develop with standard setting in CaptureNX2. Lateral CA removal is ON, loCA removal made no difference.
If you compare this to the results from the AF-S 35/1.4G here you'll notice several things:
- loCA is more prominent in the 85/1.4G than in the 35/1.4G
- bokeh is a bit different esp. if you look at the oof-rendering in front of the focal plane (on the right of the image). The 35/1.4G renders more homogeneous than the 85/1.4G. That might be only true at magnifications close to 1:10, so I'll look into this under real-life conditions again.

Following up on the loCA angle I did a series of test-shots to
(1) give you a comparison of loCA across different apertures and
(2) make the comparison across different large aperture Nikon primes namely the 35/1.4G, the 50/1.4G and the 85/1.4G

So here's the result for the AF-S 85/1.4G showing a crop at below 50%:

Clicking through the image gives you access the 100% crop. Processing was done in CaptureNX2 with standard setting meaning lateral CA removal was ON, loCA removal=OFF. I was quite astonished to see the greenish (background) and reddish (foreground) hues even at f/5.6.

The second image gives you the results from switching loCA remocal ON at 100% (again, you see the crop here at <50%):

This is much better than the (non-existing) effects on the image of the previous post! The major difference between both shots being that this shot was made near 1:20 magnification while the first one was done at around 1:10. So the effectiveness of CaptureNX 2's secret sauce for loCA removal seems to depend on the magnification and gets better the smaller the magnification is. That can be confirmed from real-life shots.

As the lens has a 77mm filter-thread I just bunged on my trusty old Canon 500D close-up filter and see what I can achieve with this.
Well, we know from the technical data that the lens can deliver a maximum magnification of 1:8.3 unaided, which I tested to be a real 1:8.1. This is nice for anything you'd need in a portrait-job but nothing for the budding nature-photog. With a 2 diopter (=500mm) close-up filter you are able to go to 1:5.9 (=85:500) when the lens is focused to infinity. The real test comes when you focus the lens to minimum focus distance (of 85cm). Then you can get down to 1:3.4 at a subject-to-front-lens distance of around 28cm.
This is already a real nice magnification for my kind of work and I'll try to find out how good the quality of this combo is.

The following image shows the results you can expect from using the 85/1.4G together with a Canon 500D close-up filter, the second image is a 100% crop from the first one:

Ice Shield:


100% crop:

Shot at f/5.6 to get at least a bit of dof, 1/250 sec, +1EV, sharpening unchanged from standard-settings. Magnification is around 1:5 in this image.

With plane test-targets strong field curvature becomes obvious, but where you hit the focus the reproduction is sharp and with good contrast even if you use larger apertures than f/5.6 or go closer than 1:5 magnification.

So in the field if you try to reduce the number of lenses that you carry with you the addition of a good close-up filter could get you into the range of magnifications that you could very well use without bringing a dedicated macro lens with you.

Oh, and btw, in case you wonder: tele-converters are a no-no with this lens as the rear-element will immediately get in conflict with the TC. So there's no way to convert this beautiful lens into a 120/2.0 or 170/2.8 lens.

An on another note: While switching between the new 35/1.4G and the 85/1.4G it is very obvious that 85mm on a DX-body is already taxing your skills to keep the lens steady. Try to shoot 1/125 sec at least if you want to avoid shake and produce sharp photos hand-held!
This in turn raises the question (again) why Nikon did not implement VRII with this lens. I can only speculate, but I think that the masses of lens-elements in such a large aperture prime that need to be moved to compensate for shake is beyond the framework for VR implementation (e.g. energy consumption, response times). This shows why Nikon needs to implement body-based image-stabilization for use with such lenses as the 85/1.4G or other fast primes.

---
This review is now finally edited, and can be found here.

If you have any question, comment or suggestion regarding my review head over here.
Und wenn es irgendwelche Fragen, Kommentare und Anregungen zu meinem Review gibt, hier gibt's die Möglichkeit auch für unsere deutschsprachigen Forums-Besucher!

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses