Camera Labs

Unpacking the new lens already revealed some interesting insights.
I will compare many points to the Sigma 400mm f/5.6 APO macro, as this is my current lens for "long" action. You can have a look at the performance of this discontinued prime in my review over there.
Now, let's cut to the meat. What were my first impressions?
- Sturdy, nothing wobbly. Nice finish, looks like EX (will see how long it lasts)
- Stiff zoom action (plus lock at 120mm!), smooth focus
- HSM means no need to switch the lens to MF, just use the focus ring The 400/5.6 has to be switched manually
- Heavy, albeit if you're already over 2kg with your body/lens combo you don't feel another 350g
- at 400mm the 120-400mm is as long as the 400/5.6 (w/o lens-shade)
- finally OS at a long lens, love it. Just shot a series of handheld shots up to 400mm in dimming daylight and will get a first impression of it's effectiveness
- focusing = good so far if it works, but had some situations where the camera says "in focus" but nothing was...
- the tripod mount is always getting in my way of handling the lens. Have already dismounted it. Sadly you have to detach the lens from the body to completely remove the collar . With the 400/5.6 you could completely unhinge the collar and detach it from the lens with out removing the body
- the front lens-cap: hurray to Sigma - they finally understood why a Nikon-cap was always much better than a Sigma-cap
- a first test with the Siemens-star seems to indicate that the lens is reasonably well centred.
- Zoom-ring in front but turns the other way round, meaning contrary to the standard Nikon orientation and also different e.g. from the Sigma 10-20mm

The lens has produced some sharp images but when it was not so sharp nothing can be said without repeating the shot from a tripod to eliminate shake. Because even with a good IS/VR/OS you should never underestimate what "residual shake" creeps into your pictures.
So this has to wait for another day.
----
Wenn es irgendwelche Fragen, Kommentare und Anregungen zu meinem Review gibt, hier gibt's die Möglichkeit auch für unsere deutschsprachigen Forums-Besucher!

Before going head over heels into the tests, lets sum up some technical data and see how it compares!

Size: 92x203.5mm at 120mm focal length extends by 63mm at 400mm focal length (all w/o lens-shade). The lens-shade adds another 90mm (!) and can be reverse-mounted on the lens for transportation. That yields a total length of 356mm at 400mm focal-length incl. lens-shade The 400/5.6 fixed focal is 300mm long with extracted lens-shade.
Weight: 1645g (incl. lens-hood but w/o tripod-collar) vs. 1294g of the Sigma 400/5.6 or 1340g of the Nikon 80-400mm VR = quite heavy
Optics: 21 lenses in 15 groups (vs. 17 lenses in 11 groups on the Nikon 80-400mm) = much too much
Closest focus distance/max.magnification: 1.5m throughout the zoom range, gives a max. magnification of 1:4.2 As with many zooms the focal length at close focusing distance is not as long as when focusing to infinity. In comparison to the 400/5.6 from a distance of 4m the 120-400@400 looks more like an effective 350mm. The 400/5.6 goes 1:3 at only 1.6m distance!
Filter-thread: 77mm = standard
IS: Yes (OS) = good Just how good remains to be seen. First impression indicates at least as good as the VR on the Nikon 80-400mm!
AF: built in focus motor/HSM, so works on D40/x/D60-bodies , easy manual-focus override.
Covers full frame/FX or smaller = normal
Comes with a nice lens-bag with "EX"-signature
Price: around 650€ new (incl. 19% VAT) = quite cheap
Front- and rear lens-cap: still cr*p on the back (mounts only in one position - vs. 3 of a Nikon cap), but good on the front
Distance information: is relayed to the camera, so the Nikon body can do all the advanced exposure-related stuff with this lens
Aperture ring: No, just like a Nikon G-lens
Lens-shade: inlcuded and revertable
Removeable tripod-collar (no idea on stability yet), easy to turn camera to portrait-mode

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

First impressions from the test-bench:
- sharpness up to 300mm looks quite good
- IQ @400mm deteriorates visibly (center and corner). Even at f/8 worse than the 400/5.6 fully open
- the OS works quite nicely. Got a handheld free-standing shot @120mm with 1/13 (!) sec quite sharp. Other shots @400mm with 1/30 (!) sec succeeded also, so these observations may be statistically significant
- contrast against back-lighting seems ok, but I'll be critically evaluating this
- fringing was not obvious so far. But again, it might just be that the situation was not critical (enough).
- bokeh looks good so far

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

Update on the effectiveness of the image stabilization system OS:
Shot a series of images (approx. 1 fps) @400mm and 1/80 sec, 1/50 sec and 1/100 sec handhold, freestanding, OS at "1".
Viewing the images fit to a 26" screen (so no pixel-peeping here) I count the following percentage of images as usable:
1/80sec: 2 out of 5
1/50sec: 4 out of 6
1/100sec: 5 out of 8
That is you get about a 50% chance to get sharp images (up to 26" magnification) with 1/50-100 sec. That seems to verify that the OS is at least as good as the VR (first version) on the Nikon 80-400mm.

The OS takes a sec to stabilize, and then works quite smooth apart from some "jitters" inbetween when holding the lens longer to the eye. If you make hefty movements, you see the OS buzzing like crazy for a short moment. I assume it centres itself to make room for the best possible correction when the movement stops. The sound is a little louder than on the Nikkors, esp. when starting/stopping which makes a slight scratching sound. And without scientific analysis it seems to suck quite some energy: I think more than the 18-200mm VR - which would come as no surprise to me.
----
Oh and as an aside: After an hour or so you will feel the difference in weight between the 400/5.6 fixed focal and the 120-400 zoom
Or in other words: The difference in weight between the two lenses is not to be taken lightly
Aside #2: Although the zoom-action is quite stiff the lens exhibits zoom creep when dangling from your shoulder. So you better use the zoom-look at the 120mm position to avoid that.

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

How long is the 120-400 really when set to 400mm?
Hehe, you think this is a silly question Well, you obviously didn't read my remark about "Closest focus distance/max.magnification" above It was to be expected
Today I did some exact measurements at 6.2m and 4m when happily shooting Siemens-stars. Now I can give you the facts about the "shrink-factor" of the zoom:
At 6.2m the zoom had 92% magnification of the fixed focal, that is it behaved like a 370mm lens.
At 4m the zoom had 89% magnification of the fixed focal, that is it behaved like a 357mm lens.
And calculating from the max magnification of 1:4.2 at 1.5m resp. 1:3 at 1.6m the zoom at its closest focus distance behaves like it has 290mm effective focal length and thus has only 75% magnification at 1.6m of the fixed focal.
A test at infinity showed less than 1% difference between both lenses!

Btw.: the EXIF-data read 5.3m on the zoom and 6.3m on the 400/5.6 resp. 3.6m on the zoom and 4.0m on the 400/5.6

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

Now let's start with the shoot-out at 400mm, beginning with f/5.6. I had two shots from each lens under identical conditions. I used AF to focus each time. Lenses were mounted to a tripod, OS was off. Distance 6.2m, which is not the usual testing-distance as you're already at approx 1:15 magnification (1:50 = 20m would be considered "normal"). But who cares...
I selected the best shot from each lens and compared the pattern in different places, remember though that my D300 is not a full-frame body, so talking about corners or borders is only with reference to the APS-C sensor. What's happening outside the APS-C frame I can't know until I get one of those Nikon D900 24MPix FX bodies

All following crops (from the same 2 images) show the 400mm/5.6 in the left half, the 120-400mm to the right. And all images are 100% crops. So we're now at pixel-peeping level

Let's look first at some pattern at the middle-right upper border (7mm off-centre):

You can clearly see the difference in IQ: the 400mm resolves around 20% better under these conditions.
(You know how to find this out? Well: measure the diameter of the gray-disc in both Siemens-stars and divide the larger by the smaller.)

Now let's look into the far-right lower corner (12mm off-centre):

The zoom deteriorates visibly, esp. in the concentric circles,while the fixed-focal shows no sign of degradation. Funnily the Siemens-star of the zoom does not deteriorate in the same way as the concentric circles do (at least according to my mm-ruler).

Next (and last) one is the upper left corner about 12mm off-centre:

While the fixed focal now also shows some slack, the zoom clearly is at its worst.

So in total I would say: Fully open at 400mm the zoom resolves 20-30% less than the fixed focal. But then: You knew that this was a tough comparison! It is more testament to the very nice performance of the old and discontinued 400mm fixed-focal

And to prospective buyers of the new 120-400mm I might add: Fret not! This is the worst-case scenario for this new lens. It should improve at shorter focal lengths and smaller apertures.
So stay tuned...

-----
Btw.: I collect all testshots with this lens in this flickr-folder

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

For now, let's stay at 400mm focal length but stop both lenses down to f/8.0 and f/11. I'll show you both the best area of the shot and the worst. So this is 4 images in total, all at 100% crop.
Here comes the f/8.0 shots, first the middle-right upper border, then the upper left corner, again the fixed-focal on the left, the zoom on the right:





Next up are the same images but now at f/11:





Interesting observation: While the upper left corner of the zoom clearly improves, the middle-upper right does not make the same kind of progress. All in all I would say that the zoom still trails the fixed-focal by about 20% at f/8.0 and 10% at f/11 in resolution.

What I find a little disturbing is how clearly you can see even (only) 10% difference in resolution. Look at the letters, look at the concentric circles and there is no doubt that the fixed focal is always a clear winner.
Now, if you don't do pixel peeping, how much is this worth? Well, if you view an image from a 10-12MPix camera full-frame on a WUXGA monitor with 1920x1200 (2.3MPix) resolution you're projecting at about 1:2=50% magnification (linear). So some clever algorithm in your computer is binning 4 original pixels from your cam together to form one pixel on your monitor. In this case, differences in resolution have to be around 20% to be as clearly seen as in our pixel-peeping examples above.
But if you're printing on DIN A3 (approx 26" diagonal) I would assume that you can discern a 10% difference in resolution easily. That's why you should sharpen your images a little before printing

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

Well, all the Siemens-stars and what-not...
You wanna see an image, taken at 400mm f/5.6?

Here's an example that also shows you the type of bokeh you can expect from this lens:

Only tweaked the curves and did some cropping. No extra sharpening apart from applying the "vivid" setting...

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

Flare, glare, auto-focus and perceived sharpness.
Now comes the crazy stuff

You know what my thoughts are after my second day with this lens?
Even at 400mm this lens might be fine, were it not for two little problems:

1. there is a huge ammount of internal flare/glare (probably resulting from those many lenses in that many groups)! That's why some of the resolution, which is there, is outshone by the nasty side-effects of flare/glare. I'll show you an example shortly...

2. Focus is a bit on the "wild side". Meaning that you cannot rely as much on AF as you perhaps would like. This one is not shown easily, so just take my feeling for it. At least I'm sure it could not be residual shake as I'm talking about shots from a sturdy tripod with OS=off and short shutter-speeds...

Together, it reminds me only of the many factors that can ruin the sharpness of your lens/body-combo:
- shake
- flare/glare
- bad focus
- stuttering IS/OS/VR/VC
- too little dof
- heated air

Now what's with flare/glare? Look at this image taken @400mm f/5.6 1/320sec:

You can clearly see the white "halo" around the single blossoms in the top third of the image. This is not a kind of "bokeh-effect" as you can see that the area of max sharpness is in the upper third!

And it's not the result of over-exposure too. See this histogram of the image above:

You see that the image has a considerable gap at the highlights and the shadows. This again indicates for me that the lens produces images that are seriously limited in dynamic range by flare/glare.

If you have any other idea which can explain this, just let me know!

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

Testing micro-focus:
After thinking again about my impression that focus is a bit on the "wild side" (others would call it "erratic") I wanted to make sure that the combination of 120-400 zoom and D300 body don't have any micro-focus problems.
So I got out Canon's micro-focus testchart (use the download-link here) and did some test-shots with the zoom and the 400/5.6.
I misadjusted the focus manually and let the lens do AF and then searched for the characteristic moire-pattern. Yep everything worked out fine through a series of multiple shots on both lenses.
So there is no micro-focus problem.

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

Testing repeatability of focus again on the Siemens-star target:
Perhaps the Canon micro-focus target hid something. So I went out to shoot the Siemens-star target from exactly the same position as yesterday again. Manually "un-focusing the lens" and then let AF find it's way. Shot at f/5.6 and different focal lengths, but mostly @400mm, to verify my findings from yesterday.
Well, what can I say: There were slight variations in sharpness but certainly not so much that you (and I) can qualify it as being on the "wild side". So I have to take this back
So let's repeat this here for clearity: Repeatability of focus on the Sigma 120-400mm OS is ok!

Now I'm getting more nervous by the minute
What is/was the reason for the fluctuations I've seen in yesterday's shots

Weeell, last thing remaining is shake
Maybe my trusty Manfrotto is not sturdy enough for eliminating residual shake when shooting an effective 600mm lens?! Maybe the tripod-mount of the lens is not as sturdy as it looks (there are some "nice" comments on what flaky tripod-mounts can do to sharpness from Björn Rörslett)?!

Funny though that the 400mm fixed-focal did not exhibit this problem. Oh boy, I have to think about further tests to find out what's really going on here!

To be continued...

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

Reshooting each and everything!
Now this is my last effort:
- tripod mounted -> reduce shake (OS=off)
- mirror-up -> even less shake
- auto-focus only once per focal length, then switch to manual for the different apertures -> thus focus shift is eliminated
- go indoors (low light) and use a flash to eliminate any unwanted residual shake
- increased target-disctance to 9m just to be on the safe side

Hope the results save me from a hysterical breakdown

I'll keep y'all posted...

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

New comparison @400mm, f/5.6-f/11
The new results look stable and far more evenly distributed across the three tested apertures than the first batch of test-shots. And differences across the test-chart were also much smaller! Therefore I'll present only one area of the test-chart, approx. 6mm off-centre at 100%. I have also uploaded the 6 full original charts in all their 12MPix glory for your further scrutiny.

As in the posts above, the fixed-focal Sigma 400mm F/5.6 is presented left, the Sigma 120-400 f/4.5-5.6 OS is presented on the right. Top is f/5.6, middle f/8.0, bottom is f/11. Here goes:







Now sharpness looks much better for the zoom than in the first test-shots. The difference to the fixed-focal is non-existing at f/5.6 and only a small disadvantage at f/8 and f/11. The zoom even has a better contrast at all apertures. Which is esp. astonishing given the much higher number of lenses and groups in the zoom.
You can also observe that the zoom at f/8 (mid/right) is better than the fixed-focal at f/5.6 (up/left)

So no nervous breakdown, yet

Here are the links to the originals converted in Lightroom 1.4.1 from the RAW files:
Sigma120-400f5.6 Sigma120-400f8.0 Sigma120-400f11
Sigma400f5.6 Sigma400f8.0 Sigma400f11

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

Some additional remarks
"What a review" you might say.
"Stay tuned. You haven't seen the worst yet" I might add.

But really: I'm totally annoyed with this contradiction between the old and the new results! But I decided to leave the old ones in, as they were achieved with methods that were comparable and thus "fair". And I'm still at a loss why the fixed-focal held up so good and the zoom went down the drain under the same conditions!

But what can we all learn from my experience?
(1) That there are many, many factors to produce critically sharp images. And those factors can easily ruin the potential that a lens has!
(2) If you see test-shots on the web that claim to prove that a given lens is crap don't haste to the same conclusion but wait for the review of a pro!

------------
To be continued with results from the shorter focal lengths...

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

Update:
Currently been shooting outdoors at 400mm against the Sigma 400/5.6 (as usual) and at 200mm against the Nikon 180mm, all at f/5.6.
Real images, no testcharts!
Using a tripod plus shooting handheld, with OS (even on a tripod) and without. Always shooting three images in a row to see the (statistical) influence of shake.
Btw. the 180mm is a f/2.8 lens, simple, fixed focal and very good. Stopped down to f/5.6 it is a very challenging benchmark. Still, shake and mis-focus can ruin the IQ of this lens.

Just some observations from that session:
- The 120-400 can match the 180mm at f/5.6 even surpass it! But the 180mm might have mis-focused some times.
- Mounted on a tripod the 120-400 showed no IQ degradation when OS was "on".
- Handheld I had the impression that the OS is sometimes "nervous" and might have added blur. I'm still evaluating the reliability of the OS.
- each of the three lenses can end up with a very good, good, or so-so result in everyday use depending on shake and focus-probs.
So the question seems to be: Are lab-tests a good approximation for field-results in this case?

I'll leave this Q open for now. But I can quote some people using the 120-400 saying that you have to find out how to get the best out of this lens. And that might just be the case with many lenses that long. Remember that 400mm on an APS-C body is the equivalent of 600mm on film/FF-bodies. Shake is prominent and dof very limited in this class of tele-lenses.

But let me add another image I captures yesterday with the new lens to get away from pixel-peeping and technicalities.
This was shot through a screendoor against the evening sun (@200mm, f/5.6, 1/125sec, -1EV):

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