Sigma AF 100-300mm 4.0 EX APO HSM IF Review

[Thomas]

Just was the lucky winner of the bidding war for a used copy of this much praised zoom. The big issue for me is: If the zoom is of comparable IQ at 300mm f/4.0 to the fixed focal Sigma 300mm f/4.0, I'll sell the latter for the much needed flexibility in framing with the zoom.

So expect some results posted here!

[Thomas]

Here are the main facts from the catalog including some comparisons to the fixed focal Sigma 300mm f/4.0:
Size: 92x224mm = 3cm longer than the 300mm fixed focal
Weight: 1455g w/o caps and collar but including lens-shade (115g) = quite heavy (250g more than the fixed focal)
Optics: some 16 elements in 14 groups = that's much more than a fixed focal has, but for a zoom it sounds quite reasonable
Closest focus distance/max.magnification: 1.8m / 1:5
Filter-thread: 82mm = expensive
No IS = very bad
AF: built in focus motor/HSM, so does work on D40/x/D60-bodies manual-focus override by turning the focus-ring
Covers full frame/FX or smaller = normal
Comes with a nice looking lens-bag
Price: around 550€ used / 1000€ new = not dirt-cheap
The front- and rear lens-cap are cr*p (as usual with a Sigma lens) It might be that a new copy has the new pincer-type front lens-cap
Distance information is relayed to the camera, so the Nikon body can do all the advanced exposure-related stuff with this lens
Aperture ring = yes, just like a Nikon D-lens, moves in 1 stops up to f/32
Supplied lens-shade reverse mountable for carrying
Tripod-collar can be removed without dismounting the lens , easy to turn camera to portrait-mode

Alternatives:
- The Nikon AF-S VR 70-300mm 4.5-5.6G IF-ED (see Gordon's review): again a very good choice including VRII. And a pretty reasonable investment. But you loose one stop at the long end plus some IQ in the corners (might get worse with FX-body). Cheaper than the Sigma.
- The Nikon AF VR 80-400mm 4.5-5.6D ED: Nice lens (see my review here) but a bit long in the teeth: no AF-S, no VRII, antique AF/MF switching not the best at 400mm f/5.6. More expensive than the Sigma.
- Unless you buy a Sigma (with OS) or Tokina (w/o OS/IS/VR) variant of the 80-400mm. But the Tokina AT-XD 80-400mm 4.5-5.6 is not recommended (see here).
But all those alternatives lack the constant f/4.0 aperture

[Thomas]

Lens arrived today in mint condition. First test-shots seem to confirm that the lens is also optically good. Which once again proves that you can get good used lenses on eBay.
I immediately filled in all the question-marks in the previous post and the results of the missing information were always positive (except for those darn lens-caps)

The finish and feel of this lens is excellent. Focus- and zoom-ring run smooth. interestingly this is a non-extending design, so even when zooming the lens stays at the same length. It also has no zom-lock, but after testing the zoom-action once holding the lens down and then holding the lens up, you can feel no difference in the torque that you need. So it seems that only a relatively lightweight group of lenses has to be moved around for zooming. And even if you deliberately shake the lens, the zoom marker will not move: Excellent!
As to the position and rotation of the zoom- and focus-ring: Orientation of the zoom-action is identical to the Nikon 18-200, or the Nion 70-200. The zoom-ring is in the middle of the lens and the focus-ring in front. This is the opposite than with the 18-200mm but the same arrangement like with the 70-200 Nikkor zooms. Orientation of the focus action is reverse to the standard Nikon scheme. But as the zoom is what I mostly need that is not a major issue.

[Thomas]

Here's a first shot into the dull-grey summer-sky developed with Lightroom from RAW with only little enhanced contrast.

Original (shot at 300mm, f/4.0, 1/250sec, ISO 200):


100% crop (warning: pixel-peeping ahead!):


That counts as "sharp" in my books. With the better RAW-converter of CaptureNX and a little more contrasty lighting this would look very good.

[Thomas]

Finally the sun is back again. This was my chance to try reflections of the sun on water to test fringing. Here are two 100% crops of shots at f/4.0, the first at 100mm the second at 300mm.



No purple fringing to speak of under this most challenging condition

[Thomas]

Moving on to the dreaded Siemens-stars
Below you'll find a series of image-pairs taken with the lens set to a focal-length of 300mm: left/1st is the 100% centre-crop, right/2nd the 100% corner-crop and if you click on either image you have access to the large original in it's full 4288 x 2848 glory.
We are starting with f/11 at the top and go down via f/8, f/5.6 to f/4.0.
Warning: massive pixel-peeping ahead

300mm f/11:


300mm f/8.0:


300mm f/5.6:


300mm f/4.0:


All shots were taken from a tripod, so simply ignore the "handheld" designation

The test-shots confirm that the center of this lens at 300mm is very good even wide open. But in the corners you see a different story: really good corner quality (approx 80% of center resolution) can be achieved at f/11 but below the IQ deteriorates considerably.

You can also see some CA at contrasty edges, but not much.

Corner results improve a bit at 200mm and are much better at 100mm focal length.

[Thomas]

Can you see this IQ deterioration in real life images? You bet!
See the following shot of 7 geese in a row at 300mm f/4.0:


The following 100% crops are first the two left-most geese followed by the center-geese:

[Thomas]

Well I put this down to an individual misalignment. Or (again) sloppy QC by Sigma
This is really a pitty as the lens otherwise is superb.

Apart from one thing that really annoyed me with a macro shot recently:
I was trying to get a structure at sharp focus that was at my feet. But although I'm not small I couldn't get the required 1.8m between my eye (=approx sensor level) and the subject to get the 1:5 magnification that the lens is capable of You might say: "Ok, step back!" but that was unfortunately not an opportunity as the object was flat on the ground and stepping back would lead to (a) a tilted angle of perspective (b) problems getting the whole of the flat object in optimal sharpness. So the ideal vantage-point would have been exactly (perpendicular) above the object at 1.8m distance. But I was not tall enough...
Lesson learnt: Never buy a lens for (near) macro work that has a closest focus distance of >1.5m
Both 300mm fixed focals from Sigma and Nikon focus below 1.5m. The mighty Sigma 400/5.6 goes to 1.6m and that really is the end of the line (at least for me).

[Quatar]

When i bought this lens i was really happy with it. Very fast zoom lens and capable of delivering sharp to very sharp images.

The main problem i encountered, was that you really have to be close to your subject. Distances further than 8 meters and the images starting to get soft. This problem was even worse with the 1,4tc on it.

This was the main reason to sell the lens and buy a prime.
Bought the Nikon 300mm F4 AF-S and this baby is way better in every aspect! And i really was a Sigma fan...

[Thomas]

Hmm, interesting observation, Quatar - and a pity!
Because I think that the lens (design) has potential and that your copy doesn't look good beyond 8m is another case of individual quality problems imho. In my test-shots things far away were looking excellent (see image of the spire).
This underlines only that one should test a new lens extensively after purchase and that one should buy from a place without prohibitive restocking fee.

[viktorz]

When i bought this lens i was really happy with it. Very fast zoom lens and capable of delivering sharp to very sharp images.

How much mega pixel may it shoot?

[Thomas]

As it works on full-frame bodies a good copy should deliver good results even on a 36MP D800 (with a little stopping down).

[drein]

Hey Thomas,

nice review. I saw in your sample picture and your test chart, that you always tested the lower left corner. On http://slrgear.com/reviews/showproduct.php/product/336/cat/31 the lens was tested for Canon and their sample had a clearly weaker left side. Would be interesting if your lens suffers from the same problem.

Also I read a couple times, that the lens (like many sigma, I had the 2.8/70-200 myself...) suffers from front focus, which can be correct on bodies like your D300.
Have you checked if your sample also suffers from front focus?

Last but not least, I'm wondering if you gave that lens a test drive on a FX body?

I'm asking all these questions, because I'm trying to decide between the Sigma and the older Nikon AF 4/300, non AF-S. And both lenses are difficult to get information of...

Best,
Peter

[Thomas]

Hello Peter, and welcome to the friendly Camera Labs forum!
To enjoy your stay here please have a look at the house-rules!
----
Regarding your questions:
- Once I find out a lens is decentered I have to decide whether the defect is too big to proceed with testing or whether results "from the other side" may still give some useful insight.
- As to front-focus. I can't remember whether I had to dial in any micro-AF-adjustment. But naturally I#m acutely aware that bad focus can ruin an otherwise excellent optic.
- this test was done befor I had access to an FX-body. So sorry, I can't give you any results from this lens outside the DX-circle.
My personal opinion is that IF you can get a well-centered copy of the Sigma 100-300/4.0 that it's well worth the money. I'm a bit hesitant to recommend fixed focals above 150mm because it normally is much harder to "zoom with your feet" at the distances that you normally try to reach out for with tele-lenses. I find fixed focals often too limited in their use in the field. That being said: The Nikon 300/4.0 certainly is a good optic.