Sigma 18-35mm F1.8 DC HSM "Art"

Bokeh

For our final image quality assessment of the Sigma 18-35 "Art," we look at the one quality that you can't objectively quantify. Japanese for "blur," bokeh in a photographic context has come to be defined as the out of focus portions (you know, the blurry parts) of an image. When aiming for shallow depth of field to isolate the subject in focus, the quality of the out of focus background can surely make or break an image.

With a lens whose signature feature is its aperture, the ability create shallow depth of field will surely be a critical component of that feature's worth.

In order to effectively assess the bokeh, we used the same focal lengths and three lenses as the sharpness page, focusing on a (fake) flower in the midst of a heavily wooded area, notable for their very busy backgrounds - branches, tree bark, leaves, etc. It's very challenging for a lens to render it in a smooth, creamy blur. While the wall of Christmas light bokeh was a huge hit in our HD 55-300 WR review and we would prefer to use that, sadly that won't be revisited for this series of tests because it's, well, summer. And when it's summer time, we don't prepare for visits from Santa like those people Down Under...

Finally, in order to do some sort of analysis on the Sigma's rendering of highlights, we have another section dedicated to it, although a lot of the foliage samples, particularly the wider angles, have some out-of-focus highlights for you to examine.

Foliage Comparison One - 18mm and 16mm

As it says in the heading, for our first focal bokeh comparison test we set the two widest lenses to their widest settings. We chose to set the DA* lens to 16mm as opposed to normalizing the two to the Sigma because we wanted to show the difference between the their minimum focal lengths and what the bokeh looks liked for each (as well as the accompanying field of view difference). The following settings were applied across all the sample images during this iteration:

For your analytical pleasure, we've included the entire aperture range from wide open (lens dependent) to F11, as the majority of your shots will be within that aperture range to mitigate diffraction. Especially when attempting to achieve shallow depth of field.

Clicking on any of the below images will allow for an enlargement.

	DA* 16-50	Sigma 18-35
F1.8	N/A
F2.2	N/A
F2.8
F3.5
F4.0
F4.5
F5.6
F8.0
F11

Foliage Comparison Two - 20mm

Now moving to 20mm in order to add Pentax's latest Limited lens.

Clicking on any of the below images will allow for an enlargement.

	DA* 16-50	Sigma 18-35	HD 20-40
F1.8	N/A		N/A
F2.2	N/A		N/A
F2.8
F3.5
F4.0
F4.5
F5.6
F8.0
F11

Foliage Comparison Three - 31mm

No change to the settings other than normalizing all the focal lengths to 31mm.

Clicking on any of the below images will allow for an enlargement.

Foliage Comparison Four - 35mm and 40mm

Our final woodland test sets the Sigma to its maximum zoom. Naturally, the two Pentax zooms were set to the same. Just like in the sharpness test on the previous page, we also checked what the Limited Zoom at its max zoom.

Clicking on any of the below images will allow for an enlargement.

F1.8	Sigma 18-35
F2.2	Sigma 18-35
F2.8	Sigma 18-35	DA* 16-50
F3.5	Sigma 18-35	DA* 16-50
F4.0	Sigma 18-35	HD 20-40, 35mm
	HD 20-40, 40mm	DA* 16-50
F4.5	Sigma 18-35	HD 20-40, 35mm
	HD 20-40, 40mm	DA* 16-50
F5.6	Sigma 18-35	HD 20-40, 35mm
	HD 20-40, 40mm	DA* 16-50
F8.0	Sigma 18-35	HD 20-40, 35mm
	HD 20-40, 40mm	DA* 16-50
F11	Sigma 18-35	HD 20-40, 35mm
	HD 20-40, 40mm	DA* 16-50

Highlight Bokeh Test

One of the other chief concerns (if not the chief concern) regarding bokeh is out of focus highlights are rendered, hopefully in the form of perfectly circular and evenly lit orbs. There had been some reports regarding "frog-egg" or "donut" type bokeh, with a distinctive inner and outer circle. We wanted to get to the bottom of that and see how our two copies of the lens reacted.

Probably the easiest test we could think of that certainly would create tons of equally sized OOF highlights required the gridded metal lid of a pick-up truck's toolbox. Slightly off topic, but if you want a backdrop for your macro photography that is essentially nothing but rows of these out of focus Christmas tree lights (and Yuletide is but months away because you live in the proper hemisphere...), then this is a great DIY hack.

Anyway, here's how the Sigma 18-35 "Art" rendered its bokeh at 18mm. We found there to be no difference between 18 and 35mm, so we'll just post the one focal length.

	Entire Scene	100% Crop
F1.8
F2.8
F4.0
F5.6
F8.0
F11

As you can see, for the first three stops or so, it had no issue with uniformly lit out of focus highlights, and then the "frog eggs" became apparent. We think that if there had to be a definitive rule to go by when using the lens, it would be to use care when these highlights are present, namely when stopping the lens to F5.6 and slower. Also, if the highlight "balls" are really really tiny, then our bets would be on the frog-egg bokeh showing up rather than not. Here's a 100% crop of some backlit trees at F2.0:

But, just to show that it's rather unpredictable, here's another example at F3.2 where we couldn't find anything but attractive highlights. Click on it to view a larger version.

Verdict

We've already given our verdict and deduction regarding the rendering of highlight bokeh immediately above, however we wanted to sum up the foliage comparisons. And a terse summation this will be.

Simply put, the 9-rounded blades of the Sigma do a phenomenal job at blurring the background and foreground in a creamy blur. Juxtaposing the Sigma with the FA Limited lens, considered by many diehard Pentaxians as the gold standard for bokeh and rendering, we couldn't really distinguish between the two. The DA* and DA Limited Zoom lenses didn't fall particularly behind in the bokeh department, but the dramatic blurring occurred before F2.8. In the case of the HD DA 20-40, at 35mm and on, it doesn't get a chance to shine until F4.0.

If there was one thing that really caught our eye regarding the bokeh, it was how it did an absolutely stunning job rending out of focus flowers. We were hard pressed to find a lens that produced such a deliciously creamy blur. Here's an example at F1.8, which you can click on for a larger version:

At the end of the day we don't have any significant reservations regarding the Sigma's bokeh. The only nuance that forced us to write "any significant" versus just "any" reservations is the frog egg bokeh, identified by many user reports and confirmed by us. This is especially true if the highlights are going to be very small, either because you stopped down your lens quite a bit or by the nature of how small/far the light sources are. It's unfortunately a characteristic of the lens you should be aware of.