Rokinon 8mm vs. 10mm vs. Sigma 8-16mm vs. Pentax 10-17mm Ultra-wide Showdown

Sharpness

Many would argue that if there is just one attribute in a lens to pay overt attention to, it would be sharpness and the maximum resolution the lens in question is capable of. This is especially true in our pixel-peeping, megapixel-counting, measureabating world of modern photograph.

Normally, "star charts" are used to assess a lens' resolving power, yet because of the extreme fields of view and the combination of rectilinear and fisheye lenses, we opted for a brick wall as a more practical test.

The below settings applied for the entirety of this chapter's comparisons:

And without further ado, the comparison with both the lenses set to their widest fields of view as well as a Pentax vs Sigma shootout at the middle and ends of each zoom's range.

Widest Focal Lengths - Center Sharpness

For the first comparison showdown, here are the four lenses at their widest setting and 100% crops taken from the center:

	Rokinon 8 FE	Pentax 10-17 FE	Sigma 8-16	Rokinon 10
F2.8
F3.5
F4.0
F4.5
F5.6
F8.0
F11
F16

Widest Focal Lengths - Corner Sharpness

Again, the same four lenses, but this time their corner sharpness from the same scene as above:

	Rokinon 8 FE	Pentax 10-17 FE	Sigma 8-16	Rokinon 10
F2.8
F3.5
F4.0
F4.5
F5.6
F8.0
F11
F16

Pentax vs Sigma - Middle of Zoom Range

In addition to the series above, we wanted to see how the Pentax fared throughout the rest of its focal lengths. As close to the middle we could get to, here's 13mm for the Pentax zoom. The Sigma is juxtaposed alongside it to see how it performs at its central marking (12mm):

Pentax vs Sigma - End of Zoom Range

The same as above, but this time with their zoom rings turned all the way in to 17 and 16mm, respectively:

Verdict

Right from the start at the widest field of view in the center proved at a minimum adequate across the board. The two fisheyes were a bit soft wide open and the Pentax zoom especially was slow to make a marked improvement. The Rokinon 10mm proved a bit soft at F2.8, however just like in vignetting, the drastic improvement in quality within half a stop (F3.5) was more than remarkable. In fact, it pretty much peaked right there. Finally the Sigma 8-16mm was astonishingly sharp from wide open, however it appeared to quickly lose that advantage, succumbing to diffraction quicker than the others.

Staying as wide as possible but moving to the edge of the frame, the clear winner was the Rokinon prime although the Pentax was not too far behind. The Sigma zoom was the most disappointing, however there's a legitimate reason for that, which we will explain at the very end of this page. And then once again the 10mm prime performed admirably, especially after F3.5.

Moving onto the central setting of the two zooms, the Sigma pulls ahead in the center and especially the corner, which deals with the forthcoming "explanation" we alluded to. The max zoom setting for each is more of the same The Sigma having the edge in the center but there being no contest in the corners as the Sigma displays far greater resolution.

Field Curvature? Or Simple Math?

The term, officially called "Petzval Field Curvature," carries the late physicist's name and refers to an optical aberration where a flat subject (i.e. the brick wall used above) cannot be brought into uniform focus across the image frame. In other words, and the most common example, is focusing on the center whereas the edges lack the same sense of being in focus and crisp. Of course there is the expected degradation between the center and the edge of any given lens, however field curvature is much more severe than the expected difference between the two. This undesired effect stems from curvature in the physical glass within a lens, and as we all know, these lenses have some seriously curved optical elements. Should you wish to learn more about FC, Photography Life has an excellent and succinct article on it that we recommend.

Anyway, so while they all do exhibit it to a degree (and a specific degree that is beyond our means to officially and accurately test), we discovered something far more compelling when it came to explaining why something in a uniform plane would appear out of focus. And the good news is that it is a very simple concept - distance from the camera.

We found that the main issue, and this is especially true with the fisheyes, is that the field of view is so extraordinarily wide that chances are, whatever is in the center is not the same distance away from the camera, and thus will find itself outside the plane of focus (the "band" of sharpness between the near and far limits of the depth of field). Specifically, the object in the corner will be behind said plane of focus. The easiest way to describe this would be a drawing compass that many of us may not have seen since grade school:

With the pin fixed to a central position (which would be the camera), the black line being drawn back and forth by the pencil is any given focus distance. Finally, the blue line is any flat object (i.e. a wall) that the camera would be focusing on. As you can see, the blue wall continues on to the left and to the right however the further you travel along it, the further you distance yourself from the plane of focus.

Here's another way to graphically demonstrate this "phenomenon" with the dark blue dot being the focus point on the subject (brick wall):

Remaining equidistance from the camera the entire time is the curved blue line, which represents the focus distance and the green band conveys the depth of field or plane of focus. Notice how both arch away from the wall. Finally, the other dots show an increase of distance from the camera despite the wall of course being flat and in the same plane the entire time. What's more is that even on the series of dots to the right, where there's no change in "elevation" between them, the distance is still changing. In the context of the lower right thumbnail, the distance to the blue focus dot was 12 feet (3.65 meters) whereas the edge of the wall by the dark red dot was three times that distance! Oh, and the red dot over the left side's telephone pole (which is actually in front of the wall)? Between four and five times as far away from the front of the lens.

We discovered this accidentally when trying to devise how to run this page of testing. After finding large brick wall, it still wasn't tall or long enough to impress either fisheye lens. The only way the wall would have filled the entire frame, even with the Rokinon 8mm, would be placing the tripod about 4-5 feet (1.5 meters) away from the wall. And now we know that wouldn't have worked because the corners would be 20+ feet (6+ meters) away.

A small aside we wanted to point out was while this concept is talking large subjects, the same concept applies to the "Focus and Recompose" method, where photographers use the center focus point and then readjust the framing while not touching the focus. This is problematic because if the depth of field is extremely shallow, such as in portrait lenses, then the edge of the frame, where you place a subject's eye, for example, may now be beyond the plane of focus.

Anyway, because the corners of the building didn't extend into the corners of the frame, we rotated the camera using our ballhead 90 degrees (just "to the left") and without changing our focus distance. This is how the two scenes looked for the fisheye lenses and where we grabbed our center (green) and corner (light blue) 100% crops from:

And this would explain why the Sigma zoom lagged so far behind the two fisheyes during the corner evaluation in the widest focal length whereas the roles reversed with the Pentax lagging significantly behind the during the Middle and End of the focal length comparisons, which did not require us to rotate the camera to get some bricks in the corner of the frame.

The Bottom Line

As you can see, when dealing with lenses of such a wide field of view the main problem wil not be field curvature as much as depth of field.

Notice how in the third and forth series (Pentax vs Sigma), the Pentax corner performance actually improves all the way from wide open to F16, which should have raised an eyebrow at first glance. Of course diffraction is affecting the image from well before such a tiny setting, but more important is simply the increase in depth of field that comes with the smaller aperture.

So at the end of the day, to effectively capture a scene in all its crisp pixel-perfect glory, don't immediately write off these lenses (and others in the UWA category) as having excessive field curvature - the issue could be poor depth of field management on behalf of the photographer. Sometimes there's not much that can be done about it other than advanced techniques such as focus stacking (merging multiple frames of the same image where focus is set to different locations to maximize foreground-to-background depth of field).

The final chapter before our Samples and Conclusion pages, a look at these lenses' focusing mechanisms and abilities is next.