Tuesday, September 7, 2010

Photography Case Study: “Who You Callin' Turkey?”

From the barnyard, here’s a kiddo with a rooster sunning on his shoulder. This was one of my last newspaper assignments.

Image details.
Nikon D90, 18-105VR, 25mm, 1/200 sec.,
f16, ISO 200. Program exposure mode with matrix meter and built-in flash at default value. Auto white balance, neutral scene mode, zero exposure compensation. VR on.

Technique and camera settings.
While children can be the devil to shoot, stumbling around the barnyard with 20 or 30 of them inspecting the garden, the tractors and the animals is sure to provide some opportunity. Here is no different. Another photo, of a girl feeding some pigmy goats, might have been the winner if the lighting had been different.

This photo’s central characteristic is the interplay of light and shadow, as well as considering compromises in lines and angles. Shadows are subdued here but provide depth and contrast in the boy’s face.

Figure 1 is a 200% actual-pixels enlargement of the boy’s hair and the rooster (click for full size). It is taken from the unprocessed original file. As you can see, even at f16, sharpness is decent enough even though diffraction has begun to soften things.

Why f16, one might ask. There’s plenty of depth of field here. I used f16 because I had to. ISO 200 is the lowest setting on the D90, and 1/200 sec. is the highest shutter speed the camera can shoot with its built-in flash. This is called flash sync speed. Interestingly enough, the best Nikon in wide use today on this front is the D40, which can sync at up to 1/500 sec., which would have allowed me to shoot at about f9 in this instance.

(There is a lower ISO setting on the D90, called “Low 1” or “Lo1” in the menus, which is advertised as “equivalent to ISO 100.” But the drawbacks of using a setting admittedly beyond the peak performance of the camera outweighed the drawbacks of shooting at f16 this time. If I had thought to bring my external flash with me, a Nikon SB-800, I also could have shot with a larger aperture, because the D90 – and many other Nikon mid- and high-level bodies — can sync with the 800 at higher shutter speeds, at the expense of flash power.)

The use of fill flash and the considerations implicated therein is the major technical takeaway from this photo.  I had to weigh the variables. If the lighting had been even brighter and I was looking at, say, f22 at 1/200 sec., I would probably have dropped to ISO Lo1. I may have tried a shot without any flash, which the program exposure mode would have probably shifted to something along the lines of f8 at 1/1000 sec. Fortunately, while f16 is beyond the normal realm of apertures I prefer to shoot at, it isn’t yet beyond my comfort zone. Circumstances also help: while the 18-105VR is a lackluster lens, the 24-30mm range is where it’s at its best.

It was a very bright day, obviously, when I took this shot. I should have planned better, taking that SB-800 with me to enable shooting at more moderate apertures with fill flash. A neutral density filter (something to darken the image) would have helped, too, although I don’t own one. It was fortunate I was not shooting the 17-35 f2.8 that day, because diffraction becomes crippling sooner – with f13 or even f11 being the sharpest aperture – than with the 18-105.

I wish I had the same photo without fill flash to illustrate how important it is in a scene like this. Without it, the rooster’s feathers would have zero detail and the boy’s face would look considerably different, under harsher lighting.

Shot technique.
Move around and remain thoughtful of the sun’s position. Being thoughtful doesn’t mean taking every photo from the same perspective. The shadow here provides needed depth and contrast. For another shot, having the sun right behind my back may have been better. Sometimes a silhouette, with the subject between the sun and camera, is best.

Kids are always moving, so I needed to be, too. There were a few shots I missed, because I couldn’t get into position before the subject had gotten bored with whatever they were up to and moved on.

Post processing.
I didn’t do much in Lightroom except confront a difficult decision on cropping and straightening and some color work.

You may have noticed that the ground isn’t quite level – actually it isn’t near level. In the original, below, you see that it is closer but still not right. Why did I rotate the image away from level?

The problem is that light pole next to the boy’s head. It must have been leaning because lens distortion should be at its minimum at 25mm and even if it were at its worst it wouldn’t cause an effect like this. Even though the post must be leaning, it remains too distracting if not righted, even at the expense of the horizon. If I were now so-inclined, I could Photoshop the pole out completely in a few minutes, but that wouldn’t be photojournalism. The exaggerated angle might even make for a more dynamic composition, as I don’t find the horizon as distracting but it is nevertheless subconsciously noted.

The white balance is unchanged from the camera default. I’ve added quite a bit of recovery to give the highlights some added depth, as well as some fill light to brighten the boy’s face. Those two adjustments threw off the rooster’s feathers, so I dialed in +2 blacks.

Clarity is reduced 5% to help soften the otherwise jarring edge between the rooster and the sky and to smooth over some diffraction-induced artifacting in the boy’s hair. A medium contrast curve was applied, as was moderate sharpening and masking.

Color correction is twofold, both in the blue channel. Saturation is increased while luminescence was reduced, to restore the punch late-summer Wyoming skies are known for. The neutral scene mode I often used in the D90 for newspaper photos did a good job of maintaining skin tones but often sucked all of the drama out of the sky.

As I mentioned, this is a study in contrast, light and shadow. It helps that boy and rooster cooperated handsomely, both focusing on the same thing (and not the camera at that!).

A modicum of preparation helped, although a better effort on this front (bringing along the external flash) would have helped more.  I wish the shingles on the barn roof were a slightly different color than the boy’s skin and hair, but if they had been black it would have been too much. And that damn light pole.

Finally, I should have gelled the flash. It’s pretty clear that the lighting on the breast of the bird is coming from a cooler source than the rest of the light, now that I look at it. Unfortunately, that sort of minute color correction is very difficult to do in post.

Image grade.
Original: C-
Post: B+

Poor planning (again, not having an external flash) caused the original image to be less than it should have been. Normatively, I would have found a spot without the leaning light pole of Red Lane in the frame. The shadows may have been too perpendicular, but perhaps if I had moved about a step, or a step-and-a-half to my left (about where the boy is looking) I could have gotten rid of the pole and put some sky between his head and the roof of the barn.

Post processing brought about half of this image’s potential back, with increased contrast and crispiness, but the white balance error between the flash- and sun-lit portions of the rooster are detrimental. Some sharpness was coaxed out of the pixels, too, with better color. It’s a start.

Next time…
A surprise! I suspect I’ll have a guest blogger filling in as I expect to be swamped by the end of the week. If it pans out, fantastic, and if not, I’ll make the time. Until then…

No comments: