After seeing the quality of images I have been getting over the last year or so with natural light and the pop-up flashes from my current compact cameras (OMD and X20), I decided I should check out how much difference a decent off-camera flash set-up would make for portraits.
I picked up the Fujifilm EF-42 speed-light unit, a generic Cannon compatible synch cable, and a LumiQuest LTD model hand-holdable softbox, which folds up and slips into just about any case big enough for a laptop. The photo above gives you an idea of how tiny the X20 actually is. I think the flash weighs more than the camera once batteries are added!
Since the Fujifilm X series cameras are particularly known for their out-of-camera (OOC) JPEGS, I thought I should also compare the camera processed images to a post-processed raw image using one of my basic workflows. I didn’t do any detailed touching up of the photos, only global adjustments to exposure, contrast, saturation, etc.
In the following pairs of photos, the pure B&W one is the straight out of camera JPEG, while the duo-toned one was processed from raw.
Shot #1 below was taken with a physical orange filter on the stepper ring under the hood of the X20. I was curious to see if this would make any difference in addition to the software color filtering I normally use during conversion to B&W.
The OOC JPEG was converted in camera using my standard settings which are BW-y, NR -2, Sharpening -1, Shadows -1, Highlights +1. I suppose I could have left off the yellow software filter and left it neutral, but my standard settings didn’t seem to affect it negatively even with two passes of light filtering. If anything, I think the optical filter may have kept the highlights under better control and maybe two passes of filtering softened the overall skin tone a bit more than I would normally, but I am not sure.
The .RAF file of the same capture was imported and then adjusted for exposure and white balance in Aperture, the contrast was increased a bit and the highlights/shadows tweaked before editing in CS6. The Topaz deNoise plug-in was used for preset “light raw” NR with a setting of 0.33 for added grain.
Finally, the Alien Skin Exposure plug-in was used for the B&W conversion, in this case without a filter (I usually use yellow or orange, but neither seemed to make a difference here compared to a neutral conversion). I also added a bit of highlight brightening using the luminosity curve, and subtle vignette effect. The whole thing was duo-toned with a warm platinum hue at the end. I don’t always do this, but for this set of pictures, it added just a bit of warmth that I think was lacking otherwise.
The straight out of camera image was not bad. I like the manually processed version better, but much of it may be personal preference. The main IQ impact seems to be from noise reduction, which you can see in the shadow areas at 100% magnification (click through the image below to see it at full size).
Since the in camera NR tends to overdo it, setting NR to the lowest level is a trade-off between too much noise and unnatural looking skin that starts to look plasticky with higher settings. To me, that is the main difference, the graininess from the camera NR is not as subtle as from the Topaz plug-in, and it tends to have some blob-type artifacts that aren’t particularly pleasant, but I don’t notice these in small enlargements without looking really closely.
The other pairs of pictures below are similar in style, but I took off the orange filter, and processed both in camera and in Photoshop using yellow filtering in the monochrome conversion process. For portraits, I don’t like to sharpen, and I have never needed to with this camera for people pictures.
This combination of lens, sensor, and processors results in very nicely sharp photos. I can’t complain about this at all, even compared to the OMD (surprisingly, I prefer the X20 at higher ISO ~800 even though it should theoretically have worse low light performance compared to the micro-4/3 sensor based camera). All of these were taken at ISO 100, and they definitely have more noise than the OMD would under similar conditions.
Adding a small light box really improves portrait taking without much cost or effort. This basic technique makes even a pocket-sized camera produce decent portraits. I have to thank the Strobist’s Lighting 101 post for the idea.
The main difference is that I upgraded to a slightly larger softbox than he recommended, which allowed me to move the light source a foot or so out farther without the light getting harsher.
As I understand it, for easily transportable lighting, the key to remember is that the bigger your diffuser/reflector is, the further out you can move your light source from the subject. It seems like roughly the diagonal width of the box is how much room you have to work with. Small softbox = head shots, larger umbrella = torso distance, and if you bounce off a wall or ceiling then you can get flattering light on an entire body and even groups of people.
Another reason I did this test, was to see if I can avoid G.A.S., or Gear Acquisition Syndrome. There have been some excellent reviews recently of the big brother to the X20, the X100s. I am trying to convince myself that I don’t need the bigger sensor and prime lens – so far it is working.
Here are some links to the best new reviews I’ve seen for the X100s. The Strobist site has one, Zack Arias has another (awesome portraits as examples!), and Nick Devlin writing for Luminous Landscapes has the other. All very tempting.