I recently made some test shots in anticipation of a planned series of photographs. I’ve often used Photoshop’s Photomerge process to stitch panoramas and to make composite images — with my Janion Building, Victoria, BC, 2015 as an example.
Though the method has, in one form or another, been around since the earliest days of panoramic photography, it has recently been associated with photographer Ryan Brenizer, who has popularized the technique, also known as “bokeh panorama,” by integrating it in digital form into his portrait work — which is exactly what I intend to do (models may apply via my contact form).
The “Brenizer Method” by stitching a number of images together, mimics the characteristics of large-format photographs, including limited depth of field and high resolution. Besides aesthetic considerations, this technique holds other attractions: it can be practised with lightweight, smaller-format equipment … and it’s cheaper, much cheaper.
It’s especially attractive to those of us who have no desire to return to the darkroom, or $50,000 entry fee to, say, the 100 megapixel Phase One XF *
Subject & technique
To find a suitable subject, I visited a nearby park, representing remnants of the once-abundant Garry oak ecosystem native to this part of “Cascadia.”
I took along the Fujifilm X-Pro2 mirrorless camera, fitted with the Fujinon XF 50-140mm f/2.8 R LM OIS WR lens, and the Nikon D800, coupled with AF-S NIKKOR 70-200mm f/2.8G ED VR II.
The Fujinon 50-140, with the X-Pro2’s APS-C sensor is nearly equivalent to a 70-200 on a 35mm “full-frame” sensor (76-213mm to be precise).
My workflow was this: I set up my tripod** with Swiss Arca-style clamp (FLM SRB-40) on an FLM 38 FTR head. Both the Nikon and Fujifilm cameras are fitted with Arca compatible L-plates — the Nikon with Sunwayfoto PNL-D800R, the Fui with the (considerably more expensive) Really Right Stuff plate.
These make it simple to switch and orient cameras (though I used the revolving collars on the respective lenses to set cameras to vertical “portrait” position).
With lenses zoomed to their longest focal length and aperture set at f2.8, I made 6 exposures, using all manual settings, including focus, starting with the lower trunk of the tree — that is, I made 3 images of the lower part of the view, then tilted the camera(s) up to capture 3 more sections, making sure to align and overlap all exposures by about 25%, sides and top/bottom. In this case, I shot JPEGS.
Post production was limited to importing files into Lightroom, opening in Photoshop and using Photomerge, before some minor tweaks. After cropping to 8X10 ratio the images end up around 100 megapixels.
Photoshop performed admirably, merging branch structures flawlessly.
That’s really all there is to it the technical details. Let’s compare results.
The X-Pro2’s jpegs are great, right out of the camera. No surprise there. Both images, to my eye, contain wonderful detail in the bark and other elements in the range of focus. They both exhibit that “large-format look” admirably. But the Nikon wins the battle of the bokeh.
Here’s where Fujifilm’s 23.6mm x 15.6mm (APS-C) X-Trans CMOS III (24.3 million pixels) sensor can’t quite match the Nikon 35.9 x 24.0 mm CMOS (36.8 million pixel) sensor. It’s smaller size restricts the amount of softness we can achieve. The bare branches of the winter oaks in the background show this effect clearly: the Nikon image renders the out-of-focus branches smoothly (it’s very creamy!) while the fuji equivalent is sharper.
Both make wonderful prints. I began by pulling a couple of 11X14s on Epson Legacy Rag.
There are a couple of other reasons I’m not yet ready to liquidate my full-frame DSLR gear, in spite of GASD (gear acquisition spectrum disorder). Add the results of this test to my list.
*Yes, I realize a sensor-to-sensor comparison can’t really be made.
**Strictly speaking, a tripod is not de rigour as long as shutter speed is fast enough to hand hold. Nonetheless, a tripod will help visualize overlap. I used the pan function on the head to control those parameters.