Ever since I discovered Lois Connor's vertical 7"x17" banquet camera work, I've been interested in making long, skinny portrait pictures. I haven't done much about it yet (so it's currently a project for another day). But Michael's thread motivated me to explore it a bit further in the interest of documenting another use for flat stitching.
It's a bit of an edge case because you need some specialized gear, but versions of this are possible even with more conventional shift tools.
With lots of resolution available, we can easily make photos in unusual aspect ratios like vertical 7x17 simply by cropping. I don't like doing that because it means throwing away hard-won pixels. That matters a lot less with cameras like the GFX 100S than it used to, but I'm a bit stubborn about these things. Plus my camera is a 50R.
With enough shift room, we have another option. My Toyo VX23D can manage 70mm of front rise or rear fall. You need a lens with an image circle of around 120mm to use all that movement, and I have several: my Fujinon GX lenses, which were made for the GX680 system.
Here's a quick example from a trail near my house. The lens I used here is the Fujinon GX 125mm f/5.6. This is the full 70mm of shift, with the camera in landscape orientation to create the maximum number of pixels. Cropped to the 7x17 format, it's 7,934 x 19,269 pixels -- plenty for a large print.
A quick test of stitching together landscape images made over the full 70mm of rise/fall available on my Toyo.
There's actually more image circle available than I can use. The limitation is not the Toyo. There's more shift room available by moving both the lens and the camera. The hard limitation is the design of the GFX 50R.
These are the six source images. After setting up the lens and camera at 35mm on the standards (right in the middle of the shift range), I took the camera up to 70mm; rear rise is like front fall -- so the first image is the bottom. Notice the black bar at the top of the first image. I then dropped the camera to make the rest of the pictures; rear fall is like front rise, so we're going up as the sequence progresses. At the 0mm mark on the rear standard, where the camera is at the bottom now (last picture in the set), notice the black bar at the bottom of the picture. Those black bars are the part of the image that is blocked by the lip of the cavity in the body of the GFX camera that contains the sensor.
Source images for the 7x17 picture.
Fortunately, Lightroom simply ignores the black bar when stitching the images, which is why you don't see it in the finished picture.
Now this is all well and good if you shoot a scene like the one I used here, where my shooting position is in about the middle of the elevation of the scene. It works because the image circle is large enough to allow a 35mm rise and a 35mm fall on the rear standard (with the lens raised to 35mm); this is what's happening in the first diagram in this set.
The picture at the top of this post was made using the setup in the first diagram.
Of course, this doesn't work if the ground is flat and the subject goes up. Think of a tree on flat ground. You want some ground and the whole tree. Setting up the camera as I did for the image at top would create a lot of unwanted foreground. What I'd want in this case is the second diagram. Instead of going up and down around the neutral position, I would want to go only up. Alas, this doesn't work with the camera in landscape orientation because there's not enough image circle. Additionally, the sensor cavity will block a lot of the final few images, which creates a risk of the stitch not working.
The solution for cases like this (or their reverse -- where all the interesting parts of the scene are
down rather than up -- think standing at the rim of a gorge) is the third diagram. If I use portrait orientation, I have just enough image circle, and the sensor cavity blocking issue is manageable. I'm using the short edge of the GFX sensor in this case, but that's still a lot of information even on a GFX 50R.
If you read this far, you might be wondering, "Why not just shoot up and down as needed and stitch it all together?" Of course you can do that. But you really need to try it to see why it's not ideal. In a nutshell, composition is harder because you can't tell in camera how much of the imaged area will be usable, plus the shape of the object you're photographing will change as you point the camera up; think of what happens to a building when you point the camera up to get the whole building in the frame.
It's not that you can't work around these issues. I just prefer getting it all sorted at the front end -- with confidence that the composition is going to work.
