Panoramas are often seen as a landscape photographers tool but they have many other uses and I like to use them in all sorts of settings, not least in woodland. Two or three vertical images stitched together can make for a very detailed 1×1 image for example and whilst the resultant image does not look like a panorama as we know it the methodology is exactly the same.
The image here though takes a more traditional approach and stitches multiple images to create a super wide panorama without the need to crop out a large chunk of the image top and bottom which would have been the case if we’d used a super wide lens to capture everything in one frame. Indeed, this particular image is a 180° panorama and even a fisheye would have been hard pressed in this instance!
For less ambitious stitched panoramas I will typically shoot 4 or 5 frames, overlapping each by around a third and using the camera handheld. Practice has helped me in this as the camera needs to pivot around the same point (on all axes) to avoid large alignment shifts which result in having to crop deeper into the stitched image. Given that I was looking at a 180° panorama for this image I shot my frames with the camera firmly mounted on a carefully levelled tripod. The initial image (below) was created using the merge function in Photoshop and is shown exactly as produced. Note the blank areas where no data was captured and how small they are; this means that the tripod was allmost levelled perfectly but not quite!
The panorama is made up of 13 individual files as can be seen here and Photoshop has applied masks to each so that the final image is comprised of a little of each individual frame. You can see the progression of the lens as it was moved between each shot.
Once I was happy that it was properly aligned I simply flattened the file to create just one layer with the raw panorama ready for processing in the ordinary way. As I set white balance, aperture, ISO, focus and shutter speed manually I tend to stitch the files first and post process afterwards.
Levelling the tripod is important as it means that you maximise the usable area of each frame. Overlapping each sequential image by between 35% and 50% gives the software the maximum material to work with and creates a better stitch. I always shoot an extra frame either side of my intended area – if my intended scene extends from B to E for example I’d shoot frames from A to F inclusive to ensure that my intended outer edges of the image are fully covered.
Dave's Place 