 |
| Zinc smelter Hobart Tasmania Panasonic Lumix FZ300 multi image panorama |
In the good ol’days of film photography if we wanted to make a panorama photo we had to crop the top and bottom off a standard image or use a dedicated panorama camera.
Some of these used standard double perforated 35mm film and exposed a length of film equivalent to two standard frames to make the 24 x 65 mm wide format. One such was the Hasselblad X-Pan which was actually made by Fujifilm. We also had a range of huge cameras from Linhof and Fujifilm which used medium format rollfilm behind an interchangeable large format lens giving a variety of frame sizes around 6x17 cm.
In the digital era those panorama film cameras have gone the way of the dodo to be replaced by computational technologies using standard digital cameras and multiple images.
Stitching multiple digital images
Several cameras offer in camera auto pano compositing which can be very effective. However this post is about creating the composite image in Adobe Bridge/Photoshop post capture.
Subject selection
We want to pick a subject which is mostly static. Some in-camera pano programmes and to some extent Photoshop can deconstruct a small amount of localised subject movement to produce a clean output but for the most part movement is the enemy of multi-stitching panoramas.
We also want to think about subject brightness range, too much of which will make for a difficult exposure solution.
We want to avoid placing one or more subject elements close to the camera. These cause parallax shift in relation to more distant subject elements as the camera is swung around, making the stitching process very difficult.
Fine foliage especially with vertical elements such as branches are also difficult as they tend to confuse the stitching software leading to multiple imaging.
Camera work
Pretty much any digital camera and lens can be used provided it supports fully manual exposure and focus and has a level gauge visible in the viewfinder. Ultrawide lenses are probably best avoided as the distortions inherent in images made with these lenses can make the compositing process difficult.
A tripod can be used but is not necessary in most circumstances.
The camera can be tilted up or down a small amount but must be kept very close to level with the horizon for every exposure. This is why an easily visible level gauge is so desirable.
The camera can be held in landscape or portrait orientation. The compositing process lops quite a bit off the top and bottom of each original frame so portrait orientation can sometimes be the best way to go.
Raw capture gives the best results with the greatest opportunity for achieving quality output.
The camera must be set to manual aperture, manual shutter speed, manual ISO sensitivity and manual focus. We need to experiment to give a firing solution compatible with the conditions prevailing, depth of field required and avoiding camera shake.
It is desirable to establish a routine for this work. So we will always make the exposure sequence in the same direction be it from left to right or the other direction.
Photoshop can also merge multiple images which overlap in both the horizontal and vertical dimensions.
We want to overlap each frame by about 30%. This does not have to be exact.
It is generally useful to make several runs of exposures, including some with more or less exposure, particularly is the light changes or something moves during the process. We can then review the images in camera or in Adobe Bridge or both and select the cleanest sequence.
Post processing
* Upload the files to Adobe Bridge and select each file of a well exposed sequence.
* In Bridge go to > Tools > Photoshop > Photomerge
Now the programme prompts us to select a Layout. We can try Auto at first and that sometimes gives a good result but sometimes not. The alternatives are Perspective, Cylindrical, Spherical, Collage and Reposition. Sometimes one of these gives a satisfactory result when the others do not.
* Check Blend images together, Vignette removal, Geometric distortion correction and Context aware fill transparent areas.
Press OK and wait while Photoshop does its thing, eventually producing a merged composite image.
* Flatten layers
* Now we can edit the image in the usual way and save it as PSD, JPG or other file format.
Messups
These are fairly common.
a) for off verticals and other shape issues, Photoshop > Select all (Ctrl+A) > Edit > Transform > Warp. This enables us to push and pull parts of the image at will.
Deselect when done (Ctrl+D.)
b) for minor double imaging and minor faulty stitching try Clone stamp and similar tools and/or Content aware fill.
End