Finally, back to working on the course! The wedding was really fun, but took a huge slice out of my time which coupled with working 60 hours in 5 days last week, meant that I had no time for anything else other than sleeping.
I completed the photographs for this project a couple of weeks ago, on a sunny Sunday morning, too sunny as will become evident in the shots I captured. Creating panoramas or collages out of multiple shots is now a pretty commonplace activity for most people, enabled by powerful tools in Photoshop, this is one I completed many years ago of Sipadan island, stitching a bunch of handheld shots together
Looking carefully you can see that hand holding such a series does not work, there are unreconcilable errors in the stitch on the left of the shot. I have not, however, used this technique to create more than simple and often silly compositions, until, that is, I bought a tilt shift lens and realized the potential this could bring to wide angle shots. My 24mm TS-E can be shifted horizontally or vertically by 12mm, moving the frame up or down, effectively doubles the effective sensor size of the camera, moving from 21MP to roughly 40MP. Because only the lens is moved the perspective is retained and software can very easily stitch the resulting image together. Work is still needed as at the extreme ends of the shift vignetting becomes a major issue.
Following the wedding I went back to the church and using an even wider 17mm TS-E was able to create the following composite of the church interior:
This was made with 3 separate images (although 2 would have worked) and formed a 6500 x 5000 pixel image that was pretty much sharp across the entire image. The 17mm TS-E is a wonderful tool for indoor architectural shots.
For the principal element of this project I spent a few hours at a favorite location the Friedensengel (or angel of peace), an Italianate monument close to the river Isar. As the location is elevated I wanted to try and create a roof top collage of Munich and so chose a time when the sun was in the East, wrong! The last time I was there it was winter and there was no foliage
In mid-summer the trees completely obscured the view, so I decided to make the best of things and turn my camera around to shoot the monument itself. The problem I had was the sun was just above the rail in the following shots forcing me to keep the camera down. This is not such a bad thing, but it meant I had an excess of foreground space in the shot.
Both of the above photographs were made using a horizontal shift from the same location, which I then stitched together into the following composite:
The great thing with this technique is that the image is truly rectilinear with little or no distortion from the lens. Alternative the lens can be rotated and a collage can be made from two portrait shots:
In this case, I could not have taken this shot otherwise as just behind me was a large tree that meant I could only capture this image (at 24mm that is) by positioning myself close to the monument.
Such techniques are excellent for creating very precise images of landscapes or architecture, however, for the truly wide or rotational collage another technique is needed. Once again I have used my 24mm TS-E, but this time simply as a manual focus prime, not using the shift capability. Mounted on a tripod I rotated the lend through 180 degrees taking 5 shots in total. This time the image is severely distorted, the fountain on the right and left of the frame is behind the camera position. One key aspect to this type of shot is to keep the exposure constant to avoid banding when merging the images. For this I adopt an average exposure having taken several readings and then fix the resulting under and over exposure in software afterwards - doing this with film would be a nightmare.
By this time the sun was starting to create a problem so I shifted camera angle and shot the road junction immediately in front of the above shot. Again using the 24mm and this time with 6 exposures I created the following:
As above this is a 180 degree turn of the camera. Two things are apparent in this shot. First of all the range of exposure was much greater and I could not lift the resulting shadows without creating a very unnatural looking effect. Secondly this shot had movement in it, the cars. I had to time each exposure so that the cars were not moving, especially in the areas where I felt the frame stitches would be made.
For these collages, I mounted the lens as best as I could, but really had little idea how to set the nodal point to the axis of the tripod, I do not have a pano head and am not interested in buying one as I am not greatly interested in these images. Care also needs to be taken to ensure that the key elements of the shot are captured within the central area of the image as the first output from Photoshop looks like this
Clearly a major amount of the image will be lost in the crop.
The text suggests making a further collage using 20 or more frames, I don't quite see the point of this unless a supremely large image is needed. However, I was up to the challenge, so switched my 24mm for a 100mm prime and produced the following of the fountain:
This uses 21 separate shots, each at 21MP and 25MB, for a potential maximum dimension in the 400MP range, assuming perfect alignment. As can be seen flare became an issue, however, this was more of a technical test than an artistic work. When I first tried to merge the images as full size TIFFs photoshop gave up in disgust, I had to downsize the images to 1440 pixel wide JPGs before it would play. Clearly there are technical limits to this type of image composition.
This weekend I watched a DVD documentary of the creation of a composite photograph by Andreas Gursky, a German photographer that I very much admire. He wanted to create a composition using the hanging baskets in which miners store their work clothes. These hang from the roof of the changing area and each miner can raise or lower the basket. As is common with his work Gursky wanted to create something far larger than that which actually existed, creating a spatial composition. The documentary follows how he imagines the work and follows the shooting of multiple frames to build the scope of the final image.
This is a further extension of what is proposed in this project, taking real images and composing them together into collage to create something else, not false, but not real either. What I did find odd about the whole process, is that whilst Gursky directed the shoot and pressed the shutter button, the actual building of the final image was done by a technician. Gursky told him what he wanted, but did not create the image. This is similar to Jeff Wall, although in Walls case he does not event operate the camera, his role is more akin to a movie director than what I have come to think of as a photographer. It is an interesting question, who actually creates a modern digital image, the guy who thought up the idea, the person who hit the shutter release, or the technician that pulls all together in software.
‘Hamm Bergswerk OST’
Either way I am very much impressed by the final image and much of Gursky's other work. Another artist employing large numbers of assembled photographers was David Hockney, however, he approached his art from a completely different direction. Influenced by Cubism and the problem of representing 3 dimensions on a 2 dimensional canvas, he took multiple snapshots (100's) and assembled them as a single composition enabling a scene or person to be viewed at once from multiple angles.
I thought I would have a go at this, but applying my own take on it. I have created a single composition out of 40 separate photographs, once again working around the Friedensengel and the fountain beneath it. I have attempted to reassemble the photographs in roughly the correct position relative to each other, but with a deliberate disregard for scale and for perspective:
Each image was shot with a 100mm tele and so can only take in a small element of the scene. I have not overlapped the photos as Hockney did, preferring to use a geometric grid, more in line with my mathematical background and desire for symmetry. The problem with the resulting image is that it is too flat, it needs more contrast and colour to work, I think a better scene could be a fairground or a shopping center, where there is a much greater range of tonality and contrast.
Collages are an interesting tool for the creation of photographs that require a much larger print than can be accommodated using the camera's native resolution, or in places such as the church I started with where the widest angle lens is still not wide enough to capture the scene. Beyond this I find less utility in stitching very large numbers of photographs together, although interesting work has been done with hyper detailed images of cities taken from elevated points. The following site contains a 26 Gigapixel photo-composite of Paris, a quite incredible image.
I do not really view this as a photograph, more as a marvelous record of the structure of a city at a specific point in time. At 354,159x75,570 pixels it is not going to be printed very often and exists as a web site. Perhaps this is the future for panoramic imagery, hosted in cyber space and never finding a home on paper. Gurskys photographs are huge in dimension, frequently several meters wide, however, at 300 dpi this image would be 30M wide and 6M high.