10 August 2012

Going Big


In all my static images (i.e. images of things that stay still such as non-animal macros and landscapes) I'm starting to make the move to stitched images. In other words, I'm creating my compositions out of more than one camera frame. In this way, I can boost the resolution of my images, allowing for really huge enlargements. It also allows me much more flexibility to create different formats (panorama and square for instance) without cropping out much of the resolution my DSLR can muster. For example, the image above was created from four vertical frames stitched together in Photoshop CS5 for a final image resolution comparable to a 35 megapixel camera, rather than my 18mp Canon 7D. The images were taken with my EF 180mm f/3.5L macro, with the camera on a tripod, manual focus, and settings of f/16 for 0.5 seconds at ISO 100.

Note that this only changes the resolution of the image. The sensor pixel size (and it's effect on the quality of the image) remains the same as my single frame Canon 7D images. So the resolution is comparable to a 35 megapixel camera, but not necessarily the quality. Nonetheless, an image created using my usual attention to factors that affect the image quality (shooting from a tripod, using mirror lock-up, critical manual focusing using Live View, etc...) will produce images that can be blown up much larger than my single frame images. For example, the original version of the image above could easily produce a print 2 feet tall by 4 feet wide, and could even go much bigger.


Producing these stitched images is pretty easy, given some patience and good technique. First, I envision my desired composition, then I grid that composition out into frames (partially determined by the aspect of the focal length I will be shooting with). The multiple frames don't have to stretch in only one direction either. I could have a grid that consists of two rows and four columns. Again it depends on my composition, the focal length I'm shooting with, and the perspective I want to maintain for my composition. I currently don't use any special equipment particular to panoramic photographers such as a leveling tripod, a panoramic head, or a tilt-shift lens. Although I may eventually move in that direction as I continue with this work. Fortunately, most modern stitching programs are able to deal well with the parallax and other artifacts, I just need to be prepared to trim off quite a bit of the final image at the edges and provide extra space on the edges of my composition/grid scheme to account for that. When designing my grid, I plan on overlapping each image by about 40-50%. This gives plenty of info for the stitching program to work with and helps avoid some of the parallax issues as well.


This stitching method is not restricted to horizontal compositions only either. This normal aspect ratio vertical image of a raindrop on an arrow-arum leaf was created from four horizontal frames, yielding an equivalent resolution of 32MP.

Once I've determined my composition and rough grid, I attach my camera to the tripod and use Live View to achieve critical focus on the most important part of the scene. I then leave the focus on manual and don't adjust it for the remainder of the process, particularly when shooting macro to avoid changing the magnification. I then move the tripod so I frame the first grid point. I shoot in RAW, so I don't worry about setting the white balance; everything gets adjusted in Lightroom later. I shoot using the 2 second self-timer and mirror lock-up. Then I move systematically through the grid, being sure to overlap each of images by about 40-50%. Locking down the tripod before each shot.

Back on the computer I import all the RAW images into Lightroom (I'm still using 3). In the first image of the series I adjust white balance, noise reduction, and clarity and sync those settings to all the images in the series. If the most important part of the image is in another frame in the series, I make sure that the white balance works with that subject. Then I export all the images to Photomerge in Photoshop CS5. I have used Hugin in the past, and it is a great program with more flexibility, but I find in most cases that Photoshop makes great merges of my images with less fuss and this shortens my workflow. Once Photoshop finishes the merge, I use the crop tool to cut off all the nubs and regain the originally envisioned composition.  Then I do my normal processing in Photoshop with color and contrast adjustments.

When the wildlife remains pretty still, you can also use this process for animals as in the panoramic image below from Great Falls National Park. I hope this article helps you to produce some large resolution images of your own. Feel free to ask questions in the comments section. Just be aware that these huge images will suck up hard drive space and processing speed much faster than your normal images.











3 comments:

  1. Elijah, I often use stitching for landscape images I might want to print very large at some future date, I hadn't thought to use it with flora/plants. Thanks for sharing this

    ReplyDelete
  2. The texture of that leaf in close-up is really quite interesting.

    ReplyDelete
  3. I have yet to attempt this procedure but find it quite fascinating. Your photographs are exceptional, as always!

    ReplyDelete