Computational Photography
[by Jay Kinghorn]
As we close the book on the first decade of the 21st century and look forward to the next 10 years, computational photography looks to make the greatest technological impact on the craft of photography as we know it.
Computational photography is a broad, if imprecise, term most often used for any imaging techniques that expand upon the normal capabilities of a digital camera. Common examples are High Dynamic Range (HDR) photos or panoramas, the outcomes of which are digital photographs that could not have been taken by a traditional camera. Less established technologies allow a photographer to set focus and depth of field on their computer instead of in-camera.
Most examples of computational photography use multiple images to enhance the quality, or flexibility of a single image. For example, when using a high ISO setting in low light, a camera will automatically capture a quick burst of images. The image processing software (either on-camera or on the computer) compares the content of the images, separating detail from image noise. The noise is discarded and the detail preserved. Other techniques still in the lab use still photos taken at regular points during a video clip to improve the detail, tonal range and quality of video footage, or allow an artist to relight a scene in post-processing to tease out hidden detail.
These technological improvements should be embraced as they come to light, because they will allow photographers to capture, create and publish photos in new and even more compelling ways. While the tools of the future of photography are important, ultimately it’s the creativity and artistry that’s applied to them that will help people tell stories that continue to move, engage and inform.
Looking forward, a photographer’s ability to exploit new opportunities and assimilate new technologies into their workflow will be a defining characteristic of the future of photography. The profession of photography will be less about being a technician and more about being a visual artist fluent in the language of color, shape line and light who communicates across mediums with greater facility than any technician ever could.
5 Responses to 'Computational Photography'
Subscribe to comments with RSS or TrackBack to 'Computational Photography'.
Excellent post Jay. I agree 100%.
Especially about photography being less about being a technician.
Rosh
The hi ISO noise reducer technique is clever, I can’t wait to test it….
Toh,
I believe the high ISO technique is being used in the Casio EX-F1 camera. You can read about it (and other computational photography uses) at http://www.newscientist.com/article/mg20427346.700-computational-cameras-perfect-your-photos-for-you.html?full=true.
Jay
I believe that the term “Computational Photography” is about as redundant as the term “molecular gastronomy”. Anyone not using computerization in some form of image capture these days is clearly a Luddite, or has the luxury of subsisting in a retro art photographic medium. The real holy grail is increased dynamic range of the camera in a single burst, the HDR fad is just that, a fad. The promise of HDR as a replacement for proper technique has been an utter failure, and the images created with this process look about as visually compelling as a graphic novel and will fade away with the star filter of yesteryear. Enhancements in noise reduction and high iso are a real step forward, and I can’t wait to see what the camera makers come up with in the future. I am still holding my breath for a camera with a native 15 stop range.
Unfortunately, some of these techniques like HDR are becoming so commonplace in everyday usage that they have almost burned themselves out. When technology is a crutch for style, the image suffers. HDR photography do not solve the problem of dynamic range, it just enables awful looking images that resemble homoerotic gladiator movies. People were excited about the star filter when it came out too…