Beyond the Camera First Match Human Vision. Then Surpass it PDF
Written by Alexis Gerard   



By affirming "you push the button, we do the rest", George Eastman created the photography industry. Today the future of that industry lies in extending its partnership with users to enable them with a true synthetic eye — an image capture device that forces no compromises compared to human vision — and with the full complement of tools needed for the coming age of pervasive visual communication.


As Eastman's successor, Kodak CEO Antonio Perez, quipped in his recent CES speech, digital cameras have been "like horse-drawn carriages with the addition of a motor instead of a horse" (or "digital analogues of analog cameras" as we at Future Image have been saying.) To be fair, that is hardly surprising: As a rule, a new technology begins by emulating another technology; then it equals it — and only later does it develop into its own medium. There are technical reasons for this: the new technology needs development and scale to catch up first to the raw performance, then to the price/performance, of the older one. There are also psychological reasons: users — and often product developers — learn only gradually to think "outside the box" of what the older technology defined as "possible."


Recently a twenty-something asked me how come his digital camera produced blurred images in low light. I looked at his pictures and pointed out they were superior to what he would have obtained with a film camera. Then the irrelevancy of my statement hit me: He had no film-based reference — he had probably never used film. He was comparing the performance of his camera to the only standard that made sense to him, the only standard that makes sense today: his eye.


The eye, or to be more accurate the human visual system, is the performance standard that digital camera designers should have in their sights for the immediate future. There's a lot of work to be done just to equal it. Our best cameras do not have our ability to see fine detail (therefore, don't expect the resolution race to end anytime soon), or to hold detail at both ends of high-contrast scenes. They do not match even our limited ability to see in color in low light. In anything less than full daylight, cameras produce blurred images if they are not held completely steady, or if the subject itself moves. (We are starting to see progress in preventing camera blur with a flowering of anti-shake technologies, but nothing yet that helps with subjects that won't hold still.)

And there are less obvious ways that cameras don't match human vision: cameras don't have peripheral vision. They don't perceive and record depth and dimension information. And they don't record anywhere near as much "metadata" as we associate with our visual memories.


Rising to these capabilities is a huge challenge. But ultimately, whether we realize it or not, we as users expect nothing less from our cameras than to equal what we see with our eyes — and we won't be satisfied until we get it.




It's also a huge opportunity because, in this dawning age of images-as-language, the integrity of visual information — how well it reproduces our perception of reality and the richness of context associated with it — who, what, when, where and why - are what determines the value of that visual information.


Which points the way to the real target for image capture engineers: Remarkable as it is, the human visual system is designed to enable us to function and survive in the natural world. It does not take into consideration the existence of computing devices, the internet, televisions — all the artifacts and infrastructures of a technological society. In fact it doesn't assume any society at all. But the ways we use — and the new ways we want to use — the images we capture is highly influenced by and dependent on all of these. And that's where the money is.


We are now in an imaging economy where the historical consumables and service-based value propositions, which were also the industry's primary monetization models, have largely migrated into the initial purchase of the camera itself: film to sensor, print to screen.


The recipe for future success centers around understanding how customers want to use images after they've captured them — and beginning, at the capture stage, a partnership to optimize their experience of these downstream uses as well as the capture experience itself. In order to accomplish that, generating sophisticated metadata [information about the picture] based on location awareness and scene analysis is necessary. But is not sufficient. Partnering with the user involves understanding both their intent and their system constraints, and also assisting them with future-proofing. By this, we mean:


Intent: What is the user trying to communicate? Will that intent be best served by a single frame, or by a short motion sequence? How can composition, focus and lighting be optimized to the user's purpose? These skills, which we at Future Image call "visual eloquence," are currently the domain of the pros, just as, until the advent of automatic cameras, achieving accurate exposure was beyond the skill set of the average user. And just as before that, until the advent of roll film and develop/print services, the entire realm of photography was accessible only to a small minority. Today, anyone can operate a point-and-shoot camera with satisfactory results. It's time to build that next level of professional skill — "visual eloquence" — into the capture process.


System Constraints: Will the picture be displayed on a 60 inch high-definition TV, or on a cell-phone screen? If it's the latter, not only would it be wise to create a low-res version for transmission, it might also be advisable to zoom and crop so that the key elements are clearly visible on a small screen. Or perhaps a dynamic file that pans and zooms across the image should be generated.


Future-Proofing: One of the key findings to come out of the research on my book "Going Visual" is that the lifecycle of images extends long after their original use, and to uses that are not predictable at the time of capture. It turns out that one of the keys to the imaging industry's "holy grail" — extensive re-use of pictures — is not just the ability to find a picture when you need it, but the option to do what you want to do with it now, which may be quite different than what you wanted to do then. A year later, I'd like to display on a 60-inch high-definition TV that same image I originally sent to a cell phone. Oh, and Uncle Bob, who was just outside the frame when I took the picture? I want him in it now. Film photography, rooted since its birth in the constraints of scarce physical materials, is all about editing time and space at the moment of capture. The digital age by contrast is about having it all, and editing for particular uses at the time of those uses.


Alexis Gerard is co-Author of "Going Visual" (John Wiley & Sons 2005) and Founder and President of Future Image Inc, the leading independent center of expertise focused on the convergence of imaging, technology and business and hosts of the Mobile Imaging Summit executive conferences



Share This