Tom Maier, Research Fellow, Color and Imaging Science

Education:

I went to college at MIT and have a BS in Chemical Engineering. I earned my PhD in Chemistry from the University of Illinois. At the time, the Chemistry Department at MIT wouldn't accept its own undergrads, so I had to search out another school for my post-graduate work. I spent four years in Champagne, Illinois. Think of Boston, and then think of landing in a place where all you see is cornfields for 100 miles in every direction. You could never lose sight of those cornfields!

Job Responsibilities:

My main area of expertise is Color Science. Much of my work involves computer or mathematical models of how an imaging system (film, hybrid, or digital) records, manipulates, and produces color images. It really comes down to two things… if Kodak makes a change in one of its films, I can predict what the subsequent color change within the image will be (I call this a forward analysis) or if we want a specific color in an image, I can tell what to change in the system (within the constraints of the system) we have to make (I call this a backward analysis). Essentially if there is an image originating in a photographic system and there is a need to change either the color of that image or something within the system that will produce a different color, I work on those kinds of problems. Color is based on both the light coming to our eyes (as it was transmitted through the film or reflected off the paper) and the processing our brain does to give us the perception of color. As such, I work on both the analysis of the light and the analysis of our perceptual abilities.

In this capacity, I serve as a consultant for a variety of R&D projects - vs. being 100% committed to any one particular project.

For example:
The SMPTE DC28 Digital Cinema Standards team several years ago requested more information on visibility/perceptual effects. In other words, what you can and cannot see up on the screen - especially as it relates to color standards.

So, think of yourself as the digital projector. If I'm the data source and I send you three numbers - maybe 300, 350 and 500 - what color are you going to display on screen when you get those numbers? There has to be some standard of encoding, so that given these three numbers, the resulting on-screen colors are predictable.

And what if I change the numbers by one code value - going from 300 to 301? We want to avoid contouring, or the visible step from one color gradation to the next, up on screen. As such, a one-code value change has to result in each individual color step being less than what audiences can distinguish or perceive. Once all the steps are put together, the picture should represent a smooth gradation of color, like that of film.

I designed an experiment for the SMPTE DC28 standards group that determined how many steps would be needed between black and white, so that each individual color gradation is indistinguishable. Digital Cinema Initiatives (DCI) funded and ran the experiment. So, for example, if it turned out to be 1000 steps needed between black and white, we would have to encode by a 10-bit number. If we needed 4000 steps, then it would be a 12-bit number, and so on…

We organized two test groups of people, who were asked to sit in a theater and look at color bars. I assumed that people would just say yes if they were asked, "Do you see a color bar?" I mean, nobody wants to look stupid. Instead, we set up the test such that either a vertical or horizontal bar was shown on screen, and the audience was asked, "Which do you see, a horizontal or vertical bar?" By making the bars less and less distinct, we could determine the threshold of what people were able to see.

Our first test group was comprised of 'experienced' people…SMPTE members, cinematographers, and studio folks. Our second set was considered 'non-experienced.' We actually approached people on Hollywood Boulevard and offered them $25 to come look at color bars in the Entertainment Technology Theatre. (Unfortunately most of the second group had to be disqualified because they either couldn't follow instructions or they had trouble seeing in a dark theater.)

We had hoped the results would point to a 10-bit encoding solution, because the Cineon System - which has essentially set the industry standard for postproduction - is 10-bit. Remember, this was in the 2003 - 2004 time frame. All post-production computers and programs are currentlywere designed to be compatible with the Cineon System. Well, the results ultimately pointed to 12-bit. It's a matter of the car not fitting in the garage anymore. And, if you can't park outside, it means you have to build a new garage. The industry has been able to build equipment for the 12-bit encoding and D-Cinema is widely used now for the distribution of movies.

The Kodak Journey

I started at Kodak in June of 1964, while I was still in college. Because it was part of the summer program, I worked for 12 weeks each year. Some days I spent all my time washing cans in Buildings 52 and 57. Sometimes I would clean floors and walls. Some days I actually spent doing production work. My last two summer internships were spent in the Kodak Research Labs, where I was formally hired in 1971 after I finished college.

My first permanent position was in the Color Photography Division. After five years, I moved to the Instant Photography Division. I was there for five years as well…and then I joined the Color Science Group. After about five years, I became group leader of the Color Science Group.

When the lab was split in half, I became a lab head of the half with the color science in it. That was from about 1993-1996. But, I never enjoyed being a lab head. It was so administrative…all about trying to get funding…and a lot of competition.

In 1997, I came to EI. I was hired to work on PreView, and at the same time to serve as a resource on color and image science questions. I have really enjoyed my time in EI. I like the people with whom I work, I like the problems I work on, and I like the fact that all of EI is so successful.

Most Proud of:

I am very proud of being married to my wife for almost 40 years, and of having raised two kids together. I'm also very proud of being a grandparent. And, I am proud of what I have accomplished in my career here at Kodak.