Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

What is 'color science' and how does a 'color scientist' go about their job?

Most have a degree in math/physics/computer science/engineer ... with a emphasis on sensor technology. Normally takes a master or a degree with a bunch of experience.

Good example is Graeme Nattress at Red ... nice little video on his background is http://www.macvideo.tv/camera-technology/interviews/?articleid=102238 (Graeme i think has a degree in math and a bunch of plugin computer science experience).
 
The trick on color science is that there's a fudge factor involved, plus a lot of trial and error. I know the people who developed the original 3D LUTs at Kodak (one of which now works for Filmlight), and it took years and years of work to get the point where they could predict and control color spaces, going from one to the other -- say, from Rec709 HD video to P3 film or vice-versa. My frustration was that these black boxes were close but not exact, and there were always subtleties that would change during the translation. I have no doubt that Red's color science was done over time, just determining a log value that could be similarly controlled and adjusted, while still allowing lots of range for manipulation in the final color process.

I think the color scientists are much smarter today than they were 10 years ago. What's sad is that conversion to film -- formerly the most difficult part of the equation to predict -- has become relatively unimportant. As one example, I just read an interview with DP Peter Deming on Oz The Great and Powerful, and he casually mentioned that only 400 35mm prints were struck for that film. The rest were all digital releases (DCI XYZ color space).
 
Marc Wielage said:
I know the people who developed the original 3D LUTs at Kodak (one of which now works for Filmlight)
What's sad is that conversion to film -- formerly the most difficult part of the equation to predict -- has become relatively unimportant.
Click to expand...
Hence, that's why Peter Postma (that's the guy) no longer just purely color science engineer, but the FilmLight product manager.
 
I was browsing the EFILM website and they were talking about a 'de-contrast LUT' http://www.efilm.com/spotlights

What would that be exactly? and how would that differ from winding down the contrast in your grading suite? Where can one look for 'creative' LUTS like this?

Is matching digital sensor data to film the be-all and end all of color science?



Thanks for your answers. The Graham Nattress videos were very inspiring.
 
Marc Wielage said:
Glenn Kennel's book Color and Mastering for Digital Cinema is a good place to start, but it's very dated (2006). Alexis Von Hurkman's book Color Correction Handbook: Professional Techniques for Video and Cinema is more for colorists than color scientists per se, but it includes several discussions on color space, color management, and color science.

Steve Hullfish's book The Art and Technique of Digital Color Correction also covers some of the same territory.
Click to expand...

Guess Brett indeed asks about "color grading". There is also another old usage of term "color science" that has to do with how the eye and brain perceives colors. Thinking of what Graeme does, term color science makes a lot of sense, as the question is very much about matching the output of the sensor with the observation of the eye-brain. In my eyes term color grading has also an artistic flavor. The one that follows from adjusting colors to support story telling.
 
jake blackstone said:
Hence, that's why Peter Postma (that's the guy) no longer just purely color science engineer, but the FilmLight product manager.
Click to expand...
And Peter's a terrific person, too -- extremely knowledgeable, patient, and equally conversant with both film and digital. Very bright guy.

Brett Harrison said:
So you're saying Lauri that there's a fine line between color science and color grading?
Click to expand...
My opinion is that color science is about the process, the math, and the physics behind why and how it works; color grading is the artistic and technical decisions on how to get the results. One might compare the guy who designs the car and tunes the engine, vs. the guy who drives the car. Neither can get anywhere without the other. But you can be a great color scientist without necessarily knowing how to deal with the artistic side, and I know colorists who are actually non-technical people but are successful because they understand how to deal with people and how to interpret the client's requests and get it on the screen.

People use the term LUT too often to describe a tool designed to get a look, and to me, that's not what it's all about. To me, a LUT is designed to take something intended for one color space or gamma range and adapt it for another -- for example, to take a flat RAW file and push it into Rec709 color space. It doesn't "correct" the image so much as it prepares it. The colorist then has to make the artistic decisions on brightness, contrast, color, and other adjustments that take it much further, often with a lot more subtlety than just an overall LUT.

Quantel's Digital Intermediate Factbook is a bit dated, but it's a free guide that goes into a pretty good explanation on what Color Science and Color Management are all about -- separate and distinct from color correction. Filmlight has some technical papers at this link that also go into color space technology, and while it's specifically on their own Truelight LUT, a lot of what they talk about can be applied to other products. Noted DI consultant Steve Shaw also has some good technical papers on his Light Illusion website.

There are also some general technical papers on color science, ACES, the DCI, and other concepts and practices on the Cinematic Color website.
 
Brett Harrison said:
So you're saying Lauri that there's a fine line between color science and color grading? I guess in the end it's all down to the subjective impression, and in that sense it's not a science - at least not until we can understand the neurological correlates in the brain of the experience of a given color out of billions of colors, and even then, we may not know if the subjective experience matches up with that - it's a classic philosophical question actually, but I am getting away from the point now!
Click to expand...

Brett, yes. I've found several old books in university library from the 50's and 60's that are titled or talk about "color science". The term obviously has pointed to a certain field of science. Understanding these scientific issues is in my view a necessity to the existence of modern image processing systems. For, say, there has to be a common agreement what is meant by green, red, and blue. Just think about the consequences if every camera manufacturer and software vendor had their own interpretation of the name of colors.

The way I see it is, as a camera manufacturer Red designs and manufacturers sensors to capture light. Then they want to make sure while they translate the light measurements into digital data the colors not only match accurately with the existing standards but also the data reveals in full power the subtle nuances in the colors.

When we as end users process our footages in post, we play within the standards. Especially, we do not try to adapt light measurements into any standards, but instead we do our things --tweak color data-- within the commonly agreed standards. This is why I see a difference between color science and color grading.
 
Marc Wielage said:
...
My opinion is that color science is about the process, the math, and the physics behind why and how it works; color grading is the artistic and technical decisions on how to get the results. One might compare the guy who designs the car and tunes the engine, vs. the guy who drives the car. Neither can get anywhere without the other. But you can be a great color scientist without necessarily knowing how to deal with the artistic side, and I know colorists who are actually non-technical people but are successful because they understand how to deal with people and how to interpret the client's requests and get it on the screen.
Click to expand...
How I think about it is color science is a very wide field the reaches outside of the entertainment industry. For example with the new LED lights, they have found by aligning the LED frequencies to the eye ROD frequencies and the rgb Cone frequencies, the over all luminescence level can be drastically reduced (so they are involved in billions of dollars per day savings in energy cost, for lighting things like homes and buildings). [i btw find the entire hybrid rod/cone color theory papers that are being written totally fascinating]

On the film industry side active/ambient 3D lighting allows a re-entrant filming experience like we are hearing about in recent films (where they are using cameras and 3d software to project light into a room which is either viewed directly by the audience or filmed for further processing/grading). I think long term green screens and most of the current lighting approaches will go away, as 3d aligned(using actual camera positions with the project) & projected active lighting takes over.
 
You must log in or register to reply here.
Back
Top