Color Correction

Shoot for the editor or shoot for the colorist?

I have been following with much interest the ongoing efforts by expert RED users in learning how to use this camera and let it live to its promise of digital-film-for-the-indies. Most unexpectedly however, it also triggered in my mind some lateral thoughts about ordinary videographing – how can we shoot video towards a more rewarding and fulfilling picture experience.

This opens up a whole range of possibilities for the hobbyist or the apprentice videographer alike.

The first obvious observation about RED is the complexity and cost – RED requires considerable (maybe unanticipated?) expertise and effort to establish a workflow. This all drills down to the second obvious observation that – contrary to video cameras – RED does not produce ready-to-use images. By purposeful design instead, the DSLR-like sensor outputs raw data which requires specific out-of-cam software to execute tasks that are ordinarily in-cam, such as white balancing or handling highlights (knee). For further reference, read these posts:

But then, with some lateral thinking, is it really a valid assumption that video cameras should output ready-to-use images? After all, in the film industry color grading has long been a natural part of the production process and, in a way, RED is only extending similar requirements into the digital world. Once past the initial learning curve, maybe people will get used to the postproduction constraints of RED and find that a familiar, natural step.

Similarly and by extension, check out Kevin Shaw’s informative website about colors, especially the PDF addressing the key question of When to Color. These paragraphs particularly caught my attention – quote:

“There is a myth that shooting video to begin with removes the necessity for color correction; scene matching and style are still needed. It is true that electronic capture has an inherent dynamic range that’s closer to video, but the downside is that this dynamic range is insufficient to capture exteriors and many available light situations. As a result the colorist must optimize clipped whites (or at least diminished white detail), crushed blacks and the flatter linear contrast of video originated material. […]
The color style and effect very often requires a collaborative effort between the DP and colorist. Some projects are shot straight and the decision on color is left for later, but often a DP will expose and light with color correction in mind. This entails good planning and communication, but almost always produces better results. […]
In conclusion, color correction – like editing – could be done in camera, but it takes a lot of time and effort and still produces an inferior result. Ideally, it is refined continuously throughout post-production, with the finer and more elaborate tweaks happening after everything else.”

Well, that is thoughtful insight! Okay, RED is not film and it is not an ordinary video camera either. But here is my point: how about “ordinary” camcorders, e.g. HDR-FX7? Should we shoot video “for the editor”, i.e. with the intent to minimize postproduction work down to the simplest edit – at the expense of overblown highlights for instance if the scenery has too much contrast? Use Auto White Balance, betting that colors will be right on?

Or should we shoot video “for the colorist”, i.e. focus right from the start on the final intended look and choose that color / gamma correction are an integral part of the video workflow? This includes underexposing if necessary (only as much as needed to optimize highlights), and use fixed or preset white balance even if colors look wrong right out of the camcorder (fixed WB is easier to correct in post with a proper color target).

I guess this depends on how much effort you are willing or allowed to put in postproduction. With my experience as a hobbyist with no budget or constraints, I vote for the latter option (shoot for the colorist), and will post supporting examples or tutorials when time allows.