I just spent around an hour on the phone with one of my clients. You may know him: Kendal Miller. He’s 1/2 of FreshDV.com although Matt Jeppsen (DP on this project) spends most of the time blogging there I think. Anyway, they asked me to come on board as colorist for their latest music video project. As is should be more natural, that turned to asking me what I need to finish the project well which turned into some pseudo-post supervisor role. So I think I’m going to blog the whole thing as it happens. Failures, screw-ups, epic wins and all.
The project is for the Christian metal band Shaken for their new single “History”. They’re in the up-and-coming category. We first talked about it a few months ago and I was able to provide input on wardrobe and art that would enable the coloring process to go much smoother. Then at NAB, Matt and I spent some time discussing technical, post and workflow.
The Pre-Grade Discussion
Today I talked with Kendal at length about the final look of the film. Working in a different locale I think is going to become more of the norm but presents it’s own challenges. So I typically try and watch through the piece before I talk with the designated decision maker. At this stage, I like to start out by asking questions based on desired audience take-away, emotive responses overall and/or certain key scenes. That naturally moves towards discussion about what the director and DP were going for during the shoot, what should be preserved, what needs to be minimized, changed or taken away, what looks to avoid or pursued and how far we want to push certain things. This usually gives me a really good start on the grade while it’s in process of editing.
The Tech and Workflow
Where we’re at right now: the footage is all with me. It was shot RED MX (.RDC files) and Canon DSLR (5D, 1D and I think maybe 7D). Major workflow considerations were the online final finish, low-impact offline editing and color pre-visualization.
Starting with the online finish, I want to be able to grade the master files- this is especially the case with the Red material, but of course, it’s not editable in it’s RAW form by the editor. The DSLR stuff, I could take or leave in it’s native H.264 but it must got through the qtChange process so I get workable reel ID and timecode information.
This brings us to the second consideration of editing. The editor is on a laptop and it isn’t the latest workhorse machine so we need something that plays nice. He’s grading in FCP so a conversion to ProRes for all the footage makes sense here
Which leaves us at our final consideration: pre-grade visualization from a distance. I want the DP to have the ability to mess with some basic elements of the footage and send me something he likes. I also want the editor to be able to work with something that’s at least had some contrast thrown back in from the Red .RDC state. I also want to be able to work from the DPs initial grades in the final finish.
After we discussed these options, we settled on Cineform. It’s really easy on lower-spec systems, native in every application and First Light allows us all to work from the same spec grades via shared database. And like ProRes, I’ll just leave the DSLR footage as Cineform files for the final grade and conform to the original .RDC Red files in Resolve. I briefly toyed with the idea of converting the Red to Cineform RAW but didn’t see an advantage with an extra CF RAW to 422 conversion and Resolve doesn’t read Cineform RAW conversions natively anyway (which is weird since SI and CF are native to Resolve). Cineform is a great way to eliminate some of the multi-format hassles and unify a given workflow.
There are distinct differences and obvious limitations between the DSLR and the Red footage. For those mostly obvious decisions, Kendal informed me that they’re going to be limiting as much as possible the amount DSLR footage used in the final edit.
I currently have all the footage and am performing batch renders on all the footage via Cineform’s ReMaster, will pull it into First Light and do a pass on it with the database set to a designated Dropbox folder. I’ll then ship all the (at this point, I’ll refer to it as offline footage) to drives supplied by Kendal and shipping them to the Jon (editor), Matt, and Kendal this week. Matt and/or Kendal will sync up their First Light database folders to mine so I can see on my files what they tweak on theirs. Looks which I’ll then recreate and tweak in Resolve.
Yes, those grades will show in Resolve if I decided to leave the First Light grades on the footage but I don’t particularly care for First Light’s primary grading controls- no offset/printer lights controls and no L/R/G/B curves which is typically where I start with each grade. And besides, I’m going to revert back to the .RDC files for the Red footage anyway.
I will be receiving updated XMLs as the edit progresses so my grade will evolve with the edit and happen concurrently saving us all a chunk of time which is cool. Following the edit lock, Kendal will be flying to my studio here for the final assembly, approval and master finishing.
Project Posts Moving Forward
My plan is to put up a post after each grading session I spend on the project and talk about the actual grade, the challenges and how all that fits with Kendal’s overall vision for the music video. I’ll be including some screen shots, why I’m doing what I’m doing, the thinking behind it and how that meets the goals of the director based on our previous conversation(s). It should be a pretty fun series of posts and I do hope to be consistent so hold me to it!