Blog Entry: This One is Blue (Behind the scenes)
A funny thing happened on the way to making H-Hour—we made a Kickstarter video. Now, making movies is not my favorite endeavor but it’s not the most loathed either. What I really like to do is make games. There are sometimes little movies contained within those and I’ve certainly made more than a few little movies over the years but those were entirely digital. Mocap for animation and a vocal performance from actual humans, sure, but not really live action shot on location where anything and absolutely everything can go wrong. And for the most part, I had a team of specialists who were very good at their disciplines leaving me free to write and direct. Which really, is quite enough to do. This time, however, was very different.
Our Kickstarter video took about one month from start to finish. I focused primarily on the creative side while our company president, Tom, took on the role of producer. He managed to line up the location, many of the props, a team of volunteer actors, and of course, authentic military equipment and firearms.
I handled most of the creative side of things and along with Tom, was one of the primary actors. A local photographer, Steve, handled lighting and cinematography. All told, we were a three man crew most days, with our intern Kevin helping out some of the time. We all wrangled props and of course, my favorite part, I sloshed commercial grade theatrical blood onto anything that looked promising. Here’s a practical tip: spring for the good stuff. We used two kinds, one non-toxic and suitable for pouring into the mouths of unwitting but good natured extras, and a second sort, apparently toxic, which was labeled “do not ingest.” In hindsight, drinking toxic blood might have been preferable to inhaling all the dust, desiccated bird excrement, and possibly crumbling asbestos and lead paint fragments.
The location was superb, everything you could ask for in a setting in which game developers were fictionally being held hostage. While nominally maintained by caretakers, there’s evidently no budget to repair the roof and birds have nested—heavily—throughout the building. So there’s excrement and feathers everywhere. I eventually gave up on using disinfecting moist towelettes and took the attitude that all the filth was just going to enhance the realism in the footage. I probably should have composited in the filth using After Effects because I became very ill after the shoot. For three days my eyelids were nearly swollen shut and I couldn’t stop sneezing. I was convinced that I had contracted an emergent new strain of SARs. Fortunately I was able to keep working through this malady.
Again, I really couldn’t ask for much more in a location. This place was vast and resplendent in its ponderous march towards complete decay. Huge strips of institutional green paint were peeling from the walls, and while dirt was prevalent and most fixtures had long ago been stripped away, we kept finding beautiful hexagonal tile work in the bathrooms and showers. It’s the kind of work you can’t even pay to have done today. All told, it was the sort of place I generally love to explore but we had a tight deadline and had to focus on work. Unfortunately there was no electricity at the location so we had to rely on generators. This added hours to the shoot because the machines had to be lugged around and repositioned. Naively we tried to place them as far away from the filming as possible reasoning that a modest amount of generator hum could be processed out using Adobe Audition and that most of the takes would be clean enough to use. It would have an indie feel, we told ourselves.
Hah, hah. Hah. Not funny really. Even a relatively quiet generator is FREAKING LOUD. After spending a few hours trying to remove the background noise the audio was far worse for it. Adobe’s tutorials cheerfully promise that you can remove repetitive noise from your footage and this is true to an extent but nothing substitutes for clean audio in the first place. And on top of the generator noise the place was much like an echo chamber or more precisely, a warren of echo chambers of varying shapes and sizes, each with radically different acoustic signatures. In the end this meant almost none of the live audio recordings were usable and all of the performances had to be looped in “the studio” (otherwise known as my home office). Of all the audio from the location shoot, I was able to use Tom’s rifle fire and some shotgun fire.
At this point it’s worth mentioning that the firearms you see in the video are real as are the bullets they were firing. In order to be able to shoot guns in a building you need something called a bullet trap. Bullet traps are 500 pound steel monsters and while inanimate, they are cantankerous, unforgiving things that, if they had hearts, their only desire would be to topple over and crush your pelvis. Really indispensable for live fire performances but best avoided in daily life. Seriously, bowling with anvils would be easier than moving these things around and we were fortunate that our extras were onboard to maneuver the bullet trap up two flights of stairs and around architectural obstacles.
We decided to get a jump on shooting the first night, which was of course, a night shoot. Which means it’s dark. This complicates matters so we hauled the generator up to the roof. The idea was Tom inserts into the building from the roof by rappelling over the edge of the roof, down the side of the building, climbs through a window, and knife kills a terrorist in a room there before moving down the hall to the line of prison cells. Of all those good intentions what you will see in the video is Tom walking down the hall plus a b-roll shot of him climbing down a ladder that leads to the roof.
Tom was a good sport and actually did rappel down the side of the building. He climbed along a treacherous ledge and up to the window. But when I reviewed the footage the next day, it was so underexposed that it was unusable. We had even shot the first scenes with Tom and I together in the cellblock and that was all red and green dancing sawdust. So here’s the lesson: light everything so that it’s well exposed and add the mood in post. When I was a film student, that was not possible, so I quickly learned that shooting in digital formats, while very flexible, doesn’t always play nicely with traditional filmmaking technique.
Speaking of “fixing it in post,” that went quite well. Surprisingly. After three exhausting days on the set, I was happy to sit down with the raw footage, split it up into takes, and do some color grading. I was also anxious to explore what I could do to make myself look less like a filthy hobo. Well, less like a meth addict. During the shoot I didn’t sleep more than two hours a day and my skin was exhibiting an allergic reaction to something in the air. I looked…unhealthy.
I’d done my research and decided to use Red Giant’s Magic Bullet Suite. They chose the name well. It is magic. Of primary interest to me was Cosmo, their skin smoothing toolset; Looks, an extremely powerful and flexible color grading package; and Mojo, which basically adds visual punch to footage.
In the end, I was able to create a “blockbuster” look with an overall blue-green tint and crunchy blacks without too much headache. And thank goodness for Cosmo which transparently took years of presumed meth abuse off my face. The only downside was processing time for the HD footage itself. The tools were intuitive and responsive, and made iterating on a look painless. As you can see in the image below, the before and after is dramatic.
You can read about Magic Bullet Suite here: http://www.redgiant.com/products/all/magic-bullet-suite/
The software is more than reasonably priced for what you get, and I highly recommend it.
Editing was very straightforward and done in Adobe Premiere. No surprises there. I’ve been using Premiere for decades now and along with Photoshop it’s one of the primary software tools I use in preproduction and production.
Thanks to meticulous storyboarding, all of the takes were usable and cut together without issue. Of all the lessons I learned in film school, that’s the one that has served me best in my career: storyboard everything. Since I didn’t have an artist on hand to help me, I did them myself using Poser Pro and a few characters I downloaded from DAZ. Here are some frames from the boards. I’m the character that looks the most ridiculous which, of course, is appropriate since I look ridiculous in reality. Tom gets to look authentic for the most part though. I simply didn’t have enough time to monkey around with my character model to make him look more idealized. But having said that, Poser Pro is outstanding for creating storyboards fast. I am not an animator (obviously) and didn’t have a lot of time to spend on the boards, but the results conveyed what we had to capture on video and saved me from a tragic breakdown during editing. Here are a few of the boards we shot from. I really like how I managed to capture “the essence of capricious sauntering” of Tom’s character in shot 23. He’s totally like that in real life.
What took the most time was painstakingly recording all the dialogue, testing it against the takes, and then either cutting it up and adjusting the audio by hand to match the performance or just rerecording the lines until they were “close enough.” The final results were a decent but far from perfect match between audio and action, but I see all the flaws. Again, there wasn’t enough time to really achieve perfection with this aspect of the video, so I had to let it go. Plus I learned that my lip motions don’t always correspond to any sort of intelligible speech, which made the process even more of a challenge. As a person who doesn’t like to look at or listen to himself on video, this was a singularly painful process and I guess I’m over it now. But I am not sure I would recommend it to anyone.
In all, I think the video turned out okay. It’s full of subtle and sometimes not so subtle messages and inside jokes. Also it’s less violent than I’d intended but the realities of live action shooting—realities such as requiring actual people to play parts and with whom coordination proved improbable—meant that a lot of that delicious brutality never got past the storyboard stage. And I wish my performance had been better but then again, I’m not really an actor—I just play one on Kickstarter. Hopefully the message the video was intended to communicate comes through and people understand our philosophy and goals a bit better. If that’s the case, when I look back on the process of making this video I’ll remember it fondly.