The guy we originally wanted to play the hero for our zombie series fell through, so we decided to try out a girl instead. Her name is Phet. Even though she’d never handled a gun or sword in her life, she jumped in with both feet and gave it her all. Thanks Phet!
Our second attempt at a zombie effect. The broken jaw is from the The Walking Dead and the background is from the set of War of the Worlds at Universal Studios.
We had a hell of a time trying to animate this effect. Tracking the face was one thing, but we also had a hard time making the effect look real. When the zombie moves, his jaw should sway from side to side. But since his jaw is only an image, there’s no way we can move it around. Just too many moving parts to make it look right.
Next time, we’ll dumb things down. Instead of a broken jaw, we’ll paste a simple flesh wound. This should make it easier for us to animate and make it look real.
This, boy and girls, is the real deal. A movie-quality, silicone-based zombie mask by the amazing Rusty Slusser. Frighteningly real. Easy to apply. Unquestionably cinematic.
Too bad this bad boy costs $869.00. Hard to justify shelling out 900 clams for just one mask, even if looks this good.
Sigh. Post zombification is harder than we thought. It’s a bitch tracking a face, especially a moving face. Problem is that we filmed the actor without using any high-contrast reference points (i.e. green dots, ping-pong balls, or lasers) on his face. Without these reference points, After Effects has a tough time tracking an actor’s movements. That’s why the zombie effect jumps all over the place. It’s also why we had to cheat and apply camera shake and motion blur to the footage (to fudge the lines and mask the jitter).
Despite shooting a very simple test (just a dude walking toward the camera), we basically had to resort to trickery to get what we want. That’s fine for now, but what happens when we shoot the real thing and our zombie turns his head, looks up at the sky, or does the moondance?
Next time, we do it right. Reference points. Lots of ‘em. Fire up the lasers boys…
It’s been a while since our last post. For some reason, it always takes time for our creative juices to reach any kind of critical mass. Slowly but surely, we seem to be gravitating toward some kind of zombie movie. Not just any zombie movie — a post-apocalyptic zombie movie (just what the world needs). We spent the better part of the month fleshing out a new storyline. We call it Samurai Dead.
Rather than shoot a traditional short film, we decided to go for a series. A series of Very Short Episodes (VSEs). 2 minutes or less. Each VSE will start with quick action and end on a cliff hanger. Hopefully, this bam-bam format will hook our audience and compel them to watch the next episode.
Hold on a second — I’m getting ahead of myself. Before we even start dreaming of our first episode, we need to figure if we can even create a zombie. No way we gonna make a zombie movie without any stinkn’ zombies!
In searching for a way to “zombify” our actors, we’re hoping to avoid the most obvious path: Makeup. It’s too time-consuming and too cumbersome. Plus, none of us know any good make-up artists. Instead, we’re hoping that — with some practice — we’ll be able to achieve a post “zombification effect”. Something like this:
As cool as this clip is, one potential problem we noticed is that the actor remains still throughout the entire clip (no head turns, facial movements, etc). This is fine for a short sample clip, but it’s probably not going to cut it for a movie (under real life shooting conditions). We’ll need to research and test before we find out if post-zombification is truly feasible for a bunch of novices like us.
For weeks, it’s been torturing us. Despite having the right equipment (a 1080/60i capable camera) and the right plug-in (Twixtor Pro for Adobe AE), all of our slow motion footage turned out looking like this:
Hard as we tried, we couldn’t get rid of the motion warping. Eventually, we figured out the culprit: When you shoot 60i, the frames are interlaced (sandwiched together), effectively capturing only 30 frames per second — not nearly enough for good slow motion. Rather, we had to shoot in 60p. Progressive scanning, unlike interlaced, captures all lines of each frame sequentially, which gives you true 60 frames per second.
In addition to 60p, we realized we had to shoot at a much higher shutter speed (1/4000) and a much low ISO (200 or below). With all these changes, we finally got some half-way decent slow motion:
Warping is still a problem, especially when the subject is moving fast, but there’s little we can do short of buying a camera that films at 120p. I think they start at $15K. Maybe next year…
For the longest time, I’ve had a hankering to shoot an absurdly long fight scene (like the one in Old Boy). Tired of waiting, I gathered the boys together and whipped up a quick excuse for a fight.
After this filming experience, I have a whole new level of appreciation for all action & kung-fu movies (even the bad ones). Fight choreography is hard damn work. It all comes down to physics: Working out the relative positions and actions of all the players is surprisingly tedious and complicated (and I only had three fighters to deal with). Storyboarding the action helps, but it’s nearly impossible to fill in every moment. Some level of improvisation seems unavoidable.
One thing that seemed to help us was breaking down the fight scene into smaller parts. This gave the actors a chance to learn the movements and it gave me a chance to fill in the missing moments. Here is a progression of the fight scene from the first take to the final take (with lots of mess-ups in between):
The end result is certainly not perfect, but it’s not bad for a first attempt. As a director, I’m definitely going to need to brush up on my choreography skills. And the boys will definitely need to put in more time at the gym…
Jack, our effects guy, has gone gaga over 3D modeling. Rather than using the same old special effect packs that everyone else is using, he wants to design his own 3D effects so we can apply muzzle flares, bullet hits, blood splatters, and explosions in a more dynamic way. He also wants to create 3D objects like buildings, drones, and spaceships so we can blow them up in our movies.
Here’s a sample of what he’s done so far:
Don’t ask me how he did it. My eyes usually glaze over when Jack talks technical. But, from what I gather, he used a really cool modeling software called Sculptris to mold a bust from a digital clump of clay (how cool is that?). From there, he used a powerful (and free) animation tool called Blender to track the bust to the model’s head movements.
It’s all incredibly cool stuff. Personally, I can’t wait for the day we’re able to apply all this to a real movie. I think we all feel this way. That’s why we keep Jack in cage and feed him bananas. So he’ll work harder, damn it!
Normal people aren’t supposed to have wet dreams about video equipment. Thank God we’re not normal. Top of the wet dream list? The Panasonic AG-AF100. Lust factor: 11.
This sick puppy is the perfect hybrid between DSLR and prosumer video camera. It does interchangeable lenses. Shallow depth of field. Variable frame rates (1080 60i and 720 60p). XLR audio inputs. And built-in zebra, bars, and OIS. Plus, it has a little foldable LCD screen that we missed oh-so-much!
The only drawback is the 4:2:0 color compression, which makes it less than ideal for our green screen work. Fortunately, we’ve found a work around: The Atomos Ninja box — a Apple ProRes Field Recorder.
It’s an external monitor that records 4:2:2 color, 10-bit uncompressed HD video AND takes cheap storage!
God, our dreams just got wetter…
We had huge plans for this shoot — an epic battle scene filled with guns, bullets, super-cool slow-motion, and amazing choreography!
This is what we got instead:
Thank God it was just a test — otherwise, we’d've taken our airsoft guns and shot each other in the face. There was just too much to handle: Follow focus. Manual zoom. Dolly action. Panning. Split-second timing. We went into this with such high hopes. To fail in this way, was just plain humiliating.
Fortunately, there were some take-aways:
Take-away #1: Fuck this shot. We’re not ready for it.
Take-away #2: We need better equipment. It’s cinchy blaming bad footage on your equipment. However, in this particular case, it’s at least partially true. We filmed with the Canon t2i, which served us well (especially being able to shoot @ 720/60p), but it’s tough shooting an intricate shot with a barebone DLSR. Focusing and zooming on such fast-moving action on a small, immobile LCD screen is hard as hell. We’re either going to need to rig up our DSLR (shoulder mount, follow focus, external monitor, etc) or buy an actual video camera (we’ve been drooling over the Panasonic AF100).
Take-away #3: Small projects. Rather than struggling with big-time shots, we decided the best approach is for us to meet every week and focus on one, simple shot and one simple effect (i.e. jumper effect, slow-motion, freeze frame, explosions, etc). As we improve and expand our body of experience we can build up to more and more complex shots — just like the one we just attempted.
Oh well. No one said this was going to be easy…