Improvising Visual Effects
COW Library : Adobe After Effects Techniques : Mark Allen : Improvising Visual Effects
Last year Mark Allen was asked by legendary hairstylist and photographer Robert Lobetta of Sebastian to create a visual accompaniment for his Aurical Angel photoshoot. The budget was limited and the pre-production time lasted only two days from first concept to standing on the set.
Further complicating pre-production, the location changed dramatically at the last minute. As a result, nearly every aspect of the short film needed to be improvised - including the visual effects.
As visual effects become more ubiquitous in projects of all sizes, impromptu effects are not uncommon on almost any set. Mark is a director who not only has been a visual effects supervisor, but also an improvisational actor. We asked him to discuss his thoughts about improvising visual effects in the context of his experience working on "Fallen."
With such a rush into production, what had you prepared before the shoot?
I needed one cornerstone to work with. Since Angels are often perceived to be taking care of mortals, I decided to make the story about growing up - moving from the child to the nurturer; so this child would end up nurturing of these fallen angels. That's pretty much all I knew going in.
I asked the art department for some things which implied childhood: a Teddy Bear, a shopping cart (because for a child this is actually a mode of transportation), and maybe one or two other things and the rest had to be made up on the day - including the visual effects.
The only effect I knew I'd have to have is an angel falling at some point. Probably several. We actually had a big trampoline because the original idea was going to have these angels bounce on a trampoline and we'd catch them right as they were about to land, but that was nixed because it's apparently very dangerous to land on your back on a trampoline. Thus - time to improvise.
What I did instead was have the models balance on just one point of their body - whether it was their knee on a pad, or their heel, or the small of their back. If they had just one point contacting the surface, it would be hard to sense that once they were composited.
At one point in the piece, the Angels land. Hard.
These models were all courageous people, but I didn't think their management would appreciate us dropping them from 20 feet to create a nice landing. I could have shown the child discover the Angles on the ground, but it's always better to show the dramatic events rather than the aftermath.
At one point when I was composing the first shot of a model having landed in the hallway it became very clear that there was a simple solution.
The falling trick is really all about the wonders of motion blur. One thing to always keep in mind when thinking about visual effects is how much you really end up seeing the element you're working on.
If you're doing a 3D spaceship, for example, and it's just going to zoom through the frame - why would the texture artist spend a day painting the perfect texture when it's really just going to be a blur? That same concept plays here. Someone falling is really going to just be a blur in the frame if they're falling fast. Knowing that, I figured I had some leeway.
So, for this shot of the Angel in the hallway, I realized that there was no reason I couldn't just cut her her out and drop her because the motion blur would save me - she'd only be falling for a few frames at most.
Faking the landing, however, would be an issue for animation, so I just asked the model to sit up and let herself fall to the ground. That's all I needed. Now I had every position I would need for the aftershock.
One should always try to visualize the entire effect when on the set. If there's a bright light, you really should try to at least put a light in there for reactionary lighting. If a body falls, you are right there with a body - so if you can't get the fall, at least shoot the aftermath of the fall.
One of the most iconic shots in the piece is when the child touches the wings. The effect is subtle, but makes a big difference.
Originally I was going to have the wings in the frame hanging like they'd been caught on something, but there was nothing magical about them. So, we just suspended them in the space and hit them with a light to pump them up to imply they were floating.
The grips were trying to hide the rig that held them up, but time was short, so I just said that it didn't matter and we'd remove the rig later.
Since we were already doing an effect here, we ended up lifting the wings out of the shot, painting the background and then floating the wings back into the scene to give them a little more life and magic.
We added a little wind to the edge feathers and pumped additional light into the wings to make them even more the center-point of the shot, and more like a bright and shiny object that the child would want to touch.
Now, from an effects standpoint, the key here is that we shot the wings in the scene knowing the actor's hand would cross over them, and intentionally not bothering with a green screen because it would spoil our lighting.
The most important thing was to get the majority of what we wanted in the camera, and then the effects from that point were just an enhancement.
In addition to improvising your visual effects, you run an improvisational acting workshop with actress Kelly Hu and have also written about improv. What are some of the most important rules?
Don't Negate. If you negate, you're stopping the scene from progressing because you're afraid of where it will go.
Don't Ask Questions. If you just ask questions, you're putting all the creative responsibility on your partner because you're afraid to present something yourself.
Don't Judge. If you're judging the performance or yourself or your partner, you're not living the moment of the scene. You're trying to escape from the moment.
Find the Game. Find the dynamic relationship between the characters. It's often the unspoken joke or secret that everyone knows including the audience, but no one is saying.
What stops actors from finding the game is if they are focused only on their own needs and their own world. Likewise, a visual effect can be as beautiful as any effect ever before seen, but if it is too focused on its own goal of being beautiful and doesn't fit with the rest of the scene, there will be no dynamic play.
Those are rules for improv acting. Can you suggest some rules for improvising visual effects?
Tell Your Story. Make sure your effects are helping tell your story or adding production value If it's just a neat idea, you might be risking too much liability and expense.
Collect Elements. When you've committed to an effect, make sure you collect elements for the artists as well. If you are going to replace something in the shots, scan around and see if there is something you could shoot or photograph to replace it with.
The biggest key to capturing elements could almost be its own rule: Observe Lighting. If you are going to want something to float in the room, then shoot it in the room - even if it's on a rig. That lighting is what is going to make it sit nicely in the final composite more than any other element.
Don't Assume You Know. Unless you are an effects artist yourself, don't assume too much about what would make things harder or easier, because what usually makes a shot really difficult is something unforeseen. Make sure you get what you want out of the shot.
For example, a director might think that a locked off shot would help the effects department, but the whole show is a moving camera - and then you get to this one locked off shot. It's nearly impossible to sell any effect when it's so different from the flow of the show.
Stage it Dynamically. With visual effects, you are always trying to hide your effect. You want the moment to feel natural and not staged or premeditated.
Often directors will get intimidated by the idea of a visual effect and frame their shots as a tableaux which looks like everyone is waiting for the big effect to happen. It's very self conscious and the audience picks up on this and the element of surprise is lost.
Basically, don't change the flow of your shooting style for the effect or you'll be drawing too much attention the effect.
Green is not a Magic Color. There is one instance when you want to use green-screen - when something is going to cross over an element which will be added later.
For example, if you're going to replace the screen of a mobile phone and nothing passes in front of the phone, there is no reason to use a green screen. In fact, by letting the screen be black, the compositor can use the reflections from the phone's glass to help sell the reality.
Now, if there is going to be some hair hanging in front of the screen as we look over the shoulder - yes, use green screen, because that will help to key out the hair. If nothing is passing in front of the area, you are only risking green light spilling on your set and actors by putting a green screen in there.
What would be the single guiding principle any director should follow when faced with an effects emergency on the set?
Focus on your story, then follow your instincts.
Don't make choices out of fear. This not only applies to improvising visual effects and improvisational acting, but it really applies to all creative and non-creative life activity. The more time you spend questioning yourself, the more time you're wasting.
Sometimes the people who have had a tremendous number of accomplishments have also had a lot more than the average number of failures - but if you're goals are clear and you're making even slightly educated guesses, then the odds are on your side. It's just a matter of volume.
The worst thing a director can do on the set is be frozen by fear while unsure about something. If I'm directing an actor and I give them a note which just simply doesn't work, I have no problem in saying "totally ignore than note, it was wrong, let's do this instead." It's okay to be wrong as long as you keep moving.
Same with visual effects. Even if you had 10 years of experience with visual effects, you aren't going to be able to predict all the thousand possible problems a shot might have - so just make sure that the effect has value and purpose for telling your story and it will have been worth doing.