/untitledfictionproject: A workflow for structured improvisation
Note: Although this post is about my experience with “Untitled Fiction Project 1.0”, an unscripted dramatic web series, I believe the process can be applied to any short or long-form fiction project that relies on improvisation.
“Untitled Fiction Project” was born as an experiment in minimalist fiction filmmaking. The original plan was to make a handful of 5-10 minute short films that all shared the basic premise of “two people in a room” (two actors, one location and one continuous scene). I wanted to create a space where I could concentrate on directing and feel free to experiment without the usual time and budget constraints of a film shoot. In fact, I decided that I would have no crew but myself, a minimal camera & sound setup (Canon 5Dm2 rig), no “movie lighting” (only ambient light), a location (borrowed) and the actors (brave students from the Gloria Gifford Conservatory). And of course… No screenplay.
Being a sometimes “one man band” filmmaker, I felt perfectly at home with most of the rules. But “no screenplay” was frightening. A screenplay gives a fiction project direction. Without a screenplay… What is there? Still, “unscripted” fit perfectly with the minimalist concept of the project and it had the possibility of adding a freshness that would be unique. The first short became an experiment in finding an unscripted workflow and my actors (Jeffrey Casciano & Lauren Plaxco) were key in helping shape the process. In fact, I enjoyed the process and it’s outcome so much that I decided to keep the storyline going. The short became Untitled Fiction Project, Episode 1.
Now, I wasn’t looking for complete improvisation since I didn’t want to compromise visually (if I don’t know what the actors are going to do, where do I put the camera?). The process should have the freshness that improvisation can bring along with the visual variety of traditional film coverage. I decided to work under “structured improvisation”, where a basic storyline is agreed upon before the shoot, but the detailed actions and dialogue aren’t discovered until rehearsals or during the shoot itself. Here is our workflow in a nutshell:
Pick a theme: I actually started the process with only one actor, and together we decided what were the general themes we were interested in. Once we picked a theme, in this case “an affair”, we found the second actor.
Develop characters: Once I had both actors, we developed backstories for their characters and came up with possible storylines. The actors on their own time developed extended backstories for their characters and for the relationship between the characters.
Lock down the storyline: The closest I came to writing was putting together a short description of what could happen in the scene. Once we had this general storyline we could lock a friendly location, find any needed props and set a date for the shoot. Here is the 3 paragraph storyline that resulted in Episode 1:
They come home from a dinner party. They’ve both had a bit to drink and are in great spirits. She opens a bottle of wine as they make fun of the guests at the party. They kiss. She begins to undress playfully. He doesn’t respond to her teasing. He has something to tell her. He’s cheated on her.
At first she thinks it’s a joke. It’s not. Then she puts on her jacket and goes for the door. He stops her, he has to really put some effort as she is hell bent on leaving. He holds her tight. She snaps and hits him as she breaks down crying. There’s nothing he can do or say.
She asks him for details. He refuses to answer as the questions grow more explicit. She is about to leave for good when he suddenly drops his facade. He begs. They embrace. He grabs the bottle of wine. They drink straight from it. He doesn’t try to convince her of anything, but after a while, he makes her smile. And laugh. She’s not leaving him.
Rough scene blocking: Once on location on the day of the shoot, I went over the general storyline with the actors and we came up with a rough blocking for the whole scene that was refined through a series of run-throughs, where some ideas for dialogue were also added to the mix.
Break up scene in segments: We divided the full scene in smaller segments, looking for natural breaking points (like the actors moving to a new part of the location or a change in direction in a conversation). The idea was to tackle each chronological story segment on it’s own, and not move on to the next segment until the one before it was completely done. In the case of Episode 1, it was divided in 10 segments:
- Hallway: They enter the apartment and walk to the kitchen.
- Kitchen: She changes clothes as he checks his phone.
- Living Room: He sits on the couch. She serves wine.
- Living Room: She sits next to him on couch and kisses him.
- Living Room: He gets up and paces. He has something to tell her.
- Kitchen: She grabs her purse and heads out the door. He goes after her.
- Hallway: She tries to exit the apartment but he stops her.
- Kitchen: She wants to hear details of the affair.
- Kitchen: He serves himself a drink.
- Bedroom: She sits in bed. He comes towards her.
Rehearse/shoot by segment: This is where all the backstory work really pays off as we constantly dipped into it to find details that would enrich the scene. We usually ran at least a couple full rehearsals of each segment before we shot to come up with more dialogue, actions or blocking ideas. The actors found it helpful to keep notes of their dialogue as we rehearsed so by the time we started shooting they had put together a dialogue list of the segment. I picked the wider shots to start with so we could continue to finesse the segment and usually only when we came to the close-ups would the dialogue finally be locked. Once we shot out the segment, we took a short break, and started the next segment. And on and on until the full scene was finished.
In general, I would say that structured improvisation shooting takes 1.5 to 2x as long as regular scripted fiction. In “Untitled Fiction Project 1.0” this was somewhat offset by the fact that I shot documentary style (handheld & ambient light only). The average 4-6 minute episode took about 8 hours to shoot. This style also had a big repercussion in the editing room since it pretty much guaranteed a mountain of footage. I had an average of 2.5 hours of raw footage from each episode or about a 30:1 shooting ratio. Still, it facilitated an incredibly fast turn-around from idea to finished product (as fast as 4 days). And it was damn fun.
