Post-Mortem: 2015 Providence 48 Hour Film Project

Analyzing and documenting our entry in the 2015 Providence 48 Hour Film Project

My 48 Hour Film Project team, Shoot the Moon Films, participated in the first Providence 48 Hour Horror Film Project last year, but had never entered the regular 48 hour festival in Providence, so we decided to give it a shot. This is my critical analysis of the processes used to create “Buck’s Bed & Breakfast” (IMDb).

WARNING: There are some details of the film which may function as spoilers, so don’t read below the fold if you haven’t seen it yet.

Day One: Preproduction, Kickoff, Writing, and Planning

After deciding to participate in the project this year, we collectively decided that we weren’t particularly interested in going the safer route with most of our production decisions. We put much more effort into ensuring good audio (including putting together mic body packs for wide shots – just in case) and lighting. For the first time, we asked for a Best Boy to work with our lighting designer, and gave her much wider latitude on lighting decisions.

Unfortunately, we were unable to work with Aaron, our usual sound wizard, for post-production (he had won the Sound Design award at the 48 Hour Providence Horror Project last year), and the engineer we worked with on The Gift of Fate, Ben, was also unavailable. At what seemed like the eleventh hour, we were able to bring a new audio engineer, Rob, on board. I had never worked with him before (or met him), so there was a bit of apprehension on my part.

The previous 48 hour project we had done in Providence involved me driving out for the pickup event, messaging the writers the information, and driving back. (Anyone familiar with my comments about the poor state of the roads in Rhode Island knows that I’m not a huge fan of driving to Providence.)

Neither our producer, Curtis, nor I live incredibly close to Providence, so we decided to have him do the pickup and message the information to the writers so that I could help with any technical issues they might have had using the collaborative writing software.

We drew “Fish Out of Water” as our genre, and the writers were both very pleased. We tend to end up pulling dramatic genres and put together far more somber productions, even though many of our actors pride themselves on their comedic chops and our writers like writing comedic parts.

After a few minutes, the writers came up with a funny concept and, in a very unusual move, called me out to ask whether or not I thought it was feasible. (I tend to separate myself from the writing process because I don’t want any of my preconceptions or technical assessments to unduly influence the writing.)

“What if they’re all nude?”

I laughed, and told them that we needed buy-in from our actors, obviously, but that I could manage to film something which wouldn’t be obscene or revealing. Even more surprisingly, everyone was “in”.

The writers cranked out the final script in about two hours after they received final confirmation from all of the actors involved. In an act of complete solidarity, they wrote small parts for themselves in the same state of undress because, quote, “We couldn’t ask them to do something we weren’t willing to do.”

A great deal of the blocking couldn’t be done until we had all of the props, since we wouldn’t be able to see what would (or would not) cover all of the areas which would need to be covered, but I stayed up for a few more hours getting scripts printed out, sending out the PDF copy to our actors and crew, and making shot notes. I also took a bit of time and loaded as much equipment as I could, with the exception of the batteries which were charging inside.

I got to sleep by 2 A.M., a new personal best, if I recall correctly.

Day Two: Filming

I got up around 5:50 A.M., and after taking the dogs out, got back to loading the vehicle with all of our remaining equipment. We ended up carpooling up to Shrewsbury to meet Curtis and the non-Connecticut-based actors at the shooting location around 8:15 A.M., since there was an unexpected traffic situation on I-290E which slowed us down. We set up hair and makeup outside in the vehicle, and had the first two scenes dressed and ready to shoot by 10 A.M. – right on time to match our producer’s timeline for the day.

One slight wrinkle which was thrown in was that one of the principal actresses, Arianna Danae, had to leave by 4 P.M. to meet a prior obligation, so we ended up having to reorder the filming sequence to wrap her parts before she had to leave.

We skipped using the gimbal or any extremely complicated rigs for most shots to save on between-shot setup time. The two notable exceptions were the opening shot, which used my trusty Glideshot portable jib, and one drawing back shot, using a rail set and dolly conversion for the tripod.

Additionally, I had purchased a cheap SDI-HDMI converter to allow using an external monitor for the production designer and lighting designer to monitor the diegetic space, but which also came in handy so that multiple people could watch for any, ahem, “unexpected exposure”.

For the most part, I mostly used a 35mm T/1.5 lens, occasionally swapping it out for a 24mm T/1.5 for constrained space or wide shots, an 85mm T/1.5 lens for CU/ECU shots, and a 70-200mm f/2.8 IS lens for depth tracking. I tried to stay with the 35mm lens to stay away from any sharp changes in perspective between characters – unless it was necessary to tell the story.

We also ended up working with two new crew members: the versatile Mary Hronicek, who set-dressed and handled continuity, and Kaeli Black, who handled – well, just about everything else on set. They dramatically reduced the amount of time it took us to switch physical positions within our filming location by prepping the next area while we were filming or blocking in another. This was more help than normal, as our production designer was one of the principal actors, and would have normally had to do that work himself. Our lighting designer / gaffer was also acting, so having additional crew members to help with light positioning and arrangement was crucial to working under a deadline.

At around 6:30 - 7:00 P.M., we finished up shooting, and headed back down to Connecticut. It took about an hour to move all of the approximately 500 GB of raw footage and audio onto the editing station (damn you, antiquated eSATA interface!), and editing got underway at 9:00 P.M. Curtis was unable to join us for editing this time, due to scheduling conflicts in Massachusetts.

We finished our first completed rough cut around 11:00 - 11:30 P.M. I went through and did all of the major audio replacement and correction work (fixing lines which weren’t recorded properly, etc) and rendered out both OMF exports for the audio engineer and an MP4 low-res copy for our composer, Garett Schmidt. As soon as they had uploaded to Google Drive, I messaged Rob and Garett with the link, plugged in my phone to charge, and promptly fell asleep on my face on the couch.

Day Three: Post-Production and Dropoff

Woke up at 7:00 A.M. (since we got the credits done the night before), walked down the street to get a cup of coffee, and started work on color grading. I wrapped up initial color grading (with one shot worth of stabilization for a dolly shot which had technical issues) around 9:45 A.M. I rendered everything out as separate clips and reimported into the NLE software, then rendered out a scratch copy and uploaded as private Youtube video so that our producer could look for any issues.

Around 1 P.M. we got the audio mixed with score back from our sound engineer, Rob. I did a quick review of the video to make sure everything was in order before we continued on.

After watching it through again, we identified three issues: there was coloring missing from a particularly difficult multiple light color shot, the stabilization algo had malfunctioned catastrophically, and there was a non-diegetic sound missing. The first and third issues were remedied quickly – but the second one turned into a bit of an issue.

I completely retracked it and removed all non-background points twice, but both times the image “jumped” at a certain point. I ended up using (ugh) Warp Stabilizer in Premiere to smooth out the instability a little, but I blew almost an hour trying to fix that single shot.

To add insult to injury, Media Encoder decided suddenly not to want to encode the video, freezing a few seconds in. Thankfully exporting directly out of the NLE worked fine. We burned a data DVD and made a USB key copy, had Nick test it out in the DVD drive and USB port on his laptop – to ensure that it would be readable on another device – and jumped into Jon’s car to head out. (Jon, by the way, is the crew member who always ends up driving me to dropoffs whenever I have to make them, since I can’t see straight by the end of the 48.)

After getting slightly lost twice in Providence, despite the assistance of Google Maps, we arrived to drop off our deliverables at 6:08 P.M. on Sunday, giving us a little less than an hour and a half before the final deadline. There was a minor paperwork scramble, due to a miscommunication between the 48 Hour Film Project instructions and our producer, but a quick set of messages to him straightened it out. Done!

Post-Mortem Analysis

The good:

  • Writing time. We had a script done within three hours of receiving our constraints, and less than two hours after getting a concept approved. This is a testament to the efficiency of our writers, Natasha Darius (my lovely wife) and Nicholas Magrey. They managed to put together something subtle, yet somewhat outrageous at the same time.
  • Comedic timing. Between Arianna Danae, Jonathan Demers, Curtis Reid, Debra Leigh Siegel, Jessica Rockwood, and Paul Magrey (and our two writers, of course), the comic timing of the writing came out beautifully. Some people can’t do comedy very well, but I’m proud of everyone who acted in this one.
  • Monitoring footage. Using the SDI-HDMI converter with a monitor was a smart move; it allowed everyone to monitor what they needed to monitor without causing a traffic jam behind the camera. It also allowed actors to see whether or not they were on camera, in certain situations.
  • Few, if any, specialty shots. We didn’t spin up the gimbal, snorricam, shoulder mount, or any other specialty rigs besides a single jib shot in the beginning and an ill-fated dolly shot in the third act. Some of this had to do with pulling comedy as a genre – and it almost certainly forgoes any sort of consideration for a Cinematography award of any sort, as that seems to be relegated to fancy stuff – but we focused far more on telling a funny story than trying to astound the viewer with slider shots, drones, or passing gimbals through windows. Truth be told, some of the light and grading work in setting up shots to work over a moving light temperature (the sun, as it changed position in the sky) was pretty complicated and artfully executed, but it’s not anything wowing. That being said, I wouldn’t change any of it.
  • Organization. Using the racking system in the vehicle and keeping hair and makeup separate from the rest of the shoot made organization far easier than it had been in past. We brought more equipment than we needed without compromising our ability to use other pieces because of the amount of storage space we had.

The not-so-good:

  • Equipment Preparation. Entirely my fault, but we hadn’t done a full functional test of the rails and dolly in quite some time, so we would have noticed that we didn’t have the proper shimming to keep the system from moving, as well as requiring some additional height to clear the grass for a completely smooth run. This would have completely eliminated the need for stabilization in post (which I hate), saving me time and sanity in that process.
  • Cross-talk and blocking time. We tend to run a pretty “familial” set, and so we tend not to be harsh or commanding when dealing with actors. This fomented some issues with timing, as we ended up waiting on some people for hair and makeup who could have been done earlier, waiting on people to stop cross-chatting during shooting, and waiting on people to be available for blocking.
  • Periodically messy camerawork. Some of my focus pulling was not completely accurate. I’m not sure whether I should have just had my 1st AC pull focus for me or whether I just needed to use the loupe more extensively on my EVF, but reviewing the final footage now still shows some of that. There is also a single shot where there are a few minor hot spots on the footage.
  • Tunnel vision. I wasted a fair amount of time trying to correct something (an annoyingly unsteady rail shot) rather than either:
    • Planning the shot differently (zoom lens, gimbal, etc)
    • Cutting losses much sooner in the grading/post process and editing around the undesirable footage or using different methods as soon as it became immediately apparent that it wasn’t functioning properly.
  • Lack of on-set schedule. Even in a time-constrained environment like a 48 Hour Film Project, we could have set up a scene schedule at the beginning of the day so that we would know how far behind schedule we were falling. This could have served to both motivate and keep the cast and crew focused, as well as allowing more effective time management during scenes.

Overall, I’m happy with the way Buck’s Bed & Breakfast turned out, even though I can (as always) see the technical flaws still left in.