Making of: Spy Games (Post-Production and Reshoots)
When last we left our wild weekend, it was 3AM Sunday and we’d finished shooting about seven hours later than we had scheduled for.
3AM Sunday put me at 43 hours since waking up on Friday morning, with only the 45-minute nap Saturday morning to break up my waking hours. Although I was actually surprisingly energetic and alert during the last lap of the contest, I must confess that very little of what happened in the initial post-production phase managed to impress itself on my memory. But I’ll try and recall what I can.
I don’t remember the process of assembling the rough cut at all, but what I do remember is that we released Ryan to go get some rest, to come back in the morning when we had the necessary plates ready for him.
Because we’d shot on the RED, we couldn’t work natively with the footage in Final Cut Pro, nor in After Effects (though this latter case has since changed). The RED camera produces 4K footage in a proprietary format, “r3d” files, but also generates Quicktime-compatible “proxies” that can be imported and edited by Final Cut. We cut the short together via the 1K proxies and then rendered high-quality versions of only the footage we used out of REDcine, conformed via Crimson.
In editing we discovered another crisis: the sound was completely unusable. We had been recording audio directly to the RED via XLR, and it was low-fidelity and crackly as shit. (Many of the early RED sound boards were bad, but it has since been rectified in later cameras and early cameras, like mine, were later given free upgrades to the new sound boards.) So once the film was cut, we had to call our Ronald Donnellson, and our original Tanya, back in to re-record their lines.
Meanwhile we were rendering shots for Ryan to work on. Initially we were doing the effects work at 1080p resolution, so that we would have an HD master. But although Ryan did the tracking and compositing with rather stunning speed, we were getting bogged down in the previewing and rendering, and I realized that since the contest wanted only an SD delivery, it was foolish to be wasting time that way. So instead of round-tripping through REDcine — which involved exporting an XML from Final Cut, importing the XML and footage to REDcine, and then rendering the footage from REDcine — we could simply render plates directly from the 1K proxies out of Final Cut, speeding up the entire process considerably.
But even working on sound while Ryan did the visuals, it became clear as the deadline ticked nearer that, with render and travel time to the drop-off factored in, it was simply not going to get done in time to be included in the competition. So we set the final renders and called it a day, since there was no reason to kill ourselves to try and finish when it wasn’t going to happen.
But I found out that the film could still be eligible for screening, and for an “Audience Favorite” award — if I delivered it within 24 hours after the competition. So after a good night’s sleep, I dropped in the finished renders, burned a DVD, and delivered the film to the contest.
The film screened, people liked it, but it didn’t win. (This one did, and well-deserved I think.)
That might have been the end of it, had the film been the kind of film you usually see as 48 Hour Film entries. But collectively we felt that the film was interesting, and generally of a good enough quality that it was worth putting a finer polish on it and releasing it as its own thing.
Priorities for this new version: an HD finish, a better sound mix, and a new opening scene.
The original opening scene between Daniel and Tanya (which you can read in the script linked in the first post on the film) was pretty bad. “There’s no crumbs, it’s potatoes”? Fuck me. This is why you don’t shoot a first draft. I’m not officially a “professional” screenwriter (no one has paid me yet), but I’ve been doing it for 11 years, I’ve gotten close to being paid, and I like to think I’m pretty good. And I’m still dropping turkeys like that on my first pass.
The problem, really, was trying to make too big a thing of the potatoes, because we had to include them. So the scene became more about Tanya trying to get him to eat the damn potatoes than actually showing the relationship between the two of them. There was no chemistry between Anthony and the actress who originally played Tanya; in fact Daniel seemed almost hostile toward her, and she was giving him the crazy eyes like she’d poisoned the food or something. All of which undermined the rest of the story since we were unlikely to buy that he would go to so much effort and danger to get her back.
After cutting the film together I thought that the second time we saw the potatoes — when they’re kind of trashed as a small sign of some kind of struggle — was a more elegant way to use them in the story. They’re subtle and they also indicate a story point, rather than being all up in Daniel’s (and our) face. So we wrote a new scene, taking out the potatoes and focusing more on their relationship with each other:
As you can see, the scene as written was pretty threadbare compared to the way it turned out in the film. The idea was that it was just a basis of the main points for the actors to hit. I wanted them to improvise a bit, so that the scene felt natural and playful.
When we were scheduling the reshoot, we did call the original actress for Tanya to see if she would be willing to come back, but she responded that she “felt she had devoted enough time to that project.”
Alright, well, you’re out then.
Fortunately Tanya only appears in the one scene (plus the pair of “tied to the chair” inserts in the final scene), so it’s not like it screwed the whole movie to have to reshoot her stuff with a different actress. We were going to be reshooting almost all her stuff anyway, and adding the inserts to the schedule was not a big deal.
So we put out a call for actresses and we came across Arden, who plays Tanya in the final film. It was actually a tricky thing to cast the role because we needed to find a girl who was:
2) Comfortable with partial nudity (even though it was on Anthony’s part and not hers, you might be surprised how many actresses are uncomfortable with that)
3) Comfortable with kissing on camera, and really going for it rather than just a peck on the cheek (we needed to believe this couple was deeply intimate)
4) Willing to do this for free.
Fortunately Arden was a pro, and once she had read the script and seen the original cut of the film, she was confident that we weren’t making a softcore porno and she graciously gave us an afternoon of her time.
This was still well before the RED sound board upgrade program was announced, let alone implemented, but we knew that the HVX sound system was actually very good. So we wound up doing a dual-system shoot, in which we would shoot with both the RED and our HVX, have a clap at the beginning of the take for sync, and then in post sync the HVX take to the appropriate RED take.
As I mentioned before, now that I had done a “practice run” of the film I understood a little better what I wanted to do stylistically with it. I like handheld cameras when done properly (Children of Men: yes; The Bourne Whatever: no). Aesthetically I enjoy it for the “reality” it can add, and practically it’s much easier to “run and gun” and get some really cool shots — although this is less true with the RED fully decked out and topping 50lbs.
But knowing that we had time and control to work, I made a conscious stylistic choice to only introduce the obviously handheld camera after the point in the story when the television starts speaking to Daniel, shooting strictly from a tripod up to that point. To get artsy-fartsy about it, the idea was that that moment “destabilizes” Daniel’s whole world, and the abruptly unanchored camera, introducing itself with a dramatic re-framing and focus pull to the television, reflects that.
I wanted the handheld feeling to continue through the rest of the film, but as I mentioned previously, footage from earlier in the original shoot was shot more conservatively than footage later. The sequence in the first cubicle (starting 4:28 in the final film) was all shot locked-off. To make it match the rest of the film, a random motion was digitally added to the frame.
We also re-set the opening scene as taking place at nighttime, since so much of the rest of the original film had wound up being shot with darkness outside of the windows. This meant the early-morning exteriors of Daniel sneaking around became day-for-night exteriors instead. (Somewhat unsuccessfully, due to the brightness of the sky.)
With the new scene cut and no ADR necessary, it was just a matter of conforming a 1080p master out via Crimson/REDcine, redoing the necessary effects at HD resolution, color correcting, and rendering the final product. Which, naturally, took me the better part of a year to get around to.
There’s no real reason. I mean there were other projects (Sandrima Rising, for one) to occupy my time, but when I really bore down on getting Spy Games done it only took me about a week. So it was basically just laziness. There was a lot of back and forth between programs that I knew I would have to do — out of Final Cut to Crimson, out of Crimson to REDcine, out of REDcine to After Effects (or Nuke), After Effects (or Nuke) to Final Cut, Final Cut to Color, Color to Final Cut, and possibly repeating a stage here or there if it didn’t turn out properly the first time for whatever reason — for each shot. Considering that the world wasn’t exactly waiting for the film, I frankly just didn’t wanna.
But the actors were waiting for the film, and considering they weren’t being paid I owed them their footage and a finished product in exchange for their time, at the very least. So finally I locked myself in my room and made myself get the thing out the door. Anthony pitched in with the sound, helping me record some new foley to beef up the audio mix, and finding some really fantastic royalty-free cues to use for the project. The original version also used royalty-free music, from the same library in fact, but they were chosen more arbitrarily and didn’t work nearly as well as the final version.
The exception is the music over the closing credits. When we originally did the film we got some music from the composition libraries of our good friends and top-notch composers Gordy Haab and Kyle Newmaster, who did the music for RvD2 among other things. Kyle sent me some tracks at the beginning of the project that we might be able to use based on our “Spy” genre choice, and I chose the Pink Panther-ish cue heard in the film. Juxtaposed with the weird ambivalence of the film’s closing, I liked the cool, smoky groove as the tone to leave the audience with. In the original version, I also used some of the other cues, but those were replaced with royalty-free stuff.
The last artistic (rather than technical) choice to be made was the title cards over the opening credits, which clearly marked this project as having been created for the 48 Hour Film. There’s an art to lying-without-lying; it’s how lawyers and marketing firms make their money. Basically what you do is you say something that is technically true, but phrased in such a way that the audience will almost certainly misinterpret it to mean something else — something preferable. You haven’t lied, but you knew they would misunderstand and make particular assumptions. Crafting those assumptions is the art.
So when the film says that it was “originally produced for the 48 Hour Film Project,” that’s absolutely true. But the mistaken assumption that I know the majority of the viewers will make, is that the version they are watching is the same version that was made in 48 hours. Because if the viewer thinks the film they’re watching was made in 48 hours they’re more likely to be impressed.
You might call that dishonest, I just call it marketing. If someone asks me directly on YouTube exactly how much of the film came out of the 48HFP, I’ll answer them directly and honestly. Likewise, I’m being totally candid here. It’s just if people want to take the film at face value without looking any further, it’s not like I can control the assumptions they make, right? Nudge nudge, wink wink.
To be fair, I did also want to flag it as a primarily 48 hour film for practical reasons previously mentioned: inconsistency in lighting, not-so-successful day-for-night, the bizarre insistence on those damn potatoes, etc. Most of the audience probably doesn’t notice that stuff, but I wanted to make sure those that did had reason to assume that I did, in fact, know the difference between my ass and a hole in the ground, and had it not been for the time constraint these things would have been rather more polished.
Still, I went back and forth on including the cards until we had found the musical cue that opens the film. The strong downbeats of the Spanish flavor basically demanded some kind of opening titles, leading to the humorous (at least that’s the intent) fade-up on possibly the least spy-like behavior imaginable. But all the important credits were already at the end of the film and I wanted to keep that cue there. So that’s what eventually helped me make the call.
I think that the introductory cards, combined with the Spanish-ish music, wound up giving the film a retro-by-way-of-Tarantino feel, which I think is just the right tone to set going in. But they do also preclude me from entering the film into other festivals, and perhaps even from including it on my reel if I want to seem professional, so I may have to create an alternate version that uses different cards to evoke the same feeling.
Well, I think that’s everything on the topic of Spy Games. I hope you guys enjoyed looking under the hood with me. It was a rough experience, but a rewarding one, and I’m very pleased with the final product. It’s definitely not the kind of film I would have made without the bizarre constraints of the competition, and I think that’s why I like it. If you’re interested in filmmaking — either getting your feet wet or keeping your chops up — I would highly recommend participating in this contest when it comes around to your area. I’ve known a lot of people who have and all of them were glad they did.
- Spy Games was early in my working relationship with Anthony, before we had decided that we were going to officially become writing partners, so it was done as-written without much of a second set of eyes and no time for me to put it in a drawer and take some time away. Now instead of having to take time to distance ourselves from drafts, we can just hand them to each other. He tells me my dialogue’s shit, I tell him his dialogue’s shit, we fix it for each other and it stops being shit, and we have a grand old time. ↩