November 21, 2018

How a Filmmaking Team Found a New 'Cinematic Language' to Make 'CAM'

The only thing scarier than being filmed is the inability to shut the camera off. 

An erotic thriller complete with a subplot of misplaced identities that would make Brian De Palma blush, Daniel Goldhaber's CAM takes the real-life implications of online sex work and douses it with genre familiarities that burn up the screen. Often identified as an independent act in which the performer engages in an e-commerce supply-and-demand between anonymous adults, the world of "cam girls" is both alluring and unsatisfying; if the naughty stakes can continually rise, what will we be able to get off on when the demands reach its peak? 

Telling the story of Alice (Madeline Brewer), a young woman who performs live, on-camera shows in which her fans decide (and pay in tokens) what she should do to herself for the pleasure of her online viewers, CAM  begins as a story on the unending competitive nature of internet success. The de rigueur of the day is extremity, and as Alice, in a scene that introduces us to her line of work, knows full well, the only way to keep fans engaged is via life-or-death situations, often resulting in the sick pleasure of viewing faux suicides performed live on-screen. 

We're all competitive to some degree, but to what end? As Alice, posing as "Lola," her stage name, gains further success, she unexpectedly gets closed out of her account. But the show, complete with a doppelgänger (also played by Brewer), must go on, and Alice is confronted with the uncanny situation of coming face-to-face with someone who looks and sounds exactly like her, someone who is claiming Lola's identity and stealing all of her fans. When Alice's personal life begins to crumble, she takes matters into her own hands, and CAM,  walking a fine line between drama and thriller, social issue study and otherworldly thriller, becomes an almost literal interpretation of the price of fame and confronting the beast within one's self.

As the film is now available to stream on Netflix, director Daniel Goldhaber and editor Daniel Garber spoke with No Film School about the origins of the project, putting together a film reliant on multiple screens, and the challenges involved in creating an astute mise-en-scène for the set of an online cam girl. 

No Film School: Daniel Goldhaber, you recently tweeted that yesterday marked the third year anniversary of yourself and Isa Mazzei agreeing to make this movie. What were the initial conversations like and what has the past three years been like trying to get this movie out into the world?

Daniel Goldhaber: That picture was taken the night Isa and I had the very first conversation about doing CAM as a genre film. A year prior to that, I had shot a bunch of porn for her to sell on her own camming show, which is [slightly referenced to] in CAM, that a lot of cam girls are also pornographers of themselves or that they do porn with other cam girls to sell on their shows. That's another way to generate income and brand.

NFS: Because you had known each other since high school, right?

Goldhaber: Yes. So I had done that with her and we had started talking about doing something in this world regarding a sex worker agency. We really wanted to bring an audience into that psychological reality and nonfiction wouldn't really allow for that. We also wanted to make a film that would be seen by a lot a people and it's even harder to make a doc that's going to be seen by a lot of people that can engage with sex work in that way.

About a year later, we had the idea of, "What if we did this as a genre film?" That unfolded during the conversation [I tweeted about yesterday] and we then spent two months brainstorming the story.

I went to Sundance in 2016 and pitched the opening scene. We knew what the story was going to be, but I didn't want to tell anybody because I knew that it could be very easily appropriated for any social media thriller. I just went and met a bunch of people and told them the opening scene of the movie, and I could see how excited people were by it (and by Isa and what she was doing with the project) and that was the moment where we were like, "Okay, there's a film here, and we're going to be able to make it because we can pitch this."

We finished the first draft in April of 2016 and we financed it by September.

"I also think that the fundamental thing that the genre offers the subject matter is familiarity. A sex worker with agency is a very unfamiliar character to see in a movie. When you're engaging with that through a genre lens, the familiarity builds an immediate bond, just by the very nature of how the rhythms of the thing work."

NFS: How did approaching the subject from the perspective of genre (thriller/horror, etc.) free you up to explore the cam-girl lifestyle more in depth?

Goldhaber: It's interesting because usually when we're asked this question, it's "Why a genre film?" and not "What did genre bring to the project?" That's very interesting. A narrative film offered the opportunity to explore a number of cinematic languages that we wouldn't have had the opportunity to experiment with in a documentary.

I also think that the fundamental thing that the genre offers the subject matter is familiarity. A sex worker with agency is a very unfamiliar character to see in a movie. When you're engaging with that through a genre lens, the familiarity builds an immediate bond, just by the very nature of how the rhythms of the thing work.

Beyond that, you're never quite sure where the stakes are coming from, and in CAM, the stakes come from Alice's loss of agency on minute-25. But the trick is that means that the audience has to, on some level (subconsciously or not), recognize that in the previous 25 minutes, she had agency. Even if that's happening in retrospect, they will be like, "Oh God, I want to go back to the good times," and that's really valuable because that's having them recognize a full mindset that I don't think most audiences have ever been confronted with.

Beyond that, genre is really fun, and I think it can express things through certain conceits that you'd have a lot of difficulties expressing otherwise. I think David Cronenberg is the master of that, finding a reality or a really difficult, complicated theme about technology and humanity and finding a way to express that through cinema, through the kind of aesthetic of genre within a narrative context. I think that it can also change the way we're talking about the relationship between the digital and the body and violence. I think genre offers really exciting opportunities.

Daniel Goldhaber's 'CAM,' courtesy of Netflix.

NFS:  Daniel Garber, how were you brought onto the project and how, given your experience editing nonfiction (and often archival) films, did CAM present a challenge for you?

Daniel Garber: Well, in terms of how I came onto the project, I go way back with Danny [Goldhaber] and Isabelle Link-Levy, our creative producer. We went to college together and I've been making films with both of them (in various ways) over the years. It was a sort of natural consequence that I ended up working on this project.

My background is really more in nonfiction. I edited The Reagan Show previously, and even though it seems like a completely different project than CAM, in many ways, I think that there are some really significant overlaps. For instance, they're both films that ultimately have to deal with performance, and one is about the performance of the present scene and the other is about performance of gender and sexuality over the internet. Even though these two performances are very different, featuring very different messages and very different ways of communicating and doing, we're still ultimately reckoning with the way that people have an offscreen persona and an onscreen persona. In the case of CAM, there are kind of three different characters that Maddy [the lead actress] is balancing.

The other thing that was so intriguing to me about CAM is that, like The Reagan Show, it's a challenge teaching the audience how to watch the film. When you're working with purely archival material, as is the case with The Reagan Show, the audience doesn't have anything to grab onto. You have no voiceover and you have no talking heads to explain the way into the film. It's just the use of montage to create meaning, to tell audiences what is happening, and inviting them to make their own conclusions based on the way that the shots are put together.

The same thing had to happen with CAM, where we have everything that's shot with the A-cam or the Alexa (and we have all of these interludes where we end up inside of the webcam or the website and you're watching messages and emojis scroll by) and had to teach audience members how to watch those things and how to pull the right meaning out of those things. It was a real challenge. We had to develop a language to make up.

Goldhaber:  In a number of ways, [Dan] Garber and I learned to make movies together. We both came to film school, to the Harvard film program, having done student work in high school, but we didn't really know what movies were. Because we had to really find a new cinematic language with this movie, I think it was actually valuable because we were really starting at Ground Zero here, but we were already really practiced with starting at Ground Zero as filmmakers together. It was really cool being able to learn something again, together, after having gone through it once before.

The coolest thing about the film is that when you watch it (or I think, when most people watch it), they don't realize that on a formal level, on a shot-by-shot level, much of the film is formally experimental, that there are cutting patterns and the types of shots that we haven't really seen in movies before—especially to represent the internet—and it is on some level basic montage.

I think the opening scene is a real testament to Dan's edit, that within five-and-a-half minutes, he has to teach you what camming is, who this character is, what this insane space is, and then subsequently turn that all on its head and have you feel threatened when nobody is in the room with her, simply because emojis are looking progressively more and more threatening. The sound is more and more threatening, and then in the last 10 seconds, you have to have all of that be convincing enough (and empathetic enough) that you can then again flip it on its head and say "Gotcha." I think that that's immensely difficult to do, and it's really cool, I think, that he manages to pull so much out of what is really "a girl screaming at a television."

"There was then a lot of narratively specific stuff that we needed, like an emoji to get an erection so hard that it explodes..."

NFS: Did you have discussions about the emojis that are used in the chat room? Some are quite graphic, some are deliberately immature, etc. 

Garber: Some of the emojis were licensed, but many of them were actually commissioned for the film and actually made by my partner.

Goldhaber: A chunk of the emojis, all the Web 1.0 ones, came from a website called Animation Factory, which was this insane entrepreneur in the early Web 1.0 days that wanted to create a GIF of everything, and so they had this huge studio of artists just making GIFs nonstop. Eventually, the bubble hit and the thing closed. Some couple in New Hampshire bought it (I think it was in New Hampshire!) and then we licensed it from them. That provided our filler stuff.

There was then a lot of narratively specific stuff that we needed, like an emoji to get an erection so hard that it explodes....That was the stuff we made with Elena Lee Gold, and that was really awesome because it was using this language of the internet that we all grew up with and that we all love but to tell a story in ways that we so rarely see the internet come to life.

NFS: The film is filled with primary, neon-like colors that fill Alice's cam environment, both within her bedroom and in the cam house she visits to record a special live show with a friend. How did you come to create such an alluring, sensual look to bathe the film in?

Goldhaber: We always wanted it to be a fantasy space, to reflect the fact that the camming world is where Alice feels like she can express herself, where she can be herself, and where she can also build her own world. That's also true of the camming clubhouse, where everybody is building their own niche, their own little character, their own little space. There were a lot of different formal influences for that.

A big influence on the pink room was a movie called Pink Narcissus that, as far as I understand, was commissioned as a gay porn film and used by the director to make a kind of aesthetic phantasmagoria. It was something that, to me when I saw it, I was like, "That's exactly what our pink room needs to look like." Isa's line in the script had been, "It's a teenage girl's fever dream," but I think we had to find a way to express that through the aesthetic of porn, but also in a way that was usable as a space.

One of the other things that Emma Rose Mead, our production designer, and I looked at was the Red Room in Twin Peaks, thinking of that as this portal into another dimension, and then being like, "Well, in some ways that's kind of what Alice's pink room is. It's this magic room in her house that is almost infinitely expansive." We looked at that and then looked at it again with Katelin Arizmendi, our DP, on just trying to bring that fantasy to life with all of the tools and tricks that we had at our disposal.

Daniel Goldhaber's 'CAM,' courtesy of Netflix.

NFS: And in Alice's "real world," the film is lit much more naturally lit and it stands out, as the day-to-day reality of Alice's life feels more drab-like and soulless.

Goldhaber: We didn't want it to feel like it was from a different film. I think you notice that there are still crossovers in design and in the overall color palette, but yeah, we still wanted her life to feel somewhat aesthetic, but we wanted it to definitely contrast. We also wanted there to be layers to it.

It was really important that Alison's mom runs a makeup salon, as that's very closely linked to performative femininity and the work that Alice is doing, and so we wanted more color in the salon and in Alison's mom's house than there necessarily is in Alice's house (which is the space that is the least curated because Alice just doesn't have the time to deal with it). We were thinking about the different levels of curation of space and trying to make sure the color and the richness of the color was reflecting that.

NFS: Many scenes are reliant on multiple-screens being fully active at the same time. With both a rapidly impatient chat room and the screens that feature the women performing for the camera, it's almost a sensory overload. Were those scenes fully realized in the edit?

Garber: A lot of that stuff was scripted out in advance and there was a lot of playing that went into production. I was actually present on set and so I saw just how much of this had already been worked out even before my work got started in earnest. Isa had scripted all of the cam-shows and there was an understanding of how each one of these shows would unfold. Maddy was always looking at the right place in the frame in part because you would actually have the site up and running on the TV screen itself, even though the screen was, of course, ultimately replaced by something else. We would either make tweaks to the show or we'd experience issues with the way it was displaying on the screen itself.

I think Dan wanted it to feel like a live interaction, and that's in large part due to Isa's work prepping it with Isabelle, our producer, who was the main wrangler of the show on set. One of the biggest challenges in editing this film was that the cam-shows were kind of infinitely variable, given how you can replace the screens (even though we have all of these close-ups of the screens) and insert different messages into the chat box. There was actually a lot of rewriting that ended up happening in post. Both the very first scene and the climax, when she gets her final face off with Lola, didn't actually work until very late in the edit. There was a lot of construction that needed to happen. Isabelle, the producer, was very involved in tweaking those scenes and getting them onto the same level as each of the others.

Goldhaber: But Garber was on set to literally edit material and turn it around for playback because, for example, we actually edited [Maddy pretending to] shoot herself, sent it to VFX, had the blood put in, and then played it back for Maddy live on set. We had to do that with the climax too, as we had to cut the movie and make performance decisions about what was gonna be up on the screen [in real time].

Additionally, when we were doing all of the pink room shooting and camming scenes, Garber needed to be on set to think through the coverage, to work on mock-ups to make sure that the coverage we were shooting would actually make some degree of sense when we took it into the edit. I think that was the hardest thing for us to wrap our heads around, and I think, looking at the movie and trying to take it back in time, there's no real precedent, on a shot-by-shot level, for how to do this. Essentially, every single screenshot in the film (and there are 700 of them) was a conversation between me, Garber, Isa, and Isabelle about what it needed to be. Each one of those is individually created, and I think it could feel like "Oh, they're just screen grabs" but if you look through how each one of those is guiding the story, you'd realize that they are individual shots. Figuring out what each one of those would be was a trial-and-error process from the beginning to the end because we didn't even know how it was going to work.

There are two examples I love to use. One is the bars on the side of the frame, a choice that was late-breaking in the edit. We had already gone through one of our "three forms of picture lock" on the movie because we needed three forms of picture lock on this movie. We had already been through one, but I had decided (for reasons that I don't understand) to make the webcam 4:3, and not exactly even 4:3, but some weird, boxy aspect ratio for the webcam site because that just felt right to me. Once we accepted that and were cutting the webcam footage, we either had to show too much of the webcam or we had to crop in so much that we're cropping out too much of the actual image. Because of this, we had to have some context of the bar (on the side) to cut into it.

This actually ended up being an extraordinary tool. We realized that we were able to create spatial relationships between the webcam and the website that we wouldn't have thought about otherwise, where if you have a little bit of the bar on the right, it almost becomes an over-the-shoulder shot and you get the sense that Alice is engaged with the guy. You don't think about that otherwise. You start going through the film, and you start realizing, "Oh my God, we have this tool in our toolbox that we didn't even know existed, and we've been editing the movie for eight months."

Because we had been using the webcam stuff as temp, we started rewriting the scenes and started showing them to people. We'd show the scene to people with bars on the right or no bars on the right, and people were going to get a read from it. We were like, "Holy crap, we can make the story better." The edit took a year, and it was the hardest part of making the movie because we had to use every tool we had, down to how much the effects were going to go into any individual scene.

The other one I really love, because there's actually a continuity error that shows how late breaking this idea was, was the voting aspect present in the climax of the film. That was not in the script. We initially had a different way of demonstrating the voting. It was like, tips of 50 versus tips of 51 and it made no sense, visually. We didn't expect that.

Garber: It was too much reading. We needed something that people could just glance at and immediately understand what was happening.

Goldhaber: We had also set up this bright blue aesthetic by that point, and so we realized, "Oh, the voting has to be [via the emojis of] hearts and teapots. Done!" The way that that works in the edit, however, is that you don't want to really sit there and be like, "Hearts and teapots." It still has to happen quickly. I forget how we reversed the order, but the original order was hearts for Lola and, of course, teapots for Teapot. We did it, cut it, but we just couldn't get the voting to make sense for people. We were like, "Why isn't this making sense?"

We then switched the order, and I forget how exactly, but we had already sent the screen to the visual effects team to comp on; they couldn't redo that. The actual cut-in is different from what's on the screen, in terms of which is first because we realized that if we switched the order of what came first, the voting would totally register for people (because it's just a dumb lizard brain thing). There were a lot of those moments in the film where suddenly it starts working and you don't know why, but you have to try it 500 times to get it there.

"When you cut to the webcam, it has to feel like a very intentional decision of saying, "This device is watching her." However, when you start cutting it in too often, it starts to feel like you're cutting to it because you don't have any coverage!"

NFS: The camera that Alice uses to record herself while performing has its own kind of presence. It's filming her, yes, but it's also surveying her choices in a way that feels voyeuristically unclean. The presence of the camera is shot and edited to give it the stature of an unapproving HAL 9000. How did you both work to enhance the "eye" that brings online visitors into Alice's world?

Goldhaber: Garber definitely kept the movie from having way too many of those shots. I kept wanting to cut to the webcam again and again, and he was like "People get it. There's a webcam in the room." [laughs]

Garber: Part of it is that this is one of those cases where, if you use it more sparingly, it will have more meaning. When you cut to the webcam, it has to feel like a very intentional decision of saying, "This device is watching her." However, when you start cutting it in too often, it starts to feel like you're cutting to it because you don't have any coverage! It starts to feel very overused.

Goldhaber: We had provided a sound effect to associate it with. We gave it the sound effect of a mini DV camera rolling, which is, of course, not what a webcam sounds like. Webcams don't make noise, and yet it's one of those sounds where it brings it to life and you wonder "Oh, what is this? I know that this thing isn't mechanical, but it's making a mechanical noise. It must be a thing. Oooh, it's spooky."

Daniel Goldhaber's 'CAM,' courtesy of Netflix.

NFS: When we see the "doppelgänger Lola," she's often pixelated, as if she's a soulless entity that's in an endless state of buffering. What considerations were made, both in the visual presentation and the edit, to make the "other" Lola, relegated to a world of computer screens, appear different than the Alice attempting to reclaim her identity back?

Goldhaber: I haven't talked about this yet. I'm so excited! Do you have the quiz show app, HQ?

NFS: I know people that play it, yeah.

Goldhaber: We were very into HQ at one point in the edit, when it first became big. The edit would literally stop so that we could play HQ every lunch.

Garber: But we never won.

Goldhaber: We never won. But early on, it was at that point where its popularity was exploding and it would buffer really badly and would datamosh. We had actually thought that datamoshing, for our film, would be too cheesy because it was unrealistic. And then we kept taking these screenshots of these amazing HQ datamoshes that were so creepy, like people's eyes ending up on their chin and stuff. We were like, "It's hard not to use this thing because it's real!"

We actually made sure that everything we did with datamoshes were produced through a real webcam lens. All of the datamoshes in the film were naturally produced through the way that real datamosh buffering happens. Our visual effects guy would do 10 of them and run them through the process, and I would pick the best one, or we would cut between a few different ones. There would then be a couple of glitches in the movie that Garber actually made as well, because we would have fun in the edit by masking and doing wacky stuff, and then some of those made it into the [finished film]. There's a few where Lola's blinking too quickly and there are fun little weird easter eggs in the movie...

Courtesy of Daniel Garber.

NFS: In the film's climax, the amount of screens present are doubled, tripled, quadrupled, and extended beyond. I imagine this would present a logistical challenge. What were the most complex aspects of putting together the big showdown between Alice and Lola?

Garber: Well, how many hours do you have? This is a big question! That was a huge, huge scene and a huge undertaking for everyone. I should send you the diagram [displayed above] because I had to construct this complex diagram [for the climax]. When we were shooting it on set, we realized that the webcam would have a delay of a few frames. We reasoned that if you have all of these screens within screens, then you would have to build in both this natural webcam delay where they're losing a few frames every time you go farther back, and on top of that, there's this weird mirroring effect that's happening because of the physical mirror and because of the electronic mirroring that's happening in the webcam.

I had to produce this diagram, this guide for our VFX guys, showing exactly how each layer needed to be flipped inside of the screen, and how many frames it needed to be lagged behind the previous layer.

It was incredibly complicated. We had to send one of the shots to get rotoscoped so that Alice's hair was just in front of the screen, still in place on the screen, and they had to basically take the webcam, put it inside of the website, put the website inside of another screen, and then have it placed inside of another screen! It was this huge recursive thing.

Goldhaber: Another thing that was a huge challenge for us was animating natural website interactions. It usually actually looks pretty hammy, and so we had to do all of that stuff because none of that actual functionality worked in the fake site we built, so a lot of the film was actual screen-capture and other interactions were animated. You then get to this thing where you're all through that and you're like, "This still isn't quite hitting." You're like, "What if we do out of focus?" Then you're like, "Well, how do you do that?"

There are so many moving pieces and the film was a teeny indie. Dan was coordinating all of this and editing the movie and it was a very small effort, but I think it's really impressive. The movie has 1,000 visual effects shots in it, and I think Garber had to deliver all of them by hand.

Garber: I had great help too. I had an assistant editor, Jon Branden, who was a really valuable component. He was also largely responsible for managing all of those VFX shots, making sure that those shots got delivered on time because we had to give the VFX team the Alexa Plate, the screen capture of the website, the Lola webcam, and the Alice webcam. All of those had to be cut to exactly the right length, with the right number of frames on the heads, so you could build the lag. It required this incredible amount of organization and meticulousness.

Goldhaber: Because the movie was so low budget, we literally didn't have the money to pay for an extra frame of rotoscoping. They literally were having to count the frames and make sure that, yeah....One of the issues we had was a few lighting gags for the movie, and we hadn't fully corrected them on set for Alice's personal webcam. It was just a goof. There's a slight lighting discontinuity in the webcam that we had to fix because the whole scene is meant to play continuously. We had to send all of the webcam footage out to be pre-graded before it was color corrected due to all the effects going on.

It was a nightmare and it took three months, three months that none of us will ever get back. Again, it was on this limited budget. Usually, you're all working in one house together and these are natural conversations that are unfolding, but on an indie scale, it's massively complicated on just a communication/logistical level.

One of the best moments of making this film for me was when we'd been editing the movie for at least nine months, and I still had not seen the climax. We went into the sound mix and we were mixing the film to finish it, and we had a blind faith that the climax was going to work, and I got to watch it for the first time [with VFX}, and it was like, "Oh my God. It makes sense. We did it."

NFS: That's a great response to have. Because it if it doesn't work, you go, "Well, okay....."

Goldhaber: If it doesn't, we're fucked. We were out of money and out of time, and that's it.

Garber: We're never allowed to make a movie again. [laughs]     

For more information on 'CAM,' click here