In fiction writing programs around the world, one of the most common credos that teachers impress upon students is to not rely on dialogue to advance the story. That’s narration’s job. Dialogue defines and develops characters, adds perspective—even creates tension between people—but it certainly shouldn’t do the heavy lifting of pulling a story from beginning to end.
2017 Peabody-nominated fictional podcast Homecoming ignores that directive. Driven entirely by overheard conversations and their surrounding sounds, Season One of Homecoming drops listeners into a story without a narrator, leaving them to find their footing between the interactions of the characters: A soldier has returned from war in Iraq (Oscar Issac); his caseworker (Catherine Keener), interviews him at an experimental soldier re-acclimation facility; her boss (David Schwimmer) calls her from an airport with distracted, hard-to-hear demands. Through a non-linear plot that’s peppered with raw conversations, on-scene background noise, and an ominous soundtrack, listeners try to make sense of this mystery.
Instead of awkward, it’s alluring.
Mark Henry Phillips is the producer, recordist, editor, sound designer, and composer of Homecoming. In this interview, he talks about how he tied everything together.
“RX is such a big part of my sound designing and mixing process that I almost don’t even think about it when I use it.” —Mark Henry Phillips
You were the producer, recordist, editor, sound designer, and composer of Homecoming. What was your process like? What was it like to be all of those people?
It was a lot of work! But the cool part was that it was a more integrated, intentional process than if all those jobs had been split up between five or six people. I really could help bring the story to life, having been involved from before the scripts were fully written all the way to the final mix. The first part of the process was to work with Eli Horowitz (the writer and creator) on the scripts to make sure they would work in the audio-only format. A lot of time that meant coming up with creative solutions, because things in the script just wouldn’t be possible to convey in sound-only. Then during the recording process I was able to get extra things that I knew we’d need to make the scene work. If it was a scene of a character walking, I tried to get some sort of breathing or muttering or whatever so it didn’t just sound like fake footsteps. During the recording process I felt like a cinematographer—Eli handled the direction of performances and I handled the sound elements of the performance.
One of the biggest roles I had was as the editor—in the film editor sense. Going through all the takes and combining them to make one believable performance was particularly tough with only audio—there was no great visual performance to fall back on. But since it was audio-only, it meant I could get super fine-grained in the edits—taking one word from this take and one word from that take. You can’t do that with video editing because that would look crazy. Because I was also the sound designer and composer and mixer, I could keep in mind all the other tools I had to make the edit work, and I think it led to a more natural and effective final product.
Homecoming creator Eli Horowitz said in a 2016 New York Times article, “I wanted a whole plot where the actual meat of the story was the conversation.” In your words, how does everything else—the background sounds, the music—keep the listener engaged throughout the conversations? What role do they each play?
My major goal with the project was to make something that sounded new and didn’t feel stagey or cheesy. I didn’t want it to feel dramatic, like a recording of a play. Something I realized early on is if the characters were just sitting in a silent environment talking, the listener's brain would stop picturing them as real people in a real place doing real things. They’d become these disembodied heads just delivering lines. It would feel artificial. So I always wanted some sound to subconsciously make the listener picture the world of the characters. This way, hopefully, they wouldn’t picture the production process with actors.
For the interview scenes between the therapist and the patient (Catherine Keener and Oscar Isaac), that meant adding background noise (like a fish tank) and subtle foley (movements on the sofa if I wanted to imply one character was uncomfortable, scribbling notes if the therapist was interested, or putting the pen down if she wanted to really connect with him). For the phone call scenes, I worked to make the background noises almost a part of the scene—an angry cab driver in Dubai, people at an airport in Detroit, etc. Those scenes felt so much more real with background noise. And then in the scenes that were “out in the world,” I tried to work with Eli and the actors so that there was always some action happening that had an audio cue that the listener would be able to picture. It’s subconscious, but all of it adds up to making the conversations feel real.
From The New York Times article: “After an early pilot of the podcast seemed full of dead air, Mr. Horowitz worked in new soundscapes throughout the series. For one scene, Mr. Phillips created a rolling, creaking Ferris wheel that sounds as if it were carrying Ms. Keener and Mr. Schwimmer in and out of the bustle of a street fair.” What kind of conversations did you have with Eli Horowitz about creating new soundscapes? What was his direction like?
I felt like every scene needed periodic sound cues to keep the scene visually alive in the audience's mind. But I also thought that having a static soundscape for a long scene would start to make that ambience feel canned. So I worked with Eli on the scripts to add in some movement and variation. The ferris wheel was actually my idea. Originally that scene was just them walking along the boardwalk. I thought that having them transition into the ferris wheel and then using different parts of the ride in the script would really paint a picture in the listeners heads. I tried to add background transitions in other scenes too: having the waitress move from one section of the diner to another, or walking into a retirement home from the parking lot. I thought having these PT Andersen-style tracking one-shots would really help create the environment.
The other conversation I tried to have was about perspective. As someone who does a lot of post work in film, I’ve learned just how important perspective can be in making a scene work. Especially since the listener didn’t have any visual reference, I thought it was important to have a very clear perspective. Are we walking with Heidi? Are we in the phone? So I was always asking, “Where are we listening to this from?” And then I’d record and design the scene accordingly.
The audio for most podcasts is recorded in a studio. For Homecoming, you recorded some of the audio on location, outside. What challenges did that present?
Noise! I spent a couple days trying to find quiet locations in Brooklyn to record our outdoor scenes. It was essentially a sound-location scout. It was remarkable how hard it was to find a quiet spot! We ended up in Greenwood cemetery, which was pretty quiet but it was right in the flight path to JFK/Laguardia, and ultimately I felt like recording in a cemetery would be a little weird. We ended up recording some scenes in Prospect Park, Gimlet Media co-founder’s back patio, and a park in Red Hook.
There was so much noise and extra sounds in all of those locations, but by using RX and field recordings to mask certain sounds, I was able to make them work.
The other real fun location stuff was getting some original ambiences. For the diner scenes I went to a couple diners in Westchester and did a lot of the walkthroughs that the lead character would do. There’s no way to navigate through a pre-recorded ambience, so if the listener was going to be moving through the space with the main character, I had to do those walkthroughs too. That was super fun. So was walking through a retirement home in Yonkers. Again, there’s no way to fake that with canned FX!
This Vanity Fair article describes the scene recorded in a field in Red Hook, New York that was disrupted by an adjacent electronic music festival. Were there any sounds or scenes in particular from Season One where RX was particularly useful?
Ha! The most helpful tool on that scene was Catherine Keener—she actually went over to the sound guys (who were doing a soundcheck even though it was 8:00 a.m. on a Sunday) and begged them to give us twenty minutes to do a little more recording. But there were plenty of beeps from trucks backing up that I had to remove using RX Spectral Repair.
How did you use RX on Homecoming? Which modules did you use, and how did you use them?
There were tons of air conditioners, noises, etc that I had to denoise. And then the lavs that needed to be cleaned up. Sometimes I had to De-clip stuff, or if an actor was off-mic I would make a copy of that line, De-noise the crap out of it, and double it with the original (that’s a trick of mine to give some more volume to an off-mic line). RX is such a big part of my sound designing and mixing process that I almost don’t even think about it when I use it.
Mark Henry Phillips | Photo by Alex John Beck
Also from the Vanity Fair article: “Horowitz was so committed to avoiding any ‘inevitable staginess’ that he handed his actors a plate, utensils, and food for scenes in which they were supposed to be eating. How do you strike a balance between maintaining an “unfiltered” sound and creating a listenable, enjoyable experience?
That’s an interesting question. I felt like having imperfect sound was the key to having it feel real. Even if that meant it got too imperfect and I had to De-noise or clean up the audio with RX. If it was just two actors standing in front of large diaphragm mics, I thought it would sound like a cartoon. The other thing was having the actors perform real actions—you sound different if you’re walking or bending over. There’s no way to fake that. So having the actors do that stuff—even if that meant they were getting off mic—was super important to me.
A lot of the scenes were supposed to sound like a phone call and a personal recording of a therapist session. For those scenes I tried to find the right balance and make it sound like a phone call or a crappy recorder but also not too unintelligible or too harsh. That was a hard balance to strike, and sometimes it probably veered too far in one direction and too far in the other. But overall I think it was probably right in the middle.
In that Vanity Fair article, Horowitz called Gimlet Media’s plans for fictional podcasts, “the tippiest tip of the iceberg.” Where do you see the future of audio in story-driven podcasts going? What else is possible that’s below the tip of the iceberg?
I totally agree. I’m super excited about what could come next and trying to think about new ways to push the medium forward. When I worked in public radio, a co-worker of mine—Tony Field—used to say that radio is the most visual medium. His point was that listeners had to create the visuals in their heads and by having them create the visuals (instead of the filmmakers, for example), it could be even more visual than TV or film. Keeping that in mind, I think there’s just tons of new ways to think about storytelling and how to create a visual world inside the listeners’ heads. Podcasts like S-Town, to name one of the most recent, are showing us how you can tell amazing stories with sound that really wouldn’t be as effective or engrossing if they were told in print or film, and I think that’s the case for narrative fiction too. Audio storytelling could combine elements of novels, films, TV, and radio in a totally new way to make an amazing new genre. I don’t think it’s fully happened, but I’m excited to keep working on it and see what happens over the next couple years!
Anything else you’d like to mention about working on Homecoming?
For the theme song, I used another favorite iZotope product: Vinyl! I used an old out-of-tune upright piano to try and evoke an old memory (did everyone play one of those as a young kid, or was that just me?). But then I wanted the quality of it to sound like it was degrading and falling apart to mirror the plot in the story about memories disappearing. So I used some tape plugins to add a warble, crosstalk, etc., and then iZotope Vinyl to add even more warble and degradation. Also the noise from Vinyl was really important—at the end of the opening theme I did a huge gain boost in the final half second or so to give it a whooshing effect. That really worked, so then I replicated that using the noise from every scene and just cranking it as a way to transition between scenes.