For this month’s theme, I interviewed Kate Bilinski who was a editor and mixer on the second season of the Serial podcast. Kate is a sound designer and re-recording mixer. Outside of the Serial podcast, she makes noises for film and television, including Todd Solondz’s ‘Wiener-Dog’ and the Emmy-winning documentary, ‘We Could Be King.’
Serial is a serialized award-winning podcast, hosted by Sarah Koenig that reveals “one story – a true story – over the course of a whole season.” It’s (rightly) hugely popular and is one of the few podcasts I know of that most people (even those who don’t listen to podcasts) have heard of. Saturday Night Live even did a parody of the podcast, and one of my other favorite podcasts The Sporkful did an episode in the style of Serial.
If you’re not familiar with the podcast or its second season, I suggest you check out the “About” page on the Serial website before reading this interview. Also, the term “tape” is used in this interview and it refers to anything that’s collected from reporting whether that’s an interview or archival footage.
Designing Sound: How did you become involved with the Serial team?
Kate Bilinski: It was a bit of a roundabout connection. My background is mostly post sound for film. Ira Glass produced Mike Birbiglia’s films, which were mixed at a studio where I frequently freelance. My understanding is that they reached out to the staff there to see if they had any recommendations for mix engineers. My name must have been thrown into the pot and I was contacted to see if I had any interest in applying. I went through a pretty standard cover letter, audio test, interview process, and after one nervous Friday afternoon that felt very similar to waiting for college acceptance letters, here we are.
DS: What did the team consist of and what was your role on season 2?
KB: It’s a pretty small core group. Sarah, Julie, and Dana are the original producers, well-known names from Season One. Then there were four new hires, myself included, filling the roles of researcher, digital editor, community editor, and music editor/mix engineer. That was the main team, but Ira Glass was our main advisor and there was constant support from the This American Life crew.
My role as music editor/mix engineer really consisted of handling all the audio components of the show. When a script was ready I would record with Sarah and then cut all of her voice over, as well as assemble the full episode with selected tape. Then I would score each episode with a library of cues created for the season, editing as needed. When that was approved, I would finally do a full mix of the show before it was uploaded for distribution.
Prior to the season drop, I acted as more of a music producer, interfacing with our two composers, Nick Thorburn and Fritz Myers, to come up with an aesthetic for the season. Fritz remained active through the season to write additional cues whenever I was in a bind.
DS: What are the stages involved in creating an episode of Serial?
KB: Prior to my involvement, there were months and months of reporting and interviews. With that material the team developed with a general outline for the season and evaluated what future reporting would need to happen. Sarah would then write a script for each episode. Each episode had at least three “Edits”, which were basically live performances of the script. Notes would be given and revisions would be made. This was an important time for me to assess tape selections and how much clean up work would be involved.
When the script was approved, Sarah would record her voice over, directed by either Julie or Dana and myself. At the same time, the script was also handed over to Fact Checkers who would comb through it for every detail. Pickups would be recorded if they had additional notes. I would set to work getting VO, tape, and music in place. Julie and I developed a two-review schedule where we would first approve pacing and music cues and then I would do a mix pass and we would do a long final review before it was uploaded.
DS: Presumably you had a lot of material in the way of interviews and news footage from multiple sources. How do you manage and organize that material?
KB: Serial had one server where everything lived. Any audio collected was organized by subject and date, but it was also completely transcribed, often with minute:second annotations. These logs allowed us to search for a topic or keywords in text and quickly find it in interviews that were sometimes several hours long.
DS: What was the schedule like, were you completing episodes well in advance of them going up online?
KB: Well, no. Haha. We were as active as active productions are, in part because the case was current and we had to assimilate and respond to new reporting. When we launched the season, there was a possibility that Bowe’s case would be thrown out. Then the Monday before the second episode was set to air, the Army announced that they would take Bowe’s case to Court Marshall and so new writing was required. Given this dynamic, writing was often being revised right up until the Tuesday or even Wednesday before the Thursday release. Reporting through the entire season made for great journalism, but there was no wiggle room in the production schedule, ultimately leading to the biweekly release.
DS: What role did Sound Design play on Serial season 2?
KB: Serial holds itself to strict journalistic integrity so I was not allowed to add any sound that didn’t actually happen. It would be inaccurate and misleading to recreate scenes from library sounds. My sound design was limited to the archival material and music choices. For example, the opening scene for Season Two is Bowe’s rescue. All of the sound behind Sarah’s VO is from the YouTube video of that day, cut to time against her narration. This technique was used again during congressional hearings, when the hush and murmur of a courtroom was woven around around Sarah’s narration, like an ambience, timed to hear specific pieces of tape at certain moments.
DS: What role did music play in Serial season 2?
KB: Music was really used as a tool to move the story along. It could point out moments that required the listeners’ full attention or transition thoughts, but I had to be very careful to not evoke inappropriate emotion or sympathy from Sarah or anyone else. Similarly, it was not always necessary to score action-filled moments, the way you would in a film, since they were gripping on their own. Learning to score Serial was definitely a learning curve for me.
DS: You mainly worked in Pro Tools, what did your session layout consist of and did you use any other tools?
KB: My layout was directly pulled from my film organizational scheme – maybe force of habit but it seemed to work well. My top VCA consisted of dialogue tracks – Sarah’s VO, then any of Bowe’s tape, followed by interview tape, then any archival footage from news broadcasts, etc. Underneath that was a “SFX” VCA, which was really more of a production effects group. Finally, there was a Music VCA. Since I was constantly repurposing cues, I almost always worked with full sets of stems, spit out per instrument, which allowed for maximum flexibility. All of these groups had their own bus but processing was very minimal.
iZotope RX was obviously frequently used, mostly as an audiosuite plugin. Denoise, Decrackle, Spectral Repair are staples of my arsenal. I’ve also become a huge fan of the Fabfilter Q2 when you are in a bind for matching dialogue – also just as an EQ in general!
DS: Were you working to any loudness standard, how did you evaluate the mix and what were the specifications you were working to?
KB: Since the world of podcasts is still fairly new, the rules are not nearly as rigid as they would be if you were mixing for TV or Radio. In some ways, it was left up to my discretion. iTunes streams at -16 LUFS and that seems referenced fairly often for Internet/Mobile masters. However, I always kept my delivery around -18 LUFS, leaving headroom since there were various streaming platforms. Serial also uploads 128 kbps stereo .mp3, which I know varies by podcast.
DS: Did you have any surprises?
KB: Mixing for a podcast provided a different set of challenges for me because your audience has a very different listening environment than say seeing a movie in a 5.1 theater. This meant I was constantly monitoring on four different platforms. I would work on my studio monitors and quality headphones, but also earbuds (I would often take mix notes on my subway commute) and laptop speakers. You have to assume the majority of people are listening on headphones of some sort so I found it really important to have a good grasp of that vantage point. The smaller speakers in everyday earbuds and laptops actually required I mix the music hotter than I normally would. If you take a compromised frequency spectrum and throw it in a noisy environment like a subway car, music was the first thing to go.
A big thank you to Kate for answering our questions. You can learn more about Kate and her work at katebilinskisound.com