What’s Missing is a brand new 10-part series that explores the wide-reaching, devastating impacts that occur when someone disappears without a trace. Hosted by Loren O’Keeffe, the founder and CEO of the Missing Persons Advocacy Network (MPAN), the series features candid interviews with family members and loved ones of missing Australians, going beyond the headlines to provide a deep, unflinching insight into their searches and struggles.
When we first talked about Casefile Presents, our Casefile‘s anonymous host knew that he wanted to produce a podcast that focuses on missing persons.
He was introduced to Loren O’Keeffe – the founder of Missing Persons Advocacy Network (MPAN) and knew that she was the person for the job.
We’ve decided on eight episodes that would feature stories of missing persons and chats with family members and two episodes where Loren would interview experts in the space.
With the help of Loren, MPAN’s book Too Short Stories and Casefile writer – Erin Munro and after many iterations and re-writes, eight full episodes were ready to record.
The structure would be as follows:
Each episode would start with a scripted story read by Loren and family members (in most episodes). This segment would be produced similarly to Casefile, with background music, nice reverb and polished mix, both Loren and the guest would read from the script where possible.
After the story, Loren would invite the guest for an unscripted, intimate chat and talk about topics such as missingness, grief, loss and what happens to families of long term missing persons.
After the chat, we go back to the well-produced story/outro, then roll the credits.
This structure would be the same for all eight episodes, the other two would be interviews with experts.
After the writing process was finished, we had to book the studio. Me being in the UK and the rest of the team in Australia, it was challenging to find a suitable place as I wasn’t there to physically check it.
We only had a week to get everything done – so 10 interviews in total plus story/narration parts.
On some days we’ve scheduled just one session, but on several days we had two sessions in a day – morning and afternoon.
The studio was located in the heart of Sydney. A recording room was set up with six microphones (Rode Procasters) around the table, as well as lighting and cameras.
The tracking room was behind the big glass door, and the tech staff set us up with monitoring headphones.
As I stepped into the studio, I did notice a few things. The studio was in the office building, with other offices next to it, separated by a thin wall. There wasn’t any acoustic treatment on the walls, and there wasn’t much isolation at all.
I sensed straight away that this could be a problem, but the first two sessions were on the weekend, so at least the start was good.
The studio also appeared quite cold with AC running constantly. This is ok for short sessions, but during longer days it can be quite problematic.
The tech staff was accommodating and professional, made us and the families feel comfortable and looked after. We also had a green room upstairs where we could greet the guests and debrief after the sessions.
The sessions would look as follows – we would greet the families, explain the process, Loren would also explain how she would run the session.
We’d start with reading from the script for the story, and during that, myself and the anonymous host would take notes for the re-does and pickups.
When this was all done, Loren and the guest could relax and had an informal chat that we recorded for the episodes.
The studio was a podcast in the box kind of a place. However, I wanted raw, uncompressed and unfiltered files. All the sessions were nicely labelled with the day and part of the day, so I could easily find the audio.
I bought a hard drive for the occasion, we made several copies to be sure, said goodbyes to Loren, and that was that!
Some people will already know that I work on templates as this is what I learned at my previous job as a dialogue editor at a movie studio. I have one template that I use for my projects and adjust them accordingly, so I did the same for What’s Missing.
I knew that Loren would feature in every episode so her track, or preferably two tracks – story parts and chat which needed to be in What’s Missing template. The rest of the tracks would be somewhat similar with some adjustments on EQs and compressors depending on a person featured.
Before the editing, I decided that the story parts will have reverb on the dialogues to make them pop a little bit. The reverb would then be off during the intimate chat making the sound slightly different when the episode moves into that segment. I dropped the original audio into the folder. I created folders for all 10 episodes as well as sessions for pickups which we recorded at the end of the week.
Draft 1 means a structural and flow edit, cutting out the mistakes, creating the story, getting the pickups and editing the flow of the chat plus the outro. So you’d get the whole picture of the episode however with no clean ups, no mixing, and very basic levelling.
Draft 2 would be more complicated. Draft 2 meant that the feedback notes were incorporated, the dialogues were cleaned up with iZotope Rx, and also the story featured the score. However, the episodes would still lack the proper mix, but it would sound close to the master.
After that, I would incorporate the final notes and do the mixes and final masters. At this stage, there should not be any further changes – however, we could still do tweaks and adjustments if needed. The team would be notified at every step of my production, so the process was transparent and organised.
The first structural edits were a time-consuming process. First of all, a lot of pickups, re-records and additional parts were all over the place so getting them in place took time. As I said before, I took notes but still, I had to scrub through the pickups session to find the relevant audio.
The other process was editing the chat, this took time because, during structural edits, I was not only editing out the uhms and aahs but also cutting down parts that were off-topic or repeated. I was working on my monitors ADAM A5X in Pro Tools, in my templates.
When the team was reviewing the first cuts, I started working on in-depth cleanups of the dialogues. That meant applying processes within iZotope RX – the audio restoration software. I use RX every day, both for manual cleanups as well as with module chains and global processes.
Because the dialogues were recorded with dynamic microphones, the gain wasn’t there, and I had to pull up the volume quite a bit. Once I did that, all the issues of space not being isolated were audible – recordings were noisy, background noises were there, and the noise floor wasn’t great. I’ve done my best to tackle that, but if recordings were better, it would be much easier to do the job.
Next was the bleed and outside noise. As I mentioned before, space wasn’t isolated. Because of the nearby offices, there was a spill and in places quite audible – laughter, chatter, the phone ringing. Dynamic range and general recording issues were also tricky to fix. Coughing, sniffling, mistakes, clothes moving, and of course moving off-axis were at times very problematic. Again, in that instance, the intelligibility comes first and quality second, so as long as the dialogue was audible enough and understandable, it was a good enough job for me.
The first thing that comes to mind when scoring a podcast is the main theme. Loren knew an artist called Jess Ribeiro and wanted to use her song Wildflowers as the theme to What’s Missing. She thought it fitted perfectly – the lyric and the music. We’ve contacted Jess, and she agreed to license her song for the podcast. Besides the final master I’ve also requested the instrumental version which we used for various trailers, promos and credits.
As mentioned in previous episodes, the structure of the 8 main episodes would follow the same flow: the story – chat – outro and credits. My idea was to make the story and the outro a polished mix with background music, in a similar way, I do it for Casefile. The chat with the guest wouldn’t have any music in it.
My initial plan was to write eight musical cues, one for each episode and then as I do for Casefile – play and mix with the textures of each cue during the scoring. For writing the cues I used my M Audio Hammer keyboard with my virtual instruments software from Spectrasonics and Spitfire.
By this point, I knew the episodes well and the feeling I wanted to convey with it. I opted for piano, some strings but also rhythmic elements and guitar in places. The music was softer and more melodic than what I usually write for Casefile.
After that initial scoring, I committed each MIDI track to audio with respective instruments, placed it on instrument audio tracks and ran through iZotope Neutron Mix assistant to get a quick premix of the instrument STEMS.
Once I had all the instruments on audio tracks, I mixed them to the narration. I use Icon platform M+ for my mixes. Each track corresponds to a single fader, and I simply ride the volume automation to mix the instruments together.
After the first mix of the score I’ve exported the episode and sent it to Casefile host and Loren to review and make further notes.
At this stage, I would say that the episode is 90% completed – the editing, scoring and first premix are done the rest will be just making it all sound great and final tweaks.
The way that my templates are set up is that I have mixing plugins on each channel as well as mastering suites running at all times. This allows me for efficiency when working on podcasts. It means that I can hear somewhat mixed and mastered so when I get to the actual mixing stage, most of the work is already done.
For What’s Missing I had a simple set up with four mono narration tracks, score track, a stereo bus with reverb on it, stereo mixing bus and master track. All of the dialogue tracks were sent to Stereo Mix bus where a bit more EQ, compression and saturation was applied to glue them all together. The stereo audio track with the score was also sent through the mix bus.
And then on the master track I have Ozone 9 with EQ, compression and limiting set up to -16dB LUFS as well as a few other modules such as Imager and Exciter.
Because the dialogue was evened out with Leveler during editing and limiting is on as well I don’t really need to run automation on it. If there is a clip that needs fixing, I do it with a clip gain. I tweak music automation a little bit, but really I focus on making the dialogues and overall mix work as good as they can. I run the first pass on my mixing AKG K712 Pro headphones – that’s the first mix.
Then for the second time, I have several earbud headphones, and I listen to the whole episode once again switching the earbuds throughout to hear how it sounds on cheap consumer headphones. That’s two mixing stages that I always do.
Mastering is more of a QC process, but I do make final tweaks here and there. I listen to the episode again on AKG K712 Pro headphones, and I also have a codec preview set to MP3 192 or 128 kbits. This allows me to hear how the exported audio will sound with the lossy compression.
Once I listen to the episode for the last time, I select all tracks and bounce it down. For What’s Missing, I’ve also exported the audio as MP3 with the bitrate 128kbits. I reckon it was good enough for the quality and the size of this podcast.
We worked really hard to get it all down for the release, and we wanted the podcast out during the Missing Person week, which happens in Australia every year the first week of August. That meant getting the designs, website, marketing, RSS feed and everything else in order. The team worked together to make the process as smooth as possible.
While they were busy with sorting that out, I was making short 1-minute trailers for each episode as well as the extended 6-minute trailer to be featured on Casefile feed, our YouTube channel and other places as well.
Everything was ready to go, and we’ve released the first episode on Friday 30th of July with other episodes coming out every Tuesday after.
Loren’s wife MK had done a fantastic job on the look of the podcast as she designed the logo and artwork. Paulina took care of the website, podcast uploads and release as well as other graphics. The Host and Loren had their hand on marketing and driving the awareness to the show.
All went great and without a hitch and I’d like to thank everyone who took part in this fantastic project.
But going back to the production process. What did I learn? What would I do differently?
In hindsight, I would take more time selecting the right studio and make sure they are what I’m looking for. Recording wise, I would probably have the signal a bit louder at the source and of course, get the electronics out of the room. Also making sure the temperature of the room is comfortable for everyone, especially during the long days.
Editing and production?
There are lessons there for sure. It was the first big project outside of Casefile, and I think that initially, I thought I will apply the same workflow, same processes and it will work fine. So I overplayed my hand here a little bit. Doing it again, I would spend more time adjusting the mix, working on getting the ambient tone right and deciding on the noise floor for the background fill. Also checking the noise on earbuds a bit more.
I was happy with the music and story parts, I would spend a bit more time on the chat and sorting some issues with that.
Overall, it was a challenging project, both in a subject matter and production. However, everyone was pleased with the outcome, and it was an amazing experience for everyone involved.