Designing the Frightening Sound of ‘The Dark Pictures Anthology’ – with Audio Director Barney Pratt

Supermassive Games has released four games in their narrative-drive horror/survival series The Dark Pictures Anthology, one game per year since the first installment Man of Medan in 2019.

Each game is a different sub-genre of horror, has a unique score, and hosts a new cast of characters. And since the creative cycles for two games overlapped every year, Audio Director Barney Pratt had to do a lot of project juggling to keep them on schedule.

Here, he compares their approach to sound on each release, talks about how advancements in technology (like spatial audio) influenced their approach, what went into creating each game's unique score, the character sound design, and several creatures featured in the games. He also shares details on the sound of the latest installment, The Devil in Me , such as recording custom sound, mixing in real-time, and so much more!

---

Interview by Jennifer Walden, photos courtesy of Bandai Namco; Supermassive Games; Barney Pratt

---

Please share:

• More

In this massive interview, we take a look at Supermassive Games interactive anthology series of survival horror games: The Dark Pictures Anthology (published by Bandai Namco Entertainment).

Barney Pratt – who was nominated for a 2021 NAVGTR Award for ‘Sound Editing in a Game Cinema’ for The Dark Pictures: House of Ashes – was the Audio Director on all four games in The Dark Pictures Anthology, which includes Man of Medan (2019), Little Hope (2020), the aforementioned House of Ashes (2021), and the recently released The Devil in Me (2022).

Here, Pratt compares their creative and technical approach to the games’ sound over the years and across the anthology, how they handle overlapping creative cycles (since they released a title every year), how the emergence of spatial audio improved immersion and player experience, why they opt for in-game processing and mixing in real-time, how they approached the music for each release in the anthology, how they tackle character design for these narrative-driven horror games, and what went into the sound design for some of the creatures. Plus, Pratt dives into the details of creating The Devil in Me , from capturing custom recordings, keeping the sounds specific to a specific sub-genre of horror, some technical and creative challenges they faced on The Devil in Me, and much more!

The Dark Pictures Anthology: The Devil In Me – Official Launch Trailer | Supermassive Games

The Dark Pictures Anthology has had a release every year since 2019. What’s changed in your approach (creative and technical) from that first release to this latest one?

Barney Pratt: Great question. Our approach is ever-evolving. We haven’t changed our core approach to uncompromising cinematic experiences, but ‘under the hood’ we are always looking for incremental gains, from the latest sound design techniques or implementing the latest tech and processes. In fact, some of our systems have had complete overhauls.

Every project is such a different audio experience that we barely reuse any assets.

Creatively, we have so much freedom on The Dark Pictures and a lot of trust from the game directors. Every project is such a different audio experience that we barely reuse any assets. The different locations, time periods, horror subgenres, characters, and narratives all dictate this change. We still need to fulfill all of the core horror emotions for the player: fear, dread, and the odd jump scare, but we look for a fresh palette every time, something relating more closely to the narrative and this helps keep it fresh and exciting for us.

Technically, far from having a linear post-production approach, our procedural systems are always changing and developing. Our character foley systems had a complete rework for The Devil in Me (aka, TDIM) due to the expanded gameplay and exploration features, which also included a breath system reacting to a character’s exhaustion and stress levels. We have run a full 3D spatial audio setup for the last two games, which just sounds amazing, and these systems, of course, always need to produce the subtly nuanced and detailed cinematic results for the Anthology.

We look for incremental gains on the Supermassive cinematic audio DNA, advancing systems as much as possible while reserving the ability to go fully bespoke and create standalone cinematic moments.

Because you’re releasing a new game every year, there’s an overlap of creative cycles as you’re finishing/releasing one game and starting work on the next. Can you talk about that process? What are the challenges of working on two games at the same time, and how do you make that process less stressful/more successful?

BP: Every overlap has been quite different. You need to get a look-ahead into the next project, which needs to happen no later than your longest lead time before you and the team are meant to start on that project.

With the majority of our pipelines in place, our longest lead-time is music composition, so I need to have the music direction clarified, the first set of compositions documented, and a composer window all aligned to hit the start date of the next project.

With the majority of our pipelines in place, our longest lead-time is music composition…

Slightly shorter lead times exist for foley assets, design of new game features, and engine upgrades – all of which need to happen for your team to be the most productive.

We have had both extremes. First, the situation where the next project is very developed and the whole audio team can be tasked up immediately. Similarly, we have had situations where the project might not be ready for all of us straight away, in which case we are able to dig into the backlog, develop features, and experiment with new tech such as spatial audio.

Spatial Audio was introduced in the third release, and it’s also used in this fourth one, The Devil in Me. On the creative side, how are you able to take advantage of spatial audio to improve player experience? Any specific examples you can share in this latest game? On the technical side, can talk about using Wwise for spatial audio?

BP: In 2015, on a PSVR1 launch project we envisaged a suite of audio features that we didn’t even have names for, things like ‘directional reverb tails,’ ‘environmentally responsive early reflections,’ etc, and although too far ahead of the curve, we found that these features were being developed. For a long time after they became available, I have wanted to get the spatial audio tech into our pipelines, but we were pragmatic and waited for it to settle down in terms of CPU usage, especially on old / 8th gen consoles.

I have wanted to get the spatial audio tech into our pipelines, but we were pragmatic…

House of Ashes (aka, HOA), was the perfect project to bring in those features. It had long underground tunnels and huge man-made, reflective, cathedral-like spaces, so we had the perfect environments to roll out spatial audio. Adding portaling, diffraction, reflections, and volumetric emitters was a paradigm shift; the results were a stunning and seamless reality of sound as you moved around the caves.

A showcase use of this tech in HOA was the sound of a dying NPC, luring the player away from a large room and along tunnels, and yes, we played some tricks to optimize the experience but the sound of that dialogue and all the diegetic sound chaining through all the spatial elements was just stunning.

…it’s so ‘real’ that the third-person spatial experience translates very closely into VR.

The player audio responds to the geometry of the environment in subtle and seamless ways. The localization of sound, distance, and directionality are unparalleled; it really is the glue in the audio experience. Of course, you need to decide how you want to use it, but it’s so ‘real’ that the third-person spatial experience translates very closely into VR.

Interestingly, as I write this, we are adding the final polish to Switchback, a VR game based in the Dark Pictures universe, which allows us to extend the spatial featureset into greater use of 3D audio and ambisonics, which, of course, we will fold back into the cinematic tech and DNA to further enhance the Anthology.
 

What about in-game processing? Were you using any for this latest release? If so, what did you feel worked best with in-game processing as opposed to baking the processing into the sound?

BP: All sound in the Dark Pictures games are processed/effected at runtime. This is to improve the cohesiveness of the experience, the consistency of project sound design, the final mix, and to help ease the finaling process.

Localization…is key to a character-driven narrative, and we want to match the same quality in multiple languages…

Localization, for example, is key to a character-driven narrative, and we want to match the same quality in multiple languages, which is totally achievable and efficient with runtime processes on speech. We simply wouldn’t be able to finish a game to such a high quality without this approach in such a short schedule.

All processing, effects, spatialization, attenuations, occlusion, and mixing is live and interactive. This is essential to allow cameras to be moved at any time, and the sound will just ‘work.’ Or, when we affect all diegetic sound during a montage moment, improvements to individual sounds feed into those affected moments as well. The tech is so feature-rich nowadays that we have no excuses. Broadly speaking, it’s a hybrid mix of game/film, enhancing intimacy in close-ups, and has a sense of voyeurism for a longer lens.

Because The Dark Pictures Anthology games are narrative-driven horror games, the characters (and actors who voice them) are crucial to the games. Can you talk about your sonic approach to character design? What goes into each character’s sound palette? What’s your approach to the implementation of character assets? What are your character systems and foley systems like for these narrative-driven games?

BP: We invest heavily in character sound. We also use the same procedural systems in exploration and sequence leading to a consistent experience between keyframe and mocap, so they need to work for a full range of movement. Far from having a linear post-production approach, our systems are always expanding and changing and the team at Molinare do a fantastic job of creating assets for these procedural systems.

…the team at Molinare do a fantastic job of creating assets for these procedural systems.

We overhauled the foley systems for TDIM, using some great in-house tech that was developed initially for the vampires in HOA that we were able to leverage in the human characters. Vampires, quadra-pedal, were developed with 12 emitters calculating various movements and distances, whereas human characters run a smaller 8 emitters.

We have separate structures for footsteps, scuffs, lower and upper cloth, and a few other areas, offering fantastic subtlety and believability.

We also included a breath system reacting to a character’s exhaustion and stress levels.

We also included a breath system reacting to a character’s exhaustion and stress levels. These systems always need to offer subtly nuanced cinematic results.

Over time we have had a lot of interest from the wider game audio community for advice on their character systems, and one feature – our ‘50% center biased panning’ – closely but not exclusively constrains character sound to the center speaker. It’s now a feature in other third-person AAA franchises.

A huge compliment to the team and the systems was winning a ‘Best Audio in a Game Cinematic’ award for Little Hope. We were up against some amazing games using linear cutscene pipelines. Surprisingly for most people, ours are created with nearly all procedural audio.

The music is unique to each release. Can you talk about the different music styles and instrumentation you chose and why it was important to switch it up for each game?

BP: We have had so much fun with the music, choosing a different approach for each game. It embodies so much that’s unique to that exact story, it characterizes people, places, eras, back stories, the core narrative, and, of course, player emotions. It’s nearly impossible for the same music to fit another infinitely complex narrative.

It’s nearly impossible for the same music to fit another infinitely complex narrative.

Thus far, the Anthology has taken us to Salem in 1692, Chicago in the late 18th Century, to 1970s USA, and through to present-day and into sci-fi. We’ve covered the South Pacific, Iraq, New England, ancient Sumerian temples, psychos, demons, and aliens. For each project we’ve done something new, linking narratives with historically accurate instrumentation representative of the stories we’re supporting, a lot of which has been experimentation for us, in how we can hit all of the horror emotional cues with completely different sound palettes.

It is a huge collaboration with Jason Graves, who has composed the series thus far. We work well together to grow ideas.

for’HOA,’ we developed a really strong and simple pitching motif that injected fear in the player.

Man of Medan (aka, MoM) was entirely in 3 / 4 time signature, representing the ocean waves.

For Little Hope (aka, LH), we chose a lonely, minimal score using instruments synonymous with early white settlers in New England, often choosing aged or detuned elements.

For HOA, we developed a really strong and simple pitching motif that injected fear in the player. It could be played on any instrument, and existed in the warfare, vampire, and alien sections of the game as we switched from orchestral to synth-based instruments as the narrative exposition progressed.

And for TDIM, we combined an original score with a heavy nod to composer Bernard Herrmann with orchestral and operatic pieces during the key scenes.

Regarding the cinematic leitmotif concept, we’ve played with melodic motifs, pitching motifs, repeated motifs (both running through the music and as overlays), identical motifs in terms of instrumentation, and evolving motifs following the character arc.

Changing the music styles as dramatically as this…keeps it fresh for the audio teams and the player

Changing the music styles as dramatically as this not only benefits the narrative, it keeps it fresh for the audio teams and the player, and it helps separate us from other game series where each experience is so different.

Anyone that has played our games will be familiar with the track ‘ODeath,’ or ‘A Conversation with Death,’ which we always use in the opening and closing titles. Every game has a different styled rendition of this amazing work, and again, we try to match these to the narrative closely.

Something we have found after LH was the audience wants a relief after the intensity of the horror, and so the versions that worked best offered an energy release, or humor, and we took this to a new level for TDIM commissioning a modern, gospel-influenced barbershop quartet rendition from the great Signature Quartet.

O’Death sung by Signature Quartet and Du’Met’s mannequins | The Devil in Me – End Credits

TDIM was pretty unique even for us, with its use of operatic and classical music within the OST. We built a system that had runtime control via three RTPCs of how diegetic/non-diegetic a piece was to match the intensity of the action, how much additional processing it received, a kind of light dreaminess, and how much distortion we wanted to add to nasty it up. All of these parameters needed to be controlled by the game at runtime in order to achieve the precise processing of music at precisely the right time, within the variable timings of the interactivity.

The Devil in Me – Audio Highlights | Supermassive Games

 
[tweet_box]Designing the Frightening Sound of ‘The Dark Pictures Anthology’ – with Audio Director Barney Pratt[/tweet_box]

Looking back at some games in this series, there have been demons and vampires. What have been some of your favorite creatures or supernatural beings to design for the series? Can you talk about what went into the design of their sounds?

BP: For the vampires in HOA, we spent a lot of time experimenting with the clicks, aiming for something original compared to a few other ‘clickers’ out there. Mexican frog toys with homemade varied diaphragm sizes, bubbling death whistles, tasers, vegetables, and other organic manipulations.

The Dark Pictures Anthology: The Devil in Me – Dev Diary: Making of the Vampire Sounds

I discovered one technique while playing with my kids. On a wet autumnal day, with the back brake of a scooter trapped under my shoe, slowly tilting a scooter down from a wheelie, I got a rubbery clicking that reminded me instantly of the gristly vampires.

No single experiment gave us ‘the click,’ but we ended up with a huge emotional range that we could mix and match to express communication between vampires, and we often overlapped each technique to further express the vampire’s state of mind.

The premise of The Devil in Me is “a group of documentary filmmakers receives a mysterious call inviting them to a modern-day replica of serial killer H. H. Holmes‘s “Murder Castle.”‘ Did you do any custom recordings, or take recording trips for TDIM? If so, where did you go, what did you capture, and how did you capture them? How were those sounds used in the game?

BP: Absolutely, we prioritize location shoots to get an original sound for every project, and TDIM was no exception. The ‘Murder Castle’ hotel locations lead us to seek out a stately home that had been preserved close to its original Victorian and Edwardian state. We found the perfect location for capturing impulses, recording ambience beds, and we got a multitude of doors and creaky elements from broken plumbing and old electricals. We decided early on to use this identical source material for both the 1892 and present-day hotel replica to join them narratively.

For MoM, some of the team organized a trip onto a huge steel-hulled navy vessel about the same size as the Ourang Medan, recording anything that moved and again impulsing the many varied spaces from which we based the IR’s for the Ourang Medan itself.

We found the perfect location for capturing impulses, recording ambience beds, and we got a multitude of doors and creaky elements…

For MoM, we also took a trip in a large fishing boat off the south coast, rigging the boat as you would a car recording session for engine, exhaust, cockpit, lower/upper decks, bow, side laps, etc. We were met with a squall when we left port, and although we nearly lost one of the team overboard, the recordings were fantastic and it was great to capture some wild weather.

We have used various recording setups over time and more recently one of the team put together two great Sennheiser MS and split stereo setups. They cover most bases with some great top-end clarity, and we expand for any given situation with Sanken, DPA and Neumann mics.

Recording sessions like this give some great source material for a rich and unique-sounding game, and it also allows the sound team to develop a really deep feeling for the highest quality end result.

The Devil in Me has been compared the Saw in the manner of heavy gore. What are the differences between the horror sub-genres for each game and what did it take to fulfill that with sound?

BP: The biggest audio difference I have found in horror sound is between a real and an unreal threat. A real threat is more of a thriller and has recognizable sounds, while an unreal threat lends itself to a much wider palette of abstract sound relating to the source of that threat, its imagery, or behavior.

Supernatural threats typically dictate breathy, airy content, moving unnaturally fast, hard to locate, and are often derived from fuller sounds to fit the narrative.

The biggest audio difference I have found in horror sound is between a real and an unreal threat.

Psychological horror sound needs to reflect the thought processes and mental state of the primary character, suspicion, confusion, or intensity for example.

Little Hope was an amazing mix of both supernatural and psychological and we used a variety of airy bass tones to create that sense of weighty supernatural dread, lots of airy reversed and stretched voices to represent the supernatural echoes of the mind as the primary character slipped between the real and unreal worlds.

We are able to subvert and mislead the audience’s expectations through their familiarity with these groups of sounds…

HOA, with the real threat of the vampires, used very little abstract sound, relying heavily on the familiar motifs for the music to create that dread.

For TDIM, we favored a very tonal resonant metallic process in much of the abstract audio which was clean enough to keep the thriller aspect, representative of the Saw-esque machinery, the resonance of fear left by the psycho, yet clean enough to work well with the diegetic elements.

We are able to subvert and mislead the audience’s expectations through their familiarity with these groups of sounds and how they relate to the subgenres, adding breathiness to something that turns out to be real for example.

What were the most challenging levels in terms of sound in The Devil in Me? Why? What were some of the challenging aspects of sound there (creative and technical)?

BP: TDIM had the mechanic of a directional microphone and you were able to explore some levels using this microphone – what an opportunity! We were even able to arrange a whole bunch of narrative-related sounds gathered from throughout the game, recorded by our media-savvy psycho into this level to lead, confusen and scare the player.

We simulated the expected behavior of sound through the in-game microphone…

We simulated the expected behavior of sound through the in-game microphone, its pickup pattern (going for hypercardioid), sensitivity, and even mic handling noise. The overall result was fantastic, quoted as “like a horror ASMR” from NME. The player is able to explore entire levels using the microphone at will, and in one particular level, in complete darkness. The player has to rely on the directionality of the mic to locate sounds and progress the game.

In terms of sound, what’s been your biggest success on The Dark Pictures Anthology games, and why do you feel that way?

BP: For me, it’s been the consistent quality we have produced as a team, so I’ll hand it over to them for more detail:

Andrej Smoljan: Absolutely unique to The Devil in Me is the directional microphone mechanic. Getting to recreate something so close to our hearts was a great opportunity. Designing vampire vocalizations in House of Ashes was exceptional, with such a wide range of emotions, subtle snarls, and energetic screams – really helping to elevate them as a realistic threat.

This systemic approach not only improves the accuracy of those sound effects, but it also reduces the time we spend on implementation…

Hugely successful more generally is the repurposing of our procedural foley systems to trigger and dynamically modulate a wide variety of sound effects and vocalizations. This systemic approach not only improves the accuracy of those sound effects, but it also reduces the time we spend on implementation and that takes the heat out of finaling a game.

Rich Sauberlich: Helping to bring the visually stunning environments to life through sound with dynamic interactive ambience systems has been a lot of fun and has added immeasurably to player immersion in the locations of The Dark Pictures. For House Of Ashes, we designed a tremor, debris, and wind system to fill the game’s dark, claustrophobic, subterranean caves and tunnels where distant rumbles would result in close, detailed dust and debris falls. In Little Hope, we implemented a dynamic wind system that gently stimulated the eerie forest trees to keep the player on edge, making them listening intently for any danger lurking among the trees.

Hugh Waller One of our biggest successes has been the evolution of the procedural foley, with a near fully procedural foley system for our characters in the Dark Pictures games, improving quality and workload. The current systems include footsteps, cloth movement, accessories, guns, moveable objects, and props. It delivers believable, dynamic, and reactive foley for actions like climbing, fighting, and running. These advances help maintain consistent quality in both traditional exploration and cinematic sequences. Our procedural foley systems are crucial for providing dynamic, flexible, and believable characters that drive the narratives and allow players to create stronger emotional bonds. We are proud of the results so far and excited about the future of these systems.

We found a location, organized a recording session, and captured some beautiful sound design source material.

Ollie Campbell: One of the biggest successes for me, in terms of sound, came in our most recent game, The Devil in Me. From the outset, looking through the concept art, you could almost hear how this location would sound before any of us put pen to paper. We found a location, organized a recording session, and captured some beautiful sound design source material. The journey from location scouting to hearing our finished recordings in-game was incredibly rewarding. For me, the evocative and nuanced sound library that we created really transformed the sound of the game and brought it to life.

One of my favorite sets of recordings from this location became the basis of a procedural system. We performed a few long takes of dropping, rolling, scraping, and rattling various objects in the dusty attic of the old stately home we’d found. These were captured off-axis, or several rooms away from the source, and then played back as part of the ambience beds at random intervals as the player explores the interiors. These creepy intermittent sounds give the notion of movement and activity ‘off-screen’ elsewhere in the hotel, which reminds the player that they are not alone, and are being constantly stalked and pursued!

BP: And, we’re starting Season 2: Directive 8020…

The Dark Pictures Anthology: Directive 8020 – Official Announce Trailer | Supermassive Games

A big thanks to Barney Pratt and the sound team for giving us a behind-the-scenes look at the sound of The Dark Pictures Anthology and to Jennifer Walden for the interview!

Please share this: