Designing and Implementing Game Audio for ‘Dying Light 2 Stay Human’ — with Techland’s Sound Team

Members of the Techland audio team – Wojciech Siadak, Tomasz Gruszka, Filip Hajzer, Edyta Mosińska-Duralska, and Michał Bendnarek – talk about their sound work on Dying Light 2 Stay Human – performing and recording zombie vocals, bridging the main character's human state and 'infected' state, creating dynamic ambiences for an open-world environment, capturing recordings of desolate locations and weapons impacts, building bespoke systems in Wwise to handle crowds, movement, zombie night chase situations, ambiences, breaths/efforts/screams, combat, weapons, using the multi-position mode of Wwise to simulate sounds coming from windows and doors, and so much more!

---

Interview by Jennifer Walden, photos courtesy of Techland; Warner Bros. Interactive

---

Please share:

• More

There are two things that players loved about Techland’s Dying Light first-person survival/horror game: the zombies and the parkour. With their latest release Dying Light 2 Stay Human, Techland delivers more of both! Set in the new location of the city of Villedor, players can avoid ground-dwelling zombie hordes by using their parkour skills to leap across the rooftops. And there are over 500 pieces of upgradable gear along with weapons that players can craft and customize to help fend the zombies off. All of this added up to a laundry list of audio assets that needed to be designed and implemented, and new systems needed to be built to handle these elements in game.

Here to talk about the game’s sound design and implementation are Techland audio team members Wojciech Siadak (Senior Sound Designer), Tomasz Gruszka (Lead Sound Designer), Filip Hajzer (Sound Designer), Edyta Mosińska-Duralska (Senior Audio Level Designer), and Michał Bendnarek (Audio Level Designer). Also on their sound team at Techland were Tomasz Siadak (Senior Sound Designer), Monika Wielechowska-Kotarba (Sound Designer), Marcin Paul (Senior Audio Tester), and Szymon Strauss (Senior Audio Producer).

In this huge interview, they discuss their approach to creating zombie sounds and building systems for the zombie hordes in-game, recording parkour sounds and creating movement systems, recording and designing weapons and combat sounds and creating systems to handle those in-game, recording and crafting distinct ambiences for different locations in the city and using City Builder to help trigger those ambiences, using Interactive Music Hierarchy in Wwise to build a breath system, using the multi-position mode of Wwise to simulate occlusion, integrating Wwise into Techland’s custom-built C-Engine, and so much more!

Dying Light 2 Stay Human – Official Gameplay Trailer

What do you feel is unique about the sound of the zombies in Dying Light 2 Stay Human? What went into creating their sounds?

Techland Senior Sound Designer Wojciech Siadak

Wojciech Siadak (WS): The basis of our design for the zombie voices was the assumption that zombies are actually people, only infected. This immediately sets the framework in which we wanted to move.

With a very wide range of different possibilities available, it is worth establishing such a framework at the beginning of production. The sounds would be as organic and believable as possible. It was clear that the basis for their creation cannot be strong audio processing, but the right source material. We also consciously decided not to use animal sounds.

In the game, each behavior of the infected is underlined by sound. So, in addition to the most important audio events – such as attacking, the reaction to pain, and death – they also produce a lot of other sounds when walking, running, suffering in the sun, etc. In other words, the infected make sounds all the time and this is what makes them different from human opponents, whose breathing we do not hear in the game. For example, for a biter (i.e., the basic opponent in Dying Light 2 Stay Human), there are 45 different audio events. Voices can change depending on time (day or night) or place (exteriors, interiors, and dark places). With four sets of voices (2 male and 2 female), this resulted in a total of almost 750 audio files. And that’s just one type of infection. There are 12 unique types of zombies in the game.

In the case of 5 types of infected, I played the role of a voice actor. I would advise any sound designer facing the task of designing zombies to give it a try as well. It gives you a lot of experience. It also helps to work with other voice actors later.

…I was practicing inhaled screams while driving and almost passed out due to hyperventilation.

In the beginning, it was not very easy because I had never done it before. I haven’t recorded anything for a long time. I’ve been experimenting with my voice and looking for interesting sounds. It turned out to be quite a dangerous task when, for example, I was practicing inhaled screams while driving and almost passed out due to hyperventilation. The hardest thing, however, was probably overcoming my own inhibitions, which were preventing me from giving 100% of myself, especially at the beginning.

The basis for creating the sounds was non-destructive pitch shifting (changing the playback rate). I stuck to that rule to keep the sounds as simple as possible. Screams consisting of many layers can be unintelligible and incomprehensible in the game. Therefore, if possible, I did not use sound layering. The sounds were created on a single track in the DAW, but obviously, they were heavily edited. In the case of some infected, this turned out to be insufficient. So, I added a layer of sound texture to the main vocal line.

Screams consisting of many layers can be unintelligible and incomprehensible in the game.

I have a trick when creating screams consisting of several layers, especially if these voices are placed in low registers where the human ear is less sensitive to pitch. In the benchmark, I temporarily increase the main playback rate of the session, thereby increasing the pitch. Then, I check if the voices blend well and make the necessary changes.

An interesting idea is also something called reverse voice-acting. I happened to find a very interesting zombie roar in the library. It was great for one of the infected’s attack sounds. Unfortunately, there was only one and it couldn’t be used to create a full set. I began to analyze it and increase the pitch so as to hear the voice that was the basis for creating this sound as naturally as possible. I learned how to repeat that scream, and then I recorded several versions of it, so I got the whole set. Of course, the final result was different from the roar that inspired me, but I think it can be an interesting way to stimulate creativity.

On the technical side, it is, of course, worth using microphones with extended bandwidth. Even the omnidirectional ones work well if the recording takes place in a damped room. I have noticed that sound from omnidirectional mics tends to be more open and natural compared with directional mics.

DYING LIGHT 2 Sound Level Design

Players can run into massive crowds of the infected, especially at nighttime. And I read that there’s a new crowd system that allows zombies to behave more dynamically. They have an evolution cycle in the game, so they systematically evolve and change based on particular circumstances. What did this require in terms of sound? How did it impact the sound team’s responsibilities?

WS: In Dying Light 2 Stay Human the main focus is on people. But of course, the infected are still a major part of the game’s soundscape. Even if they are only harmless biters wandering around the streets of the City.

Of course, in order to set the right tone for the events in the game and emphasize the gameplay, it’s not enough to simply voice the individual infected. The game features multiple audio systems to ensure it sounds the way a city overran by zombies would sound. The systems have two equal goals; they increase the intensity of the gaming experience and provide information to the player about the gameplay.

It automatically detects groups of zombies, places an emitter next to them, and plays the appropriate sound…

One of the first created systems was the zombie crowd. In short, it generates an ambience of zombie horde around the player. It automatically detects groups of zombies, places an emitter next to them, and plays the appropriate sound, changing it depending on the size of the group, distance from the player, and time of day. The system additionally takes care not to generate too many emitters and not to place them close to each other.

To support one of the main designs of the game – that the street is for zombies and the rooftops are for people – we introduced two additional systems. We didn’t want the player to feel too safe standing on the roof. As soon as he or she gets close to the edge, they will hear the reaction of zombies sensing a potential victim – screams and movement below, something falling and being knocked off by a passing infected. Similarly, the scream will be triggered if the player chooses to leap between buildings.

…the most important and largest system is the one that supports the night chase mechanic.

However, the most important and largest system is the one that supports the night chase mechanic. This is one of the most important moments of the day/night cycle. It is initiated most often by Howlers patrolling the night streets, which inform other zombies about the player’s position with a prolonged scream.

The chase consists of 4 phases of increasing intensity. The main factor modulating this system is the player’s distance from the pursuit group. If the zombies are close, we hear their actual screams and footsteps as well as close-up sounds of running or gasping.

However, if the player manages to move away from the group and their real sounds are no longer audible, the system starts generating special sounds of distant pursuit, screams, noises, falling objects, and running zombies. This way, the player is fully informed that the chase is still ongoing and he or she won’t feel safe yet. This is especially useful if the player decides to mute the music.

Parkour has been a favorite aspect of both Dying Light games. I read that the City’s more vertical design will affect parkour, giving players access to more extensive movesets and movement options. What did this require in terms of movement sounds (was it mostly foley? How many unique movement sounds did you have to create?) What are some aspects of sound that help the parkour to feel satisfying?

WS: The player’s movement, if I remember correctly, was one of the first things we started prototyping during the pre-production stage of the game. We were wondering what parkour actually sounds like. We wanted to establish a reference point for further work.

It was also an opportunity to move around outside and get a bit of fresh air, which is always a good idea! We used a GoPro camera and 4 miniature DPA microphones, which we attached in various configurations on the parkouring person.

We used a GoPro camera and 4 miniature DPA microphones, which we attached in various configurations on the parkouring person.

Later, in the studio, we were looking for the most interesting configuration. In the end, most of the target movement recordings were made by mounting two microphones directly on the shoes and another pair on the ears.

The player’s movement was almost entirely recorded outside. I recorded the majority of the elements on an old military compound. For two hours, I jumped on everything I could: walls, fences, lamps, barriers, etc. I recorded a very rich, organic sound base, from which the assets for the game were later created.

…the player’s steps…change in volume, pitch, and equalization depending on the speed of movement.

All in all, there are over 150 different player movement events in the game. Everything is fully interactive and is subject to constant modulation. For example, the player’s steps (although they basically consist of two types: walking and running) change in volume, pitch, and equalization depending on the speed of movement. This gives a very good effect. For example, when walking very, very slowly, we hear the sound of a gently placed foot instead of a typical gaming step. Modulation also occurs during the stealth mode, where we dynamically reduce the initial transients of footsteps, reflecting that the player is trying to move as quietly as possible.

We also tried to include as many details as possible in the sounds of movement to support gameplay. Some systems are certainly not consciously registered by players, but I hope they have an effect on their subconscious. For example, when climbing, as we move along the wall, the player’s stamina drops. Once it reaches zero, the player falls off the wall. To aid this system with sound, I added the shuffle of boots against the wall. However, this is only audible when the stamina drops below half, showing that the player’s movements become more and more clumsy due to fatigue.

What about the breaths? What did it take to implement all those character sounds into the game?

WS: The premise of the game was to emphasize the player’s physicality as much as possible. We knew that breathing could help show it. We could use it to show that Aiden is tired or that his infection is taking over. Therefore, actor Jonah Scott (who voiced Aiden) recorded most of the breaths and efforts for three levels of fatigue (low, mid, and high). And then we re-recorded the entire breath system for the critical infection state. This resulted in over 1,500 audio files supporting the effort and scream system.

…we re-recorded the entire breath system for the critical infection state.

Of course, we wanted to capture the physicality in Jonah Scott’s voice recordings. The recordings were usually made in sequences, so if we recorded the jump effort, Jonah performed the sequence: breaths – jump – breaths, not an isolated single effort.

We also wanted to capture the physicality by asking Jonah to move as much as possible while recording. So, for instance, the attacks were recorded with arm swings, while recording climbing, he was hanging on a bar that was in the studio at the time, and before recording the “tired” version of the assets, he was doing push-ups or squats.

The most difficult element was the transitions and dependencies between the individual elements of the breath system.

The implementation of the system also looked interesting. The most difficult element was the transitions and dependencies between the individual elements of the breath system. For example, breathing will behave differently if the player stops after a light run, and differently after an intense sprint. But this required too much work from the developers, so we quickly tried a different approach. We built the system using Interactive Music Hierarchy in Wwise. We only asked the developers for basic information about the player, i.e. whether he is standing, walking, running, sprinting, jumping, landing, etc. Based on this, we have built a fairly extensive and stable breath system.

There’s now a paraglider that the player can use. What did this require in terms of sound?

WS: Indeed, this is a brand-new mechanic that wasn’t available in the first Dying Light. Working on this system was very satisfying. It was not as extensive as most of the other player systems, but it is very impressive.

…the implementation was not so easy, because the system had to be built from scratch.

The first sound that was quite easy to implement was opening the canopy of the paraglider. I wanted to show the physicality of this sound with a lot of intense punch. When creating it, I wanted to give it an interesting character, so that it could become one of the iconic sounds of Dying Light 2 Stay Human.

…with hundreds of different sound systems, it’s not worth complicating the sound unnecessarily.

When it comes to the sounds of flying on a paraglider, the implementation was not so easy, because the system had to be built from scratch. Again, we decided to build the system in Wwise, based on basic gameplay information (e.g., flight direction and speed). These parameters are managed by the state machine in Wwise, which properly blends flight sounds. Each change of direction is additionally emphasized by the player’s delicate effort and the tension of the paraglider ropes.

It’s worth remembering that in a game as big as Dying Light 2 Stay Human, with hundreds of different sound systems, it’s not worth complicating the sound unnecessarily. Simpler systems tend to be more resistant to breaking and easier to maintain and develop later on.

One of the nice features of this system is that when it rains, you can hear the drops hitting the canopy of the paraglider.

The player is able to upgrade gear and craft weapons as well as add weapon mods. This game has over 500 gear pieces. Wow! Can you talk about that in terms of sound, and what that meant for your responsibilities as a sound designer?

Techland Lead Sound Designer Tomasz Gruszka

Tomasz Gruszka (TG): In Dying Light 2 Stay Human, we have a large number of weapons that can be modified in many ways. This gives us the ability to create dedicated sounds for each modification. However, we had to keep memory constraints in mind, which challenged us to create sounds that would suit each type of weapon.

Ultimately, we wanted a satisfying sound, reflecting a specific modification. We have several different ones to choose from, such as toxic or electric. Each of them is a hand-made installation from items collected in the game world. What it sounds like is a combination of the sounds of these items and the devices that support them. For example, the Freeze mod sound contains a compressor sound that builds up a charge of freezing substances. The effect of its action and the enemy’s reaction is also very important and specific to each type of modification, e.g. cries of pain differ depending on whether the modification was fire or electric. All these elements make up the final effect.

A timer counts down to when Aiden turns into a zombie if he doesn’t get the needed exposure to UV light… and he goes into “zombie-mode.” How did you handle this in terms of sound?

Techland Technical Sound Designer Filip Hajzer

Filip Haizer (FH): As the player’s immunity falls, we progressively replace Aiden’s normal voice assets with the “infected” equivalents. As the player’s immunity reaches zero, he or she is slowly losing health. In most cases, this will result in Aiden dying, but there are some moments in the story where Aiden goes into “fury mode” instead. We consider those moments to be the verge of humanity – Aiden is still a human being but his infection takes over, giving him superhuman strength.

During those moments, we replaced all of his voice assets with a “zombified version,” which is acted out by Jonah in a special way and processed by pitch shifters by the audio team. This was a great start, but we quickly realized it’s not enough to portray a human being losing control over his whole self, so we went a little further by adding an extra layer to all impacts of Aiden’s attacks during the fury, used a state to lowpass the environment and NPCs and added a little reverb to some sounds. All those elements combined allowed us to create a pretty believable impression of losing control over one’s body to the THV virus.

What went into the sound of the weapon impacts, blocks, and the resulting gore? Also, how did you approach the sound of a lethal blow versus a regular or glancing blow?

TG: When we started working on the strikes in combat, we set the goal of creating very realistic effects. As in Dying Light, we did many hours of baseball bat and machete sessions, recording hits on meat and bones. After the recordings, the meat and bones were donated to the Wrocław Zoo as tiger feed.

…we did many hours of baseball bat and machete sessions, recording hits on meat and bones.

The recording session delivered a powerful base of sounds with natural beats, tearing, breaking, and cutting bones. This was our starting point; we spent a long time searching for something interesting that expressed the details as best as possible.

However, using pure, realistic sounds does not produce the satisfying effect we are already culturally accustomed to. All these organic elements gave the base layer, which we combined with additional sounds that emphasize the whole effect. Our approach combines hyperrealism and abstraction.

With separate layers of meat punches, bone breaks, and blood spatters, we were able to manipulate the amount of gore in a fight quite easily, and the final effect is quite juicy.

…brutality is not only reflected by hits, but also by the reactions of opponents who feel pain…

In addition, brutality is not only reflected by hits, but also by the reactions of opponents who feel pain when moving and screaming in a specific way. This enhances the final effect.

Blocking with weapons is a naturally inherent element of combat that occurs very often. For blocks, a very simple mini system of playing two sounds simultaneously was created: the sound of the weapon that the player is holding and the sound of the opponent’s weapon.

The recording session for blocks was an interesting experience. Recording a clean hit (e.g. with a metal pipe against a hard surface) did not give a satisfactory effect. They were lacking power and the suggestion of a clash of large forces. I started recording the sounds with one of the historic shotgun microphones, which gives the recorded sounds a pleasant tone.

Finishing hits are distinguished from regular ones by additional layers and a mix change. Usually, there is more gore on the finishing hit. It can also be a head crash or decapitation.

What are some of your favorite weapons in terms of sound, and what went into the sound of those weapons?

TG: As in the previous part, in Dying Light 2 Stay Human, the machete has probably become the most popular weapon that I spent a lot of time working on, so naturally I had to become familiar with it. It is a weapon that can cut in several different ways, from a simple blow to the body to severing a limb or head.

the handle of the machete accidentally came loose. These ugly and undesirable sounds…became an inspiration…

I recorded the machete and its metallic hums many times in the studio. During the recording session for the collision of weapons with surfaces, the handle of the machete accidentally came loose. These ugly and undesirable sounds at the beginning became an inspiration for recordings of destroyed weapons. If the machete is destroyed in the game, we will hear the sounds from this recording session.

What was the most challenging weapon to design in terms of sound? Why?

TG: First, we wanted to achieve a satisfying result for a single weapon. There were dozens of them to create and we wanted to keep the ability to quickly iterate them, which was a big challenge. A modular system was created in Wwise. It consists of many components, including clean hits, gun noise, bone or skull fractures, tearing of clothes, and gushing blood.

A modular system was created in Wwise. It consists of many components, including clean hits, gun noise, bone or skull fractures…

Added to this was a group of abstract sounds covering exceptional situations such as power attack. What we hear in the game is a real-time mix of multiple layers. These elements randomly change the volume and pitch, as well as the playing time, so the player will hear a slightly different sound each time. Randomization was crucial to reduce repetitiveness.

Additionally, layer control allows us to change the balance, sound quality, and character of weapons easily. With a traditional approach, any change to the overall feel of combat would involve redesigning a large number of assets.

Thanks to the use of a modular system, we optimized the size of combat assets in memory, making space for other elements of the game.

What are some aspects of sound that help combat feel satisfying?

TG: Initially we analyzed in detail many areas that could have an impact on immersion and combat satisfaction. Coordinating a lot of elements, such as strikes, efforts, animations, combat pace, music, and mix allowed us to achieve the final effect.

For most one-handed weapons, a group of light, heavy, and finishing hits has been created. This variety adds realism to combat. Stronger hits are characterized by more impressive sound, followed by a broken bone or bigger blood splash. At the same time, it gives the player a sense of strength and satisfaction from a successful hit. Also, the swing itself can be weaker or stronger depending on the hit type or the passive and active skills.

Hitting the body, we will hear more sounds of tearing clothes, whereas a hit on the head will sound juicier…

The point of impact also affects the sound. Hitting the body, we will hear more sounds of tearing clothes, whereas a hit on the head will sound juicier with an added layer of a cracking skull. A lot of work has been put into exceptional moments, such as cutting off limbs or heads. These situations are rewarding for the player, so there is a lot going on in the sound as well.

Player damage requires a very close perspective to feel like we’re actually taking the hits. The player can hear that the hit was made with a metal pipe or a wooden stick. Additionally, the damage sounds are related to the player’s health. When it’s low, we can hear stronger blows and ringing in the ears.

An indispensable element of combat is the efforts of the player and opponents.

An indispensable element of combat is the efforts of the player and opponents. In the case of the player, we did lots of recordings, looking for sounds that would reflect emotions and effort as realistically as possible. It shows the power of aggressive strikes when the player is at full health and stamina, and desperation when the hero takes damage from opponents.

We took a similar approach to recordings for human opponents. We spent many hours trying to record the most authentic behaviors that capture physicality, brutality, aggression, and suffering. Matching the image and sound had a huge impact on the overall experience of combat.

The system is supported by dedicated music that can react with accents at selected moments, for example after a perfect block.

Dying Light 2 Stay Human takes place in a new main location, the city of Villedor. Overall, what are some sonic characteristics of this city?

Techland Senior Audio Level Designer Edyta Mosinska-Duralska

Edyta Mosińska-Duralska (EMD): Villedor is the last city – the habitat of humans in a dead world. In the ambience of this city, we wanted to get quite contradictory emotions and impressions. On one hand, there’s sadness and death – the world is dead, deserted. Everyone is infected and it doesn’t take much to turn into a zombie. The world is dangerous, dark, and a constant reminder of the inevitable. The infected were human once but failed to survive. On the other hand, there’s life and hope – people still want to create settlements, civilization, and simply live, to regain at least a semblance of what they have lost, even though it is no longer possible.

In the Villedor ambience, we wanted to reconcile these two worlds. We recorded a lot of dedicated sounds and ambience specifically for this project. We visited WWII bunkers, and abandoned factories; we went to the mountains in winter, and we recorded empty forest ambience. We recorded everything that would give the feeling of a dying, abandoned world on the verge of extinction.

In zones ruled by bandits, we will hear sounds to warn us about the danger and the presence of hostile people.

However, we couldn’t forget about the “fighting” part of the City, the one that still hopes and, despite adversity, lasts. Players are reminded of the sounds of human activity, the sounds of their work, calls, voices, or laughter.

Villedor is divided into several regions and districts. Each of them can have a different owner. We also had to reproduce this in audio. In zones ruled by bandits, we will hear sounds to warn us about the danger and the presence of hostile people. We will hear them knocking over trash cans, breaking windows, or laughing ominously.

The zone ruled by Peacekeepers, i.e. a military group, will be distinguished by the sounds of marching troops, people giving orders, or preparing weapons.

The “survivors” zone, on the other hand, will abound in more “farm” sounds – people working, talking, and sounds of farm animals.

Our goal was to make ambience matter…

When creating ambience for Dying Light 2 Stay Human, we put a lot of emphasis on making them as immersive as possible and harmonizing with the gameplay. We often asked ourselves, “What should the player hear at a given moment to get the most out of it?” Our goal was to make ambience matter; it’s not just noise, or a background for other sounds. We wanted the ambience to inform the player about location, district owner, and whether it is safe there or not.

Ultimately, there are 220 ambient VO types in the game recorded for 9 languages…

We thought for a long time about the best way to convey information about safety level and came to the conclusion that we will focus on voice-over. Probably nothing informs about a threat better than pleas for life heard somewhere in the ambience. The idea is simple, but very demanding because each VO needs to be recorded for all languages in which the game is released. Ultimately, there are 220 ambient VO types in the game recorded for 9 languages, and each of these 220 types is recorded in three or four versions.

DYING LIGHT 2 Audio Level Design

The game uses City Builder (a system that “generates” the city, including audio). What did this mean for you in terms of your sound work or sound assets that needed to be created for this city building system?

EMD: City Builder automated our work on open word audio. The tool that was created for Level Designers and Level Artists was also very useful for the audio department. When we developed a whole system of audio generators and a workflow on the map, the sound of the City implemented itself, and the audio team only had to check critical places.

When we developed a whole system of audio generators and a workflow on the map, the sound of the City implemented itself…

This was possible thanks to a thorough analysis of what the city of Villedor is made of – what types of streets, squares, parks, lakes, etc. there are. Then we grouped similar elements and prepared presets for these groups (i.e. a list of audio settings). Thanks to this, we could immediately assign appropriate audio properties to each new element on the map.

This required the cooperation of all the people who worked on the Dying Light 2 Stay Human map and it took a while to develop a common front, but we managed to reconcile the needs and requirements of all departments. Automatic distribution of ambient sounds has been developed for both interiors and exteriors.

50% of our work is creating systems and settings. If all works, designing the sounds themselves is a pleasure.

This is a good moment to emphasize the importance of the technical background in the audio team’s work. Many people are often surprised that the work of a sound designer is not only about creating sounds. In fact, 50% of our work is creating systems and settings. If all works, designing the sounds themselves is a pleasure. In addition, by having effective and efficient systems, we can devote the saved time to this “artistic” part and get carried away by our imagination.

There are 15 named Districts: Houndfield, Trinity, Horseshoe, Quarry End, Downtown, The Wharf, New Dawn Park, Garrison, Muddy Grounds, Saint Paul Island, Lower Dam Ayre, Newfound Lost Lands, Horizon Village, East Borough, and West Borough. What are some distinct sonic aspects of the different districts, or sonic differences between a few parts of the City?

EMD: When creating ambients for Villedor, we tried to focus on what is really important and what carries information for the player. As I mentioned before, we didn’t want our ambience to be something that is “humming in the background.” With this mindset, we decided to focus mainly on distinguishing two areas of the City – Region 1 and Region 2.

…we didn’t want our ambience to be something that is ‘humming in the background.’

Region 1 is the older part of the City, with lower buildings, monuments, and narrow streets. Region 2 is a modern part with skyscrapers, wide streets, and viaducts. If I had to describe these regions in two words, I’d say that Region 1 would be “wood and stone” (it refers to the history of the City, its traditions, and culture.) Region 2 is “metal and concrete” (it portrays modernity and progress).

Each region has several areas. From the player’s perspective, it is important to whom a given area belongs, its level of difficulty, and the attitude of its inhabitants towards the player.

…when we’re in survivor areas of Region 1, we will hear people working manually…

These audio aspects also depend on whether we are in Region 1 or 2. For instance, when we’re in survivor areas of Region 1, we will hear people working manually (e.g. hammering nails, and sawing wood). In Region 2, the same people will use power tools. This is another reference to the maxim “Region 1: wood – tradition,” “Region 2: metal – modernity.”

Everywhere, the street level will have dark, ominous drones in it…

There are also aspects of the ambience that are not subject to these rules, yet the sonic world of Villedor is very coherent at the same time. Everywhere, the street level will have dark, ominous drones in it, reminding you that the streets in this world are “death” and that they belong to the infected. Roofs are “life” and hope, and there we will hear all the sounds of people’s activity and the rebirth of nature. This effect is also invariably influenced by the systems that Wojtek spoke about.

…the roof and the street are completely different sound realms.

It is worth mentioning that we created ambience from two layers: an abstract one, which contains sounds that are not physically present on the map (e.g. people working in the distance), and a “real” layer or real sound sources, called auto-emitters, which are placed on the map. This invoked the need to prepare several sets of assets for real sound sources and to manage them properly, because, for example, the same tree can appear both on the roof and on the street. And as I mentioned earlier, the roof and the street are completely different sound realms.

DYING LIGHT 2 – Additional Music Abbey Road Recordings

Dying Light 2 Stay Human is an open-world game, so what are some things you did with sound to help make this world feel “real,” to feel engaging, dynamic, and “alive”?

EMD: Audio in Dying Light 2 Stay Human reacts to many aspects of the game. We have around 100 states and switches in Wwise that change and shape our soundscape. We have almost endless control over what players hear. Sounds differ if the player is on a rooftop or on the street level, if he or she is parkouring or just exploring the City, using the paraglider or binoculars, actively fighting or trying to stealth-kill bandits or the infected, what time of day it is, what weather we have, who the owner of the zone is, whether the zone is safe, and what the attitude of the owner is towards the player.

We have around 100 states and switches in Wwise that change and shape our soundscape.

We can go even deeper to more detailed aspects of the game, like, if the player is playing the main quest, whether he or she is actively playing it or exploring the city, or in what part of that quest and the whole story he or she is in.

So our mix varies depending on many, many things. And that is great as well as very powerful. We can shape player perception according to our design intentions almost like in a linear game. What sound effects should be in the foreground, eg. during the first cutscene at the Bazaar? Aiden is ill, he is about to turn, and people that see him are terrified and want to kill him because at any minute, he can become a threat.

…the church bell in the Survivor Hub, The Bazaar, strikes the hour like almost every clock in a major European city does.

During that scene, we wanted to focus on the fragility of humanity and the physicality of the player character. We decided to let the sound design do its magic here. But during the final fight, emotions are most important, so music is in the foreground there. Sounds in ambiences, like alarms, rockets being prepared, and metal rattling and squeaks are made softer and quieter to give music all the space it needs.

The details and flavors that make the City alive, dynamic, and engaging are also important. My favorite is the fact that the church bell in the Survivor Hub, The Bazaar, strikes the hour like almost every clock in a major European city does. Of course, the number of strokes corresponds to the current hour in the game.

For the new DLC called “Bloody Ties,” what were some fun opportunities for sound? There are bloodsport tournaments, a new location, Carnage Hall, and new weapons. What are some highlights in terms of sound for these?

Techland Audio Level Designer Michal Bednarek

Michał Bendnarek (MB): The “Bloody Ties” DLC focuses on a fighting tournament organized in an old opera house somewhere on the outskirts of the City. It’s kind of an MMA event-meets-ancient-gladiator battle. “The show must go on” slogan was being repeated throughout the production and sound naturally had to follow to enhance that feeling.

The fights in Carnage Hall take place all the time, whether the player participates in them or not. While traversing through the opera house, rare is the moment when you don’t hear the music blaring from the speakers or the crowd cheering for their favorite contenders. You can still hear glimpses of the opera’s former purpose, either through a specific music track or sound effect, but much like the building itself, it’s been reshaped to the gory needs of this new world and its inhabitants. For example, the gong that would’ve been used to mark the end of an entr’acte of an operatic piece now signals the beginning of a fight.

…the crowd ambience…is dynamically reacting to players’ actions, further immersing them in those bloody shows.

The base gameplay mechanic in the DLC is centered around the challenges you accept in the main arena. We wanted to make sure that players who are exceptionally efficient in combat or parkour get properly rewarded by their adoring audience and we were able to achieve that by harnessing the power and flexibility of the music system designed for Dying Light 2 Stay Human. Using the abundance of states sent by the game allowed us to create the crowd ambience that is dynamically reacting to players’ actions, further immersing them in those bloody shows.

For that same reason, instead of just reusing combat tracks from the base game, we went with a diegetic approach to the music accompanying you in the arenas. Each challenge is a spectacle meticulously directed by its creators, with the limited resources available in a post-apocalyptic world, and we wanted to stay true to that. So instead of pristine, orchestral compositions, you hear rough metal riffs, rendered with big arena reverb and the speaker’s post-process, giving it more of a “makeshift” vibe.

…we went with a diegetic approach to the music accompanying you in the arenas.

With such an abundance of sonic stimuli, the biggest challenge was to make sure the player hears exactly what he or she needs to hear during combat against dozens of enemies, with a rock song playing from the speakers and the crowd chanting your name. We haven’t had situations like that in the base game. So, even though it can get messy, we had to make sure the information is clear by taking extra care of the dynamic mixing of the soundscape.

Combat has to come first and its readability was our main concern. The landed attacks and received hits have to always cut through the mix. We also added this low-frequency heartbeat for when the player is low on health.

With the improved brutality of combat (stronger killing blows and heavy hits), we had an opportunity to further highlight the state of this world, where people would entertain themselves by witnessing these “reinvented” gladiator fights.

Dying Light 2 Composer interview

What were your biggest technical challenges in terms of sound on Dying Light 2 Stay Human? What were some solutions to these challenges?

FH: We spent a lot of time fine-tuning the sound propagation system. It was a challenging system mostly due to the open-world design of Dying Light 2 Stay Human.

For calculating propagation paths, we used dedicated in-game objects called nodes…

The City of Villedor is vast and generated semi-procedurally with the City Builder tool. We ended up using the multi-position mode of Wwise to simulate sounds coming from windows and doors. For calculating propagation paths, we used dedicated in-game objects called nodes, which are a part of every door and window prefab. Based on the position of those objects, optimal propagation paths were being calculated beforehand, creating connections between those nodes. Every time a sound is obstructed by walls, its position is changed to match the position of a node nearest to the player and is moved away from the player at a distance equal to the length of the path. If the path is angled, we move the source even further to create a realistic impression.

Unfortunately, we could not use propagation nodes for sounds that were obstructed by building edges…

Unfortunately, we could not use propagation nodes for sounds that were obstructed by building edges when the player roams the streets. To solve this problem, we decided to only change parameter values sent by the system when the listener and audio source are both outside, leaving the emitter position unchanged.

That was a promising base for the propagation system, but we needed to further simulate sound traveling through the environment. To enhance the system, we created a source reverb feature that made it possible to inject the reverb of the room from which the sounds are coming. Additionally, to help the player find exits from any given room, we used the previously mentioned propagation nodes to play exterior ambience with 3D positioning.

Techland created a new engine named C-Engine to power the game. Was this a good fit for the sound team, too? How did you incorporate Wwise to achieve what you desired in terms of sound?

FH: The C-Engine and Wwise were a perfect combination of tools for us. Thanks to our amazing technology team, we were able to create integration with Wwise in a very custom way, which paid off in relation to optimization and, later, the creation of many needed gameplay features.

C-Engine’s prefab-based design allowed us to easily track the implementation of most sounds…

C-Engine’s prefab-based design allowed us to easily track the implementation of most sounds and keep a project structure that is easy to maintain and track. When working on a project with such an impressive scale, keeping your content in an organized structure is the real key to success. In this respect, Wwise also proved to be effective. We created a small amount of sharesets within the Wwise session that were used for sound design. Because of that approach, shareset iterations were very quickly implemented across the entire game.

Wwise provided us with an ocean of possibilities for dynamic mixing, advanced ambience implementation, quick prototyping of new audio systems, and, of course, amazing interactive music integration. Having this many possibilities could be a curse, but in our case, it created a very exciting work environment. It made everyone in the audio team determined to search for fresh solutions.

What are you most proud of in terms of your sound work on Dying Light 2 Stay Human?

EMD: I am most proud of using City Builder for open-world audio in our game. It’s a mechanism that saves us a lot of manual work. I am also proud that we managed to create audio that reacts to almost every aspect of the game, with the main storyline; it’s, let’s say, ‘made-to-measure.” We know exactly where in the main story the player is, and even whether he or she is actively engaged in completing quest objectives. This translates to a lot of possibilities to interfere with the mix of the entire game at selected moments.

I am also proud that we managed to create audio that reacts to almost every aspect of the game…

FH: I am especially proud of how all systems (audio and music) complement each other and support every game mechanic we have. Having this many interactive and dynamic elements within the mix was challenging and we had to be sensible not to make them conflict with each other. We managed to create a clear mix that does not overwhelm the player with too much information at one time yet highlights every important thing around and about the player. Working on a mix like that was an awesome adventure and I consider our result a huge success.

WS: For me, more important than the sound effect we achieved in our game is the team that we managed to assemble during the production of Dying Light 2 Stay Human. From two people in the beginning, the team grew to eight people. We were taken care of by the amazing producer Szymon Strauss and the experienced composer Olivier Deriviere. I think we’re a very cohesive team.

A big thanks to Wojciech Siadak, Tomasz Gruszka, Filip Hajzer, Edyta Mosińska-Duralska, and Michał Bendnarek for giving us a behind-the-scenes look at the sound of Dying Light 2 Stay Human and to Jennifer Walden for the interview!

Please share this: