Mirrored Phantoms
Team member: 21
Production time: 3 months
Engine: Unreal
Team role: Audio Designer
Prepare to be consumed by Mirrored Phantoms, a chilling descent into a decaying mannequin factory where reality itself has fractured. This first-person psychological horror experience preys on your deepest fears, immersing you in an unsettling atmosphere where the line between what is real and what is imagined blurs with every haunting step on your journey to escape.
How did we design the audio？
Overall mood
Despite being a horror game in a manikin factory setting, the art set has some skin texture due to the fact that it's manikins that are cursed and the art set of the cursed dolls, and there are some fleshy items such as eyeballs in the art of the levels. At the same time, we wanted to express the backstory through the human voice, so in the overall sound design, we preferred to use the sound of living creatures in combination with the sound of real machinery to convey the conflict between reality and the spooky atmosphere.
Audio design tool
After talking to the programmers, since not all programs use the Fmod engine, I decided to combine the sound system that comes with Unreal with Fmod for this project. For complex sound systems, such as protagonist sounds, and monster sounds, I used Fmod to create a more immersive sound experience, while for simpler sound systems, such as one-time-use sounds and animation-controlled sounds, I used Unreal’s MetaSound system.
Communication
The workflow we use is that when the level is not yet in production, I will create the more important one-off and item sounds according to the needs of the art and level designers, and when the level is in the white box, I will ask the level designers to give me a video of the critical path in the level, and then I will create the sound content that we want the player to experience based on what the video shows. At the same time, I will make a video for the level designers and programmers on how the sounds should be played.
Design Examples
Enemy
Our enemy is a multi-headed cursed monster and has multiple states, five in total, from Present to kill. It was designed both to make sure it was recognizable enough and to make sure that each state and the switching between them was communicated properly. I achieved this by switching between whispers and screams of varying frequencies and then progressively augmenting the player's heartbeat sound with the Sense of oppression.
Level
Unique audio in each level is based on the need of the level. For example, the goal of the designer of level 4 is to convey to the player the concept that they need to go down, so I raise the pitch to fulfill the concept of choosing the right direction while the player is going down. Another example is the part in the second level where the player needs to see the monster in advance.
Cutscene
In order to save resources on the art, we tried to use cutscene sounds to convey the direction of the plot.
ED
In order to craft a more complete experience, I created this ED based on the backstory and narrative.