Understand different needs.
We reviewed low vision, blindness, and color-vision differences. This helped frame visual impairment as a range of experiences, rather than one uniform set of needs.
MARSHINIMANSION
An audio based video gameFor all thrill seekers
Marshini Mansion / Game Demo
Listen to the mansion.
Choose a mode to activate the keyboard.
Explore the seven original floors, then take on three new floors by sound alone.
Scroll through the introduction, then choose to play the game or explore the case study.
Marshini Mansion / Inclusive Game Design / Winter 2024
A thriller for all.
Marshini Mansion is an audio-first horror game designed for blind and low-vision (BLV) and sighted players to share the same thrill. Our team built a new game around sound and the physical keyboard, making a common experience the starting point for accessibility.
- Design Team
- Thomas Chang
- Marcelle Brooks
- Cayde Bruce
- Deanna Kelly
- Instructor / Course
- Professor Marshini Chetty
- Inclusive Technology
- UChicago
- Outcome
- Audio-Based Video Game
- Final Film
- People’s Choice Award
My Contributions
- User-Centered Design Research
- Level & Interaction Design
- Sound Design
- Asset Integration
- Film Production
7Original Level Designs
32Original Sound Effects & Recordings
1Immersive Soundscape
2Stages of Participant Research
01 / Introduction & Reframe
Design for the experience
we could have together.
The course brief asked us to apply a user-centered design process to an interactive system for an underserved or marginalized community, then communicate the work through a website and short film. We chose gaming with BLV communities.
Our preliminary research suggested that making a game operable could still leave players with very different experiences. Adaptations could change the pace, mechanics, or feeling so much that the experiences were hard to relate to across communities.
We reframed the challenge around a shared experience: how might sighted and BLV players experience the same suspense, decisions, and discovery? We would build an audio-first game from the ground up, with the same information available to everyone.
About the Inclusive Technology course briefStarting Question
How can we make a game accessible to BLV players?
Our Reframe
What new game could both communities experience in the same way?
Accounts of video games adapted for accessibility in our preliminary research raised two concerns: sharing experiences with sighted friends, and finding the complexity and excitement players wanted.
“…makes it difficult to talk to sighted friends about games.”
“…they lack the complexity and adrenaline of more mainstream games.”
03 / Final Deliverables
The game.
The world within it.
The final project brought together the playable Python / Pygame prototype, its sound library, a project website, and a short film. The original game contained seven levels; this portfolio demo adds three more.
My level design used the same keyboard surface to introduce new combinations of routes, objects, traps, and monsters. Difficulty grew through the situations players had to interpret, while the physical frame of reference remained consistent.
Seven original levels.Routes and hazards change across a familiar surface, all within one world.
Increasing difficulty.New combinations of routes, objects, and threats demand more careful decisions.
Real stakes.One wrong step into a trap or monster, and it’s game over.
Sound leads the way.Distinct cues carry the state, the feedback, and the suspense.
My Sound-Design Work
Make the feedback
belong to the world.
My background in film and documentary work helped me approach sound as both information and atmosphere. Early versions sounded choppy: cues did not layer well, and the breaks in timing disrupted the illusion of being inside the mansion.
I researched foley techniques and sound design, then processed the sounds with effects, reverb, and echo. I worked on transitions, layering, timing, and pacing so that footsteps, threats, environmental sounds, and narration could feel like parts of the same space.
I used plucked metal coils for the tripwire and hyena recordings for the Wackergarl. Normal, wet, and muddy footsteps gave movement different textures. I generated the original voiceovers with AI, then applied sound design in post-production to fit them into the world.
From a source sound to
an immersive environment.
Listen to the original step.wav, then the processed wall2.wav. Processing gives this wall cue a sense of space and a cohesive tail, helping it belong to the larger environment.
0:00
0:00
Different sounds. Specific meanings.
These are processed recordings from the original game. Each cue has a job within the larger soundscape.
0:00
0:00
0:00
0:00
0:00
0:00
0:00
A place, an action,
and something approaching.
Turn individual layers on or off to hear how they work together. This mix illustrates the roles of selected original recordings, rather than replaying a specific floor.
Final Project Film
See the work
come together.
I created the five-minute video summary using early generative AI tools as part of our project deliverables.
04 / Conclusion
Inclusion can create
a new kind of play.
The response to the prototype was encouraging. My central takeaway is that inclusive design can create new experiences: it can change what a game is and who gets to share it.
Testing also exposed a steep learning curve. Players with less gaming or puzzle experience had more difficulty understanding how to begin. Feedback informed improvements to the introduction, but onboarding still needs a more thoughtful approach.
We were asking players to learn an unfamiliar game paradigm. That makes the learning challenge understandable; it also makes teaching the experience part of the design work.
“fun”
“challenging, but in a good way”
Two excerpts from reported participant feedback · original project evaluation
Build a better tutorial.
Create a safe space to learn the controls and recognize audio cues before introducing danger. Test the tutorial with players who have different levels of gaming and puzzle experience.
Co-design with BLV players.
The prototype sessions used blindfolded participants. Further evaluation with blind and low-vision players is needed to examine orientation, cue recognition, and the balance between challenge and confusion.
More levels. New games.
Expand the mansion with new levels and environments, then explore new games built around the same tactile keyboard and audio-first mechanics. Keep the shared experience at the center.