ALBUM
5 tracks, 17 min
Role
Project Lead · Design
Team
1 Project Lead · 1 Artist · 1 Dev
Made with
Illustrator · Photoshop · AfterEffects · Logic Pro
Domicile – a narrative game about two siblings navigating sudden violence, in search of parents who never returned. How do they make sense of a world that changed overnight?
TRACK 00 · OVERVIEW
The Problem
With current affairs, news media is increasingly saturated with conflict, injustice, and suffering – leading to widespread compassion fatigue.
TRACK 00 · OVERVIEW
The Problem
With current affairs, news media is increasingly saturated with conflict, injustice, and suffering – leading to widespread compassion fatigue.
How might we design experiences that counter compassion fatigue to foster empathy?
To answer this question, we need to understand where existing formats fall short.
How might we design experiences that counter compassion fatigue to foster empathy?
To answer this question, we need to understand where existing formats fall short.
The Gap
The Gap
News use statistics to show casualties, but we feel less empathetic for numbers.
News use statistics to show casualties, but we feel less empathetic for numbers.
Repeated exposure to crisis content with no way to intervene, is what drives numbness.
Repeated exposure to crisis content with no way to intervene, is what drives numbness.
Doom scrolling trains automatic scrolling reflexes, and so crisis content gets scrolled past easier.
Doom scrolling trains automatic scrolling reflexes, and so crisis content gets scrolled past easier.
HYPOTHESIS
Games are a stronger medium for empathy because players become active participants – they experience as the character, making choices that have consequences.
HYPOTHESIS
Games are a stronger medium for empathy because players become active participants – they experience as the character, making choices that have consequences.
…So what character should the player embody?
…So what character should the player embody?
TRACK 01 · CHARACTER DESIGN
Human vs. Animal
We contemplated the suitability of human versus animal characters.
The Human Approach – an immediate no, as it's politically loaded. The goal is to raise empathy. In such matters, it's inevitable that biases cloud judgement. Players experience the game with prejudice already set in mind.
The Animal Approach – Many games use personified animals to explore morals, e.g., Undertale, Night in the Woods. Animals are also inherently adorable, providing shortcuts to human empathy.
So, we opted for the latter approach.
Now, which animal for which character?
2 points were considered: 1) can I actually draw the animal?, and 2) does the match-making make sense? The protagonist in particular took multiple rounds of trial and error to land (ft. my Google search history).
|
Sea otters were a compatible representation. Their "raft dynamic": hand-holding to prevent separation, doubled as an empathy hook. When the protagonist's family is forcibly driven apart, the impact hits like a gut punch because this separation goes against their very biology.
Cast in tow.
Graphic Art vs. Pixel Art
That said, we lacked a clear visual direction. Graphic art had a distinct flair, but with 2 artists, inconsistency in style would break immersion, and a broken narrative fails to evoke empathy. Pixel art, however, offered some level of standardization. So, we pivoted.
From Graphic Art to Pixel Art
How does one with zero experience go about making 96 sprites?
First, identifying the tool.
Second, translating from vector to pixel.
For me, this process was like trying to parse sign language – a hard-to-grasp mind-shift. After floundering on my own, I turned to AI.
Third, refining the animation cycle.
The output provided an excellent reference. While, the idle poses were relatively straightforward, the left-right walk cycle was tricky to conceptualize, and required multiple iterations.
And finally, the completed sprites.
TRACK 01 · SIDE B
Defining Voice
With sprites completed, voice design was next. I assessed different approaches to voice SFX, in terms of feasibility.
After some deliberation, I chose the third approach.
Abstract speech was largely inspired by Inscryption (Daniel Mullins Games, 2021), which uses synth blips, and modified audio to convey unique personality. P03 (Inscryption's primary antagonist), in particular, expressed an authentic range of emotion – proof that the approach, whilst limited, wasn't a "lesser" option.
First step, sourcing sounds in Logic Pro.
I searched sample libraries for high-potential sounds, and collated 8 samples (half of which are shown below).
Then came mapping audio to corresponding characters.
C4-C5 Range
C3 Range
C2-C3 Range
C6 Range
Speech into Sound Translation
This process was like translating vector to pixels, and was quite a process to wrap my head around. Whilst the approach lacked language to express emotion, rhythm, pitch, and volume still remained. So, rather than trying to convert every spoken syllable into sound, I focused on these variables to strengthen emotional weight.
Neas
"Oh no! My journal!"
Kairo
"Neas… Stop stalling by the window. It won’t make them come faster."
Tusker
"It ain’t opening. No one is in the house you said?"
Orion
"I saw a little’un leave moments ago. Parents are gone, little’un gone… It should be empty."
Arc 1 dialogue was completed but how do you scale for the remaining 60 lines?
To avoid repetition, I focused on efficient iteration. I reused existing audio assets where appropriate, and whenever a scene called for a specific emotion, I simply adjusted the source MIDI notes, re-pitching and shuffling keys into new variations.
Neas
"… Okay, yeah. Because we’re telling the truth right?"
Kairo
"Even still, it’s too dangerous… We should… We need to think things through."
Neas
"? But they took away daddy… And daddy is not a bad person. He’s good!"
Kairo
"It means only a short while! That must mean dad’s on his way back!"
TRACK 02 · LEVEL DESIGN
Layout Design for Arc 1's Stealth Sequence.
With the cast finalized, I moved onto map design, where Arc 1 needed renovation.
This arc hinged on a stealth sequence and so layout architecture needed to be prioritized. The arrangement of furniture had to accommodate real hiding spots. It was also important that the house felt lived-in, to make Domicile believable and immersive – especially since that's what makes emotions land.
I took my teammate's initial layout and assessed it against this requirement, among others.
After the second floor map redesign, I realized the first floor map hadn't been delivered, so I mocked up a rough iteration using current and archived assets to clearly communicate the design.
The final version took multiple iterations. One recurring issue was tile-mapping, where extra tiles left too much open floor. I coordinated with the dev to refine room dimensions and finalize minute placement details. The map became populated enough to facilitate fluid play, with a better lived-in quality, given the available assets.
Layout Design For Arc 2's Exploration Mechanic
Layout of Arc 2's map was also crucial, since this arc's core mechanic hinged on exploring an already raided neighbor's house – where players discover Domicile's worldbuilding.
To begin, I blocked out the space with rough shapes to test placement. I scoped the neighbor's house to a single-floor layout. A two-story layout, like Arc 1, would have diluted the already limited interactable objects, leaving lots of dead space.
Object Distribution
I then planned out the distribution of interactable objects in a way that encourages players to walk around and interact with the entire map. Content-heavy items (highlighted in orange) were deliberately spread out to one per room to avoid burdening players with extensive text and ruining pace.
Randomly angled furniture would've sold the post-raid narrative but it was too geometrically complex for a pixel, top-down view. So, for the final layout, I changed to weathered textures and grid-aligned assets that looked knocked over, looted or broken. Even so, the layout still read as too organized to be at a post-raided state.
TRACK 03 · INTERACTION DESIGN
Layout design handled effective player navigation, next was refining interaction design at multiple levels.
The prologue, being the first point of interaction, demanded intuitive design above all, as it sets expectations for the experience.
Competing Feedback Loops
The biggest problem was a lack of event sequencing. We had initially assumed players would read before acting but instead, they immediately started trying to interact, triggering multiple events to happen at once.
Why the Prologue matters.
The Prologue's journal entries show the innocence and normalcy of the protagonist's day to day before war. When this routine life is shattered by violence, it evokes empathy because a childhood is interrupted by something abnormal.
Introducing Linear Event Sequencing
Since the prologue's emotional weight depends on players reading the entries, I gated player input to ensure that the text finishes before the interaction cue is prompted. The new sequence: Left Text Animation → Prompt → Drag & Drop → Right Text Animation → Continue.
Next was refining Arc 1's stealth UI cues.
Polishing the Visuals
In Arc 1, players must escape without getting caught by raiders, who detect players two ways – sight and sound. The problem was with the visuals, rather than interaction: the presence of sound rings, off-centered vision cones, and z-index bugs. It made the system look unpolished and like it was using placeholders.
By benchmarking existing games that use the same mechanism (e.g., Invisible, Inc., Stealth Master: Assassin Ninja), I removed the sound ring – the detection mechanic still intact, and applied a transparent gradient to the vision cone, which read better than a block color. I also replaced the static player icon with an animated marker so players could instantly identify their own position.
Last up, Arc 2's exploration UI cues.
The previous iteration had a couple issues. First, the magnifying glass icon that appeared near interactable objects worked, but didn't tell players what to press. Second, our dev disabled mouse and trackpad input to simplify interactions. But players, who were used to games that supported both input types, found themselves gravitating towards a disabled control, almost like a phantom limb.
As a result, we reinstated the mouse and trackpad input, and consolidated the proximity prompt to an "F" key prompt to avoid any possible confusion.
TRACK 04 · TAKEAWAY
64% of players reported a heightened awareness of the topic – achieving meaningful progress toward the original goal.
A project this nuanced meant a lot of iterations. I had expected to nail each decision on the go but later learned to embrace the process.
Building a game in 8 months, starting as an amateur in almost every discipline, was a whirlwind, and gave me more perspective on my range as a designer than any project could have.
Want to know more? Let's chat!
ALBUM
5 tracks, 17 min
Role
Project Lead · Design
Team
1 Project Lead · 1 Artist · 1 Dev
Made with
Illustrator · Photoshop · AfterEffects · Logic Pro
Domicile – a narrative game about two siblings navigating sudden violence, in search of parents who never returned. How do they make sense of a world that changed overnight?
TRACK 00 · OVERVIEW
The Problem
With current affairs, news media is increasingly saturated with conflict, injustice, and suffering – leading to widespread compassion fatigue.
How might we design experiences that counter compassion fatigue to foster empathy?
To answer this question, we need to understand where existing formats fall short.
The Gap
News use statistics to show casualties, but we feel less empathetic for numbers.
Repeated exposure to crisis content with no way to intervene, is what drives numbness.
Doom scrolling trains automatic scrolling reflexes, and so crisis content gets scrolled past easier.
HYPOTHESIS
Games are a stronger medium for empathy because players become active participants – they experience as the character, making choices that have consequences.
…So what character should the player embody?
TRACK 01 · CHARACTER DESIGN
Human vs. Animal
We contemplated the suitability of human versus animal characters.
The Human Approach
An immediate no, as it's politically loaded. The goal is to raise empathy. In such matters, it's inevitable that biases cloud judgement. Players experience the game with prejudice already set in mind.
The Animal Approach
Many games use personified animals to explore morals, e.g., Undertale, Night in the Woods. Animals are also inherently adorable, providing shortcuts to human empathy.
So, we opted for the latter approach.
Now, which animal for which character?
2 points were considered: 1) can I actually draw the animal?, and 2) does the match-making make sense? The protagonist in particular took multiple rounds of trial and error to land.
|
Sea otters were a compatible representation. Their "raft dynamic": hand-holding to prevent separation, doubled as an empathy hook. When the protagonist's family is forcibly driven apart, the impact hits like a gut punch because this separation goes against their very biology.
Cast in tow.
Graphic Art vs. Pixel Art
That said, we lacked a clear visual direction. Graphic art had a distinct flair, but with 2 artists, inconsistency in style would break immersion, and a broken narrative fails to evoke empathy. Pixel art, however, offered some level of standardization. So, we pivoted.
From Graphic Art to Pixel Art
How does one with zero experience go about making 96 sprites?
First, identifying the tool.
Second, translating from vector to pixel.
For me, this process was like trying to parse sign language – a hard-to-grasp mind-shift. After floundering on my own, I turned to AI.
Third, refining the animation cycle.
The output provided an excellent reference. While, the idle poses were relatively straightforward, the left-right walk cycle was tricky to conceptualize, and required multiple iterations.
And finally, the completed sprites.
TRACK 01 · SIDE B
Defining Voice
With sprites completed, voice design was next. I assessed different approaches to voice SFX, in terms of feasibility.
After some deliberation, I chose the third approach.
Abstract speech was largely inspired by Inscryption (Daniel Mullins Games, 2021), which uses synth blips, and modified audio to convey unique personality. P03 (Inscryption's primary antagonist), in particular, expressed an authentic range of emotion – proof that the approach, whilst limited, wasn't a "lesser" option.
First step, sourcing sounds in Logic Pro.
I searched sample libraries for high-potential sounds, and collated 8 samples (half of which are shown below).
Then came mapping audio to corresponding characters.
C4-C5 Range
C3 Range
C2-C3 Range
C6 Range
Speech into Sound Translation
This process was like translating vector to pixels, and was quite a process to wrap my head around. Whilst the approach lacked language to express emotion, rhythm, pitch, and volume still remained. So, rather than trying to convert every spoken syllable into sound, I focused on these variables to strengthen emotional weight.
Neas
"Oh no! My journal!"
Kairo
"Neas… Stop stalling by the window. It won’t make them come faster."
Tusker
"It ain’t opening. No one is in the house you said?"
Orion
"I saw a little’un leave moments ago. Parents are gone, little’un gone… It should be empty."
Arc 1 dialogue was completed but how do you scale for the remaining 60 lines?
To avoid repetition, I focused on efficient iteration. I reused existing audio assets where appropriate, and whenever a scene called for a specific emotion, I simply adjusted the source MIDI notes, re-pitching and shuffling keys into new variations.
Neas
"… Okay, yeah. Because we’re telling the truth right?"
Kairo
"Even still, it’s too dangerous… We should… We need to think things through."
Neas
"? But they took away daddy… And daddy is not a bad person. He’s good!"
Kairo
"It means only a short while! That must mean dad’s on his way back!"
TRACK 02 · LEVEL DESIGN
Arc 1 – Stealth
With the cast finalized, I moved onto map design, where Arc 1 needed rennovation.
This arc hinged on a stealth sequence and so layout architecture needed to be prioritized. The arrangement of furniture had to accommodate real hiding spots. It was also important that the house felt lived-in, to make Domicile believable and immersive – especially since that's what makes emotions land.
I took my teammate's initial layout and assessed it against this requirement, among others.
After the second floor map redesign, I realized the first floor map hadn't been delivered, so I mocked up a rough iteration using current and archived assets to clearly communicate the design.
The final version took multiple iterations. One recurring issue was tile-mapping, where extra tiles left too much open floor. I coordinated with the dev to refine room dimensions and finalize minute placement details. The map became populated enough to facilitate fluid play, with a better lived-in quality, given the available assets.
Layout Design For Arc 2's Exploration Mechanic
Layout of Arc 2's map was also crucial, since this arc's core mechanic hinged on exploring an already raided neighbor's house – where players discover Domicile's worldbuilding.
To begin, I blocked out the space with rough shapes to test placement. I scoped the neighbor's house to a single-floor layout. A two-story layout, like Arc 1, would have diluted the already limited interactable objects, leaving lots of dead space.
Object Distribution
I then planned out the distribution of interactable objects in a way that encourages players to walk around and interact with the entire map. Content-heavy items (highlighted in orange) were deliberately spread out to one per room to avoid burdening players with extensive text and ruining pace.
Randomly angled furniture would've sold the post-raid narrative but it was too geometrically complex for a pixel, top-down view. So, for the final layout, I changed to weathered textures and grid-aligned assets that looked knocked over, looted or broken. Even so, the layout still read as too organized to be at a post-raided state.
TRACK 03 · INTERACTION DESIGN
Layout design handled effective player navigation, next was refining interaction design at multiple levels.
The prologue, being the first point of interaction, demanded intuitive design above all, as it sets expectations for the experience.
Competing Feedback Loops
The biggest problem was a lack of event sequencing. We had initially assumed players would read before acting but instead, they immediately started trying to interact, triggering multiple events to happen at once.
Why the Prologue matters.
The Prologue's journal entries show the innocence and normalcy of the protagonist's day to day before war. When this routine life is shattered by violence, it evokes empathy because a childhood is interrupted by something abnormal.
Introducing Linear Event Sequencing
Since the prologue's emotional weight depends on players reading the entries, I gated player input to ensure that the text finishes before the interaction cue is prompted. The new sequence: Left Text Animation → Prompt → Drag & Drop → Right Text Animation → Continue.
Next was refining Arc 1's stealth UI cues.
Polishing the Visuals
In Arc 1, players must escape without getting caught by raiders, who detect players two ways – sight and sound. The problem was with the visuals, rather than interaction: the presence of sound rings, off-centered vision cones, and z-index bugs. It made the system look unpolished and like it was using placeholders.
By benchmarking existing games that use the same mechanism (e.g., Invisible, Inc., Stealth Master: Assassin Ninja), I removed the sound ring – the detection mechanic still intact, and applied a transparent gradient to the vision cone, which read better than a block color. I also replaced the static player icon with an animated marker so players could instantly identify their own position.
Last up, Arc 2's exploration UI cues.
The previous iteration had a couple issues. First, the magnifying glass icon that appeared near interactable objects worked, but didn't tell players what to press. Second, our dev disabled mouse and trackpad input to simplify interactions. But players, who were used to games that supported both input types, found themselves gravitating towards a disabled control, almost like a phantom limb.
As a result, we reinstated the mouse and trackpad input, and consolidated the proximity prompt to an "F" key prompt to avoid any possible confusion.
TRACK 04 · TAKEAWAY
64% of players reported a heightened awareness of the topic – achieving meaningful progress toward the original goal.
A project this nuanced meant a lot of iterations. I had expected to nail each decision on the go but later learned to embrace the process.
Building a game in 8 months, starting as an amateur in almost every discipline, was a whirlwind, and gave me more perspective on my range as a designer than any project could have.
Want to know more? Let's chat!