Everything in Particle is black, white, and a single point. We made it that way deliberately.
The work mode is austere on purpose. There is a number, a verb, and a counter. Nothing else competes for attention. When you sit down to make a particle, the screen disappears into the work — which is the point.
But there is a debt that accumulates with rigor. When the work ends, the system has nothing to give back. The screen stays as bare as before. After a thousand sessions, that bareness stops feeling like elegance and starts feeling like flatness. Reduction without contrast is not minimalism. It is barrenness.
This article is about the moment we decided that wasn't acceptable, and the discipline we are committing to so we don't ruin it.
#The decision
We built Afterglow. After a completed work session — and only when the break timer is actively running — the screen blooms into a slow, generative shader. The reductive black-and-white world steps back, and a different visual world steps forward for the duration of the break. When the break ends, the bloom recedes and Particle returns to its monochrome.
The break is the only moment we use color. That isn't a phase or a default we'll iterate on. It's a rule. Color in Particle is reserved for the time you've earned through real work — the time the system has nothing to ask of you.
#Three approaches we considered
Before we settled on this, we sat with three alternatives long enough that each one almost won.
Stay pure. The cleanest version of Particle is pure. No animation, no bloom — work and rest both lived in the same austere palette, and the user supplied any meaning. We gave this idea more time than the other two because it's the safest. It also turns out to be the wrong answer. A reductive system that only reduces, with no moment of release, doesn't read as discipline. It reads as cold. The break becomes a gap in the work; the user fills the gap with their phone, their email, their mind racing about the next thing. Reduction without a return doesn't liberate the rest. It abandons it.
Gamify. Confetti when a session ends. A progress bar that fills, a streak that grows. We've written elsewhere about why we'll never use streak mechanics — they exploit loss aversion to keep users engaged through anxiety, not joy. The same logic rules out unlock-based shaders, badge systems, and "the next preset opens at 50 sessions." These mechanics produce engagement metrics. They don't produce recovery. The break is too important to be a slot machine.
Atmosphere-coded UI. Each user already has a chosen atmosphere — a palette and lighting register that travels through the whole app. We tried using it for the break: tint the shader through the user's accent and glow colors so the experience stays consistent end-to-end. Code-wise it was beautiful. Felt-wise it was wrong. The atmosphere is the room you live in. It should be what you see most of the time. The break, by contrast, should be a different room — one you're invited into only after you've put work in. The atmosphere can't deliver that. It can only deliver consistency.
We chose: a separate visual world, present only during running breaks, defined by the shader's own original palette — not the rest of the app's. The break has its own color identity, and the rule is that you have to do real work to enter it.
#The thesis: color earned
The shader is what we call a color earned element — a piece of visual language that doesn't accumulate background presence, but exists only at narrow, specific moments. Four reasons we believe this works.
It honors the work. The black-and-white work mode isn't a limitation; it's a frame. A frame's meaning depends on what fills it. When the frame fills with austerity for thirty minutes and then fills with a bloom for five, the bloom is because of the austerity. Strip the austerity and the bloom is just decoration. The contrast is the message.
It models recovery. Recovery isn't more stimulation. It's a different mode of attention — the kind that happens when you let your eyes rest on a fire, watch leaves move, or stare at the ocean. Generative motion that you can't predict but doesn't ask anything of you. The brain's recovery systems aren't engaged by demands; they're engaged by being permitted to wander. Afterglow is a generative wander-prompt — a piece of visual world to look at without the obligation to interpret. Closer to what trees do for us than to anything productivity software has tried.
It maps to the research. The default-mode network — the brain's "rest-state" attention pattern — is what produces synthesis, integration, and creative recombination. It engages when external task demands drop. Most digital products fight against this; they fill the moment of release with another notification. Afterglow does the opposite. It says: you've completed a session of work. Now your job is to look at this for five minutes without trying to do anything with it. The research on rest-state attention has been quietly consistent for decades, and knowledge-work culture has quietly ignored it for just as long.
It rejects gamification at the level of grammar. Color in Particle is not a reward. It is a context. You don't earn it through performance metrics — you earn it by completing a real work session. The work itself certifies the moment. There is nothing to optimize. There is no high score. There is only: did you work, or didn't you. If you did, the color appears. When the break ends, it leaves. Nothing to chase, because nothing to lose.
#The promise
This is the part most product articles skip. Saying "we built X" is easy. Committing to a future constraint that prevents X from rotting is harder, and it's the only thing that makes a feature trustworthy in year two and year five.
So: we are not promising that Afterglow will only ever have three shaders. The right number is the number that pass the test, not a number we choose to make ourselves sound restrained.
What we are promising is the test:
No shader joins Afterglow because it looks cool, because it's trending in WebGL demos, or because we have the engineering capacity. A shader joins Afterglow only when we are convinced — by user experience, by our own use, and where applicable by published research — that it produces a recovery effect analogous to time in nature.
In practice that means each candidate shader has to clear four bars:
- Slow enough that the eye doesn't strain. The motion never asks the viewer to track or follow it.
- Colors derived from natural phenomena. Sunsets, distant fire, undergrowth, deep water, dawn light, weather. Not abstract neon, not designer gradients dressed up to look organic.
- Visual relief. Areas of low information that the eye can rest on. Not maximalism. Not a 4K wallpaper.
- Invitation, not demand. It coexists with thinking, daydreaming, looking out the window. It does not pull attention to itself.
Things this rules out:
- A shader gallery with hundreds of presets and no curation.
- Mix-and-match parameter editors so users build their own.
- Shader-based achievements, unlocks, or rotation tied to performance.
- Shaders during work mode, during pause, or during idle. Those states have different design needs and different costs.
The first shader that ships is Trailing the Twinkling Tunnelwisp — a CC0 Shadertoy port that we kept in its original color pipeline because the original colors are what make it restorative. We did not atmosphere-code it to match the user's theme. The shader's earned palette is its own.
The second shader will join when, and only when, it earns its place. The third may not exist for a year. The break is too important to fill quickly.
#What we deliberately did not build
The features we didn't ship are how you can tell what a product actually believes.
- No shader during work mode. Work needs reduction, not painting.
- No shader during pause or idle. Those are interaction states. The user is making a decision. The screen needs to be readable, and the controls need to be the visual focus. The shader appears only when a break is actively running and disappears the moment it pauses.
- No user-uploaded shaders. Curation is the value. A custom-shader feature would optimize for self-expression at the cost of every other promise above.
- No rotation during a single break. A shader transitions in once, holds for the break, and transitions out. A break is not a slideshow.
- No Flow gating on the first shader. Tunnelwisp is free. The contrast between work and break is part of Particle's core — not a paywall feature.
Each of these absences is doing real work in the design.
#The door between two disciplines
The structure underneath Afterglow is simple: reduction without a reward window becomes barrenness. Reward without reduction becomes noise. The two need each other.
Afterglow is the door between them — a door that only opens after you've done the work, and only stays open for as long as you need. Most apps want the door open at all times. They confuse "engagement" with "presence" and let color, motion, and notification flood every surface. The result is that nothing feels meaningful, because nothing is reserved. Every visual gesture has been spent.
Particle is trying to do the opposite. We're keeping our visual gestures rare on purpose, so the few we make can carry weight. The reduction makes the bloom possible. The bloom makes the reduction worth keeping.
The black is the work. The bloom is the rest. Both are the discipline.