Vision
Bloomalizer today is a seed. It listens to music and turns 131,072 particles into shapes that move with it. That works, and it is beautiful, and it is roughly one percent of the idea. This page is about the other ninety-nine — written early, on purpose, so the direction is visible before it is finished.
The turn inward
Almost every music visualizer points outward. It is made for a room: a club wall, a festival screen, a stream overlay, a crowd to impress. The measure of success is how much attention it grabs.
This one is at its best pointing the other way — alone, in the dark, with headphones on. Same technology, opposite direction. Not a thing to show people, but a thing to sit inside.
Everything below follows from that one decision.
Music has a feeling, not only a volume
Right now the visuals react to loudness. Something gets louder, something gets bigger. That is what nearly all visualizers do, and it is why they all eventually feel the same: a graphic equaliser wearing different clothes.
Loudness is the least interesting thing about a piece of music. A slow minor chord and a hard techno kick can hit the same meter and mean completely opposite things.
The direction is for the visuals to respond to character rather than amplitude — to how a track feels, and how that feeling moves. Psychology has a useful map here: place any emotional state on two axes, one running calm to activated, the other unpleasant to pleasant. Almost everything you might say about a piece of music lands somewhere on that plane. A genre is a region of it. A mood is a point in it. A good set is a path across it.
The ambition is that you no longer set a dial. The music sets it, and the visuals travel with you — drifting when the track drifts, tightening when it tightens, arriving somewhere by the end that you did not choose at the start.
Colour drawn from the sound
The palettes here were chosen by hand. Thirteen of them, each a deliberate combination, and the visuals borrow one at a time.
The better version is that they are not chosen at all. There is real research on this: people asked to match colours to music agree far more than chance, and the agreement runs through emotion rather than convention. Fast and major reaches for saturated and light. Slow and minor reaches for dark and blue. Nobody teaches you this; it is simply how the two senses talk to each other.
So the colours should come from the music — from its brightness and its weight, from the record sleeve when the track is known, from what is actually playing rather than from a list somebody wrote in advance.
Worlds, not wallpaper
Behind the particles there are backdrops: aurora, rain, fire, a forest treeline. They are good, and they are all flat. Each one is a surface painted behind the subject, with depth suggested rather than present. You are looking at a picture of a place.
The intention is places you are inside. Space with real depth, that the camera moves through rather than in front of — light that falls through it, distance you can feel, somewhere you could turn your head. Fewer of them, each one much further in. A cathedral rather than a poster of a cathedral.
The old geometry
In 1926 Heinrich Klüver catalogued the shapes people report seeing in altered states and behind closed eyes, and found they collapse into four families: lattices and honeycombs, cobwebs, tunnels and funnels, spirals. It did not matter much what caused the state. The same four kept appearing.
Decades later these were explained as something stranger and more beautiful: they are the visual cortex's own wiring made visible. Left without input, the system falls into its own resonances, and you see the architecture of your own seeing.
Three of those four families are already here — the tunnel, the spirals of the galaxy and the helix, the self-similar lattice of the fractal scenes. They were not built from the research. They were built because they looked right, and only afterwards turned out to be the shapes the brain makes on its own.
The fourth, the cobweb, does not exist yet. It is the clearest single thing left to build.
Breath instead of beat
The energy control runs from calm to intense. Push it all the way down and the scene slows, dims, and lets go — long crossfades, rare changes, almost stillness. That end of the dial is the most interesting one, and it is barely explored.
Follow it far enough and the music stops being the driver at all. Motion paced by breathing rather than by a kick drum: expansion on the inhale, settling on the exhale, a rhythm you set with your body instead of one arriving from outside. Silence treated as an input rather than an absence — the visuals already keep drifting when nothing is playing, which is closer to the point than it sounds.
From there: sessions with a length and a shape. A voice, if you want one. Something to rest attention on that is neither a candle nor a screen full of notifications — moving enough to hold you, slow enough to let go of.
This is the direction that matters most. A visualizer is a pleasant thing. A reliable surface for sitting with yourself is a useful one.
Journeys, not shuffles
Scenes currently rotate. A timer runs out and something else arrives, chosen with a weighted roll of the dice. It is varied, and it has no memory — the twentieth minute knows nothing about the first.
Sessions should have an arc. A beginning that settles you, a middle that goes somewhere, an end that returns. Twenty minutes with a shape to it, the way a good set or a good piece of music has a shape, so that finishing feels like arriving rather than like stopping.
What this will never do
Some of this only works if the thing can be trusted, so a few of these are permanent.
No strobing. Not at any setting, not for any effect. High-contrast flashing is a genuine seizure risk for people with photosensitive epilepsy, and every transition here has a floor on how fast it is allowed to happen. Percussion is expressed by pulses that decay, never by flashes. This will not change.
Your audio stays yours. Sound is analysed on your own machine and is not uploaded. If that ever needs an exception, it will be something you switch on deliberately, off by default, and stated plainly — as track identification already is. See the privacy page.
No account, no ads, no tracking, no cost. Nothing to sign up for and nothing following you around. Something meant for stillness cannot also be harvesting you.
Where it actually is today
To be clear, because the rest of this page is about the future: mood and genre awareness, colour derived from the music, three-dimensional worlds, breath pacing and session arcs do not exist yet. None of it is built.
What exists is real and worth your time — eleven shapes, eleven backdrops, thirteen palettes, twenty-five scenes, reacting live to whatever you play, and able to name the track while it does. That is on the about page, and it is all honest.
This page is the compass, not the map. It is here early so the direction is legible while there is still time for it to change.