| Metric | Value | Copy |
|---|---|---|
| {{ row.key }} | {{ row.value }} |
| Priority | Operator cue | Rationale |
|---|---|---|
| {{ cue.priority }} | {{ cue.cue }} | {{ cue.rationale }} |
A Matrix-style screensaver is really a motion illusion. Falling glyph columns, glowing heads, and dark afterimages create the feeling of an endless digital rain, but the effect only holds together when density, speed, and contrast stay in balance. This simulator gives you live control over that balance instead of locking the scene to one preset.
The package is aimed at scene shaping, not passive wallpaper playback. You can choose the palette, glyph set, rain direction, density, speed, glow, flicker, wave motion, noise, and fullscreen preview state, then read the result back through a live canvas, a metrics table, a compact chart of current runtime signals, an operator-brief table, and a JSON snapshot.
That makes it useful for ambient backdrops, demo booths, stream intermissions, coding-themed kiosk loops, or display testing where a static image feels too dead. A restrained setup can sit behind text for a long time, while a heavier scene can push toward a cinematic control-room wall.
The important judgment happens in motion, not in one dramatic still. The preview can pause, reset, and request fullscreen, but the runtime readouts matter just as much as the color choice. A scene that looks impressive for one second can become muddy or stutter-prone when it has to loop continuously.
It is still a synthetic visual effect, not a film asset or a performance benchmark. Column positions are reseeded when the animation resets, delivered frame rate depends on the browser and device, and one advanced control, Scanline strength, currently influences the guidance layer more than the pixels themselves.
Start by deciding whether you want a calm background or a hero scene. The quickest baseline comes from Theme, Character set, Density, and Drop speed alone. If that first pass already feels too busy, no amount of extra glow or flicker will rescue it.
Once the baseline works, shape readability before texture. Character scale, Head brightness, Glow intensity, and Tail persistence decide whether the rain reads as crisp lanes or as a soft wash. Only after that is settled does it make sense to add Random flicker, Background noise, or a stronger Column motion profile.
The most common wrong assumption is that a busier scene is automatically better. In this slug, heavy density, high speed, and strong turbulent sway can push the preview into a less readable state even before the frame rate drops. When that happens, the problem is usually motion pressure, not a lack of visual effects.
Before treating a scene as finished, check the live preview, then read Measured FPS, Quality scale, and the three Operator Brief cues together. That combination tells you whether the scene is actually stable enough for a long loop.
The preview is a browser-side canvas simulation built from column state rather than from prerecorded video. Each column tracks a drop position, a phase value, a drift factor, and a held glyph index from the selected character pool. Density and character scale determine the effective font size, and that font size in turn determines how many columns fit across the current canvas width.
Motion comes from several layers working together. Rain direction flips the vertical travel sign, Drop speed scales how quickly each column advances, and Column motion adds horizontal offset. In Straight mode there is no lateral sway, in Sine wave mode each column oscillates around its lane, and in Turbulent mode the offset combines multiple waves for a rougher track. When a stream crosses the visible edge, Respawn frequency controls how aggressively it is reseeded.
The visual texture settings do not all work in the same way. Glow intensity changes shadow bloom, Head brightness controls the brighter leading glyphs, Tail persistence and Background dim change how much of the previous frame remains visible, Random flicker modulates stream alpha, and Background noise adds low-opacity pixel speckle. Scanline strength is different: it feeds the guidance logic and exported scene state, but the current renderer does not apply it as a separate scanline raster pass on the preview canvas.
Runtime management is as important as the scene settings. The requested cadence is quantized to 30, 45, or 60 FPS, and the page measures delivered frame rate once per second. If actual delivery falls well below the target, the package steps Quality scale down and rebuilds the canvas at a lower effective render density. If delivery recovers, it steps the scale back up. The preview loop also pauses while the document is hidden and resizes itself when fullscreen state changes.
| Control family | Primary effect | What to watch |
|---|---|---|
| Theme and Character set | Change palette and glyph vocabulary. | Visual tone and legibility. |
| Density, Character scale, and Columns | Change how many lanes fit on the canvas and how large each glyph feels. | Readability and runtime load. |
| Drop speed, Rain direction, Column motion, and Wave strength | Set the perceived motion profile of the rain. | Measured FPS and the first Operator Brief cue. |
| Glow intensity, Head brightness, Tail persistence, and Background dim | Shape contrast and trail depth. | Whether lead glyphs stand apart from the tails. |
| Random flicker, Flicker rate, Background noise, and Scanline strength | Feed the noise and guidance layer. | The second and third Operator Brief cues. |
The output surfaces should also be read narrowly. Matrix Metrics is a live snapshot table, Matrix Stream Chart compares a small set of current numeric readings rather than a time history, and Operator Brief is a three-row heuristic summary of motion pressure, noise load, and contrast balance. Those cues are useful because they expose the package's internal scene heuristics, but they are guidance, not hard errors.
This slug does not ship a dedicated rendering backend. The rain effect, charting, fullscreen state, and exports are handled in the browser session after the page assets load, and the current settings are mirrored into the page URL so the same scene can be revisited or shared.
The cleanest workflow is to shape the live preview first and use the metrics only after the scene already feels close.
For fair comparisons, keep one baseline scene and change only one motion or texture family between resets.
The most important numbers are Target FPS, Measured FPS, and Quality scale. Target FPS is only the requested cadence. Measured FPS is what the browser actually delivered over the last window, and Quality scale tells you whether the package already had to soften the render path to protect motion.
The other views should be read as support material, not as proof that the preview is perfect. Matrix Stream Chart is a current-state comparison, not a performance history, and Operator Brief is a rule-based advisory layer. A good-looking scene with poor runtime numbers is still a weak loop, and a clean chart does not guarantee that the preview feels alive at full size.
Set Theme to Classic green, Character set to ASCII letters, Rain direction to Downward, Density to Medium, and Drop speed to about 0.75x. Leave Wave mode on Straight with its default strength, keep Glow intensity near 45, and keep noise layers modest. A healthy run here usually holds Measured FPS near the chosen target, leaves Quality scale close to full, and produces a calm Operator Brief rather than an aggressive warning.
Switch to Violet grid, Katakana style, Upward rain, Ultra density, 2.80x drop speed, Turbulent motion, and Wave strength 85. Add Random flicker, Background noise, Background dim 60, and Scanline strength 70, then target 60 FPS. This is exactly the kind of setup that can flip the first cue to Lock stream vector and the second to Manage visual noise. If Measured FPS drops and Quality scale falls below full, the scene has crossed from dramatic into performance-limited.
Keep a balanced scene, then raise Scanline strength from 10 to 90. Matrix Metrics and the JSON payload will reflect the higher value, and Operator Brief language may shift because the noise load changed, but the preview itself will not suddenly gain a separate CRT line mask. That is a renderer boundary, not a bad input.
The rain animation runs in the browser, and this slug does not ship a dedicated rendering backend. The practical sharing behavior is the URL: current settings are mirrored there so the same scene can be reopened.
Target FPS is the requested cadence. Measured FPS is the delivered cadence over the recent sampling window. Canvas size, density, motion load, and device capability can all pull the real number away from the target.
Reset rebuilds the column field, reseeds motion phases, refreshes noise dots, and clears the recent frame counters. The scene settings stay the same, but the exact rain pattern starts over.
Not always. Fullscreen changes the preview area, which changes Canvas size and can either improve the feel or increase the rendering load. Check Measured FPS after the resize instead of assuming the larger view is automatically better.
Because the current preview renderer does not implement a separate scanline overlay from that control. The value is still carried into metrics, JSON state, and the guidance logic, so it is not ignored, but its effect is advisory rather than a standalone raster layer.