AFTER DARK / FOLLOW THE LIGHT
Free explorer

CHOOSE A RHYTHM

REAL TESS DATA
A SINGLE POINT OF LIGHT

β Doradus

+2.601% from its median

Hover or focus a difference label to trace its cycle. Click, Enter or Space holds that cycle; activate it again to release the hold. Hover can preview another cycle. The Following cycle indicator marks the held cycle. Use its A, B and C buttons to switch the held cycle without changing measurements. Unavailable measured comparisons are disabled. Show all cycles or Escape releases highlighting and keeps your comparison. Tab moves between labels. Highlighting never changes measurements or observation time.

Hover or tap any separate panel or the stacked chart to choose a cycle position. Left and right arrows move the guide within the current view; Shift moves farther. Home and End select the visible cycle-position edges. The same selected observations stay highlighted as you scroll into the stacked chart. In the stacked chart, Pin holds the current position as a reference while you explore another. Differences are current minus pinned in percentage points, for the same cycle. Unpin removes only the reference. When a cycle is being traced, Escape first shows all cycles while keeping the comparison; otherwise Clear or Escape removes both. Readouts use the nearest displayed observation and its reported one-sigma uncertainty, without interpolation. Missing observations stay empty. This does not change the observation clock.

ILLUSTRATIVE MAGNIFICATION
Size, colour, depth & texture are illustrative.
01 / THE LIGHTGlow contrast amplified
Scroll changes the view, not observation time.—

01 THE LIGHT

The night
has a pulse.

A light that rises and falls.

Scroll to unfold its light

02 THE TIMELINE

Give the light a timeline.

139days of observations

Zero is the median brightness. Positive values are brighter; negative values are dimmer.

03 THE CYCLES

First, meet three cycles.

9.843days per cycle

A, B and C are example windows from different times. Ranges show days since the first observation. They are chosen for coverage; gaps remain.

Hover or tap a panel to compare matching cycle positions. Values come from nearby observed dots; gaps stay empty. The selection follows the dots as you scroll. Use Clear or Escape to remove it. Explore a new position here or in the stacked chart with hover, tap, or ← →.

04 ALIGN THE CYCLES

Now, bring them together.

Your selected dots keep their colours and measured brightness as they come together. Scroll on to bring back the other observations, then add their summary line.

05 THE PATTERN

The rhythm survives the stacking.

Choose a position, select Pin, then explore another with hover, tap, or ← →. Dashed links join each cycle’s two measured points; their labels show the brightness difference in percentage points. Hover a label to trace one cycle. Click to hold it; the Following cycle indicator confirms the hold. Use A, B or C beside it to follow another cycle. Show all cycles or Escape resets the highlight while keeping your comparison. Unpin keeps the current selection.

Open the free explorer ↑ Back to the light

A full observation window takes 32 seconds at 0.75× speed.

BEHIND THE LIGHT

The same real observations.

Scrolling changes the explanation and arrangement. It does not advance the observation clock. Use replay or the time slider to change the measured light shown by the glowing illustration.

The star’s magnification, surface, colour, position and depth are illustrative. The light-to-chart movement is a presentation transition, not a scientific trajectory. Chart axes appear once the points reach their measured coordinates.

The timeline and final chart display up to 6,000 uniformly index-sampled observations per target; shorter source files show every retained observation. The cycle introduction temporarily isolates three coverage-selected example windows from those same points. Their panels use the same brightness scale; only horizontal positions change as they align. The remaining displayed observations return before the summary line appears. The observation clock can select a measurement outside the three example windows; that point reappears in the full chart. The white line is a median in each of 240 phase bins, calculated from every retained measurement for that target. It is a summary, not a fitted physical model. Original gaps and reported measurement uncertainties are preserved. Different targets use different brightness scales and glow amplification.

The three-panel comparison uses a shared cycle-position guide. Each highlighted dot and brightness readout comes from the nearest displayed observation in that window, with its original reported uncertainty. Dots stay at their actual measured positions; values are not interpolated onto the guide. If the requested time falls inside an original observation gap, that panel shows no measurement. The selected tuple from each cycle is held fixed as the panels merge; scrolling never chooses a different measurement. In the settled stacked chart, hover, tap or arrow keys choose a new cycle position directly; the same three example windows supply its nearest real observations. The zoomed transit view uses its visible phase range for pointer and keyboard input. The shared guides converge, and coloured rings keep the selected dots visible even where they overlap. A selected gap stays empty. Clear or Escape removes the selection; choosing another target resets it. Pin captures the current phase and its three original observation tuples. New pointer or keyboard input changes only the current selection. For each cycle, the displayed difference is 100 × (current relative flux − pinned relative flux), in percentage points of brightness change from the same target median. It is not a percentage relative to the pinned value. Differences use the unrounded source values, before display rounding. If either endpoint has no observation, its difference is unavailable. Dashed coloured links are comparison guides between the two exact measured coordinates, not interpolated observations or a fitted light curve. Thin leaders connect their difference badges to the guide midpoints. Badge placement can move for readability; the measured points stay fixed. A guide and its on-chart badge are omitted if an endpoint is missing or outside the current view; the header readout remains available. Hovering or focusing a difference label traces that cycle’s observations, endpoint markers and connector while the others dim. Clicking the label holds the cycle; activating it again releases the hold. Hover can preview another cycle. Escape first shows all cycles and keeps the comparison. This changes only visual emphasis, never the selected measurements, their coordinates or observation time. No uncertainty for the difference or significance is inferred. Unpin removes only the reference; Clear, Escape and target changes reset both. A pin outside transit zoom remains unchanged and is labelled outside the view. Comparison leaves the separate observation clock unchanged.

The charts use the observations supplied with this experience. Exploring them does not change the source measurements.

CHOOSE A RHYTHM

Different lights. Different rhythms.

Follow a pulsation, an eclipse, or a planet in passing. Each light opens the same journey through its own observations.

Brightness is relative to each target’s median; chart scales and glow amplification differ.