Three dots. One retina.
A very precise brain.
This test asks one simple question: does the middle dot sit on the same line as the other two? Your retina receives the dots; your brain compares their positions.
The eye’s
light sensor
The retina is a thin, light-sensitive layer at the back of the eye. Its central area—the macula—provides the sharp vision used for reading and recognising faces.
At the centre of the macula, tiny cone cells sit close together. They sample the light pattern rather like a camera sensor, but a screen pixel and a retinal cell are not the same thing.
Schematic—not anatomical scale
Your brain joins
the dots
You are not trying to identify a tiny dot. You are comparing where three dots sit relative to one another.
The brain combines information from many retinal cells. That pooled signal can reveal an alignment change finer than ordinary letter-chart vision. This unusually precise position sense is called hyperacuity.
Why retinal change
can bend a straight pattern
If the macula changes shape, light from a straight row may land on a shifted pattern of retinal cells. The brain can then perceive a bend, bump, or wave even though the object itself is straight. This symptom is called metamorphopsia.
A straight row is usually perceived as straight.
A straight row may appear bent or displaced.
Distortion can occur with macular conditions such as age-related macular degeneration, macular pucker, or a macular hole. But difficulty on this app can also come from blur, attention, calibration, or the screen.
This app demonstrates the principle. It does not map the retina, test each eye separately, or diagnose retinal disease.Distance changes
what reaches the eye
A nearby pixel looks larger than the same pixel viewed farther away. The app therefore asks for your device type and approximate viewing distance.
Typical pixel densities are used for phones, tablets, and computers. The result is deliberately labelled as an estimate because actual screens and browser rendering vary.
It quietly finds your
screen-relative limit.
Easy correct answers quickly reduce large offsets and move toward your likely discrimination range.
Near that range, two correct answers make one small step finer; one error makes one small step easier.
Later turning points in the fine staircase are averaged. Aligned trials check response reliability. The result remains descriptive, not diagnostic.
Sources & further reading
This experience demonstrates psychophysics. It is not validated as a clinical instrument and cannot diagnose retinal, neurological, or eye disease.