Blind Spot Mapper

What is the blind spot?

The blind spot is the part of each eye’s visual field that falls on the optic disc, the place at the back of the eye where the optic nerve leaves it. There are no rods or cones on the disc, so light that lands there is not seen. In each eye it lies about 15° towards the temple from where you are looking and 1.5° below, and it is about 5.5° wide and 7.5° tall, roughly three thumb widths across at arm’s length. Edme Mariotte reported it in 1668. He looked with his right eye at a small round of paper on a dark wall, and as he stepped back a second one, to its right and a little lower, vanished, as the dot on the card at the top of this page does.

The card works straight away. Close your left eye, look at the cross with your right, and let the dot drift: at the right distance it vanishes as it passes through your blind spot and comes back as it leaves. The rest of the tool turns that vanishing into a measurement: where your blind spot is, how big it is in degrees, and what it is on your retina.

Setting up in one minute

1. Screen size. A pixel on a screen has no fixed size: the nominal 96 to the inch is 3.78 to the millimetre, while a 13-inch laptop at its usual setting shows about 5.0 and a phone about 6.0. So the tool asks you to hold any bank, ID or loyalty card against the screen and match an outline to it. Every such card is an ID-1 card under ISO/IEC 7810, 85.60 by 53.98 mm, so the match tells the page how many pixels make a millimetre. A ruler works too. Your browser remembers the answer for this screen and nothing else. Zooming the page changes the size of its pixels, and not every browser tells the page, so if you zoom, calibrate again.
2. Distance. The tool suggests 50 cm, about 5.8 lengths of a card’s long side from your eye to the screen, whenever the screen is wide enough, and otherwise the furthest that fits. Tell it the distance you actually use. With no tape measure, sit at about that distance and switch on No tape measure: the tool then works the distance out from where it finds your blind spot.
3. Eye. Choose the eye to test and close the other. The cross sits near one edge so the dot has the rest of the screen on the side of your temple.

The test then fills the screen and takes about two minutes. Press Space, or tap anywhere, the moment the dot appears, vanishes or comes back. On a phone it needs the phone held sideways: at the nearest distance the test allows, 20 cm, it needs 11.1 cm of screen across.

Worked example: where the blind spot falls at 50 cm

An angle in the visual field becomes a distance on the screen through one right-angled triangle, the eye at one corner, the cross at another and the point at the third. A point x mm from the cross, seen from d mm away, is θ = atan(x/d) from the cross, so the point at an angle θ is x = d tan θ from it. The unit circle explorer shows where the tangent comes from.

• The centre of a typical blind spot, 15° out, is 500 × tan 15° = 134.0 mm from the cross. The test screen puts its dot there for real: with the screen calibrated it is drawn 134.0 mm from the cross, whatever the screen.
• A spot 5.5° wide centred there runs from 12.25° to 17.75°, which is from 500 × tan 12.25° = 108.6 mm to 500 × tan 17.75° = 160.1 mm: 51.5 mm wide.
• Its centre is 1.5° below the cross, which on a flat screen is tan 1.5° × √(500² + 134.0²) = 13.6 mm down. The point is 517.6 mm from the eye rather than 500, so the vertical uses that distance.
• The test reaches 22° out, the centre plus half the width plus a margin, so at 50 cm it needs 500 × tan 22° + 30 = 232.0 mm of screen, with 15 mm spare at each side.
• Its dot is half a degree across, 2 × 500 × tan 0.25° = 4.36 mm, and the cross a degree.

A 286 mm laptop screen therefore allows the test at up to 634 mm, so 50 cm fits comfortably, while a 140 mm phone held sideways allows 272 mm, and the tool asks for 27 cm.

Where the blind spot falls at each distance

The table gives a typical blind spot’s place on a flat screen, from the model the test places its dot with, and the screen width the test needs at each distance.

Every length grows in proportion to the distance, which is why the card works at any size once you find the right distance, and why a distance that is 10 percent out puts every measured angle about 10 percent out.

How the test maps it

The test finds the edges of your blind spot the way kinetic perimetry finds the edge of any scotoma: by moving a dot from inside it outwards until it is seen. It runs in this order.

1. Reaction time. Five times, a dot appears and you press. Their median is your reaction time, typically about 0.3 s.
2. Across. The dot runs out along a line 1.5° below the cross, from 8° to 24° at 2° a second, and you press as it vanishes and again as it comes back.
3. Down. It runs down through the middle of that stretch, and you press as it vanishes and returns. The middles of the two passes, each pulled back by speed times reaction time, give the centre.
4. Out from the centre. Twelve runs start at the centre, where you cannot see the dot, and move out at 1° a second, one every 30°, in a shuffled order. You press the moment it appears.

Each press comes one reaction time after the dot crossed the edge, so every run overshoots by about 1°/s × 0.3 s = 0.3°, and an uncorrected map is too big by that much on every side. The tool pulls each edge back to where the dot was one median reaction time before your press. A direct least-squares ellipse (Fitzgibbon, Pilu and Fisher, 1999) through the twelve corrected points gives the centre, width and height, and the area is that of the polygon through them.

Five times, a dot flashes for a fifth of a second at the centre of your blind spot, as Heijl and Krakau proposed in 1975. If you are still looking at the cross you cannot see it, so a press counts as a fixation loss, and more than 20 percent losses, two of the five, marks the map unreliable, the criterion automated perimeters use. Two blank trials, with no dot at all, catch presses made by expectation. Nothing is shown until the end, so no point found can hint at where the next one will be.

Why you never notice it

With both eyes open, each eye covers the other’s gap. The right eye’s blind spot lies to the right of where you look and the left eye’s to the left, so no point in front of you falls in both. Switch on Both eyes and the field chart draws the two together, and the eye diagram follows one point into each eye: in one it lands on the disc, in the other on working retina.

With one eye closed you still do not see a hole, because the brain fills the gap from what surrounds it. Switch on Filling in and the card’s dot sits on stripes: where the dot vanishes, the stripes look unbroken, though the screen has a solid dot there. The blind spot is also 15° out, where vision is coarse, and the eyes never stop moving, so nothing stays in it for long.

From the field to the retina

The eye’s optics turn the image over, top to bottom and left to right, as a lens does in the ray diagram simulator: light from the temporal field crosses at the eye’s nodal point and lands on the nasal half of the retina, and light from below fixation lands above the fovea. So the blind spot, 15° towards the temple and 1.5° down, is the optic disc on the nasal side of the retina, a little above the fovea. Panel 2 draws that crossing from above, and the thin lens calculator gives the same inversion as a negative magnification.

Panel 3 shows the retina as an ophthalmoscope sees it, from the front, which is how fundus photographs are printed. Looking in from the front undoes the left to right reversal, so the disc appears on the same side as the blind spot, to the right of the fovea in a right eye, but still upside down: the top of your blind spot is the bottom of the outline drawn on the disc. Near the back of the eye one degree of field covers about 0.29 mm of retina, because the nodal point, where the light crosses, is about 16.7 mm in front of the retina, as in Emsley’s reduced eye: 16.7 × π/180 = 0.29 mm. So a 5.5 by 7.5° blind spot is about 1.6 by 2.2 mm, beside a typical disc of 1.8 by 1.9 mm (Jonas, Gusek and Naumann, 1988). The nerve fibres that leave through the disc become the optic nerve, which the brain explorer follows to the chiasm, and the visual field defects explorer shows the blind spot sitting in every field chart as the pathway behind it is cut. The visual field defect charts draw it in each eye beside every classic defect.

What your result can and cannot tell you

The readouts compare your map with typical values: a centre about 15.5 ± 1.1° from fixation and 1.5 ± 0.9° below it (Rohrschneider, 2004), and a size of about 5.5 by 7.5°. They say whether your centre lies inside the typical range, within two standard deviations of those means, and never more than that. A result far outside it usually means the distance or the calibration was off.

To check the whole chain, map the same eye at 40 cm and again at 60 cm. Angles do not depend on distance, so the two maps should agree within about a degree. If they do not, one of the distances or the calibration was wrong.

This is a demonstration, not an eye test. An enlarged blind spot is one of the things a doctor looks for in some conditions of the optic nerve and the retina, but a screen test cannot tell a real enlargement from a distance that was out or a head that moved, so it cannot be used to look for them. If your vision changes suddenly, or you notice new blind areas, get urgent eye care.

What this model leaves out

• A fixed head. A perimeter holds the head in a chinrest. Here the distance is your word and your head is free to move, and moving 2 cm closer at 50 cm makes every angle 4 percent larger than the tool takes it to be, so the map comes out 4 percent small.
• Light levels. A perimeter controls the brightness of the dot and the background. A screen does not, and a bigger or brighter dot finds a slightly smaller blind spot, because it is seen as soon as any part of it leaves the disc.
• The blood vessels. The large vessels that leave the disc above and below cast shadows of their own, angioscotomas, which a sensitive test finds as narrow extensions of the blind spot. Twelve runs are too few to catch them.
• The eye’s length. The 0.29 mm a degree holds for an eye of average length, and model eyes differ: Drasdo and Fowler’s wide-angle schematic eye (1974) gives about 0.27 mm. A long, short-sighted eye has more retina in each degree, so its disc is smaller in degrees than its size in millimetres suggests.
• Diagnosis. The tool gives no score and no verdict. It measures, and says how the measurement compares with typical values.

Common mistakes

• Closing the wrong eye. To map the right eye, close the left. The blind spot of the open eye is on the same side as that eye, to the right of the cross for the right eye.
• Following the dot. Looking at the dot moves the blind spot with your gaze, so the dot never vanishes. Keep looking at the cross, and let the dot come and go at the edge of your vision.
• Guessing the distance. Every angle scales with it, so a guess of 50 cm at 40 cm makes the blind spot look about a fifth too small and too close to the cross. Measure it, with a tape or with card lengths.
• Skipping the card. Press Done without a card and the tool keeps the outline’s starting size, which is only a guess from the screen’s size in pixels. A 24-inch monitor at 1920 by 1080 shows 3.6 pixels a millimetre and starts at 4.2, so every dot would be drawn about a sixth further from the cross than intended, and the map would come out about an eighth too small and too close to the cross.