The double slit experiment explained with stripes and dots

By Julia Berens · · 5 min read

Hands on a beige keyboard at a vintage computer workstation, with a cyan-and-pink screen and an amber glow beneath it.

Here is the double slit experiment explained as a prediction you can check: if light from two narrow slits travels the same distance to the middle of a screen, it makes a bright stripe there. Move sideways to a spot where one path is half a wavelength longer, and the light can cancel. The same bright and dark bands can emerge when a detector records light one photon at a time. A single detection is a dot, not a stripe.

The double slit experiment explained from the center out

Picture a laser shining straight onto a barrier with two narrow, parallel slits. The slits are vertical and sit side by side. A screen stands behind the barrier. Light spreads sideways after passing through each slit, a process called diffraction. Because both slits are lit by the same laser, their waves can maintain a steady phase relationship. That lets the overlap produce stripes that stay in place rather than shifting around and washing out.

Start at the point on the screen directly opposite the midpoint between the slits. The paths from the two slits to this point are equal in length. In this symmetric setup, a crest from one slit arrives with a crest from the other, and a trough arrives with a trough. The light waves add, producing a bright center. This is constructive interference.

Now move horizontally along the screen. One path gets longer than the other. At a path difference of half a wavelength, the crest of one wave arrives with the trough of the other. The two contributions subtract. If they have equal amplitudes, they cancel completely. This is destructive interference. Repeating the comparison farther out gives alternating bright and dark bands. The bands run vertically because the path difference in this setup changes mainly as you move sideways across the screen.

Predicting where the stripes land

Call the wavelength λ. A path difference of λ makes one wave arrive a full cycle behind the other, so the crests line up again. At 1.5λ, the waves are opposite again. You can predict the next few stripes by counting whole wavelengths for bright bands and half steps between them for dark bands. There is a matching sequence on either side of the bright center.

  • Bright: path differences of 0, λ, 2λ, and so on.
  • Dark: path differences of 0.5λ, 1.5λ, 2.5λ, and so on, with complete cancellation when the amplitudes match.

Near the center, if the screen is far from the slits compared with their separation, the distance between neighboring bright stripes is approximately λL/d. Here L is the distance from the barrier to the screen, and d is the distance between the slit centers. For red light with a wavelength of 650 nanometers, a screen 2 meters away, and a slit separation of 0.50 millimeters, the spacing is about 2.6 millimeters. Moving the screen twice as far away would roughly double that spacing. Doubling the slit separation instead would roughly halve it.

The formula predicts the spacing, not the brightness of every stripe. Real slits have width, so the light from each slit spreads unevenly. As a result, the fine interference stripes sit inside a broader diffraction pattern. Some bright stripes can be much dimmer than the center. OpenStax's treatment of interference shows how the geometry leads to the spacing formula.

One photon makes one dot

Replace the screen with a detector that can record the position of a single photon detection. Send photons toward the slits one at a time, without measuring which slit each photon passes through. The detector registers one localized event at a time. After a few events, the dots look scattered. After many events, some bands have collected lots of dots while others have collected few. Under the same optical conditions, the regions with the most detections correspond to the bright stripes in the wave prediction.

The stripes do not require photons to meet or collide. Quantum mechanics assigns a probability amplitude to each alternative for reaching a particular detector position. When the slit alternatives are indistinguishable, the amplitudes add first. The squared magnitude of that sum gives the probability for a detection there. Where the amplitudes reinforce, detections are more likely. Where they cancel, detections are less likely, or impossible in the ideal case of exact cancellation. The calculation predicts the distribution of many dots, not the exact position of the next one.

Blocking a slit changes the prediction

Imagine recording three sets of dots: one with only the left slit open, one with only the right slit open, and one with both open. Each single-slit run produces a spread of detections because light diffracts through the open slit. If the two-slit result were just a mixture of those runs, adding their count profiles after adjusting to the same incoming exposure would predict the shape of the two-slit pattern. But at a dark fringe, both single-slit runs can register detections while the two-slit run registers almost none. Opening an extra route can reduce the count at a particular spot.

Keeping both slits open while reliably recording which slit each photon passes through also removes the interference stripes. The difference does not depend on a person looking at the record. A physical record that distinguishes the two alternatives is enough. In that case the detection probabilities from the two alternatives add without the interference that produced the dark fringes.

A prediction to test on paper

Mark the center and the first dark stripe on a sketch of the screen. At the dark stripe, label the original path difference λ/2. Now replace the light with light of half the wavelength, keeping the slits and screen fixed. The distances traveled to that mark have not changed. Only the unit you use to count wave cycles has changed: the old half wavelength is one whole wavelength for the new light. That mark now meets the condition for constructive interference. In the central region, the new first dark stripe lies about halfway from the center to your original mark.

For more problems to work through on paper, see GQMC training and practice problems or use the GQMC learning resources to review the physics behind your sketch.

Frequently asked questions

Why is the middle stripe bright?

With straight-on illumination in a symmetric setup, light travels the same distance from each slit to the center. The waves arrive in step and reinforce each other.

Does one photon make a whole stripe?

No. A detected photon produces a single localized event. The stripes appear in the distribution of many events.

Why are real dark stripes not always black?

Complete cancellation requires the two waves to arrive exactly out of step and with equal amplitudes. Unequal illumination or stray light can leave a dark stripe with some brightness.