Turning a Photograph Into a Grid of Dots: A Method That Works on the Kitchen Table

Copying a photograph onto a peg grid sounds like a job for software, and there are apps that will do it. The manual method is more useful, takes twenty minutes, and teaches the child something the app hides: which parts of a picture actually carry the likeness.

Start by throwing detail away

Print the photograph small, no larger than a postcard, then squint at it or hold it at arm’s length until the features blur. Whatever survives that treatment is what you can reproduce. Eyelashes, individual leaves and text on a T-shirt do not survive; the dark mass of hair, the pale oval of a face and the line of a shoulder do.

Draw the grid over the print

Rule the print into the same number of squares as your board has holes, using a pencil and a ruler. Count the rows on the board first, because guessing produces a picture that runs off the edge two thirds of the way through. Number the columns along the top so a child can find their place after a break.

Match the grid to the board you own

Different boards have different hole counts and spacings, and a grid drawn for one will not transfer cleanly to another. If you are still choosing, Lite-Brite shows the boards currently in the range and it is worth noting the grid size before you commit to a method built around it. Once you know the count, the same ruled sheet works for every photograph you attempt.

Reduce the colours before you start

Decide on four or five colours and assign each one a job: darkest shadow, mid tone, light, highlight, accent. Anything in the photograph that falls between two of them goes to the nearer one. This is the single decision that separates a readable result from a colourful blur, and it is the part software does silently.

Order of work that avoids disaster

• Place the four extreme corners of the subject first to lock the scale.
• Do the darkest and lightest areas next; they define the shape.
• Fill mid tones last, adjusting as you go rather than following the sketch slavishly.
• Leave the background empty until the subject is finished, then decide if it needs anything at all.
• Photograph the result before dismantling it, because the picture is temporary by design.

Faces are harder than animals

We are extraordinarily good at spotting a wrong face, so a portrait at low resolution rarely looks like the person. A pet, a car, a house or a favourite toy is far more forgiving and produces a better first result. Save people for later, and start with a silhouette rather than a full-face photograph if a person is unavoidable.

When the app is the better tool

For a complicated image, a phone app that posterises and pixelates will produce a cleaner map faster, and there is no shame in using one. What it will not do is teach anyone which details mattered. Use it for the second attempt, once the child already knows why the eyes went in before the jumper.