Clickory

How does a phone shrink a huge photo small enough to send in a flash?

Keep going

What else makes you wonder?

What kind of picture would be hardest to shrink at all?

Imagine every neighboring pixel changing. What pattern would the phone have left to describe?

Would “aaaaaa” shrink more than an ordinary sentence?

Pictures use pixels, but the same repeat-and-count idea can work on letters, sounds, or any ordered symbols.

How much detail can a photo lose before your eyes notice?

Try zooming into a heavily squeezed picture. Tiny blocks and fuzzy edges often reveal what was discarded first.

ages 8–11How does a computer remember things even after you turn it off? Open ages 8–11What's the fastest way to find one name in a giant sorted list? Open ages 8–11How does a song fly through the air into your radio? Open
After you watchHow does a phone shrink a huge photo small enough to send in a flash?

The short answer

One simple lossless way a phone can shrink a picture is by folding up repeats: instead of storing 'blue, blue, blue, blue', it stores 'blue × 4'. The longer the runs of the same color, the more this trick can fold, so a plain picture shrinks a lot and a busy, speckly one barely shrinks at all. Real camera photos usually need a different, lossy kind of compression too.

Try this next

  • What if you draw the messiest picture you can? Slide the squeeze row all the way to 'messy' and count the tags — does the file ever get smaller, or does it stay full?
  • What if the plain picture had a tiny speck of a different color? Imagine one red pixel dropped into the blue sky — predict whether that single speck cuts a long run in half and adds tags.
The whole story

How it works

This story models run-length encoding. A picture is a grid of tiny colored squares called pixels. If a stretch of boxes are all the same color, this method writes the color once and a count of how many boxes in a row share it. A row that goes blue twelve times becomes the single tag 'blue × 12'. Rebuild it later by stamping out that many blues. Wherever the picture has long, calm stretches of one color, those stretches collapse into tiny tags and the file gets small.

What people get wrong

Lots of people think compression is a magic shrink-ray that makes every file smaller by the same amount. It isn't. This folding trick only saves space when colors repeat. On a busy photo where almost every neighboring pixel is a different color, there are no runs to fold, so the same trick saves almost nothing — and can even add a little. How much a file shrinks depends on the picture, not on the trick alone.

The catch

The simple folding trick is perfectly safe — you can rebuild the picture exactly, every pixel — but it does nothing for busy, detailed photos. To shrink those anyway, phones switch to a different kind of squeezing that throws away fine details your eyes barely notice. That wins huge savings on any photo, but it's a one-way door: you can never get those exact discarded details back.

Questions kids ask

Why do some files shrink a lot and others almost nothing?

Shrinking works by folding up repeats. A file with long stretches of the same thing (a plain blue sky) folds way down, while a file where everything is different (confetti, or random noise) has nothing to fold, so it barely shrinks.

Is the shrunk picture exactly the same as before?

With the simple 'count the repeats' trick, yes — you can rebuild every pixel exactly. But the heavy squeezing real photos use throws away tiny details to save more space, so that version is not a perfect copy.

Does this work on more than just pictures?

The repeat-and-count idea can work on any ordered symbols, including letters or sound samples, when identical values sit next to each other. A line like 'aaaaaa' folds tiny; random gibberish barely shrinks. Real music and video formats also use other, more powerful kinds of pattern-finding.

Why can't you keep shrinking a file forever?

Each squeeze removes repetition. Once a file has no patterns left to fold, there's nothing more to remove, so squeezing it again does nothing — you've hit the floor.

Talk about it

  • Ask them: the trick stores 'blue × 12' instead of twelve blues. So what picture could you draw that this trick would shrink almost to nothing?
  • Ask: a confetti photo barely shrank with this trick — so how do you think your phone still manages to text a real, detailed photo so fast?

For grown-ups

The kid-simple trick here is run-length encoding (RLE): store a (color, count) pair instead of repeating a symbol. It's lossless — you reconstruct the original bytes exactly — and it excels on low-entropy, flat regions but cannot compress high-entropy (noisy) data, where it may even add overhead. That information-theoretic floor is why photographs use lossy formats like JPEG, which discard high-frequency detail the human eye barely registers (via a frequency transform and quantization) to win large ratios on busy images, at the cost of never recovering the original. The general law holds everywhere: compression trades predictable repetition for size, and there is no free lunch on truly random data.