Retouch a photo

Remove red eye from a photo

Red eyes, fixed in one tap.

It looks for flash-red pupils as soon as your photo opens and asks before it changes a thing.

This is the layer that most often meets an old file: JPG, PNG, WebP, AVIF, BMP and ICO decode here, TIFF where your browser supports it, and a HEIC pulls a decoder the first time one arrives.

No photo handy?

  • Free
  • No sign-up
  • Photos stay on your device
A portrait print with the eyes the red-eye check looks at marked
What this tool does, and the files it opens

Take the red out of a flash photograph. The scan starts as soon as a picture is open. Find the face first and it looks only at the eyes; either way candidates are ringed on the portrait and nothing changes until you accept them.

Open a picture and the scan runs half a second later, over the whole frame, with the Red-eye tool already in hand. Candidates are ringed before anything is final, and a pupil the scan missed is corrected by clicking it. The face finder, under Skin on the rail, re-runs the scan against the eyes it locates.

This is the layer that most often meets an old file: JPG, PNG, WebP, AVIF, BMP and ICO decode here, TIFF where your browser supports it, and a HEIC pulls a decoder the first time one arrives.

Keys on the desk: Move V · Heal J · Clone S · Smooth B · Dodge O · Burn N · Colour A · Whiten W · Red-eye R · Skin F · Shine H · Under-eye U · Face G · hold Space to see the original · Ctrl Z to undo.

Upload a photo

What the tool does to a face

The scan works on a reduced copy — around 480 pixels on the long edge, which is plenty for a blob a few pixels across and keeps the pass instant on a large file. Every pixel is tested for red dominance: a red channel comfortably above both others, and a red share of the total above 42 per cent. The flagged pixels are then grouped into connected regions and each region is judged on three things — area within a band, an aspect ratio between roughly 0.45 and 2.2, and at least 42 per cent of its bounding box filled. What survives all of that is small, round and solid, which is what a pupil is and what a stripe of red fabric is not.

The correction desaturates and darkens. Each pixel inside an accepted disc takes a neutral built from its own green and blue, scaled down, faded in from the edge of the disc so there is no hard rim. Two things are deliberately left alone: any pixel not actually red-dominant, which is how the iris around the pupil survives, and any pixel that is bright and neutral in all three channels, which is the catchlight.

Knowing where the eyes are changes what that test is allowed to be. Press Find the face and a 229,738-byte detector is downloaded once, run on this device, and asked for one thing only: a box and five points per face, two of which are the eyes. The scan then covers a disc over each of those points instead of the whole picture, and the colour test drops to 26 levels of dominance and a 38 per cent share — because the question has changed from is this a pupil to is this pupil red, and the second question is much easier. Two eye points is the smallest thing a detector can return and very nearly the most useful thing this particular layer could be given.

The confirmation step survives that change, deliberately. Five points say where an eye is, not where its pupil is or how far it is open, so the disc is generous and the shape tests still run inside it; and a layer that changed a face without being asked would be the wrong habit for this site to acquire even on the one operation with a single right answer.

The order to work in

Let the scan finish, then look at the rings before you press anything. Zoom in with 2 and check each one is on a pupil. Accepting is all-or-nothing in this release: if the scan has offered you a genuine pupil and a red bead together, you are choosing between correcting both and correcting neither, and correcting both is usually still the better picture.

Then hold the compare control. Red-eye is the one layer where the corrected version almost always looks right and the original looks wrong, which makes it the easiest place to stop paying attention. Look specifically at the catchlight and at the boundary of the iris; if either has gone soft, the disc that was found is larger than the pupil.

What the scan runs on
Scan resolution
≈480 px long edge
Search area
whole frame, or one disc per detected eye
Red dominance threshold
44 levels over max(G, B); 26 inside an eye
Red share of total
above 42%; above 38% inside an eye
Region aspect ratio
0.45 to 2.2
Bounding-box fill
at least 42%
Candidates offered
up to 6, largest first
Correction
neutral from G and B, darkened to 74%
Catchlight
kept — bright neutral pixels are skipped

What it will not do

  • There is no manual pupil tool. If the scan misses an eye — too small in frame, partly behind a lens, lit orange rather than red — there is no way to point at it and say “there”. That is the honest gap in this layer and the one most worth closing.
  • Without the face finder it cannot separate a pupil from something else that is small, round and red. The search then covers the whole frame, so a bead, an out-of-focus tail-light or a bright lip highlight can be offered. With the finder that mostly stops, because the search is inside an eye — but the finder has to find the face first, and a face turned away, very small in frame or heavily shadowed is one it will miss. The rings exist for both cases.
  • It does not correct animal eyeshine. Green and yellow reflections from a tapetum are a different phenomenon with different colour, and neither the dominance test nor the shape tests apply to them.

What people ask

Why is there no strength slider here when every other layer has one?
Because there is no partial right answer. Flash red-eye is a physical artefact — the retina, lit straight on through a dilated pupil, reflected back into the lens — and a pupil is either showing the colour of the back of an eye or it is showing the colour of a pupil. Half-correcting it produces a dull maroon that is not a real state of either. Every other control on this site is a judgement with a defensible range; this one is a repair.
Why does it ask before correcting, if there is only one right answer?
Because the question is whether the thing it found is a pupil, not how much to change it. A colour test that looks for small, round, red-dominant blobs will sometimes offer you a bead on a necklace, a tail-light out of focus behind a shoulder, or a bright lip highlight. Each candidate is ringed on the picture and you accept them together, so the thing being confirmed is the identification.
It found the pupils but the correction looks flat and dead.
That is usually the catchlight, and this layer is built to keep it. The catchlight is the small specular reflection of the flash itself, the one bright neutral thing inside a red pupil, and painting over it is the single clearest tell of a careless fix. Pixels that are bright in all three channels are left exactly as found. If the eye still looks flat, the catchlight was lost when the photograph was taken and nothing here will invent one — a painted highlight in the wrong place is worse than none.
It found nothing and the pupils are clearly red.
Press Find the face and try again. Searching the whole frame, the test has to ask for three things at once — a red channel at least 44 levels above the higher of green and blue, red taking more than 42 per cent of the total, and a blob that is small, roughly round and solidly filled — because anything looser starts offering beads and tail-lights. An eye lit orange rather than red will fail that. Once the detector has run, the search covers only the discs it put over the eyes it found, and the same test runs at 26 levels and 38 per cent inside them, because a small red blob in an eye socket is a pupil. A pupil three pixels across, or one behind an arm of sunglasses, will still be missed, and there is no manual pupil tool in this release, which remains a real gap.
Do I have to download the face finder to fix red-eye?
No. Everything on this page worked before there was a detector and still does: the scan runs on its own half a second after a picture opens, over the whole frame, and the correction is identical. The download — 229,738 bytes, once, from our model file server, cached afterwards — buys two things: the search stops offering you objects that are not eyes, and it starts finding pupils the strict test had to let go. Neither is worth a surprise on a metered connection, which is why the size is on the button rather than in a footnote.
Does this work on the green or yellow eyes in a photograph of a cat?
No. Animal eyeshine comes from the tapetum lucidum, a reflective layer behind the retina that people do not have, and it returns green, yellow or blue rather than the red of human retinal blood. The colour test here is built for the human case and the shape tests assume a human pupil. Nothing stops you loading the picture; the scan will simply find nothing.
Should I run this before or after the other layers?
It does not matter what order you press things in — the stack always runs heal, then red-eye, then skin. Red-eye sits in the middle deliberately: it is colour work and should be evaluated against luminance that no other layer has moved yet, and it needs the discrete marks already gone so a healed spot near an eye is not mistaken for one.

Read next

Photographic visual for Red Eye