You know the feeling. Someone sends you a photo from the weekend, and there is a half second where you do not recognise the person in it. The face is yours, technically. But it is heavier, or flatter, or the nose is doing something the nose has never done in your entire life. Twenty minutes earlier you looked in the mirror and thought, honestly, not bad.

So which one is real?

Neither, and that is the useful answer. The mirror and the camera are two different instruments telling you two different kinds of lie, and once you know what each one is doing, the whole thing stops feeling like a verdict on your face.

There are three separate effects at work here. They have all been measured in published research. None of them is about you being unattractive, and two of them are fixable in about ten seconds.

First problem: you have been looking at the wrong face your whole life

The mirror does not show you your face. It shows you your face reversed, left for right, and you have been staring at that reversed version every single day since you were tall enough to see over the sink.

Faces are not symmetrical. Yours certainly is not. One eye sits a little higher, the smile pulls harder to one side, the nose leans a few degrees off centre. Flip all of that and you get a face that is recognisably you but subtly, unplaceably wrong.

In 1977, three psychologists named Theodore Mita, Marshall Dermer and Jeffrey Knight ran the experiment that settles this. They photographed 37 women, then printed every photo twice, once normally and once mirror flipped. Then they asked each woman which version she preferred, and separately asked her close friends the same question.

The women picked their mirror image. Their friends picked the true image. Both groups were confident. Both groups were choosing the version of that face they had seen most often.

That is the mere exposure effect, one of the sturdiest findings in social psychology: we like things more as they become familiar, and we do not notice this is happening. You have logged thousands of hours with your reflection. Your friends have logged thousands of hours with the real thing. When you look at a photo of yourself and feel that small flinch, you are not seeing an ugly face. You are seeing an unfamiliar one.

Here is the part worth sitting with. Everyone who has ever loved you has only ever seen the version in the photo. That is the face they think of when they think of you. The one you prefer is the one nobody else has ever met.

Second problem: your phone is twelve inches from your face

This one is not psychology. It is geometry, and it is brutal.

In 2018, a research team including Boris Paskhover, a facial plastic surgeon at Rutgers, published a paper in JAMA Facial Plastic Surgery with a title that gets straight to it: Nasal Distortion in Short-Distance Photographs: The Selfie Effect. They built a mathematical model of the human face and worked out exactly what happens when you move a camera closer to it.

At twelve inches, roughly where your arm holds a phone, the base of your nose appears about 30 percent wider than it does at five feet. The tip appears about 7 percent wider.

Thirty percent. That is not a mood or a bad angle. That is a measurable change in the proportions of your face, and it happens every time you hold a phone at arm's length.

The reason is simple once you see it. Perspective makes near things big and far things small. When the camera is a foot away, your nose is meaningfully closer to the lens than your ears are, so it gets enlarged relative to everything else. Push the camera back to five feet and the difference between nose distance and ear distance becomes trivial, so your features fall back into their real proportions.

Two things follow from this, and both matter.

The first is that it has nothing to do with your camera being cheap. It is not the sensor or the megapixels or the lens quality. A ten thousand dollar camera at twelve inches produces exactly the same distortion, because the distortion is caused by where the camera is standing, not what it is made of.

The second is that "the camera adds ten pounds" has survived for decades as an explanation while being basically the wrong story. The camera is not adding weight. Short distance is expanding whatever is nearest the lens, which is usually the nose, the cheeks and the chin, and everyone reads that as a wider face.

Front-facing phone cameras also tend to use a wide angle lens so they can fit you in the frame at close range, which stretches the edges of the image on top of the perspective problem. Two effects stacking, both fixed by the same thing: distance.

The researchers put the ideal portrait distance at about five feet. Portrait photographers landed on the same answer generations ago by trial and error, which is why the classic portrait lens is an 85mm and not a 24mm. It forces you to stand back.

Third problem: a photo freezes a face that was built to move

Even at the right distance, with good light, a still photo is fighting your face.

In 2012, researchers at Harvard and Berkeley, led by Robert Post, ran a study called The Frozen Face Effect: Why Static Photographs May Not Do You Justice. They took two second video clips of twenty people, then pulled 1,200 still frames out of those same clips. Same faces, same lighting, same moments. Then they asked people to rate how flattering each one was.

The videos won. Consistently. The same face was rated more attractive in motion than in any of the stills taken from that motion.

Sit with that: the still frames came from inside the video. There was no better lighting, no makeup, no different day. The only variable was movement, and movement won every time.

The researchers offered a plausible explanation. Your brain evolved to read moving faces and it seems to average across many moments to build an impression. A photograph hands it a single frozen millisecond instead, and a lot of those milliseconds are objectively strange. Human faces pass through genuinely odd shapes on the way between expressions. Watch any video frame by frame and you will find several frames that look like a stranger having a small crisis. In motion you never notice. Frozen and printed, that is the picture that ends up on the group chat.

The authors made the connection themselves. They wrote that the effect may explain "why photography of faces is so difficult to master and why people anecdotally believe they look worse in photographs."

This is also why you look fine on a video call and terrible in the screenshot. Nothing changed except that time stopped.

The other things that are quietly working against you

The three effects above do the heavy lifting. A few smaller ones pile on.

Camera height. A phone held at chest level shoots up at you, which shortens your neck, widens your jaw and points the lens into your nostrils. Eye level or very slightly above is the flattering choice, and it is not vanity, it is just geometry again.

Light direction. Overhead light, which is most indoor light, drops shadows into your eye sockets and under your nose and chin. Light coming from in front of you, a window is ideal, fills those shadows in. Same face, completely different photograph.

Your face in repose. Most of us have no idea what our neutral face looks like, because we have never seen it. The moment you look in a mirror you arrange yourself slightly. Candid photos catch the unarranged version, and it reads as unfamiliar for the same reason the flipped face does.

The self view era. You now spend hours a week watching your own face on video calls, which is historically bizarre. Nobody before about 2020 had this much exposure to their own moving image, and there is a growing body of work on how self view affects self perception. It has raised everyone's sensitivity to their own face without making anyone happier about it.

What actually fixes it

Here is the honest list, and none of it requires buying anything.

Get the camera back. This is the single highest impact change you can make. Five feet, not twelve inches. Use the rear camera instead of the selfie camera, prop the phone somewhere, use the timer or a voice trigger. If you must take it yourself, hold the phone as far as your arm allows and crop in afterwards. Cropping a distant photo is fine. It keeps the true proportions. Standing closer does not.

Get the lens up to eye level. Prop the phone on a stack of books so it sits at or just above your eyes.

Turn to face the light. Stand facing a window, not with your back to it, and turn off the overhead. Soft light from the front, nothing behind you.

Take far more photos than feels sensible. This is the direct counter to the frozen face effect. If a single frozen millisecond is the enemy, the answer is to collect a hundred of them and keep the two that landed. Professional portrait sessions shoot hundreds of frames to deliver five. That is not indulgence, that is the whole method. Talk, move, keep the shutter going, and stop trying to hold one perfect expression, because a held expression is exactly what reads as stiff.

Give it a beat before you judge. Novel faces get more appealing as they become familiar. That is the mere exposure effect running in your favour for once. The photo you hate today is often fine in a week.

What cannot be fixed, and it is only fair to say so

No photograph, no photographer and no software will ever make your true face feel as comfortable to you as your reflection does. That preference was built by thousands of hours of repetition and it is not going anywhere.

So there is a ceiling on how much you will ever like a picture of yourself, and it sits lower than how much other people like it. Anyone promising to close that gap completely is selling something.

The useful move is to stop using your own reaction as the measure. You are the least reliable judge of your own photo in any room. If you want to know whether a headshot is good, ask someone who knows your face and does not have to be nice to you. They are looking at the version they already recognise.

Where AI headshots fit into this

Worth being precise here, because the honest version is more useful than the sales version.

An AI headshot does not fix problem one. Your true face stays your true face and it will still feel a little unfamiliar to you. Nothing changes that.

What it does address is problems two and three, which are the ones actually making the photo worse than reality.

The distance problem gets solved by re-rendering. A generator takes your close range selfie and rebuilds the portrait at proper portrait geometry, the framing and compression you would get from an 85mm lens at five feet. That is why the output often looks like a completely different photograph of the same person, and why people say it looks like them but better. It is not flattering you. It is removing a distortion that was there.

The frozen millisecond problem gets solved by volume. You get a batch instead of a single frame, so you are choosing the good moment rather than being stuck with whichever one the shutter happened to catch.

Two things it will not do. It will not make you look like someone else, and if it does, that is a bad result, not a good one. And it cannot recover information your input photo never had, which is why a sharp, well lit, eye level source photo matters so much. If you want the full version of that, we wrote how to take the selfie and lighting for your input photo.

If you would rather book a photographer, book one. A good photographer solves the same two problems with a longer lens and four hundred frames, plus judgement about you specifically that no model has. We put the honest comparison in AI headshots vs a photographer. The point of this article is not which one you pick. It is that you were never the problem.

The short version

You do not look bad in photos. Three things are happening at once:

1. You are comparing every photo against a mirror image that nobody else has ever seen, and familiarity is doing the rest.
2. A camera twelve inches away makes the base of your nose look about 30 percent wider than it is, which is geometry, not your face.
3. A still photo freezes one millisecond out of a face that reads better in motion, and most milliseconds are unflattering.

Move the camera back, get it to eye level, face a window, take a hundred frames instead of one. That fixes most of it. The rest is your brain preferring a face it has seen more often, and it is allowed to.

FAQ

Why do I look better in the mirror than in photos? Because the mirror shows your face reversed, and that reversed version is the one you have seen every day of your life. In a 1977 study by Mita, Dermer and Knight, women shown normal and mirror flipped photos of themselves preferred the mirror version, while their friends preferred the true version. Both groups picked the face they were more familiar with. Your photo is not less attractive, it is less familiar to you specifically.

Is the mirror or the camera more accurate? The camera is closer to what other people see, since it captures your face unreversed. But a camera held at arm's length introduces its own distortion by enlarging whatever is nearest the lens. The most accurate view of your face is a photo taken from about five feet away, which has neither the mirror's reversal nor the selfie's perspective distortion.

Why does my nose look bigger in photos? Because of camera distance, not your nose. A 2018 study in JAMA Facial Plastic Surgery found that a photo taken at twelve inches makes the base of the nose appear roughly 30 percent wider than the same face photographed at five feet. Perspective enlarges whatever is closest to the lens, and in a selfie that is your nose. Step back and the effect disappears.

Am I actually unattractive or is it just the camera? Almost certainly the camera, plus the familiarity gap. Three measured effects push a photo away from how you look in life: mirror reversal, short distance distortion, and the frozen face effect where still images rate lower than video of the same face. All three make the photo worse than reality rather than revealing some truth about you.

Why do I look fine on video calls but bad in screenshots? This is the frozen face effect. A 2012 study by Post and colleagues rated the same faces in two second video clips and in 1,200 still frames pulled from those clips. The videos were consistently rated more flattering. Faces pass through odd transitional shapes between expressions that you never notice in motion, and a screenshot catches exactly one of them.

What is the best distance to take a headshot? About five feet, which is what the 2018 JAMA research identified as the distance where facial proportions stop being distorted. It is also why the classic portrait lens is an 85mm rather than a wide angle: it forces the photographer to stand back. Shoot from five feet and crop in if you want a tighter frame, because cropping preserves the proportions and stepping closer destroys them.

Does the camera really add ten pounds? Not really, and the phrase hides what is actually happening. Nothing is being added. Short camera distance expands the features closest to the lens, usually the nose, cheeks and chin, and a wider looking face gets read as a heavier one. Move the camera back and the effect goes away, which would not happen if the camera were genuinely adding weight.

Do AI headshots fix any of this? They address two of the three problems. Re-rendering your selfie at portrait geometry removes the short distance distortion, and getting a batch of images instead of one frame beats the frozen face problem by letting you pick the moment that landed. What they cannot fix is the familiarity gap, since your true face will always feel slightly less comfortable to you than your reflection. Any tool claiming to solve that is overselling.

Should I flip my photo so it matches the mirror? You can, and you will probably prefer it. Just be aware that everyone else will not, since they are used to the unflipped version. Flipping also reverses any text in the frame, which looks wrong immediately. For a professional headshot, leave it unflipped. It is the face your colleagues actually recognise.