How to Look Thinner in Photos: The Camera Geometry That Actually Changes Your Face

How to look thinner in photos is mostly a geometry problem, not a body problem. A phone held at chest height, a foot from your face, with a wide front lens, renders a heavier jaw and a shorter neck than a lens five feet away at eye level. Change the geometry and the face changes.

That is the whole premise of this guide. Nothing below asks you to change your body, hold your breath or suck anything in. Every technique here is a change to where the camera is, how far away it is, where the light comes from and how your head sits on your neck — the four variables a portrait photographer adjusts before they touch anything else. They are mechanical, repeatable, and they work on any face.

Why a camera renders you differently than a mirror does

Two things happen between your mirror and your camera roll. Your eyes see in stereo and build depth from parallax; a single lens flattens that depth onto one plane. And a lens close to your face renders near features larger than far ones. Neither is about your body.

The folk phrase "the camera adds ten pounds" is a rough description of both effects at once. Binocular vision gives you a running sense of which parts of a face are nearer and which recede — the side of a cheek reads as a curve going away from you. A photograph discards that. What is left is a two-dimensional outline, and an outline of a rounded form tends to read wider than the same form felt in three dimensions.

Layered on top is perspective distortion, and this part is not folklore — it is measurable. Because the size a feature occupies in the frame falls off with distance, a nose a few inches closer to the lens than the ears is rendered proportionally larger when the whole face is only a foot away, and barely differently when the face is five feet away. At close range the front plane of the face swells, the jaw softens into the neck, and the silhouette widens. Step back and the same face is rendered with its real proportions.

So the useful question is never "how do I look thinner" but "which variable is currently distorting me, and by how much." There are five, and they stack. Fix them in order of effect size and most people are done after the first three.

Subject distance: the variable that does the most work

The single biggest lever is not the lens, it is how far the lens sits from your face. Perspective is set by distance alone. A 24mm lens and a 105mm lens fired from the same spot render identical facial proportions — the only difference is how much of the scene each one includes. Focal length matters because it changes where you stand.

There is a measured number for this. A 2018 study in JAMA Facial Plastic Surgery, by researchers at Rutgers New Jersey Medical School and Stanford, built a mathematical model of how photographic distance changes apparent facial proportions. At 12 inches — roughly a selfie held at a bent elbow — the model puts the base of the nose about 30% wider, and the nasal tip about 7% wider, than the same face shot from five feet. The same compression logic applies to the rest of the front plane: cheeks, chin, and the width of the jaw against the neck.

Five feet is the practical floor for a portrait. Beyond that the returns flatten out; classic portrait focal lengths of 85mm to 135mm exist precisely because they let a photographer fill the frame with a head from roughly six to twelve feet away. You do not need those lenses. You need the distance they imply.

What this means with a phone

Phone front cameras are wide by design — several iPhone generations, for instance, have used a 23mm-equivalent front lens, wide enough to fit two faces into a group selfie. Held at arm's length, that combination is the worst-case version of everything above. Three fixes, in order of how much they help:

• Use the rear camera, not the front one. The main rear camera is usually the better camera on the phone — larger sensor, better optics — and using it forces you to prop the phone up and step away rather than holding it out.
• Put the phone on something and back up to five or six feet. A stack of books on a counter, a windowsill, a cheap tripod. Then use the self-timer or a voice shutter.
• Zoom with your feet, then crop — or use the 2x/3x lens if your phone has one. Backing up and cropping in gives you the undistorted proportions of a longer lens. Digital zoom from the same close position does not: cropping a distorted frame keeps the distortion.

That last point is the one people get wrong most often. Stepping closer to make your head bigger in frame is not the same as zooming. One changes perspective; the other only changes framing. If you take one thing from this article, take the step backward. Our companion piece on the physics behind a selfie that actually looks like you goes deeper on the phone-specific mechanics; this article stays on the geometry that applies to any camera.

Camera height: what each few inches actually does

Height is the second-largest variable and the one people misjudge in both directions. The reference point is your eyes, not your head, not your chin. From there, small moves have specific and predictable effects, and there is a fairly narrow band where the result reads as a portrait rather than as an angle.

1. 1Below eye level (a phone at chest height, a laptop on a desk). The lens now looks up the underside of your jaw and into your nostrils. The soft plane under the chin is turned toward the camera and rendered at full size, the neck disappears, and the body — being closer to the lens than the head — is enlarged relative to the face. This is the default position of almost every unflattering photo ever taken of anyone.
2. 2Exactly at eye level. Neutral. No vertical distortion, eyes at the natural centre of attention, jawline rendered as it is. If you only fix one thing about height, fix it to here. It is also the correct setting for a webcam.
3. 3Two to four inches above eye level. The most commonly used position in professional headshots. The under-chin plane rotates away from the lens and shortens in the frame; the jawline reads as a cleaner edge; the eyes, now angled slightly upward toward the lens, occupy a fraction more of the face. This is a real effect, and it is small — which is the point.
4. 4A foot or more above eye level. Now you are in caricature. The cranium enlarges, the body shrinks away below, the chin recedes and the neck vanishes. It reads instantly as a pose rather than a portrait, and on a professional profile it undercuts you.

The practical instruction is simple: raise the camera until the lens is between your eyebrow line and the top of your head, and stop. If you are on a video call, the same rule is why every guide tells you to stack books under a laptop: the built-in webcam starts well below eye level, and raising it until the lens clears your brow line is the entire fix.

The chin-forward-and-down move (and why it fixes a double chin in photos)

This is the single direction portrait photographers give most often, and it is the direct answer to a double chin in photos. It is counterintuitive, it feels absurd while you are doing it, and it looks completely normal in the resulting frame. Peter Hurley popularised calling it "the turtle." Other educators just say: chin forward, then slightly down.

The mechanics are worth understanding, because understanding them stops you from overdoing it. A double chin in a photograph is not only fat, it is slack: the soft tissue between the point of the chin and the top of the throat has nothing to stretch it, so it hangs and the camera renders it as a second horizontal edge. Pushing the head forward on the neck — as if sliding it toward the lens on a rail — puts that tissue under tension. The skin along the jaw pulls taut, the jaw becomes a defined line, and the second edge flattens out.

Dropping the chin a fraction after pushing forward does the rest. It stops the forward push from turning into a chin-up tilt, which is the classic overcorrection: raising the chin points your nostrils at the lens and stretches the throat into the most prominent thing in the frame.

How to do it without looking strange

The move fails in two ways: done from a slumped posture it reads as peering, and done too far it reads as a tilt. Five checks keep it in the window where the camera sees a jawline and the viewer sees nothing unusual at all.

• Sit or stand tall first. The move only works from a lengthened spine — do it while slumped and you just look like you are peering at something.
• Slide the head forward, horizontally, not up. Think of a turtle emerging, or of moving your face an inch closer to a screen you are reading.
• Then lower the chin by a genuinely small amount — a quarter to half an inch. Any more and you are looking up through your brow line.
• Hold it for the shot only. It is a two-second position, not a posture.
• Check it in the frame, not in the feeling. It will feel like far too much and read as roughly nothing.

If the whole idea of arranging your face for a lens makes you tense, that tension will show up before the geometry does — and no amount of correct chin position rescues a jaw that is clenched. Our guide to posing for a headshot when you're not photogenic covers the expression and confidence side of that problem; this piece deliberately stays on the mechanical half.

Shoulder rotation and the silhouette it produces

Shoulders are the widest horizontal element in a head-and-shoulders crop, so their angle sets the width of the whole shape. Square to the lens, both shoulders are rendered at full width and the silhouette is a rectangle. Rotated, the far shoulder falls behind the near one and the outline narrows into a triangle.

Thirty to forty-five degrees is the working range. Turn your feet and torso away from the camera by roughly that much, then bring your head back to face the lens directly. The result is a diagonal shoulder line rather than a horizontal one, which does three things at once: it narrows the outline, it introduces depth into an otherwise flat image, and it puts your face at a slight angle to your body, which reads as engagement rather than as a mugshot.

Two adjustments make it land:

• Drop the near shoulder slightly. This is a convention rather than a law — some photographers deliberately raise it — but as a default it stops a lifted near shoulder from crowding the jaw and eating the neck you just worked to expose.
• Keep the head rotation smaller than the body rotation. If shoulders go 45 degrees, the head comes back most of the way — not all of it. A head turned exactly as far as the shoulders looks stiff; a head that overshoots strains the neck tendons visibly.

One caution about the crop: a rotated silhouette only survives if the frame is wide enough to show it. In a small circular avatar — the shape LinkedIn, Slack and most platforms actually display — most of the shoulder line is cropped away, and the geometry that still matters is almost entirely head and jaw. Rotate anyway when you have the room, and prioritise the first three fixes when you do not.

Lighting: separating the jaw from the neck without squeezing anything

Posing tenses the jawline. Lighting is what makes the camera see it. A defined jaw in a photograph is a tonal edge — a lit plane above a darker plane — and if your face and neck are lit identically, no amount of chin work produces that edge. Direction is the whole game here, and there are two decisions.

The first is vertical. A light source placed slightly above your eye line casts a soft shadow under the brow, under the nose and, critically, under the jaw. That under-jaw shadow is what separates face from neck. A light at or below face height fills that shadow in and merges the two into one continuous mass. This is why a window at roughly head height flatters and a phone screen lighting you from below does not.

The second is horizontal. When your face is rotated, the side turned away from the camera is the "short" side. Lighting that short side — placing the light on the far side of your face — leaves the broad, camera-facing side falling into gentle shadow. The result is a narrower, more sculpted rendering. Lighting the near, broad side does the opposite: it puts the widest visible plane in full light and reads fuller. Photographers call these short lighting and broad lighting, and short lighting is the slimming one. The exception is worth knowing: on a very narrow or gaunt face, broad lighting is the correct choice precisely because it widens.

• Free version: stand facing a window, then rotate your body 30-40 degrees so the window is on the side of your face turned away from the camera. Camera between you and the wall opposite the window.
• Height check: the light should be slightly above your eyes. If you can see your own shadow rising up a wall behind you, the light is too low.
• Avoid overhead-only light. A ceiling fixture directly above you drops the shadow into your eye sockets and under your chin as a hard blob rather than a soft edge — that shadow is doing damage, not definition.

Everything above is the short version of a much bigger subject. If you want the full treatment of key light, fill, contrast ratios and window setups in plain English, that lives in our headshot lighting guide — this section only covers the part of lighting that changes how wide a face is rendered.

What a collar does to the same face

Wardrobe is the last geometric variable, and it is genuinely one — the neckline draws a line across the same region of the frame where you just built jaw-to-neck separation. Two people with identical bones and identical lighting will read differently in a crew neck and an open collar, because the neckline decides how much neck the viewer sees.

The mechanism is straightforward. A vertical opening — an unbuttoned collar, a V-neck, an open blazer with visible lapel diagonals — draws the eye up and down, and exposes a column of neck that gives the jaw somewhere to end. A high horizontal neckline does the reverse: it cuts the frame across, hides the neck entirely, and pushes the chin to sit directly on top of the fabric, which removes the separation you engineered with light and chin position.

• Open collar or a soft V: maximum jaw-to-neck separation. The safest default for a headshot.
• Buttoned dress shirt, top button undone: works well; the collar points create their own diagonals that echo an angled shoulder line.
• Buttoned to the throat, or a crew neck: a hard horizontal under the chin. Not wrong, but it removes a tool you would otherwise have.
• Turtleneck or a heavy scarf: eliminates the jaw-neck boundary by design. Choose it for the look, knowing it costs you the definition.
• Structured shoulders (a jacket, not a knit): a defined shoulder edge gives the silhouette a clean outer line; soft knitwear lets it spread.

Colour and pattern decisions sit outside geometry but interact with it — high-contrast patterns near the face compete with the jawline for attention. The full wardrobe reasoning is in what to wear for a headshot.

Where AI headshots sit in all this

An AI headshot generator does not fix your geometry — it inherits it. The model learns your face from the selfies you upload, so if every input photo is an arm's-length front-camera shot taken from below, the model learns that rendering of you and reproduces it faithfully across the whole batch of outputs. Garbage geometry in, garbage geometry out.

That cuts both ways, and it is the most practical thing to know before you upload anything. Input selfies taken from five or six feet, at eye level, on the rear camera, in soft light from slightly above give the model an undistorted reference — and the generated results carry that through. Our guide to taking input selfies that actually work is essentially this article applied to the upload stage, and it is worth reading before you spend a session on inputs that will limit everything downstream.

Be clear-eyed about what AI cannot do here. A photographer standing in front of you watches the chin move, sees the shadow land wrong, and says "forward an inch, now down" — a live correction loop in the two seconds before the shutter. No generator has that, and on the geometry in this article that loop is exactly the thing that matters most. What a generator offers instead is variation at volume: many versions of camera height, light direction and crop from one upload, from which you pick the geometry that works, rather than getting one frame and living with it. Our honest comparison of AI headshots and a photographer lays out that trade on cost, control and turnaround in full.

How much of this an editor should be allowed to change

There is a clean line between the techniques above and what happens in software afterwards, and it is worth naming. Everything in this article changes how a lens records a real face. Editing changes the record. The first is photography; the second becomes misrepresentation somewhere along a slider, and the useful question is where.

A defensible rule: an editor may correct what the equipment got wrong, and may not alter what your face is. Lens distortion correction, exposure and contrast that restore the separation you actually had in the room, a slightly straightened horizon, a collar that bunched — all fair. Reshaping a jawline, narrowing a face with a warp tool, or slimming a neck is not a correction, it is a different person. It also fails the only test that matters for a professional photo: whether the person who meets you recognises you from it.

• Fair game: colour, exposure, sharpening, removing a stray hair or a one-week blemish, lens correction, cropping.
• Grey area: dodging and burning to deepen a shadow that already exists under the jaw. Restrained, this is just lighting done late; overdone, it is sculpting.
• Not fair game: liquify on the jaw, chin, cheeks or neck. Skin smoothing that removes texture. Anything that changes a proportion.

We treat this in depth in headshot retouching: how much is too much, which owns the editing question across all of it. The reason it belongs in a geometry article at all is that good geometry is what makes heavy editing unnecessary. Almost everyone reaching for the liquify tool is fixing a problem that a step backward and three inches of camera height would have prevented.

A ten-minute practice run

None of this is knowledge you can use until you have seen it happen to your own face. The fastest way is a controlled test where you change one variable at a time and compare frames side by side. Ten minutes, a phone, a window and something to prop the phone on is enough to calibrate for good.

1. 1Find a window with soft daylight — overcast is ideal, direct sun is not. Stand facing it, then rotate about 35 degrees so the window falls on the far side of your face.
2. 2Prop the phone at eye level, five to six feet away, rear camera, self-timer on burst or a three-second delay.
3. 3Frame A — baseline. Shoulders square, neutral head, normal posture. This is your control.
4. 4Frame B — distance only. Same pose, but take one shot at arm's length with the front camera for comparison. Keep it. It is the most convincing evidence in the whole exercise.
5. 5Frame C — rotation. Back to the tripod position, shoulders rotated 35 degrees, head back to the lens, near shoulder dropped.
6. 6Frame D — chin. Same as C, plus head slid forward and chin down a quarter inch.
7. 7Frame E — height. Same as D, phone raised three inches.
8. 8Compare A, C, D and E in a grid on your phone. Then note which single change moved the needle most for your face — it is not the same one for everyone, and knowing yours is the actual output of this exercise.

If you want to turn that into finished photos rather than a test, our walkthrough of professional headshots at home covers backgrounds, framing and the full DIY method built around this same geometry — that article is the how-to; this one is the why.

The short version

If you remember five things from this article, remember them in this order, because that is roughly the order of effect size. Each one is free, each takes seconds, and none of them requires equipment you do not already own.

1. 1Get the lens at least five feet away and crop in. This is the biggest single change available to you.
2. 2Put the lens at eye level or a few inches above it — never below.
3. 3Slide your head forward, then drop the chin a quarter inch.
4. 4Rotate your shoulders 30-45 degrees and bring your head back to the lens.
5. 5Put the light slightly above your eyes and on the far side of your face.

What all five have in common is that they change how a lens renders you, not what you look like. That distinction matters. The goal is not a smaller version of you in the frame — it is an accurate one, taken with the same care a photographer would apply, instead of an accidental one taken from a foot below your chin.