Name the shot, not the feeling: film's shot vocabulary tells a model exactly where the camera sits, and it still works when a model release breaks everything else.
- Angle is the most reliable lever. Low angle reads as power, high angle as vulnerability.
- One of each, no more. Pick one angle, one distance and one movement; stacking terms averages the shots.
- Movement needs time. A slow push-in is meaningless in a two-second generation.
- Near-synonyms differ. Handheld feels urgent and human; shaky means loss of control.
- When stuck, simplify. Composition and angle are understood more reliably than movement.
- Angle low, high, Dutch and overhead: where the camera sits, and the power it gives the subject.
- Distance extreme close-up, close-up, over the shoulder and wide shot: how close the audience gets.
- Movement push-in, pull-out, tracking, dolly zoom, locked, handheld, shaky and long take, by the effect each gives.
- Every camera move, one per row each move by name, what the camera physically does, and a line to paste.
- Speed and intensity words calm, controlled, energetic or abrupt: how the camera moves, not where.
- Focus rack focus, POV and foreground reveal control when the audience learns something.
- Lenses from ultra-wide to telephoto: how much of the world is in frame and how near things feel.
- Shallow vs deep focus a thin sharp slice that points the eye, or near and far both sharp.
- Aperture words the f-number: a wide opening gives shallow focus, a narrow one keeps more in focus.
- Composition centre frame, negative space, frame within a frame, reflection and silhouette.
- Which tools document camera control camera words, settings and presets, as Runway, Veo, Kling, Higgsfield and Midjourney describe them.
- Using it in a prompt how to combine terms, how to ask for a still camera, and what to fall back on.
Camera grammar is the shared vocabulary for where a camera sits, how it moves and what its lens does. This page has twenty-four shot types, grouped by what they actually control, then a one-row-per-move movement table, speed words, a lens table, and which video tools document camera control. This vocabulary predates AI by a century — which is exactly why it still works when a model release breaks everything else.
A generative model has seen millions of films. It knows what "low angle" means far better than it knows what "make it look powerful" means.
Naming the shot is not jargon. It is the difference between describing a feeling and specifying a camera position.
Angle — where the camera sits
The single most reliable lever, because it maps to how humans already read physical height and dominance.
Overhead and high angle are not the same note. High angle diminishes a person; overhead removes them from the situation entirely — they become one element in a pattern.
Where the camera sits sets power. Low angle makes a subject look powerful, high angle vulnerable, and overhead removes them from the situation entirely.
Distance — how much is in frame
Close-up carries a range rather than one meaning — intimacy when the subject is open, intensity when they are not. The distance sets the access; the performance decides what you do with it.
Distance sets how close the audience gets. A close-up can mean intimacy or intensity; the performance decides which.
Movement — what the camera does during the take
Push-in and long take both produce tension by different means. One approaches; the other refuses to release. Knowing which you want is the difference between a shot that builds and one that merely lasts.
Handheld and shaky are also not synonyms. Handheld implies a person holding the camera — urgent, present, human. Shaky implies loss of control. Prompting for one and getting the other is a common and fixable failure.
On anamorphic: it names a physical lens, and its signatures — oval bokeh, horizontal blue flares, a wider frame — are specific enough to ask for individually. The realism words sets out what the lens actually does.
Pick movement by the effect you want. A push-in builds tension by approaching, a long take by refusing to cut, and handheld is not the same as shaky.
Every camera move, one per row
The plain point: the movement list above is about what a move feels like. This table is the vocabulary itself: the name of each move, what the camera physically does, and a line you can paste. Meanings come from film references and from the video tools’ own glossaries, linked in each row. The prompt lines are this site’s own examples. How to order these words inside a full prompt is a different question, answered in the anatomy of an AI video prompt.
| Move | What the camera does | Example prompt line |
|---|---|---|
| Dolly in (push in) | Means: the whole camera travels toward the subject, drawing the viewer closer. 1 | Prompt line: slow dolly in toward her face as she realises what she has done |
| Dolly out (pull back) | Means: the camera travels away, revealing more of the scene. 1 | Prompt line: the camera pulls back from the lone chair to reveal the empty hall around it |
| Truck left / right | Means: the camera slides sideways, parallel to the subject. 1 | Prompt line: the camera trucks right past a row of shop windows, keeping pace with the reflections |
| Tracking shot | Means: the camera moves with a moving subject. 2 | Prompt line: tracking shot alongside the cyclist, level with her shoulder, the street streaming past |
| Pedestal up / down | Means: the whole camera rises or lowers while staying level, unlike a tilt. 3 | Prompt line: slow pedestal up from the cluttered desk to his face, the camera staying level |
| Pan left / right | Means: the camera turns horizontally from a fixed spot. 4 | Prompt line: slow pan left across the harbour, ending on a single red boat |
| Tilt up / down | Means: the camera tips up or down from a fixed spot. 4 | Prompt line: tilt up from the boots to the face of the soldier standing guard |
| Orbit | Means: the camera circles all the way round the subject. 2 | Prompt line: the camera orbits the sculpture once, slowly, against a black background |
| Arc | Means: the camera moves on a curve around the subject, not necessarily a full circle. 4 | Prompt line: a half arc around the two dancers, from her side to his |
| Crane / jib | Means: the camera rises, falls or sweeps on a mechanical arm, often for big reveals. 1 | Prompt line: crane up from the crowd to reveal the whole square at dusk |
| Handheld | Means: the camera is held by a person, so it moves and shakes slightly. 1 | Prompt line: handheld camera following him through the crowded kitchen, slight natural shake |
| Whip pan | Means: a pan so fast the image blurs, often used to jump between subjects. 1 | Prompt line: whip pan from the bride to the stranger at the back of the church |
| Zoom in / out | Means: the lens changes focal length; the camera itself stays put. 1 | Prompt line: slow zoom in on the window of the farmhouse, the camera locked on a tripod |
| Static / locked-off | Means: the camera does not move at all. 4 | Prompt line: static locked-off shot, the camera completely still, only the rain moving |
Both make the subject bigger in frame, and they are not the same shot. A dolly moves the camera; a zoom changes the lens. 4
In a dolly the background shifts as the camera passes things; in a zoom everything is magnified together and nothing shifts. If you want the viewer to feel they are moving into the room, write dolly in. If you want a sudden “look at that”, write zoom. Do both at once, in opposite directions, and you get the dolly zoom from the movement list above. 1
Speed and intensity words
The move says where the camera goes; a speed word says how. The vendors write these words straight into their own examples: Google’s Veo guide uses “slow pan left” and “slow zoom in”, and Runway says a style descriptor can carry motion speed. 4 5
- Calm: slow, gentle, subtle, barely moving. Good for tension that builds, and for short clips where a big move has no room.
- Controlled: smooth, steady, deliberate, gliding. Reads as a dolly or gimbal rather than a person.
- Energetic: fast, rapid, sweeping, aggressive. Use for action and reveals.
- Abrupt: sudden, snap, crash (as in crash zoom), whip. Use sparingly; one per shot. 2
“Slow motion” is not a camera speed. It slows the action, not the camera, so “slow dolly in” and “dolly in, slow motion” ask for different things.
Name the move, then the speed. “Slow dolly in” and “slow zoom in” are different shots; pick the one you mean.
Focus — what is sharp, and when
These three control when the audience learns something, which is why they are the hardest to specify and the most valuable when they land.
Focus controls when the audience learns something. Rack focus, POV and foreground reveals are the hardest to specify and the most valuable when they land.
Lenses: focal length, focus and aperture
The plain point: the lens decides how much of the world is in the frame and how near or far things feel. The number is the focal length, in millimetres. 6 The ranges below are for a full-frame camera; Nikon gives smaller numbers for its smaller-sensor (DX) cameras.
| Lens | What it does | Example prompt line |
|---|---|---|
| 14–24mm (ultra-wide to wide) | Means: takes in a very wide view and stretches the sense of distance, so near things look big and far things small. 6 7 | Prompt line: shot on a 16mm ultra-wide lens, the cathedral ceiling soaring overhead, tiny figures below |
| 35mm | Means: a moderately wide view: room for the setting around a person; for portraits, about as wide as most photographers go. 8 | Prompt line: 35mm lens, waist-up, the busy cafe visible around her |
| 50mm | Means: a “standard” lens, close to how the eye sees, so it looks natural and unforced. 6 | Prompt line: 50mm lens at eye level, natural perspective, documentary feel |
| 85mm | Means: a short telephoto: the classic portrait length, with a narrower view and an easily blurred background. 8 | Prompt line: 85mm portrait lens, head and shoulders, background softly blurred |
| 135mm | Means: a longer telephoto: tight framing, and the background seems pulled up close behind the subject (compression). 7 6 | Prompt line: 135mm telephoto, the runner framed tight, the city skyline stacked close behind him |
Shallow focus vs deep focus
- Shallow focus (shallow depth of field): only a thin slice of the picture is sharp; everything in front and behind melts into blur. It points the eye at one thing. 9 Prompt line:
shallow depth of field, only her eyes in focus, the party behind a soft blur - Deep focus: near and far are both sharp, so the viewer can look around the whole frame. 10 Prompt line:
deep focus, the child in the foreground and the house far behind both pin-sharp
Aperture words
The aperture is the opening in the lens, written as an f-number. A low f-number (a wide opening, such as f/1.4 or f/1.8) gives shallow focus; a high one (such as f/8 or f/11) keeps more in focus. A longer lens and a closer camera also make focus shallower. 9 The specific f-numbers are this page’s examples of “low” and “high”, not figures from the source.
Google’s Veo guide lists lens and focus terms of its own, among them shallow and deep depth of field, wide-angle and telephoto. 4 A plain description (“background softly blurred”) is a useful backup to the number, because it says what you want to see rather than how a camera would get there.
Wide lenses show the world, long lenses isolate the subject. Pair the lens with a focus word: “85mm, shallow depth of field” or “24mm, deep focus”.
Composition — where the subject sits
Centre frame and locked shot both read as "control" — and they are not the same thing.
Locked is camera behaviour: nothing moves. Centre frame is composition: the subject owns the middle. Combine them and the control reads as absolute. Use one against the other — a locked camera on an off-centre subject — and you get something considerably more unsettled.
Where the subject sits in frame carries meaning. Centre frame and a locked camera both read as control, but a locked camera on an off-centre subject feels unsettled.
Which video tools document camera control
Most video tools take camera words in the prompt. Some also document a camera control of their own: a setting or preset that moves the camera without relying on wording alone. What each vendor’s own pages say, read on 12 Oct 2026:
| Tool | What the vendor documents | Prompt or setting? |
|---|---|---|
| Runway | Documents: a camera vocabulary page with a description and example output for each term, from pan and tilt to orbit, crash zoom and Steadicam; and a prompting guide that treats camera motion as one part of a prompt. 2 5 | Type: prompt words. |
| Google Veo | Documents: a prompt guide that defines camera moves (dolly, zoom, pan, tilt, truck, crane, aerial, handheld, arc, static), angles and lens terms. 4 | Type: prompt words. |
| Kling | Documents: a camera movement control with six basic moves and four preset “master shots”; its image-to-video page also describes Motion Control for pans, tilts and zooms. 11 12 The guide does not say which Kling models the control applies to. | Type: a setting, plus prompt words (the Kling cheat sheet covers the prompt side). |
| Higgsfield | Documents: a library of named camera presets, such as Dolly In, Crash Zoom In, 360 Orbit, Bullet Time, Crane Up, FPV Drone and Whip Pan. 13 | Type: presets (the Higgsfield cheat sheet covers matching the prompt to the preset). |
| Midjourney video | Documents: a low or high motion setting that changes how much the camera and subject move. 14 | Type: a parameter (--motion low or --motion high). |
A preset or setting is the safer choice when you need the same move on every shot, because it does not depend on how the model reads your wording. A phrase is more flexible. Where a tool has both, the vendor pages above do not say which wins when they disagree, so keep them consistent.
Check whether your tool has a camera setting before you fight it with words. Runway and Veo document camera words; Kling and Higgsfield also document controls and presets.
Using this in a prompt
- Name the shot, not the feeling. "Low angle" outperforms "make him look powerful" because it specifies a camera position rather than an outcome.
- One angle, one distance, one movement. Stacking five terms produces an average of five shots.
- Combining is not always a problem. Runway says the opposite for its own Gen-4.5 model: combining terms is encouraged for cinematic sequences, and for moves through the frame it suggests describing what comes into view at each phase. 2 So treat “one of each” as a starting point when a result is muddled, not a rule for every tool.
- A truly still camera needs saying plainly. Runway’s own static-shot example spells it out in a full sentence. 2
- Movement needs duration. "Slow push-in" is meaningless in a two-second generation.
- Composition survives model changes better than movement does. If a generation keeps failing, drop to composition and angle — those are the most reliably understood.
The full prompt structure is on anatomy of an AI video prompt, and current model capability is on the models page.
Name the shot, not the feeling. Use one angle, one distance and one movement, and fall back to composition and angle when a generation keeps failing.
Questions people ask
What is the difference between a dolly and a zoom?
A dolly moves the camera; a zoom changes the lens.4 In a dolly the background shifts as the camera passes things, while in a zoom everything is magnified together and nothing shifts.
What does a low angle shot mean?
The camera sits below eye line, looking up, and the subject reads as powerful. A high angle, looking down, reads as vulnerability.
What lens should I ask for in an AI portrait?
An 85mm lens is the classic portrait length, with a narrower view and an easily blurred background.8 Pair it with a focus word, such as “85mm, shallow depth of field”.
How do I keep the camera still in an AI video?
Say it plainly, in a full sentence. Runway’s own example of a static shot says “The camera is entirely motionless for the duration of the scene.”2
QUICK CHECK · 3 QUESTIONS
Pick an answer to see if you got it. Nothing is recorded.
1. According to the page, why write “low angle” rather than “make him look powerful”?
- Models ignore any word that describes a feeling
- It names a camera position rather than an outcome
- It makes the clip render faster
Show the answer
Answer: It names a camera position rather than an outcome. The page says a model knows what low angle means far better than make it look powerful, because naming the shot specifies where the camera sits.
2. What is the difference between a dolly and a zoom, according to the page?
- They are two names for the same shot
- A dolly is fast and a zoom is slow
- A dolly moves the camera; a zoom changes the lens
Show the answer
Answer: A dolly moves the camera; a zoom changes the lens. The page says in a dolly the background shifts as the camera passes things, while in a zoom everything is magnified together.
3. If a generation keeps failing, what does the page suggest?
- Simplify to composition and angle, which are understood most reliably
- Add more movement terms so the model has more to go on
- Stack several angles so the model can choose one
Show the answer
Answer: Simplify to composition and angle, which are understood most reliably. The page says composition and angle survive model changes better than movement, so drop back to them when a generation keeps failing.
Related on this site: Art Styles for AI Prompts · First Frame and Last Frame in AI Video · Seamless Looping AI Video · Free AI Prompt Generator for Images and Video · Camera Words for AI Video
- 1StudioBinder, Types of Camera Movements in Film Explained: Definitive Guide, 2 Feb 2025, read at source 12 Oct 2026“A push-in draws the audience closer to a subject, enhancing tension or intimacy, typically with a dolly camera movement or Steadicam.” ↑a“A pull-out is a camera movement that smoothly moves the camera away from the subject, revealing the broader context of the scene.” ↑b“Truck camera movement is a type of tracking shot that moves laterally left or right.” ↑c“Boom shots move the camera vertically, typically with a crane or jib, to capture grand establishing shots to introduce locations.” ↑d“A handheld shot mimics a documentary style, adding a raw, immersive feel.” ↑e“When done quickly with speed, the fast camera movement is known as a whip pan.” ↑f“Although zooms are technically not a camera movement, they do create movement within the image.” ↑g“A dolly zoom combines a push-in or vice versa a zoom-out, altering the depth perception within the frame.” ↑h
- 2Runway, Camera Terms, Prompts, & Examples, read at source 12 Oct 2026tracking, “Camera follows alongside a moving subject”. ↑aorbit, “Camera circles completely around the subject”. ↑bcrash zoom, “Very fast, dramatic zoom in or out”. ↑c“This guide provides a reference library of camera terminology with corresponding Text to Video prompts and their Gen-4.5 outputs.” ↑d“Yes—combining terms is encouraged, especially for cinematic sequences.” ↑e“The camera is entirely motionless for the duration of the scene, with movement only occurring from the building moving within the locked perspective.” (wording re-read at source 12 Nov 2026) ↑f“The camera is entirely motionless for the duration of the scene, with movement only occurring from the building moving within the locked perspective.” (wording re-read at source 12 Nov 2026) ↑g
- 3StudioBinder, The Pedestal Shot in Film — Ultimate Guide, read at source 12 Oct 2026“the camera itself is moved up or down without tilting the lens”. ↑
- 4Google Cloud, Video generation prompt guide, last updated 9 Oct 2026, read at source 12 Oct 2026pan, “the camera rotates horizontally left or right from a fixed position.” ↑atilt, “the camera rotates vertically up or down from a fixed position.” ↑barc, “the camera moves in a circular or semi-circular path around the subject.” ↑cstatic, “the camera remains completely still, there is no movement.” ↑ddolly, “the camera physically moves closer to the subject or further away.”; zoom, “the camera's lens changes its focal length to magnify or de-magnify the subject.” ↑eits examples include “slow pan left” and “slow zoom in”. ↑ftelephoto, “narrows the field of view and compresses perspective, making distant subjects appear closer”; deep depth of field, “keeps most or all of the image, from foreground to background, in sharp focus.” ↑gcrane, “the camera is mounted on a crane and moves vertically (up or down) or in sweeping arcs”. ↑hdolly, “the camera physically moves closer to the subject or further away.”; zoom, “the camera's lens changes its focal length to magnify or de-magnify the subject.” ↑i
- 5Runway, Gen-4 Video Prompting Guide, read at source 12 Oct 2026“you might use a style descriptor to convey motion speed”. ↑a“Camera motion can be prompted for movement style (locked, handheld, dolly, pan, and more)”. ↑b
- 6Nikon, Understanding Focal Length, updated Dec 2025, read at source 12 Oct 2026“The longer the focal length, the narrower the angle of view and the higher the magnification.” ↑awide-angle, “FX format approx. 14 – 35mm”; “The shorter the focal length, the wider the angle of view and the lower the magnification.” ↑bstandard lenses, “FX format approx. 50 – 60mm”; “Standard lenses are popular as they are closest to the angle of view we humans see.” ↑ctelephoto, “FX format approx. 70 – 200mm”. ↑d
- 7StudioBinder, Focal Length: An Easy Guide to Using and Understanding Camera Lenses, 5 Mar 2025, read at source 12 Oct 2026“Wider focal lengths exaggerate the appearance of distance between objects.” ↑a“Longer focal lengths compress the image by causing the background to appear closer to the subject.” ↑b
- 8StudioBinder, What is the Best Lens for Portraits?, 7 Feb 2021, read at source 12 Oct 2026“The 35mm lens is typically the widest lens anybody uses for portrait photography.” ↑a“most photographers would agree that 85mm lenses are the most naturally suited to portrait photography.” ↑b“The 35mm lens is typically the widest lens anybody uses for portrait photography.” ↑c
- 9StudioBinder, What is Shallow Depth of Field? Definitions and Examples, 20 Feb 2025, read at source 12 Oct 2026“A shallow depth of field is the small or narrow area in an image that is in focus.” ↑a“Opening up your lens aperture to a low f/stop can dramatically decrease the depth of field.”; “The longer the lens, the more shallow depth of field you can achieve.”; “The closer the camera is to your subject, the more shallow depth of field.” ↑b
- 10StudioBinder, Deep Focus Shot, read at source 12 Oct 2026“A deep focus shot keeps all the planes of the image - from foreground to background - in sharp focus at once.” ↑
- 11Kling AI, Kling AI Camera Control Guide, 24 Nov 2025, read at source 12 Oct 2026“six basic camera movements, including ‘horizontal, vertical, zoom, pan, tilt, roll’”, and four Master Shots, “‘move left and zoom in, move right and zoom in, move forward and zoom up, move down and zoom out’”. ↑
- 12Kling AI, Image to Video AI, read at source 12 Oct 2026“Use advanced Motion Control to direct camera pans, tilts, and zooms”. ↑
- 13Higgsfield, Higgsfield Camera Controls, read at source 12 Oct 2026, page title“Higgsfield Camera Controls – 50+ Cinematic AI-Motion Presets”. ↑
- 14Midjourney, Video, read at source 12 Oct 2026 in a browser“Low Motion (default): is more likely to result in still scenes, low camera motion, slow motion, or subtle character movements.” ↑
Free printable AI cheat sheets, and AI news every week.
Sixteen one-page AI guides and checklists in one free PDF, no sign-up. And if you want to keep up, the newsletter covers what changed in AI each week, in plain English, with the sources.
SUBSCRIBE FREE ↗ GET THE CHEAT SHEET PACK