To prompt Higgsfield, pick the camera move as a preset, name the same move in the prompt, then describe the subject and setting, and keep lens, lighting and character fixed across every shot that has to match. Higgsfield publishes a camera-control guide, a preset library and blog guides on keeping a character consistent and on UGC ads, not a full prompting manual. This sheet quotes those pages and labels the rest.
- 50+ named camera presets — crash zoom, bullet time, dolly, crane, FPV drone, Snorricam. The full list is in the table below.
- Set the move as a parameter, then name it in the prompt too.
- Higgsfield's rule for a series: “Keep the camera movement preset, lens, and lighting the same across every shot”, and reuse the same movement wording in the prompt.
- Settings hold; descriptions drift. Of lighting, Higgsfield says “The preset is a firmer constraint. The text is a suggestion.”
- One face across a campaign: Higgsfield’s Soul ID is trained from at least 20 photos of one person, and Higgsfield says it keeps identity consistent without guaranteeing a perfect match every time.
- Testing UGC ad hooks: keep product, creator, preset and setting fixed and change only the opening line.
- Its shortest working prompt is one line: “Slow camera pan left across a coastal town at dusk, tripod fixed, horizon level.”
The vendor documents part of this and not the rest. Quoted lines are the vendor's; anything unquoted is labelled as practice.
The formula
Higgsfield, How to Control Camera Movement, Angles, and Lens in AI Video (first read 16 Sep 2026 as a 24 Jul 2026 post; the page now carries 9 Oct 2026 as its date), re-read at source 12 Oct 2026: “In Cinema Studio this line runs with the camera movement preset, lighting, and lens already set as parameters”. Until 12 Oct 2026 this sheet quoted an older wording of the same guide, which called the preset “Camera MoveSet Style” and a “hard constraint”.
Template:
[camera move, worded the same way every shot] [across / toward] [subject] in [setting and time], [fixed details: tripod, horizon, lens].
Worked example:
Slow dolly push-in toward a ceramicist centring clay, north-facing studio at dusk, window light only, tripod-smooth, horizon level.
Breakdown — what each part does
Camera movement preset
Chosen in the interface. This is where the camera movement is set.
Movement wording
Name the same move in the prompt, in the same words each time.
Subject and setting
What the camera moves across or toward, and where.
Lens and lighting
Set them as parameters too. Higgsfield: “A text description is interpreted fresh each time and can vary.”
Camera presets: the names Higgsfield uses
The plain point: use the preset’s exact name, then repeat the move in the prompt in plain words. Higgsfield’s camera controls page lists the presets below (it advertises “50+”); the grouping and the “use it for” column are this site’s, not Higgsfield’s.1
| Kind of move | Preset names on Higgsfield’s page | Use it for (this site) |
|---|---|---|
| Push in, pull back | Dolly In, Dolly Out, Super Dolly In, Super Dolly Out, Double Dolly, Dolly Zoom In, Dolly Zoom Out | Moving closer to a face or product, or revealing the room around it |
| Sideways and around | Dolly Left, Dolly Right, Pan Left, Pan Right, Whip Pan, Arc Left, Arc Right, 360 Orbit, Lazy Susan, 3D Rotation | Following action across the frame; showing a product from every side |
| Up and down | Crane Up, Crane Down, Crane Over The Head, Jib Up, Jib Down, Tilt Up, Tilt Down, Overhead, Incline, Aerial Pullback | Opening and closing shots; showing scale |
| Zooms | Zoom In, Zoom Out, Crash Zoom In, Crash Zoom Out, Rapid Zoom In, Rapid Zoom Out, YoYo Zoom, Eating Zoom, Eyes In, Mouth In | Emphasis and comedy beats; quick reactions |
| Following and mounted | Head Tracking, Snorricam, Handheld, Hero Cam, Robo Arm, FPV Drone, Car Chasing, Car Grip, Buckle Up, Road Rush | Action, cars, sport and first-person energy |
| Point of view and transitions | Object POV, Through Object In, Through Object Out, Flying Cam Transition, Fisheye, Dutch Angle, Focus Change | Scene changes and unusual angles |
| Time and speed | Bullet Time, Hyperlapse, Timelapse Glam, Timelapse Human, Timelapse Landscape, Low Shutter | Freezing a moment or compressing hours |
| Still and style | Static, General, Wiggle, Glam, BTS | Dialogue, talking heads, behind-the-scenes looks |
Higgsfield’s camera guide defines the basic moves in its own words: a pan turns the camera on the spot with no change in location; a dolly is “a real change in distance”; an orbit circles a centred subject and is the “go-to move for product reveals and hero shots”; tracking holds “a consistent distance and relative position” beside a moving subject. Its Cinema Studio has its own list of camera movement presets, “30+” of them, including POV, Robot Arm, Pan Left and Helicopter Shot, so a name on one list may not appear on the other.2
The same guide lists the lens and aperture settings worth keeping fixed for a matching series:
| Setting | Names Higgsfield uses |
|---|---|
| Lens looks | Clean Sharp · Anamorphic · Vintage Anamorphic · Warm Vintage · Halation Vintage |
| Aperture | f/1.4 Wide Open · f/4 Moderate · f/11 Deep Focus |
Higgsfield, How to Control Camera Movement, Angles, and Lens in AI Video, re-read at source 12 Oct 2026; Anamorphic gives “horizontal oval bokeh” and Halation Vintage a “glowing red-tinged halo”, in Higgsfield’s words.
Matching shots come from matching settings. Write down the preset, lens and aperture of the first good shot and reuse all three, word for word, for the rest of the series.
Keeping the same character: Soul ID and references
The plain point: for the same face in every shot, Higgsfield’s answer is Soul ID, an identity you train once from photos. Higgsfield describes it as “a trained identity layer inside Higgsfield’s Soul 2.0 image model”. It asks for at least 20 photos of one person, ideally recent (the last four to five months) and in consistent lighting, and warns: “Sunglasses, heavy shadows, cropped faces or busy backgrounds degrade results.” Training takes about three to five minutes, in Higgsfield’s figure, and “A Soul ID trained for images carries into video”: you animate a generated frame with a video model.3
What Higgsfield says it does not do. Soul ID “locks identity, not direction”: pose, scene and styling still come from presets and prompts. It is “stronger on close-up and mid-range shots” than on wide shots where the face is small, and it “maintains a consistent identity across generations rather than guaranteeing a perfect match every time.” Soul Reference is a different tool: you upload a reference image and it copies the look (composition, lighting, pose, mood), not the person.4
Why the template keeps to a medium close-up: Higgsfield says Soul ID holds best at close and mid range. The wider approach to identity across tools is on character consistency in AI.
Hooks for UGC ads
The plain point: a UGC ad (user-generated-content style: one person talking to a phone camera) lives or dies on its opening line, so test openings one at a time. Higgsfield’s August 2026 guide sorts these ads into six formats, each with what it needs. The opening lines in the last column are this site’s examples, not Higgsfield’s.5
| Format | What it needs, and where it runs | An opening line to adapt |
|---|---|---|
| GRWMget ready with me | Creator plus product in a routine · TikTok, Reels | “Five minutes before work, and this is the only thing I reach for.” |
| Tutorial | Product plus 3-4 usage steps · TikTok, YouTube Shorts | “Three steps, and most people skip the second one.” |
| Before / after | Two states of the same subject · Reels, TikTok | “Same kitchen, two weeks apart.” |
| Testimonial | Creator plus a specific claim · Meta feed, Reels | “I didn’t expect to keep using it after the first week.” |
| Unboxing | Product packaging · TikTok, YouTube Shorts | “Opening this on camera, first reaction, no edits.” |
| POV demo | First-person product use · TikTok | “POV: it’s Monday and the inbox is already full.” |
How Higgsfield says to test a hook: keep the product, the creator, the preset and the setting fixed and “change only the opening line”, so any difference in results comes from the hook; compare formats or presets in a separate batch. For one face across the whole campaign, the guide points to Soul ID: set the avatar up once and reuse it across products, presets and placements.6
Building motion-design ads with Higgsfield from an AI assistant is a separate workflow: Higgsfield MCP + Claude: make motion design ads with AI.
Does it take a negative prompt?
Until Higgsfield documents a negative field, describe what you want to see rather than what you do not.
Corrected 16 Sep 2026: an earlier version of this sheet said flatly that Higgsfield has no dedicated negative field, and that putting a specific person in the shot is the model's strength. Neither appears in Higgsfield's camera-control guide or its camera controls page, so both were removed.
Treat the move as a setting and its wording as a fixed phrase. Change neither between shots that need to match.
What not to do
Free or paid? What the free plan allows, who owns the output, and when you need to pay: free vs paid AI image, video and music tools.
Where this model sits among the others: which models take a negative prompt, and what each vendor publishes as its formula.
Camera language that survives across models is on the anatomy of a video prompt.
The checklist
Before you send it:
Higgsfield, How to Control Camera Movement, Angles, and Lens in AI Video · Higgsfield, Camera controls All read at source on 16 Sep 2026 and re-read 12 Oct 2026.
Added 12 Oct 2026: Higgsfield, Why Does Your AI Character’s Face Keep Changing? (9 Jun 2026) · Higgsfield, How to Make AI UGC Videos for Marketing in 2026 (20 Aug 2026), both read 12 Oct 2026.
Model versions and vendor documentation both move. Re-read the source before relying on a specific number.
Every other sheet in this set: all model cheat sheets.
Related on this site: Higgsfield vs Freepik vs Krea vs OpenArt (Freepik is now called Magnific)7 · Higgsfield Prop Sheets · What Is Higgsfield? · Higgsfield Camera Presets
- 1Higgsfield, Higgsfield Camera Controls – 50+ Cinematic AI-Motion Presets, read at source 12 Oct 2026 (preset names copied as listed).↑
- 2Higgsfield, How to Control Camera Movement, Angles, and Lens in AI Video, re-read at source 12 Oct 2026“Camera movement presets set how the camera moves, with 30+ presets available”. ↑
- 3Higgsfield, Why Does Your AI Character’s Face Keep Changing?, 9 Jun 2026, read at source 12 Oct 2026“Quality beats quantity.” ↑
- 4Higgsfield, Why Does Your AI Character’s Face Keep Changing?, 9 Jun 2026, read at source 12 Oct 2026.↑
- 5Higgsfield, How to Make AI UGC Videos for Marketing in 2026, 20 Aug 2026, read at source 12 Oct 2026 (format, platform and “what it needs” columns).↑
- 6Higgsfield, How to Make AI UGC Videos for Marketing in 2026, 20 Aug 2026, read at source 12 Oct 2026“so one face carries the whole campaign.” ↑
- 7Magnific, Pricing plans | Magnific (formerly Freepik), read at source 27 Oct 2026page title “Pricing plans | Magnific (formerly Freepik)” ↑
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