
AI DrawingTo date, it has not been the time to enter “a good-looking girl” to meet creative needs。"Photography" and ""Lens Terminology”It's the "key" that separates professional creators from ordinary users。
We can look at it in two ways:
I. PHOTOGRAPHIC TECHNIQUES: "DICTIONAL LANGUAGE" OF AI DRAWINGS
YOU CAN IMAGINE AI AS A PHOTOGRAPHER WITH INVINCIBLE TECHNOLOGY, BUT AESTHETICS AND DIRECTIONS NEED YOUR GUIDANCE. THAT'S WHAT YOU'RE TALKING ABOUTscene-by-scene script.
Composition and Perspective:
- Example of terms(b) Dutch angle (Dutch Angle Tilt), extreme close-up (high feature), low-angle shot (low angle) and over-the-shoulder shot (over shoulder lens)。
- ImportanceI said, "A girl and a cat in front of a bookstore," and AI might give you a mediocre look. But if you say low-angle shot, looking up at a girl and a cat sitting on the floors of an an old bookshop, tossing Mediterranean image in background, it's a different story. The perspective determines the psychological relationship between the audience and the role。
Light and Shadow:
- Example of termsRembrandt Lighting (Lembrandt Light), Cinematic Lighting (filmlighting), Backlighting (reverse light), Golden hour (golden hour), High-key Lighting (highlighting)。
- Importance: I will further describe “summer midday, strong light”, but if I add a dramatic rim light on the girl and cat, deep shadows contrast with bright medialanean walls (a strong reversal of the light, drawing dramatic contours for girls and cats in contrast to the bright Mediterranean wall). This immediately escalated from “a snapshot” to “an intentional work of art”. Light is emotional brush。
Dynamics and Fuzzy:
- Example of terms:motion blur (motion fuzzy), bokeh (scatter), shallow depth of field (sand deep), panning shot (scrambling focus)。
- Importance: you mentioned "the feeling of wind," which is the meaning. it can be: wind blowing through the girl's hair, shallow depth of film with cleary bokeh of background buildings, motion blour on moving leaves (the hair of a wind blowing girl, the shallowness of the light brings a soft background and a twirl in the moving leaves). this gives the static images a sense of breathing and the passage of time。

II. LENS TERMINOLOGY: "PHYSICAL HANG" FOR AI DRAWINGS
THE LENS TERM IS A MUCH MORE NUANCED "OPTIC MAGIC," AND YOU TELL AI WHAT THE VISUAL DEVICE IS。
The magic of focal length:
- Example of terms: shot on 35mm (35mm camera), telephoto lens (long-focal lens), wide-angle lens (wide-angle lens), fisheye lens (fish glasses)。
- Importance: wide-angle lens (e.g. 16 mm) produces a strong vision and deformation, highlighting the grandness of the building and the breadth of the scene; and telephoto lens (e.g. 200 mm) compress space so that the building in the background is as close to the person as it is to the person and creates a vision of reality, softness。The same scene, with different focal lengths, tells completely different space stories。
Aperture and depth:
- Example of termsf/1.2, f/8, deep focus。
- Importance: wide vision, f/1.2, extremely shallow depth of field (big halo, f/1.2, very shallow) will force all attention to the eyes of girls and cats, and the background melts into abstract light, with a sense of privacy and emotion; f/8 or deep focus (deep focus) will make the details from cat beards to faraway buildings clear, with a calm, objective, documentary sense of view。
Film and camera sense:
- Example of terms: shot on Portra 400 (photo by Portra 400 film), anamophic Lens flare (shrizzle of deformation lens), vintage lens (old lens)。
- ImportanceIt injects "soul particles" into the work. Shot on Kodak Gold 200 and shot on Cinestill 800T will give a very different colour science and a sense of particle mass -- a warm memory, and Saber's night colour. Anamopropic Lens flare (blue radiant radiance that is characteristic of a broad screen film) can immediately inject a big Hollywood sensation。

Why is it so important
1. Exact communication to reduce guesswork: AI IS THE MASTER OF SYMBOL UNDERSTANDING. YOU GAVE IT THE SYMBOL."Prompt wordThe more professional it is and the more common it is in the human creation system, the more accurate it is in the direction of “brain patches”, which significantly reduces the number of cards drawn。
2. Style and authorship: When everyone uses "Looking good, Master's Work" and you use Chiaroscuro Lighting, Gordon Parks style documentary photography, you have unique aesthetic fingerprints。
Breaking the plane, injecting the soulPhoto is “Putting in light” and “Art of the moment”. These terms will help youtime(fast door velocity)Space(focal lens)physiotherapy(optic properties) andEmotionIt's a two-dimensional image, filled with multidimensional narratives。
In shortI DON'T KNOW, YOU'RE MAKING A PICTURE OF AI; YOU'RE GOOD AT ITDirector, photographer and light guideAND AI IS YOUR MOST POWERFUL, ALL-POWERFUL FILM CREW。

There's a method of filming that's usually called visual and visual effects "Crash Zoom" Or more specifically "Cosmic Zoom" (Cosmic Focus/From Space to Micro)If it's physical motion through drones, it's often called "FPV Drone Dive" (FPV drone dive)I don't know. I'll explain it a little, because it'll be used a lot。
For current AI video generation tools (e. g. Runway Gen-3, Luma Dream Machine, Sora, Kling, etc.) this is to be generatedLarge, consistent and high-speedPrompt has to emphasize thatContinuity of spatial migrationandA dramatic change in perspective.
Here are some of the policies and specific references for Prompt:
Key Names
TO GET AI TO UNDERSTAND THIS COMPLEX LONG SHOT, YOU NEED TO COMBINE THE FOLLOWING DIMENSIONS:
- Image by Camera Movement
FPV Drone dive
Rapid decline
God's eye view to extreme close-up
Fast zoom in
- Start (Start):
Blinding light from above
Top-down aerial view
- End (End):
Macro shot of skin texture
Sunburned skin detail
Peering skin (skinned)
4. Skills to improve success rates (Tips)
BECAUSE THE AI VIDEO MODEL IS PRONE TO HALLUCINATIONS WHEN DEALING WITH A "MACRO-MICRO" TRAJECTORIES, IT IS SUGGESTED THAT THE FOLLOWING PARAMETERS OR ADDITIONAL DESCRIPTIONS BE TRIED:
Emphasis on seamless transition: Use words such as Seamless transition or Morphing to help AI understand that this is the same shot, not a clip spell。
Division generation (if a mirror fails):
If AI is unable to generate a perfect sky-to-face lens at a time, the "End Frame" function of Runway Gen-3 or Luma can be used。
The first video: Generates a dive from the sky to the girl's body。
The second video: using the last frame of the first paragraph as the starting frame, Prompt was replaced by "the lens continues to advance to the face microspecies to show skin details" and the video was extended。
Motion range control: In Runway, you can raise the Motion parameter (e. g. 8-10) or use the Zoom In option of the Cameron Motion to maximize to fit Tilt Down (slash down)。
AND HERE I'VE MADE A FEW SIMPLE TABLES ON SOME OF THE PHOTOGRAPHY TECHNIQUES AND THE TERMINOLOGY OF PROFESSIONAL LENSES IN THE AI DRAWINGS, WHICH I HOPE WILL BE USEFUL。
| English terminology (Term/Abbr.) | English | USE TECHNIQUE & AI TIP ADVICE | Attention & Common Issues |
| POV (Point of View) | First person perspective |
Skill"Through someone's eyes" is a simulation of the world. It's often used to enhance immersion。 Prompt word: |
Notice: AI OFTEN FORGETS TO DRAW “HANDS” OR PARTS OF THE BODY AND NEEDS TO BE CLEARLY EMPHASIZED (E.G you know, hands reaching out). |
| Low Angle / World's-eye view | Low angle / look / bug vision |
SkillFrom the bottom up. It makes the subject look big, powerful, oppressive or sacred。 Prompt word: |
Notice(b) Long legs when taking a picture, but easy to get a double chin or nostrils, with an angle of control。 |
| High Angle / Bird's-eye view | High angle / Overlooking / God's perspective |
Skill: Film from higher to lower. It makes the subject look small, lonely, or displays the whole layout。 Prompt word: |
NoticeIt's easy to compress a person's height and looks small. It's good for a big scene or crowd。 |
| Eye-Level | Panorama |
Skill: Camera parallels the main eye. The most natural, objective and friendly perspective。 Prompt word: |
Notice: The most conservative, but sometimes somewhat flat, and needs to be made more fun by light and image。 |
| Dutch Angle | Dutch Horn / Tilt |
Skill: Camera tilts. It creates tension, anxiety, confusion or a dynamic sense of speed。 Prompt word: |
Notice: Don't abuse it, otherwise the audience will feel “difficult” or simply a picture。 |
| Over-the-Shoonder (OTS) | Pass the shoulder lens |
Skill: Look over a person's shoulder at another person/matter. Creating a sense of personal relations or dialogue。 Prompt word: |
Notice: The shoulders of the outlook are often vague (false) and need to be controlled。 |
| Selfie | Selfie |
Skill: Simulate self-censorship action. Arms extended, face deformed slightly。 Prompt word: |
NoticeIT'S HARD FOR AI TO DRAW A FINGER FOR A CELL PHONE i don't knowCover your fingers。 |
| English terminology (Term/Abbr.) | English | USE TECHNIQUE & AI TIP ADVICE | Attention & Common Issues |
| Exchange-Up (ECU) | / Extremely close |
SkillLocal focus (e.g. eyes, lips, watches). Emphasis on detail texture。 Prompt word: |
Notice: Except for the focus, the surroundings are usually very vague (Boke)。 |
| Close-Up (CU) | Nearview / Headshot |
Skill: usually means above the chest or only the head. It's not like I'm going to have to do this。 Prompt word: |
Notice: Background is usually distorted and environmental information is scarce。 |
| Medium Shot (MS) | medium term view |
SkillOver the waist. Shows the action and part of the environment, the most common narrative view。 Prompt word: |
NoticeTHIS IS THE MOST STABLE PICTURE OF AN AI-GENERATED PERSON, WITH A LOWER RATE OF HAND AND FACE DAMAGE。 |
| Cowboy Shot / MLS | Cowview / Medium Vision |
Skill: Up to knee (from western film to show a gun). Show confidence and body language。 Prompt word: |
Notice: IF YOU DON'T WANT AI TO CUT OFF THE PERSON'S LEG, USE THE WORD full body More to ensure facial clarity。 |
| Full Body/Wide Shot (WS) | Full body / wide corner |
SkillThis post is part of our special coverage Global Voices 2011. Emphasis on the relationship between people and the environment。 Prompt word: |
Notice:AI, IT HURTSI don't know. Distance leads to low distribution of facial pixels and loss of face (spread code). Need to cooperate high resolution face Or later repaint to repair the face。 |
| Extreme Long Shot (ELS) | Vision / Grand Vision |
Skill: Very small person, very large environment. It's an expression of an epic sense, a mild sense or a view。 Prompt word: |
Notice: People are basically just a contours, don't expect |
| English terminology (Term/Abbr.) | English | USE TECHNIQUE & AI TIP ADVICE | Attention & Common Issues |
| Exchange-Up (ECU) | / Extremely close |
SkillLocal focus (e.g. eyes, lips, watches). Emphasis on detail texture。 Prompt word: |
Notice: Except for the focus, the surroundings are usually very vague (Boke)。 |
| Close-Up (CU) | Nearview / Headshot |
Skill: usually means above the chest or only the head. It's not like I'm going to have to do this。 Prompt word: |
Notice: Background is usually distorted and environmental information is scarce。 |
| Medium Shot (MS) | medium term view |
SkillOver the waist. Shows the action and part of the environment, the most common narrative view。 Prompt word: |
NoticeTHIS IS THE MOST STABLE PICTURE OF AN AI-GENERATED PERSON, WITH A LOWER RATE OF HAND AND FACE DAMAGE。 |
| Cowboy Shot / MLS | Cowview / Medium Vision |
Skill: Up to knee (from western film to show a gun). Show confidence and body language。 Prompt word: |
Notice: IF YOU DON'T WANT AI TO CUT OFF THE PERSON'S LEG, USE THE WORD full body More to ensure facial clarity。 |
| Full Body/Wide Shot (WS) | Full body / wide corner |
SkillThis post is part of our special coverage Global Voices 2011. Emphasis on the relationship between people and the environment。 Prompt word: |
Notice:AI, IT HURTSI don't know. Distance leads to low distribution of facial pixels and loss of face (spread code). Need to cooperate high resolution face Or later repaint to repair the face。 |
| Extreme Long Shot (ELS) | Vision / Grand Vision |
Skill: Very small person, very large environment. It's an expression of an epic sense, a mild sense or a view。 Prompt word: |
Notice: People are basically just a contours, don't expect |