In this time of life maps and video tools, the threshold of creation has fallen to unprecedentedly low levels, whether by producing amazing pictures with Midjourney, Nano Banana Pro, or by using Kling, Hailuo or Dream AI vision 2.0. Anyone can get a very visual impact image by entering a few simple hints。
However, as the sense of freshness fades, many people encounter an unbreakable bottleneck:
WHY IS MY AI IMAGE, NO MATTER HOW BEAUTIFULLY WRITTEN IT IS, AND HOW ADVANCED IT IS, STILL CARRIES A THICK "PLASTIC AI"
MANY PEOPLE THINK THAT THEIR HINTS ARE NOT COMPLEX ENOUGH, SO THEY'RE A CRAZY PILE OF LIGHT, MATERIAL VOCABULARY. BUT TODAY I WOULD LIKE TO TELL YOU A CRUEL TRUTH: THE IDEA THAT THESE AI IMAGES ARE FILMED IS BASED NOT ON THE PRETENSE OF A PILE OF SUBJECTS, BUT ON A REASONABLE “FOREGROUND COVER”。
THE BIGGEST PROBLEM FOR AI WAS NEVER “UNREAL”, BUT “TOO CLEAN”. IT ALWAYS TENDS TO PRESENT EVERYTHING PERFECTLY AND UNRESERVEDLY BEFORE YOU. IN REAL FILM PHOTOGRAPHY, THE OUTLOOK IS MORE THAN JUST DECORATION, IT'S A SHARP OBJECT USED TO BREAK THE AUDIENCE'S SIGHT, AND IT BRINGS THE AUDIENCE CLOSER TO THE PICTURE. AND MOST OF ALL -- THESE MASKS ARE BASICALLY TELLING THE AUDIENCE WHERE THE CAMERAS ARE。
TODAY, I'M GOING TO TEACH YOU TWO PROMISING TECHNIQUES THAT CAN QUICKLY REMOVE THE AI FROM VIDEO AND PICTURE WITHOUT A COMPLICATED OPERATION AND WITH A SIMPLE HINT。
Method I: Physical protection of the future — creation of “space sense” and spatial layers
The biggest problem with many people writing their hints is that there's only the subject, there's no prospect。
FOR EXAMPLE, IF YOU WANT TO DRAW A GIRL IN THE FOREST, IF YOU ONLY DESCRIBE THE GIRL AND THE FOREST, AI WILL PUT THE SUBJECT DIRECTLY IN THE CENTER. THIS CAN LEAD TO A LACK OF “TIME SENSE” OF THE IMAGE. IN THE REAL WORLD, THERE ARE VERY FEW INSTANCES WHERE THE SIGHT IS COMPLETELY INVISIBLE. THE LEAVES, THE COLUMNS, THE SHOULDERS OF THE ROADERS ARE ALL SUBTLY IMPLYING THE EXISTENCE OF PHYSICAL SPACE。
Wang Jia Wai's film screenshot
Core principles: Why do we need physical cover
THE CURRENT MAINSTREAM AI PAINTING TOOL IS STILL ESSENTIALLY “CHARACTERIZED” PIXELS ON THE 2D PLANE. IF YOU DON'T GIVE A CLEAR COVER, AI WILL DEFAULT THAT THE CAMERA IS A "EYE OF ALL KNOWLEDGE FLOATING IN THE AIR."。
In the real film industry, however, the studios use top-level digital cameras such as Arri Alexa35, which occupy absolute physical space when they are filmed with Cooke S4 or Panaviment C-Series. It may lie behind the bushes and hide by the door frame. The nature of the physical cover is to deceive the brain with the use of “perceived relationships” and “deepness”. When there is an extremely vague object at the top of the picture, the brain automatically calculates the Z-axis (deep) space, smashing a three-dimensional world in a two-dimensional picture。
There's a lot of different scenarios in television dramas, creating an atmosphere, communicating the environment
2. Catalyst guide: common errors where newers are the easiest to step on
FATAL ERROR 1: USE AI AS A "PHOTOGRAPH" INSTEAD OF A "FILM CAMERA."
Many people use the Nano Banana Pro or dream word in extremely detailed terms, but all about the subject:
"Choose Camera":
A beautiful girl, wearing a white dress, standing in a foreground, highly detached face, 8k, clinical lighting.
♪ a beautiful girl in a white dress, standing in the forest, full of face details, 8k, film light ♪
"The real pain":
THIS IS SIMPLY A HINT FROM THE THINKING OF “DOCUMENTARY PHOTOGRAPHS” OR “SHOWING OF TREASURE PRODUCTS”. NOW AI IS VERY POWERFUL, AND IT'LL GIVE YOU A PERFECT "BEAUTY HEAD SHOT" OR "WHOLE PICTURE."。
BUT IT'S THE BIGGEST PIT: IT'S TOO CLEAN, TOO WHITE. AI FAITHFULLY CARRIED OUT YOUR ORDERS, PUTTING THE MAIN CHARACTER IN THE ABSOLUTE VISUAL CENTER OF THE PICTURE, AND NOTHING STOPPED HER. THIS MAKES IT LOOK LIKE A BEAUTIFUL “WALLPAPER” OR “PLAY ORIGINAL”, WITHOUT A “TIME SENSE” OR A “STORY SENSE”. THE AUDIENCE CAN SEE AT FIRST SIGHT THAT IT WAS CREATED BY AI, BECAUSE REAL MOVIE TEAMS RARELY TELL STORIES WITH THIS BLIND, DRY, TOPLESS PERSPECTIVE。
The correct solution:
Don't just stare at the subject! You're going to add "where did you steal her" to the hint. Turn the term “girls standing in the forest” into “a girl standing in the forest, in the middle of an extremely defunct tropical paragus leaf, invisible”. This is the moment of the movie。
Tips:
Transformer shot, low angle shot. The most promising, primitive, deep green leaves are blurry to become almost abstract, semi-transparent, blocking parts of the picture and creating a peeking, hidden perspective. Through the gap, a lonely Chinese woman may be immersed in meditation, standing in an old, mist-covered, primitive forest. The golden light of the hour runs through the canopy of the tree and captures the small dust particles that float in the air (the size of light), in stark contrast to the low-key, dark shadows. Very shallow, delicate film particles, Kodak Portra 400 film shoot, film class orange。
Fatal error 2: Undesired shield
THE "LEAVE" OUTLOOK IS HARD ON THE INDOOR SCENE OR "GLASS WITH DROPLETS" IN THE DESERT. AI CAN PAINT, BUT IT DESTROYS THE SENSE OF LOGICAL TRUTH AND GIVES A GLIMPSE OF WHAT IS ALL ABOUT。
3. Physicological analysis: How can the foreground layer be added to the hint
Many beginners are prone to “photo-thinking” when they write hints — only the subject. Now AI (e.g. Nano Banana Pro or Dream AI) will give you a perfect "high-clean" or "big-head sticker." But it's the biggest pit: it's too clear, like a cheap studio with no story。
The real master writes "director thinking." You have to add "where are your cameras hidden" to the hint。
In this case, we're going to create a "sniffing" film. We don't take the lead, but we observe it from a highly defunct object。
[Phrasing formula]: [Industrial-grade photographic parameters] + [extremely prospective fissibility (dissemination medium)] + [median subject and action] + [background environment] + [film and colour sense]
:
Close-up on the Heavily Out-of-Focus summit of a married woman in a poorly placed, high-hotel-loby.
(Middleview close-up, ARRI Alexa 35 camera with Leitz Summilux-C shot. The edges of the texture marble pillar, which is the most foreground, and which is the most invisible, tectonic, are delicately the left-hand frame of the lens, creating a hidden, perceived perspective. In the clear picture, an elegant woman in a velvet tuxedo is making a nervous phone call in a dark, high-end hotel lobby. Warm chromosomal size light penetrates the background. Film-class depth, natural film particles, rich texture, Kodak Portra 400 film shoot, fine orange. I'm not sure
[hard nucletion: where is the high sense of the hint? _Other Organiser
From "shield" to "narrative media": we used silhouetted Edge (marble pillar) of texture. It's not just a physical obstacle, it's more like a “screen”, suggesting that the audience is hiding behind the poles. Marble's sense of texture increases the image's sense of height even when it's out of focus。
Top-level industrial lens language (Leitz Summilux-C): Leitz Summilux-C is known for its cream-like aberration and soft and natural image. To feed this model directly to AI, AI automatically retrieves the optical dispersion and colour science data at its bottom, and it becomes extremely natural, relaxed and digitally devoid of it。
WHEN YOU THROW THIS IMAGE INTO THE CASCADING OR CONCH AI TO GENERATE THE VIDEO, YOU'RE GOING TO BE ABLE TO USE IT
Just enter a short dynamic word:
“The camera is moving slightly forward, passing through the pole, and the tall woman is making a nervous phone call in the dark, high-end hotel lobby”。
AS THE CAMERA MOVES, THE LEFT MARBLE PILLAR WILL HAVE A NATURAL DYNAMIC SHIFT, AND THE UNMATCHED "NAKED EYES 3D VISION" WILL IMMEDIATELY TAKE YOUR VIDEO OUT
Method two: The atmosphere is blocking the future - breaking the vacuum and allowing the air to be seen
If "physical cover" is a real thing in front of the camera, then "air cover" is something in the air。
MANY AI-GENERATED IMAGES GIVE PEOPLE A STRONG SENSE OF “VACUITY”. IT'S LIKE PEOPLE LIVE IN A GLASS MASK THAT'S DRAINED. BUT IN THE LIGHT OF THE REAL WORLD, THERE MUST BE A DEPENDENCE ON MEDIA SUCH AS SNOWFLAKE, FOG, DUST. THIS WILL GIVE THE IMAGE AN ADVANCED THICKNESS。
Core principles: air vision and medium refraction
In the real physical world, the transmission of light cannot be 100 per cent penetrating. The air is filled with dust and small particles. The more objects are far from the lens, the more spaced air media, the lower the contrast。
Why is every frame of animated film so beautiful as to make wallpaper? He was always in the air of his image, with small light dust, petals or rain. When the light refractions in these media, there is a very strong sense of breathing。
New Hae-cheng film screenshot
The point is not for the Gatt effect, but for the physical space to be seen. When the light has a path, the image becomes a thick image。
2. Hole avoidance guide: common operations to destroy film sense
Fatal error 1: The misuse of “extreme detail” leads to a plastic sense that AI automatically erases “impurities” (i.e., air media) from the image when you constantly reinforce “8k, detail, clean sharp Focus” in the hint. It's too clear, and it's AI's heavy-seated culprit。
FATAL ERROR 2: ONLY "LIGHT" AND NOT "DUST" ONLY "LIGHT" OR "DINDAR LIGHT" BUT WITHOUT MEDIA SUPPORT, AAI PAINTS A COLUMN OF LIGHT AS IF IT HAD BEEN PAINTED HARD AT A LATER STAGE WITH A BRUSH, WHICH IS VERY HARD TO RECONCILE。
Pyrolysis: introduction of “particle sense” hints to make air
To allow AI to understand the thickness of the space, both "photo-source" and "photo-media" must be present in the hint. You can add the words "particles " , " haze " , " dust motes " , " medium sense " 。
MANY NEWCOMERS DO EMOTIONAL SHORTS OR MVS, AND THE MOST VULNERABLE AREA OF ERROR IS “ABSOLUTE CLEAN”。
I'm sorry
A young man standing by the window of a small office at sunset, Makoto Shinkai 2D anime style, highly detailed, 8k, clean clear sky.
(A young man standing by a small office window at sunset, Animation style, high detail, 8k, clear sky. I'm not sure
IT'S VERY SHARP, IT'S VERY COLORFUL, BUT IT'S LIKE PEOPLE ARE ON THE BACKBOARD. BECAUSE THE AIR WAS DRAINED AND THE LIGHT WAS NOT RETROCIRCULATED, THE WHOLE PICTURE WAS FULL OF CHEAP “WEB GAME CG SENSES” WITH NO EMOTION。
We now inject “air” and “heat” into this narrative space。
:
A young man standing by a wide window of a small, numbered office at gold hour, looking out the room.
2D ANIMATED FILM STYLE. A YOUNG MAN STOOD IN FRONT OF THE WINDOW OF A SMALL, GROCERIES-FILLED ENTREPRENEURSHIP OFFICE AT “GOLDEN TIME” (TWILIGHT) AND LOOKED OUT THE WINDOW WITH SOME THOUGHT. THE ATMOSPHERE WAS THICK AND THIN IN THE ROOM. THE STRONG DINDAL EFFECT (CHRIST'S LIGHT) PENETRATES THE WINDOW GLASS. AT THE FOREFRONT ARE SMALL DUST PARTICLES FLOATING IN SUNLIGHT BEAMS AND EMITTING WARM LIGHT. DARK SHADOWS, THICK WARM AMBERS AND SOFT AND PURPLE TUNINGS, DEEP IN FILM CLASS. I'M NOT SURE
[hard nuclear disassembly: How does the atmosphere give the image a "godly emotion"? _Other Organiser
Breaking the vacuum, seeing the path of light: In the real world, there must be a trajectories in the sun. With the addition of Tyndall effect (Dindar effect) and atmosheric haze (temperature mist), the light is no longer a hard brightness, but rather a physical column with a mass, which gently wraps up the person。
dust dust is the ultimate romance: and it's a new sea spirit. these luminous little particles not only constitute a prospect of extreme restraint, but also inject a sense of quiet, sad time。
THE "RESPIRATION" OF THE GRAPHIC VIDEO: WHEN YOU PUT THIS VERY THICK IMAGE IN THE CASCADING OR CONCH AI
Just enter a simple hint:
Animated style, men standing behind windows overlooking sunset cities. At dusk, a strong volume of light penetrates the room. The golden dust slowly floats in the light column. Men look up at the sky. Film sense, depression, smooth dynamic effects。
THE SLOW FLOW OF DUST IN THE AIR IS ENOUGH TO HOLD UP A VERY TOUCHING, EVEN DIRECT, HIGH-QUALITY, LONG SHOT OF THE VIDEO