24/11/2024, 18:20 Hailuo AI tutorial

Hailuo AI tutorial
2 more properties

Hailuo AI Text to Video Tutorial

Getting Started : What is Prompt
1. Basic Prompt Formula
2. Precise Prompt Formula
3. Tips for using Prompt Formulas
Intermediate: How to achieve top performance
Precise Camera movement
Precise Aesthetic Control
Prompt optimization: lens description and aesthetic expression
Hailuo AI Image to Video Tutorial
Image to Video Tutorial
1. I2V Basic Prompt Formula
2. I2V Precise Prompt Formula
The Advantages of Hailuo AI
Vivid Emotional Expression
Cinematic Explosion Effects
Concept Combination
Practical Application Cases - (Movies, Commercials, Promotional Videos, Short Videos)
Text-to-video Cases for Practical Application
1 Mainly-Used Cases
2 Specially-Used Cases：Montage with Similar Objects
PS：Ways to Improve Subject Stability
Image-to-video Cases for Practical Application
1 Mainly-Used Cases
2 Specially-Used Cases：Using the last frame to create long shot
3 Specially-Used Cases： Image composition and seamless blending across transitions
Comprehensive Application Case

Hailuo AI Text to Video Tutorial

Getting Started : What is Prompt
Prompt is the key to the world of generated AI, constructed from your inspiration and creativity.
Whether you are a newcomer to AI or a veteran who is already proficient in using AI to empower
productivity, you can easily master the skills of writing prompts through a clear structure and a
flexible way of expression. The vision of “one person + one AI = one professional crew” is right
around the corner.
To improve the quality of video results, we have developed two Prompt formulas of your use:

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 1/29
24/11/2024, 18:20 Hailuo AI tutorial

1. Basic Prompt Formula

Basic Prompt Formula is designed for those who are uncertain about the exact videos they wish to
create, or those who seek to use Hailuo AI as a source of inspiration. This type of prompt will offer
you a range of more incentive visual ideas.
Basic Prompt Formula= Main Subject + Scene + Motion
Main Subject: This core message of the video can be a person, an animal, another object or an
imaginary object that does not exist.
Scene：The scene describes the environmental information of the video, etc a library, a coffee shop
...... or even a fantastical fictional space.
Motion: Motion is the description of the main object's moving status in the video, such as stillness,
movement; the transformation of the environment and space, and so on.
Example
A puppy runs in the park.
A woman with an umbrella walks down a rainy street.
A small stream flows quietly in a valley.

2. Precise Prompt Formula

The Prompt Precision Formula is designed for users requiring advanced camera movement or image
presentation, aiming for a more professional-grade video production. Precise and detailed prompts
can enhance the visual quality of the footage.
Precise Prompt Formula = Main Subject + Scene + Motion + Camera Movement + Aesthetic
Atmosphere
Camera Movement: You can use common camera movements, such as tracking, zooming, panning,
handheld and other camera movements to restrict the final presentation.
Aesthetic Atmosphere: You can describe the visual style and the sense of atmosphere to get a
better result.
For Example：
A couple sit on a park bench communicating. The camera maintains a fixed shot of the couple.
The color tone of the picture is warm, the atmosphere is cozy.
A lamb is grazing in a meadow. The camera slowly pushes forward the lamb. The color tone of
the picture is natural and realistic.
A man in a suit eats noodles in a noodle shop. The camera gradually pulls away to show the
noisy environment of the noodle shop. The picture has a natural color tone.

3. Tips for using Prompt Formulas

There's no right or wrong to write a prompt. But overall, prompt writing follows 2 main basic
principles:
• The more precise the expression is, the more accurate the result will be.
• The richer the expression gets, the better the overall effect will be.

Intermediate: How to achieve top performance

Precise Camera movement
Hailuo AI is adept at responding effectively to various camera movements. It can provide
professional video users with well-defined video footage.
• Add detailed orders to the camera movement

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 2/29
24/11/2024, 18:20 Hailuo AI tutorial

For example, “Lens rotating down” can be optimized as :

“The camera descends slowly and then wraps around to the right as it descends.”
• Add detailed descriptions to the change made with camera movement
For example, “The camera shoots the giraffe from the bottom up” could be optimized as:
“The camera starts with a close-up of the giraffe's feet, slowly rising to capture the giraffe's
body, and finally stops at the giraffe's face, showing giraffe's facial expression in close-up.”
• Limit the length of the camera movement to a reasonable 5-6 seconds: avoid overly complex
shots that can lead to generation failures
The following example shows the result of adding 【precise camera movement】and 【detailed
changes in the frame】in the prompt.
Example:** The green highlight** is the output responding to the change of the prompt

Generated Video 1 Generated Video 2

0:00 / 0:05

0:00 / 0:05

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 3/29
24/11/2024, 18:20 Hailuo AI tutorial

Prompt 1 Prompt 2
The camera begins with an overhead shot of a The camera looks down on a moonlit forest
mysterious forest clearing, with a female clearing. The camera rotates around a female
warrior in armor standing alone in the center warrior in silver armor holding a longsword,
of the clearing. Her armor shimmers with a after which the camera drops down to capture
cold silver light in the moonlight, and the trees the side of the female warrior, who has blood
around her cast dappled shadows. The camera on her longsword. The camera continues to
begins to slowly descend and hover around circle, sliding past the side and gradually
her, as the perspective changes, the pulling in closer to capture the side of the
longsword in her hand can be seen trembling warrior woman's face, after which the camera
slightly, with undried blood remaining on the wraps around to her front, with the focus
blade, the camera continues to wrap around, gradually settling on a close-up of her face.
sliding from side to side, gradually drawing in
closer to capture the side of her determined
face. Her slightly clenched lower lip could be
seen, and a few loose strands of hair on her
face were soaked in sweat and plastered to her
forehead. As the camera circled around to her
front, the image became tighter, the focus
gradually coming to rest on her face. By now,
the forest in the background has blurred into a
dark shadow. The camera eventually stops at a
close-up of her front, focusing on her sharp
eyes. The moonlight hits her face, outlining
every detail of her face.

Generated Video 3

0:00 / 0:05

Prompt 3
The camera wraps around to capture a female
warrior in silver armor standing in the jungle at
night.

Precise Aesthetic Control

In addition to precise camera movement, Hailuo AI is able to accurately recognize aesthetics and
ambience descriptions in the prompt and generate beautiful graphics and dynamic designs
accordingly.
The following example shows the results of adding [Aesthetic and Atmospheric Descriptions] to the
prompt, with an increase in output quality.

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 4/29
24/11/2024, 18:20 Hailuo AI tutorial

Example 1:** The highlighted portions of the Prompt are additional aesthetic descriptions.

Generated Video 4 Generated Video 5

0:00 / 0:05

0:00 / 0:05

Prompt 4 Prompt 5
On the streets of a modern city, the camera In the modern city, the camera first follows a
first follows a motorcycle as it speeds between motorcycle as it speeds through the buildings,
skyscrapers. The wheels leave deep imprints then pans quickly right to capture the police
on the wet ground as the surrounding car in hot pursuit behind it, with the busy
buildings sway in the ever-moving wind. The streets, traffic and lights as a blurred
camera pans quickly right to catch a police car backdrop.
following close behind, its lights flashing,
reflecting the neon lights and traffic on the
street in brilliant colors. The bustling cityscape
blurs into a flow of light in the background,
and the movements of the motorcycle and
police car seem unusually hurried and tense.

Generated Video 6

0:00 / 0:05

Prompt 6
The motorcycle sped through the city with
police cars in hot pursuit.

Example 2:** The same basic prompt can be rendered very differently with different aesthetic
ambience descriptions.

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 5/29
24/11/2024, 18:20 Hailuo AI tutorial

Generated Video 7 Generated Video 8

0:00 / 0:05 0:00 / 0:05

Prompt 7 Prompt 8
The camera captures a woman sitting in a cafe, The camera shoots a woman sitting in a cafe,
the woman looks up in thought, the camera the woman looks up at the window, the
slowly moves to capture the street outside the camera slowly moves to capture the street
window, the image is gray in tone, the colors outside the window, the image shows warm
are undersaturated and the atmosphere is tones, rich colors and a relaxing atmosphere.
somber.

Generated Video 9 Generated Video 10

0:00 / 0:05 0:00 / 0:05

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 6/29
24/11/2024, 18:20 Hailuo AI tutorial

Prompt 9 Prompt 10
A male in a trench coat walks down a street in A male in a trench coat walks down a street in
a modern city, the camera follows the male's a modern city, and the camera follows the
movement from behind, the image has a male's movements from behind, with a brilliant
grayish tone and a somber atmosphere. sci-fi atmosphere and colorful tones.

Generated Video 11 Generated Video 12

0:00 / 0:05 0:00 / 0:05

Prompt 11 Prompt 12
The camera captures a seaside wilderness with The camera shoots a seaside wilderness, the
grass blowing in the sea breeze and a vast grass is blown by the sea breeze, and in the
ocean in the distance, with somber tones and distance is a vast ocean, the color tone of the
low color saturation. picture is warm, and the picture has a warm
atmosphere.

Generated Video 13 Generated Video 14

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 7/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 0:05 0:00 / 0:05

Prompt 13 Prompt 14
Close up shot of a boy reading a book at a Close up shot of a boy reading a book at a
desk in a room with strong contrasts of light desk in a room with warm colors and a cozy,
and dark, gray tones, and a tense, depressing warm atmosphere.
atmosphere.

Generated Video 15 Generated Video 16

0:00 / 0:05 0:00 / 0:05

Prompt 15 Prompt 16
Two women driving a blue Cuban convertible Two women driving a blue Cuban convertible
sedan on a seaside highway, the camera looks sedan on a seaside highway, the camera looks
down on the car from the air, following the down on the car from the air, following the
car's movement, the image has a somber tone, car's movement, the image has bright, rich
very low color saturation, and a depressing tones and a relaxed atmosphere.
atmosphere.

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 8/29
24/11/2024, 18:20 Hailuo AI tutorial

Prompt optimization: lens description and aesthetic expression

Hailuo AI's precise lens control and aesthetic understanding can help users generate more
responsive outputs

Generated Video 17 Generated Video 18

0:00 / 0:05 0:00 / 0:04

Original Prompt 17 Optimized Prompt 18

A man on the rooftop of a tall building at night Under the deep blue night sky, the neon lights
looks down at the city. of the city are reflected in puddles of freshly
rained water on the rooftops of tall buildings.
The camera slowly rises from the reflection in
the water, through the rain, to reveal a black
figure standing on the edge of a rooftop. The
camera then continues to rise to show a
sprawling modern cityscape. Finally, the
camera freezes and the lone figure standing
on the edge stares at the city, the interplay of
light and shadow creating a cold atmosphere.

Generated Video 19 Generated Video 20

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 9/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 0:05 0:00 / 0:05

Original Prompt 19 Optimized Prompt 20

The camera goes through the glass and The camera shoots a modern glass-look high-
focuses on a pensive lady in the window. rise building through a clear glass, the building
shimmering with reflected light in the sunlight.
The camera then closes in on a window, where
reflections on the glass show a bustling
cityscape outside. The camera slowly pulls
back and through the glass, eventually
focusing on a woman in a suit inside the
window. She sits in a minimally decorated
office, staring intently at a computer screen.
Her face appears pensive and introspective in
the soft light, contrasting with the hustle and
bustle outside the window. The office is
decorated in a simple and modern way,
emphasizing her focus and introspection.

Generated Video 21 Generated Video 22

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 10/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 0:05 0:00 / 0:05

Original Prompt 21 Optimized Prompt 22

In a forest covered in dark clouds, the young In a forest shrouded by dark clouds, which
boy sprints through the mud with his sword in were low in the sky, the young boy held a
his hand, his sleeves swinging in the wind. The sharp sword with a cold glint in his hand, and
camera pushes forward and lightning walked quickly on the muddy ground with firm
illuminates his resolute face. footsteps. His sleeves hunt in the cold wind, as
if adding a touch of vigor to his determination
at this moment. The camera slowly advances
from his side, and the trees in the background
sway in the wind, making it look particularly
eerie. Suddenly, a bolt of lightning breaks
through the gloomy sky, instantly illuminating
the young boy's determined and resolute face
as he looks ahead with determination in his
eyes.

Generated Video 23 Generated Video 24

0:00 / 0:05 0:00 / 0:05

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 11/29
24/11/2024, 18:20 Hailuo AI tutorial

Original Prompt 23 Optimized Prompt 24

When the boy plays the game, the character In a dimly lit room, a young boy is captured in
on the screen suddenly comes to life and close-up, his eyes glued to the glowing screen
interacts with the boy, blending reality with of a game console. The camera is slightly
the virtual. above the horizon, focusing on his
concentrated expression as his fingers deftly
manipulate the controller. Suddenly, the mood
changes as the figure of a game character
suddenly appears and breaks through the
screen. The camera has a shallow depth of
field and pans smoothly between the boy and
the now three-dimensional figure, blurring the
edges of the console and the wall behind it,
which is plastered with posters for various
games. The boy interacts with the character in
surprise, his movements reflecting the
excitement and disbelief of the moment. The
characters are vivid and detailed, jumping and
dancing around the room and playing with the
boy. The camera captures the fantastical
scene with dynamic angles and fluid
movements, enhancing the sense of wonder
and movement through a skillful combination
of slow motion and fast editing. Subtle
changes in lighting enhance the realism of the
characters' movements, with shadows and
highlights interplaying on the boy and the
characters.

Hailuo AI Image to Video Tutorial

Image to Video Tutorial

1. I2V Basic Prompt Formula
The image used in image-to-video generation appears as the first frame of the video, defining the
subject's appearance and the basic aesthetic style of the video. Therefore, compared to text-to-
video prompts, image-to-video prompts can appropriately reduce the amount of information.
Basic Prompt Formula=Main Subject in the first frame + Motion/Change
Main Subject in the first frame：The objects in the scene, including key details and environmental
information, are precisely recognized by Hailuo AI, which can identify characters, objects, and other
visual elements in the image and generate videos based on the prompt requirements.
Motion/ Change：Motion is the description of the main object's moving status in the video, such as
stillness, movement; the transformation of the environment and space, and so on.

Specific Cases：（The image was generated using MidJourney.）

Generated Video 25 Generated Video 26

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 12/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 0:05
0:00 / 0:05

Image 25 Image 26

Prompt 25 Prompt 26
In the scene, a small dog emits a blue light A blue furry creature in the scene is constantly
from its eyes. The messy clothes in front of the stirring a steaming pot of soup. Then, the blue
dog also begin to glow blue and slowly float creature blows on the pot, causing the soup in
into the air, folding neatly on their own before the bowl in front of it to freeze into a block of
landing in front of the dog. Afterward, the blue ice.
light in the dog's eyes fades away.

2. I2V Precise Prompt Formula

Adding camera details and atmosphere information to the basic formula allows for more precise
control over the video, resulting in more dynamic or distinctly stylized outputs.
Precise Prompt Formula= Main Subject in the first frame + Motion/Change +Camera Movement
+ Aesthetic Atmosphere
Camera Movement：Image-to-video generation can also precisely control camera movements.
Using the information in the image as a reference in the prompt can lead to more accurate results.
Aesthetic Atmosphere：Although the first frame establishes the visual tone and atmosphere of the
video's beginning, Hailuo AI's image-to-video generation can still adjust the visual aesthetics and
mood to produce content that better aligns with expectations.

Specific Case：（The image was generated using MidJourney.）

Generated Video 27 Generated Video 28

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 13/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 0:05 0:00 / 0:05

Image 27 Image 28

Prompt 27 Prompt 28
A cat in the scene runs quickly toward the The camera pulls back as a man runs toward it.
camera, with white electric sparks emanating Shadows of figures rapidly close in from
from its eyes. Its entire body becomes behind, growing larger and larger. As one
surrounded by electricity as it runs faster and shadow draws near, it reveals a humanoid
faster. The scenery on both sides rushes creature with a goat-like face, its features
backward rapidly, creating motion blur that illuminated by a yellowish light. The color tone
transforms into a glowing white time tunnel. of the scene shifts to an eerie palette, creating
a chilling and terrifying atmosphere.

The Advantages of Hailuo AI

Vivid Emotional Expression
Hailuo AI possesses top-tier facial expression capabilities in the current AI video field, capable of
generating diverse and vivid expressions based on emotion-related prompts. Additionally, Hailuo AI
can precisely control the intensity of a single expression and achieve smooth transitions between
multiple expressions.

Generated Video 29 Generated Video 30

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 14/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 0:11
0:00 / 0:05

Prompt 29 Prompt 30
In a café, a girl is listening to the boy sitting A woman walks toward the camera, crying,
across from her. Her expression starts off with the destroyed ruins of a city behind her.
happy, suddenly turns to surprise, and then
shifts to sadness.
(From X user: el.cine)

Generated Video 31

0:00 / 0:05

Prompt 31
A blond boy sits in a brightly lit classroom,
surrounded by many classmates. He looks at
the test paper on his desk with a puzzled
expression, furrowing his brow. Then, he
closes his eyes and begins to cry sadly. After
he starts crying, the classmates around him
turn their heads to look at him.

Cinematic Explosion Effects

Hailu AI can easily create cinematic-quality explosion effects.

Generated Video 32 Generated Video 33

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 15/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 0:05 0:00 / 0:10

Prompt 32 Prompt 33
In an abandoned factory area illuminated by An explosion occurs in an abandoned factory,
flames, a white sedan weaves through a and a massive mechanical dragon emerges
massive explosion. Burning buildings and from the ruins.
flying debris trace brilliant fiery arcs in the air.
The low-angle shot follows the car as it races
through the explosion, capturing its thrilling
speed and the dazzling light of the flames.

Generated Video 34 Generated Video 35

0:00 / 0:05 0:00 / 0:04

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 16/29
24/11/2024, 18:20 Hailuo AI tutorial

Prompt 34 Prompt 35
In a dilapidated warehouse, dust flies through The camera captures a low-angle shot of a
the air, and the atmosphere is thick with speeding car racing through the night, its
smoke and the scent of gunpowder. Suddenly, wheels spinning fiercely in the fiery glow on
explosives detonate, and intense flames and the ground. The camera rapidly zooms in on
thick smoke instantly engulf the entire scene, the car, when suddenly, a shockwave from an
with flames dancing on the walls. Broken explosion violently lifts the car's body.
wooden planks and metal debris scatter in all Immediately after, the camera spins quickly,
directions, while the camera shakes violently capturing another car emerging from the
from the explosion's impact. The surrounding smoke of the explosion, as the intense chase
light flickers and shifts unpredictably with the unfolds, intertwining with flashes of fire and
bursts of fire. darkness.

Concept Combination
Hailuo AI possesses strong combination abilities, allowing it to combine various different
materials/characteristics in any configuration and freely created the desired visual scenes.

Generated Video 36 Generated Video 37

0:00 / 0:05

0:00 / 0:05

Prompt 36 Prompt 37
A big cat with zebra stripes walks across the A horse with leopard spots walks slowly
vast desert. through the snow.

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 17/29
24/11/2024, 18:20 Hailuo AI tutorial

Generated Video 38 Generated Video 39

0:00 / 0:05 0:00 / 0:05

Prompt 38 Prompt 39
A cute little rabbit with dragonfly wings flies in A close-up shot shows a spider made of white
the blue sky. crystal crawling on a child's palm. The spider
has vibrant crystal spires growing on its back,
reflecting a spectrum of colorful light,
surrounded by the darkness of a cave.

Practical Application Cases - (Movies, Commercials,

Promotional Videos, Short Videos)
Text-to-video Cases for Practical Application
T2V Advantages：convenient and efficient production; offering diverse creative ideas.
T2V Disadvantages： difficult to ensure the stability of the subject/visual style.

1 Mainly-Used Cases
The practical use cases of text-to-video are primarily for creating empty shots, non-narrative shots,
and large-scale special effects shots.

Generated Video 40 Generated Video 41

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 18/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 1:22
0:00 / 0:47

Case 40 — "Sci-Fi Trailer" Case 41 —— "Avalanche Disaster Movie

Traditional sci-fi imagery features similar visual Trailer"
elements. By creating empty shots with these Based on the theme of "avalanche," multiple
common visual elements, we can produce sets of empty shots and large-scale scene
scenes with similar settings and aesthetic combinations are designed to make the most
styles. Using high-quality shots such as of Hailuo AI's outstanding performance in
"explosions" as key rhythm points, a tense and text-to-video generation. Through
thrilling sci-fi movie teaser is created. rhythmically strong editing, a teaser is created
with a simple storyline that ties the shots
together.

2 Specially-Used Cases：Montage with Similar Objects

Based on Hailuo AI's rich scene realism capabilities, in fast-paced montage scenes, similar prompts
can be used to generate a variety of scenes with similar compositions, allowing for seamless
transitions. This maximizes the avoidance of subject stability issues without affecting the narrative
structure.
Montage with Similar Objects — "The Sci-Fi Portal Across Time and Space"

Generated Video 42

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 19/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 0:06

The similar "sci-fi portal" image is generated in

different time and space settings, with the
shots quickly assembled together, creating a
highly dynamic response to the narrative
structure of crossing time and space.

PS：Ways to Improve Subject Stability

Based on Hailuo AI's high-quality prompt response capabilities, text-to-video can be used to
maintain specific shape consistency.
By testing the consistency of shape description prompts, stable visual elements are identified, and
these are repeatedly used to maintain a certain level of stability.
Subject Elements — Hairstyle + Age/Gender + Clothing Color + Environment
Hairstyle/Clothing: These elements stabilize the character's head image.
Age: Generates characters of a similar age.
Environment: The same environmental space, especially lighting, helps enhance the audience's
"consistency" perception of the character's image.
For example, in environments with high contrast between light and shadow, retaining minimal facial
details of the character helps maintain "consistency" in their appearance.

Cases：

Generated Video 43 Generated Video 44

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 20/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 0:05 0:00 / 0:05

Case 43 Case 44
A close-up of a brown-haired little boy A close-up of a brown-haired little boy
wearing a brown shirt, facing the camera. His wearing a brown shirt, with his face illuminated
face is illuminated by candlelight, and his by sunlight. His expression is one of surprise,
expression is a surprise. The environment is and the environment has a strong contrast
dimly lit with a strong contrast between light between light and shadow.
and shadow.

Generated Video 45

0:00 / 0:05

Case 45
A close-up of a brown-haired little boy
wearing a brown shirt, facing slightly away
from the camera, with his head tilted upward
as he walks through a dark green forest. His
face is illuminated by an oil lamp he holds in
his hand, and the environment is dimly lit with
a strong contrast between light and shadow.

Image-to-video Cases for Practical Application

I2V Advantages：Precise control of the visual subject; precise control of the visual aesthetics.

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 21/29
24/11/2024, 18:20 Hailuo AI tutorial

I2V Disadvantages：The initial workload is relatively large; the margin for error in creation is low;
creative freedom is relatively limited.

1 Mainly-Used Cases
Image-to-video ensures the stability of the video's subject information, making it suitable for
narratives with fixed subjects. It allows for the continuation of video segments from static frames,
completing the narrative function.
Examples: The segment is excerpted from the HailuoAI promotional video "Magic Wheel."
Using a set of images with the same subject, a video is generated to meet the narrative
requirements, ensuring consistency of the subject's appearance in the story's visuals. (The video
sound is post-production.)

Generated Video 46 Generated Video 47

0:00 / 0:05 0:00 / 0:05

Image Image

Plot 46 Plot 47
The little evil dragon police officer proudly The little evil dragon police officer feels
introduces the Magic Wheel to the camera. nervous about the world spiraling out of
control but still tries to maintain composure.

2 Specially-Used Cases：Using the last frame to create long shot

Currently, Hailuo AI can generate 6 seconds of high-quality video, but it is possible to reasonably use
the last frame of an existing video as the first frame for text-to-video output, allowing for the
continuation and creation of pseudo-long shots that exceed 6 seconds.(The video sound is post-
production.)
Example: The segment is excerpted from the HailuoAI promotional video "Magic Wheel."

Generated Video 48

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 22/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 0:04

Step 1
First, generate the basic action required for the video — the girl raises her hand to touch the Magic
Wheel and then lowers her arm.

Generated Video 49

0:00 / 0:04

Step 2
Capture the last frame of the video with the desired effect (see below), and use this final frame as an
image for the second text-to-video generation — the camera slowly pulls back, capturing the Magic
Wheel as it begins to glow and rotate after being touched by the girl.

The last frame of the first generated video

Generated Video 50

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 23/29
24/11/2024, 18:20 Hailuo AI tutorial

0:00 / 0:03

Step 3
Capture the appropriate frame with the best video effect (see below) and use it as an image for the
third text-to-video generation — the Magic Wheel gradually goes out of control, emitting electrical
sparks before stopping its operation.

The last frame of the second generated video

Generated Video 51

0:00 / 0:12

Step 4
Combine the three segments of video editing as described above to obtain a complete 11-second
video, showcasing the full action of the girl touching the Magic Wheel, causing its magic to fail.

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 24/29
24/11/2024, 18:20 Hailuo AI tutorial

3 Specially-Used Cases： Image composition and seamless blending across

transitions
By combining any two images with an appropriate ratio and then creating an image-to-video, Hailuo
AI can seamlessly blend the characters from one image into the other. Once the subject transitions
from the original scene to the target scene, the lighting, shadows, and color tones will perfectly
match the new scene’s setting. This offers infinite possibilities for deterministic creation in a wide
range of high-quality scenarios.

Generated Video 52 Generated Video 53

0:00 / 0:05
0:00 / 0:05

Image Image

Prompt 52 Prompt 53
A woman walks towards the green field, stands The woman walks towards the tent, and the
on the grass, and gazes at the beautiful lake. camera slowly moves to the left.
The camera slowly moves to the left.

Generated Video 54

0:00 / 0:05

Image

Prompt 54
The woman walks towards the tent, and the
camera slowly moves to the left.

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 25/29
24/11/2024, 18:20 Hailuo AI tutorial

Comprehensive Application Case

T2V / I2V Comprehensive Application
• Use text-to-video for scenes or shots that do not require subject stability.
• Use image-to-video for shots that require precise and stable subject appearance.
Example:Hailuo AI's Halloween-Themed film (Video sound is post-production).

Shot1: Generated Video 55

0:00 / 0:05

Plot
Show a chubby man in a green outfit enjoying himself indoors during Halloween, interrupted by a
knock on the door.
Tutorial
As the first shot of the opening scene, use text-to-video for production. By specifying detailed
subject characteristics — middle-aged, black long hair, green shirt, black glasses, and overweight
— establish a solid foundation for subject description in the subsequent video generation.

Shot2: Generated Video 56

0:00 / 0:03

Plot
Indicate that the knocking sound is coming from outside the door.

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 26/29
24/11/2024, 18:20 Hailuo AI tutorial

Tutorial
Immediately follow the man's head-turning motion with a point-of-view shot from his perspective,
revealing the source of the knocking sound. This is an empty shot with no subject, created using
text-to-video, slowly pushing in on a white door.

Shot3: Generated Video 57

0:00 / 0:03

Plot
The man scolds and tells the person outside the door to go away.
Tutorial
Use a close-up of the man to advance the plot. Since it’s a close-up, only some of the man's features
need to be shown. Text-to-video will be used to generate and maintain the subject's description —
black hair, overweight, glasses, green shirt… ensuring the stability of the subject's appearance.

Shot4: Generated Video 58

0:00 / 0:02

Plot
The knocking sound doesn't stop
Tutorial
Provide a close-up of the door, with the knocking sound growing louder. There is still no subject in
the frame, and text-to-video is used to generate the close-up shot of the door.

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 27/29
24/11/2024, 18:20 Hailuo AI tutorial

Shot5: Generated Video 59

0:00 / 0:02

Plot
The man reluctantly gets up to open the door.
Tutorial
The man needs to transition from his initial state in the first scene to the action of getting up to open
the door. The consistency of the subject is crucial here, and it needs to be done under the same
scene and lighting conditions. Image-to-video will be used to continue from the last frame of the
first shot to ensure the subject's appearance remains stable.

Shot6: Generated Video 60

0:00 / 0:01

Plot
The man opens the door.
Tutorial

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 28/29
24/11/2024, 18:20 Hailuo AI tutorial

Since the full subject does not appear, only partial features of the subject are shown. Use text-to-
video to generate a close-up of the man's hand reaching out to open the door.

Shot7: Generated Video 61

https://sixth-switch-2ac.notion.site/Hailuo-AI-tutorial-1446c20a98eb802c94c5d9ebcb497401 29/29