{"id":58317,"date":"2026-10-19T10:21:10","date_gmt":"2026-10-19T02:21:10","guid":{"rendered":"https:\/\/www.1ai.net\/?p=58317"},"modified":"2026-09-20T15:21:40","modified_gmt":"2026-09-20T07:21:40","slug":"%e6%8f%90%e7%a4%ba%e8%af%8d%e5%88%9b%e4%bd%9c%e7%ac%ac%e4%b8%83%e5%8d%81%e5%85%ad%e8%8a%82%ef%bc%9a%e7%94%a8-ai-%e5%88%9b%e4%bd%9c%e5%88%9b%e6%84%8f%e5%93%81%e7%89%8c%e7%9f%ad%e7%89%87%e7%9a%84","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/58317.html","title":{"rendered":"SECTION 76 OF THE AI TIP: A COMPLETE WAY TO CREATE A SHORT FILM WITH AN AI CREATIVE BRANDING"},"content":{"rendered":"

DoAI Video<\/a>At times, many people will focus on the unity of the person, the sense of the scene, the flow of action and the physical details. Of course, these are important, but what really determines whether the audience will continue to look at them is often the first few seconds\u3002<\/p>\n

A lot of videos look at each frame on their own, but it's still flat. The problem is usually not the quality of the picture, but the absence of a memory point at the beginning. The person stands in the center of the picture, the camera moves slowly, the person looks up to the camera and the title goes down\u3002<\/p>\n

There was no obvious error in this opening, but at first sight the audience was able to judge what was going to happen. The image did not leave any information to be confirmed, nor did it create a new viewer\u3002<\/p>\n

An effective opening does not necessarily require complex stories or many special effects. A more practical approach would be to change a relationship in the picture first\u3002<\/p>\n

For example:<\/p>\n

The characters did not come from their normal positions, but rose from the sand with the tables, chairs and computers\u3002<\/p>\n

Words are not simply folded into the picture as ordinary subtitles, but become visible as people pass through\u3002<\/p>\n

The cover was not removed after a second, but a natural transition from a full fish eye poster to a normal video\u3002<\/p>\n

The three opening points of this period are the same\u3002<\/p>\n

They look different in form, but they're all about the same thing:<\/p>\n

How to set up a viewing interest in the first second\u3002<\/p>\n

First, let's see what these three openings have done<\/strong><\/p>\n

Opening one: the person and office equipment rise from the sand<\/p>\n

I'll give you a normal beach\u3002<\/p>\n

The outlook is a flat sandy sand, with a middle view of rising waves and distant roaders moving along the coast\u3002<\/p>\n

Then the sand in the centre of the picture began to rise\u3002<\/p>\n

A woman sitting in an office chair, along with a computer table, monitor, keyboard, mouse and seat, slowly rises from within the sand\u3002<\/p>\n

People continue to use computers as naturally as they do in their offices when they and office equipment are fully on the ground\u3002<\/p>\n

And then the camera slowly pushes forward and pushes this anomaly further into the audience\u3002<\/p>\n

The opening change is:<\/p>\n

The way people and scenes enter the picture\u3002<\/p>\n

The opening two: after the character passed, the text became visible<\/p>\n

The image uses a fixed position\u3002
\nForeground is a blue couch with a white wall. A man in black was standing next to the wall,
\nThe body is basically in place\u3002
\nThe person placed his hands on the wall, turned his body slightly and started to work\u3002
\nAs the person continues to push, the whole white wall moves slowly to the other side of the picture, which was hidden behind the wall in the vertical title
\nThe \"dynamic mask\" is coming out\u3002
\nThe walls were not moved and the text behind them was still covered\u3002
\nWhen the wall moves to the other side, the title appears in its entirety and remains in the picture\u3002
\nThe opening change is:
\nThe relationship between the shield and the rear information\u3002<\/p>\n

Opening three: Fish-eye posters are starting to be a normal video<\/p>\n

The video begins not with a normal video, but with a complete design poster\u3002<\/p>\n

The center is a circular fish eye image of the person standing in front of the elevator with a coloured dress\u3002<\/p>\n

THE TOP OF THE CIRCULAR AREA IS THE VISIBLE HEADING \u201c3 OPENINGS OF AI\u201d, FOLLOWED BY COLUMN DESCRIPTION AND DECORATION INFORMATION\u3002<\/p>\n

Subsequently, the central round eye area continued to expand outward\u3002<\/p>\n

The title and design elements of the surroundings gradually exit, and the fish eye aberration slowly diminishes and eventually restores to a static video of normal proportions\u3002<\/p>\n

The audience saw not \"cover cuts to the face\" but:<\/p>\n

The cover opens itself and takes the audience into the video\u3002<\/p>\n

The three opening phrases correspond to three common and useful hint structures:<\/p>\n

Unconventional drawings, dynamic mask, cover transtext\u3002<\/p>\n

They are divided below\u3002
\n
\nII. First method: unconventional drawings<\/strong><\/p>\n

1. What are non-conventional entries<\/p>\n

Many of the problems at the beginning are the way people appear too conventional\u3002<\/p>\n

The way people come in from the left side of the picture, or the lens cuts in front of them, is not wrong, but it is difficult to make a clear impression\u3002<\/p>\n

The idea of a non-conventional entry is to get the subject out of the picture in the way that the audience expected, but from a position that it should not have appeared\u3002<\/p>\n

For example:<\/p>\n

The character came out of the wall\u3002<\/p>\n

Products rise slowly from the surface\u3002<\/p>\n

A sofa grows out of the grass\u3002<\/p>\n

A desk came out of the beach sand with the characters\u3002<\/p>\n

The focus of the non-conventional drawings is not just \u201csubjects appear\u201d, but:<\/p>\n

Which kind of environmental material is the subject, and in which ways it enters the picture\u3002<\/p>\n

2. Why is the beach office case attractive<\/p>\n

In the case of beach offices, the place where people were actually captured was not just \u201cpersons coming out of the ground\u201d\u3002<\/p>\n

Its more important visual relationships are:<\/p>\n

The office and the beach were put in the same shot\u3002<\/p>\n

Computer tables, monitors, office chairs and working conditions should have appeared in offices, study rooms or studios\u3002<\/p>\n

Sea winds, sand, waves and long-distance people belong to outdoor beaches\u3002<\/p>\n

When these two unrelated space logics appear at the same time, the audience can see what the person is doing at once, but at the same time it raises questions:<\/p>\n

Why would she work on the beach<\/p>\n

How did these office equipment come out<\/p>\n

This \u201cunreasonable but fully understandable\u201d relationship is the starting point of memory\u3002<\/p>\n

3. beach office case alert (seedance 2.0)
\n[unconventional drawing hint]<\/p>\n

Take Figure 1. [photographic parameters and graphic senses] Cooke Panchro\/i Classic Prime Lines, with 1\/4 or 1\/8 Black Pro-Mist radium and focal segment at the front of the camera: 35mm and 50mm focal points are used more frequently in the room than in the day-to-day use of 35mm and 50mm focal areas, with the optical circle controlled between T2.0 and T2.8. Kodak Portra 400 film color. It's called Teal & Warm Amber, no music, no subtitles\u3002
\nFixed aircraft, clear day-time beach scenes, with the prospect of a level-blank gold sand field, with a middle view of a blue wave of constant rises and rises, with roaders walking along the coast naturally\u3002
\nAt the beginning of the picture, there were no figures, tables, chairs and computer equipment in the outlook\u3002
\nSubsequently, the sand floor in the centre of the picture was slightly shaken and slowly raised, and the sand particles were squeezed and fell naturally around. A young woman, along with a complete computer table, monitor, keyboard, mouse and black office chair, slowly rises vertically from within the sand as a fixed whole. The guy threw his head off and threw the sand off his head
\nThe person is always sitting on an office chair, with his hands on a natural computer. The relative position between the table, monitor, keyboard, character and seat is fixed, unseparated and non-floating\u3002
\nDuring the lifting process, sand particles fall naturally along the legs of the table, the bottom of the chair and around the person, moving in a manner consistent with gravity and having real sand mass and weight feedback\u3002
\nThe person and the entire office equipment were completely stabilized and stopped, and the legs and chairs naturally fell on the sand. The characters continue to use computers in a calm manner, and the background waves and distant passers-by remain in motion\u3002
\nThe camera then moved quickly, moving closer from the beach landscape to the characters and eventually creating a periphery among the clear characters\u3002
\nThe structure of the face, hair, clothing, body ratio and office equipment is consistent. The sudden ejection of persons, the transparent passage through sand, the separation of tables and chairs, the floating of monitors, changes in the structure of equipment, deformation of limbs, background jumps and the sudden acceleration of the lens are prohibited\u3002<\/p>\n

\"baf6c7cej00tln9vt009xd000un00oyp\"<\/p>\n

you can just upload your character map directly with a copyance 2.0<\/p>\n

III. Second approach: dynamic mask<\/strong><\/p>\n

1. What is a dynamic mask<\/p>\n

A lot of course videos will start with a title\u3002<\/p>\n

The most common method is to simply superimpose text next to the person, or to dilute subtitles from the image\u3002<\/p>\n

While this approach can convey information, there is no real link between the title, the person and the scene\u3002<\/p>\n

The idea of a dynamic mask is to use walls, doors, curtains or other large objects to cover the rear, then move the shield through the movement of people, and gradually reveal the title, product or new scene\u3002<\/p>\n

In this case, the person placed his hands on the white wall, the body turned slightly and began to work\u3002<\/p>\n

As the person continues to push, the entire white wall moves slowly to the other side of the picture, and the title that was hidden behind the wall gradually emerges\u3002<\/p>\n

The main role of the person is to complete the driving movement, while the true dynamic mask is the moving wall\u3002<\/p>\n

The rear information was completely covered before the wall was moved\u3002<\/p>\n

When the surface begins to move, the title appears gradually from the edge of the wall\u3002<\/p>\n

The wall was pushed away and the title or product was presented in its entirety\u3002<\/p>\n

The opening change is:<\/p>\n

The relationship between the shield and the rear information\u3002<\/p>\n

2. Why does the wall drive a better picture than the usual subtitles<\/p>\n

Common subtitles are usually at the top of the picture\u3002<\/p>\n

No matter how the person moves, subtitles are unaffected by walls, characters and space structures, and are therefore prone to a clear sense of later overlap\u3002<\/p>\n

A dynamic mask that pushes the wall gives the text real access to the scenery\u3002<\/p>\n

The person first held his hands against the wall, causing a slight physical force, and then the wall moved to the other side, with the person pushing\u3002<\/p>\n

The audience will observe the wall in the direction of the person ' s arm and then see the title hidden behind the wall\u3002<\/p>\n

Throughout the process, character moves were no longer merely performances, but instead became the cause of the title\u3002<\/p>\n

In other words, the title does not appear on its own, but is \u201cpushed\u201d by the person himself\u3002<\/p>\n

This design can accomplish three things simultaneously:<\/p>\n

The character finishes a clear move\u3002<\/p>\n

There was a marked change in the landscape structure\u3002<\/p>\n

The title or product automatically enters the image\u3002<\/p>\n

Such openings would have a more spatial and visual memory point than simply diluted words\u3002<\/p>\n

3. Dynamic mask case alerts<\/p>\n

\"Push Wall Dynamic Mask Intent\"<\/p>\n

Take Figure 1. [photographic parameters and graphic senses] Cooke Panchro\/i Classic Prime Lines, with 1\/4 or 1\/8 Black Pro-Mist radium and focal segment at the front of the camera: 35mm and 50mm focal points are used more frequently in the room than in the day-to-day use of 35mm and 50mm focal areas, with the optical circle controlled between T2.0 and T2.8. Kodak Portra 400 film color. It's called Teal & Warm Amber, no music, no subtitles\u3002
\nStanding screen 9:16, about 4 seconds, fixed position, horizon, single-scenario, single-scenario, real cell phone. The bright and warm modern family living room, with a blue clothed couch on the left side, is a red brown glass cabinet full of books, with a wide white rectangular door hole in the center, with a warm brownwood floor, a light-colored wooden wall, a table and chair, a kitchen area and a dark brown wooden door with light glass windows at the end\u3002
\nThe character was replaced by a young adult East Asian woman, with a strict reference image: long, long, long black microvolts, face Liu Hai, red Chinese hair on the left side, white colour, fine five officials, new Chinese thongs with red brown and brown bouquets, short-dressed short skirts with water drops on the chest, white over knee stockings and black Mary Jane shoes on the feet\u3002
\nThe woman stood by the white wall on the left side of the door hole, remained essentially in place, facing the right side of the picture, raised her right hand and attached her hand to the edge of the white wall on the left side of the door hole. The left hand naturally leans and the body leans forward and works. She then slowly pushed the whole central rectangular door hole, the door frame and the kitchen and restaurant space visible behind the door hole in her right hand as a complete and stable structure to the right. The doorholes remain vertically intact, do not rotate, do not collapse, and the internal space elements maintain a unified view and smoothing of the whole\u3002
\nAs the door hole moved to the right, the original position of the door hole gradually revealed a wide and flat white wall. The centre of the white wall gradually appears four Chinese characters from top to bottom row \" Dynamic mask \" . The text is sculpted like a light sculptor on the surface of a white plaster wall, which is grey, with a slight internal shading, rough edges and a small amount of plaster wall ash. Text emerges as the white wall emerges, the undisclosed area does not appear in advance, and the words that have emerged are fixed on the wall and do not drift, blink or hang. Very little small amount of white dust can fall around wall moving edges and writing\u3002
\nThe door hole stopped flatly on the right side of the picture in about 2.5 seconds, and the vertical \"dynamic mask\" was complete. The woman opens her right hand, takes her arms back, stands up to her body, looks at the text of the wall, then gently lifts her right hand to the back of the head or ear, and makes a natural slight hair-dressing or doubting move, with a slight deviation. The final figure stood on the left side of the picture, with a complete \u201cdynamic mask\u201d on the central white wall, a door hole on the right, and all elements remained stable\u3002
\nThe motion of the camera, the movement of the person through the image, the body of the person as a mask, the movement of the wall to the left, the deformation of the door, the collapse of the wall, the movement of furniture, the suspension of the text, the disorder, the change of the person, the change of clothes, the disappearance of the hair, the removal of the hand from the wall, the distortion of the body and the redrawing of the background\u3002<\/p>\n

IV. A third approach: development of fish eye posters<\/strong><\/p>\n

1. How did this opening take place<\/p>\n

The third opening focus is not just a fish eye effect on the video, but a designed fish eye poster to make a continuous transition to a raw image of normal proportions\u3002<\/p>\n

The video starts with a complete stand-up poster\u3002<\/p>\n

The centre of the poster was a round fish eye image, and the person stood in front of the elevator with a coloured dress. Places the main header above a circular area, and adds a column description, a graphic symbol and other layout elements below\u3002<\/p>\n

Subsequently, the central round eye area continued to expand outward\u3002<\/p>\n

As the circular boundary moves closer and closer to the scene, the title and decoration elements of the poster exit the picture, and the fish eye aberration diminishes. Eventually, the image reverted to normal visualization of raw material\u3002<\/p>\n

Throughout the process, the person, clothing, clothing and elevator space remained continuous\u3002<\/p>\n

The viewer did not suddenly switch the poster to a video, but:<\/p>\n

The fish eye posters were gradually opened from the centre and finally entered normal\u3002<\/p>\n

2. This approach is appropriate for end frame generation<\/p>\n

This opening stage has a very clear start and end state\u3002<\/p>\n

The start state is a designed fish eye poster\u3002<\/p>\n

The end state is a normal proportion of the original figure\u3002<\/p>\n

The changes that need to be accomplished are also clear:<\/p>\n

Round eye area expanded outward\u3002<\/p>\n

Poster text and decoration information exit\u3002<\/p>\n

The curved eye of the fish gradually returned to normal\u3002<\/p>\n

Thus, such effects are particularly appropriate for the generation of the front frame\u3002<\/p>\n

After the first frame is used, the model does not have to speculate about what the image will become. It needs only a continuous and reasonable transition between two well-established images\u3002<\/p>\n

It's better to control the frame than to enter a poster and ask the model to freely generate a normal image:<\/p>\n

Personal identity\u3002<\/p>\n

The character position\u3002<\/p>\n

Clothing and props\u3002<\/p>\n

Background space\u3002<\/p>\n

The end result of fish eye changes\u3002<\/p>\n

The last frame of the video is also easier to get back to the original\u3002<\/p>\n

3. Correct production process<\/p>\n

Such openings cannot be done directly first\u3002<\/p>\n

A more stable process is to obtain the fish eye version from the original image and then set it out based on the fish eye version\u3002<\/p>\n

Step one: prepare the original<\/p>\n

Choose your own figure first\u3002<\/p>\n

Then prepare a reference map<\/p>\n

\"041173f9jln9ye00yvd000v9000s9p\"<\/p>\n

Step two: Turn the original image into a fish eye<\/p>\n

enter the original image and reference diagram to use the banana pro to generate the picture and convert the general view into a full round eye view\u3002<\/p>\n

The prompt words are as follows:<\/p>\n

Remove text 2025 and replace the characters in figure I with figure II, figure I, figure I, with a lot of clothes<\/p>\n

\"e7d00588jln9yp01o4d000pf019kp\"<\/p>\n

Step 3: Production of posters based on fish eye images<\/p>\n

Once the fish eye picture is available, the poster will be printed. I found a poster here for reference generation\u3002<\/p>\n

Reference poster<\/p>\n

the references are as follows (generated using image2.0):<\/p>\n

Imaging the style of figure II, I'm going to make a similar cover, and I'm going to change the Hongkong to the 3 openings of AI, and then I'm going to change the text at random, and it's going to have to do with the stupor planet, the ai godfather, the ai teaching, the ai dry, the loki<\/p>\n

\"92653<\/p>\n

Step 4: Generating videos<\/p>\n

When the video is generated, two pictures are uploaded\u3002<\/p>\n

First Frame<\/p>\n

Using a designed fish eye poster\u3002<\/p>\n

End Frame<\/p>\n

Use the original image of the person that was first prepared\u3002<\/p>\n

IT'S DIRECTLY GENERATED WITH A FRONT-END FRAME<\/p>\n

\"ef860a27jln9zg00c7d000qa00oip\"<\/p>\n

V. How should these three openings be chosen<\/strong><\/p>\n

None of these three approaches is absolutely better, depending on what you want the most\u3002<\/p>\n

To make people or products feel better<\/p>\n

Select an unconventional painting\u3002<\/p>\n

It changes the way the subject enters the scene\u3002<\/p>\n

In the case of beach offices, the first thing the audience remembers is the rise of people from the sand with computer tables and office chairs\u3002<\/p>\n

It's a way for people to look, for products to show, for brand world views and for space to contrast<\/p>\n

I want the title to come out of character action<\/p>\n

Select the dynamic mask\u3002<\/p>\n

It changes the relationship between the shield and the back\u3002<\/p>\n

The person is essentially standing in place, with his hands moving towards the other side of the wall. Titles hidden behind walls will become visible as the area under cover decreases\u3002<\/p>\n

This is appropriate for the curriculum title, chapter import, product name, sales words, price information and scenario conversion\u3002<\/p>\n

I want the cover to enter the body<\/p>\n

Select the fish eye poster for rollout\u3002<\/p>\n

It changes the way it connects to the image and the normal video\u3002<\/p>\n

A full fish eye poster will focus on the title and then expand the round image to a normal video, allowing for a balance between message and lens transition\u3002<\/p>\n

This is particularly appropriate for serialized content and accounts with a fixed packing style\u3002<\/p>\n

Why is it best to design only one core rule for a start-up<\/strong><\/p>\n

The opening is not as good as the change\u3002<\/p>\n

Models are vulnerable to loss of focus if the person is required in a single shot to rise from the ground, rotate the lens, explode the text, deform the scene, fly in the product and switch the light\u3002<\/p>\n

In each of the three cases during the period, only one core rule was established\u3002<\/p>\n

The beach office only addresses:<\/p>\n

How people and office equipment emerged from the sand\u3002<\/p>\n

The dynamic mask only addresses:<\/p>\n

How the person pushes the wall so that the text behind the wall is gradually revealed\u3002<\/p>\n

Fish-eye posters address only:<\/p>\n

How does a completed cover go back to normal video\u3002<\/p>\n

The more uniform the rules, the easier the model is to understand\u3002<\/p>\n

The clearer the relationship between the person, the text and the scene, the easier it is to start in a stable manner\u3002<\/p>\n

Conclusion<\/strong><\/p>\n

IT'S NOT ABOUT HOW MUCH CHANGE THERE IS, BUT WHETHER THERE'S A CLEAR VISUAL RULE\u3002<\/p>\n

The way people appear, the way the veil is moved, the way the cover is entered into the body, the relationships can also form a memory point if they are redesigned\u3002<\/p>\n

The hints need to be written, like directors, in terms of start-up status, order of change, stabilization elements and final results. The clearer the rules, the easier the model is to understand and the more stable the person and the scene\u3002<\/p>\n

The opening is not simply adding special effects, but redesigning the relationship between people, information and space. Write that clearly, ordinary images can also change from the first second\u3002<\/p>","protected":false},"excerpt":{"rendered":"

IN MAKING THE AI VIDEO, MANY PEOPLE WILL FOCUS MAINLY ON THE CONSISTENCY OF THE PERSON, THE QUALITY OF THE SCENE, THE FLOW OF ACTION AND THE PHYSICAL DETAILS. OF COURSE, THESE ARE IMPORTANT, BUT WHAT REALLY DETERMINES WHETHER THE AUDIENCE WILL CONTINUE TO LOOK AT THEM IS OFTEN THE FIRST FEW SECONDS. A LOT OF VIDEOS LOOK AT EACH FRAME ON THEIR OWN, BUT IT'S STILL FLAT. THE PROBLEM IS USUALLY NOT THE QUALITY OF THE PICTURE, BUT THE ABSENCE OF A MEMORY POINT AT THE BEGINNING. THE PERSON STANDS IN THE CENTER OF THE PICTURE, THE CAMERA MOVES SLOWLY, THE PERSON LOOKS UP TO THE CAMERA AND THE TITLE GOES DOWN. THERE WAS NO OBVIOUS ERROR IN THIS OPENING, BUT THE AUDIENCE WAS ABLE TO JUDGE AT FIRST SIGHT WHAT WOULD HAPPEN. THE IMAGE DID NOT LEAVE ANY INFORMATION TO BE CONFIRMED, NOR DID IT CREATE A NEW VIEWER. AN EFFECTIVE OPENING DOES NOT NECESSARILY REQUIRE A COMPLEX STORY OR A LOT OF SPECIAL EFFECTS. IT'S MORE PRACTICAL TO CHANGE THE PICTURE FIRST<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[149,144],"tags":[802,3149,956,5321,192,6481],"collection":[],"class_list":{"0":"post-58317","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"hentry","6":"category-jiaocheng","7":"category-baike","8":"tag-ai","12":"tag-prompt","13":"tag-6481"},"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/58317","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=58317"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/58317\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=58317"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=58317"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=58317"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=58317"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}