{"id":57324,"date":"2026-09-14T11:42:15","date_gmt":"2026-09-14T03:42:15","guid":{"rendered":"https:\/\/www.1ai.net\/?p=57324"},"modified":"2026-09-14T11:43:19","modified_gmt":"2026-09-14T03:43:19","slug":"%e6%89%8b%e6%8a%8a%e6%89%8b%e6%95%99%e4%bd%a0%e4%bb%8e0%e5%88%b01%e5%ae%8c%e6%95%b4%e5%88%b6%e4%bd%9cai%e8%a7%86%e9%a2%91%e7%9f%ad%e7%89%87%e5%85%a8%e6%b5%81%e7%a8%8b%ef%bc%88%e9%9b%b6%e5%9f%ba","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/57324.html","title":{"rendered":"HAND HANDS TO TEACH YOU TO COMPLETE THE FULL-TIME PRODUCTION OF THE AI VIDEO CLIP FROM 0 TO 1"},"content":{"rendered":"

Can ordinary people make big Hollywood movies<\/p>\n

If I told you, it's low cost, you don't have to go to the show, you can make a big movie of the Ring-style<\/p>\n

And this article, I'm going to take this whole tool out of your head\u3002<\/p>\n

After learning it, you don't just look at it, but you really know how to make a big, high-level video that belongs to you\u3002<\/p>\n

What steps are to be taken between an idea and a complete film<\/p>\n

Just recently, I'm doing something like the Ring AI short film<\/a>.<\/p>\n

Take this case today and untangle the production process from 0 to 1\u3002<\/p>\n

From how the script is written, how the characters are determined, how the scenes are done, how the mirrors are removed, how the videos are generated, how the film is produced, how the film is made, how the film is made<\/p>\n

And I'm going to do it step by step\u3002<\/p>\n

The article also contains the text of the text that I actually use, the three-viewer template for the person, and the thought and details of how the video was generated\u3002<\/p>\n

I HOPE YOU CAN READ THE WHOLE PROCESS AND START MAKING YOUR FIRST AI SHORT FILM\u3002<\/p>\n

I. Overview of Work Flows<\/strong><\/p>\n

A lot of rookies did it for the first time AI Video<\/a>The easiest mistake is to open up the tools to generate video without knowing what to film\u3002<\/p>\n

We've got a new model today. Try it<\/p>\n

See a new platform tomorrow and try it again\u3002<\/p>\n

Finally, the tool tried a circle, with a lot of points, with a bunch of scattered shots in its hand, but never a complete piece\u3002<\/p>\n

This is often where the problem arises: each tool is tried, but the production process is not linked\u3002<\/p>\n

ACTUALLY, IT'S NOT COMPLICATED BY THE LOGIC OF AN AI VIDEO\u3002<\/p>\n

First, the steps are clear, and every step is a lot smoother\u3002<\/p>\n

You can remember the following process:<\/p>\n

\"8c21d5efj00tlc4me00ard000pv00aup\"<\/p>\n

Ideas, scripts, style tests, character\/ scene\/ props, video generation, clipping<\/p>\n

The bottom line is the same, regardless of whether you're making a costume, a phantom, a city, an ad, or a documentary\u3002<\/p>\n

The subject matter changes, the model changes, but the process is largely unchanged\u3002<\/p>\n

II. HOW AM I SUPPOSED TO CHOOSE WITH SO MANY AI SOFTWARE ON THE MARKET<\/strong><\/p>\n

First of all, conclusions\u3002<\/p>\n

IF YOU'RE JUST TRYING TO MAKE THE FIRST AI VIDEO IN ITS ENTIRETY, YOU DON'T REALLY NEED TO LEARN A DOZEN SOFTWARES AT ONCE\u3002<\/p>\n

It's going to take three things:<\/p>\n

\u2022 DeepSeek\/ChatGPT: in charge of scripts, drama decomposes and the optimization of tips\u3002<\/p>\n

\u2022 LibTV: responsible for person, scene, spectroscopy and video generation\u3002<\/p>\n

\u2022 Clipping: final editing, subtitles and music\u3002<\/p>\n

Basically enough\u3002<\/p>\n

Aggregation platform: LibTV<\/p>\n

I myself mainly used LibTV to run the whole generation process\u3002<\/p>\n

\"80c40a34j00tlc4m00aod000jl00irp\"<\/p>\n

It is more convenient that people, scenes, pictures and videos can be processed in the same painting, without having to cut back and forth between different websites\u3002<\/p>\n

Double-click a canvas to create a picture or video node. Model selection columns can also switch directly to different pictures and video models\u3002<\/p>\n

\"8d680dcfj00tlc4nc002fd000pj00hfp\"<\/p>\n

For starters, this nodal canvas has one advantage:<\/p>\n

\"13a2637cj00tlc4no0076d000of00cxp\"<\/p>\n

You can see very intuitively how a video was made step by step\u3002<\/p>\n

If the material is made on other platforms, upload it directly to LibTV\u3002<\/p>\n

III. THE STUDENT: Let's get this episode straight<\/strong><\/p>\n

The script does not make it clear that the people behind it, the scenes, the spectroscopy, could all go back to work\u3002<\/p>\n

BUT IN THE AI ERA, YOU DON'T HAVE TO COME UP AND WRITE A FULL STORY\u3002<\/p>\n

EVEN IF THERE'S ONLY ONE VAGUE IDEA, WE CAN THROW IT TO AI, SO IT CAN HELP YOU BUILD THE SKELETAL SKELETAL AND THEN SHARPEN THE STORY\u3002<\/p>\n

This time, I tested the script creations using a high model of DeepSeek V4 Pro and GPT-6 Astra, respectively\u3002<\/p>\n

In this case, I prefer GPT-6 Astra\u3002<\/p>\n

It is better dealt with in the logic of the drama, in the construction of the world view, and in the desire and motivation of the person, and the story that comes closer to my expectations\u3002<\/p>\n

\"e<\/p>\n

Of course, DeepSeek V4 Pro's performance is very good\u3002<\/p>\n

IN TERMS OF MY USE, ITS ABILITY TO PLAY THE SCRIPT AND THE MEDIUM MODEL OF GPT-5.6 ARE BASICALLY FLAT, AND IT'S USED TO COMB THE IDEAL AND FRAME THE STORY, AND IT'S PERFECT\u3002<\/p>\n

People can choose models according to their needs\u3002<\/p>\n

THE KEY IS TO USE AI TO SAY, \"WHAT DO I THINK I WANT TO SAY?\" AND GRADUALLY TO SAY, \"WHAT IS THIS STORY ABOUT?\"\u3002<\/p>\n

HOWEVER, AI SECONDS OUT OF THE SCRIPT DOESN'T MEAN THE SCRIPT CAN BE SHOT DIRECTLY\u3002<\/p>\n

The first draft will continue to be deleted\u3002<\/p>\n

When I get the first edition, I'm going to go over it again, and I'm going to look at the core at three points: the logic of the story goes wrong, the image is generated, the lines are not human\u3002<\/p>\n

This time, for example, the story was set in a phantom world of Central European style, and the presence of vehicles on the bridge seemed to be somewhat counterproductive\u3002<\/p>\n

So I deleted the vehicle and replaced the withdrawal with the departure of the wounded by the guards\u3002<\/p>\n

The blasting iron ball, which had been rolled out from the bottom of the car, was replaced by a one-eyed demon, which would have made it more coherent and better suited to the world view\u3002<\/p>\n

\"153115009jlc4g001nd000ph74p\"<\/p>\n

IF THERE'S ANYTHING WRONG AND ANYTHING NEEDS TO BE ADJUSTED, YOU JUST NEED TO TALK TO AI\u3002<\/p>\n

IT SEEMS TO ME THAT IN THE AI ERA, LEARNING TO DISCOVER PROBLEMS AND THROW THEM AT AI IS AN IMPORTANT ABILITY\u3002<\/p>\n

IV. Style test: let's get this thing fixed<\/strong><\/p>\n

After the script is finalized, don't rush to make a scene. I'm used to doing a \"style test\" first -- looking for the right mark, looking for a reference map, trying out the whole project\u3002<\/p>\n

THE STYLE ISN'T MADE OUT OF NOTHING. I'LL START BY LOOKING FOR PICTURES -- MOVIES, DRAMAS, PICTURES LIKE THAT -- THAT'S ALL AVAILABLE. THROW THE REFERENCE TO AI, TURN IT BACK INTO A HINT, AND THE DIRECTION BECOMES CLEAR\u3002<\/p>\n

\"959e1415j00tlc4ow0006pd000nb00hyp\"<\/p>\n

IN THE EARLY PART OF THIS PERIOD, A NUMBER OF REFERENCES WERE REPULSED, AND THE DIRECTION WAS GRADUALLY CONCENTRATED ON DARK SCHIZOPHRENIA, THE GOTH RELIGIOUS AESTHETICS AND THE WRITING OF THE CG MASS, THE SACRED NUCLEI\u3002<\/p>\n

My approach is to put the script in ChatGPT and make it a set of concepts based on the story\u3002<\/p>\n

This step begins with a general picture of the world: what is the city of the clouds and seas, what is the structure of the East Bridge, where angels fight demons, and what is the atmosphere inside the temple\u3002<\/p>\n

\"0fe0bbffj00tlc4p90hcd000op00iwp\"<\/p>\n

Once the concept map is out, the more appropriate images of the structure, space and atmosphere will be selected and will continue to be refined into a formal scene\u3002<\/p>\n

However, in my actual use, I found that ChatGPT created maps that sometimes had a strong sense of gloom\u3002<\/p>\n

The atmosphere is good and the details of the structure, stone texture and decoration are not clear enough\u3002<\/p>\n

I'LL HAND OVER THE SELECTED CONCEPT TO AI<\/p>\n

It's a counterprompt word that describes the image, light, color, material and architectural features<\/p>\n

And put in MidJourney to generate a set of scenes and blanks\u3002<\/p>\n

\"85f979f6j00tlc4pk0006ad0001900fyp\"<\/p>\n

That's how I started. Start with ChatGPT to find what you want, then continue to refine with MidJourney\u3002<\/p>\n

From the results of this exercise, there has been a considerable reduction in the oscillation, and the desired atmosphere has been better\u3002<\/p>\n

And that's how I created this video scene asset\u3002<\/p>\n

\"0e139acbj00tlc4pu00t9d000pd00e8p\"<\/p>\n

\u25cf<\/p>\n

the dark script fantasy scene, inside the gothic cathedral, swallowed by the ancient forest, with a super-wide screen, and a film-based concept of the cg environment. the low-altitude wide-angle, looking from the cracks to the far-off altar, is deep, and the point of loss of vision is right. the outlook on the left is a giant carving angel, with his head down, his feathers broken, his long robe wrinkled, and his face crawled with moss and vines. the right-hand outlook lies in a fractured rock column, with rubble, beryllium and a small amount of white wild flowers. the mid-central moist plank road and layers of broken steps lead to a far-flung stone altar. several large and distorted ancient trees are growing from within the church, thick black trunks circled along both sides, roots torn to the ground, entangled with stones, and giant branches stretching over and forming natural arches. the tall gothian arch, the long pillars, the broken walls and the trees. the roof collapsed in large areas, with the top plunging down with vines and dead branches. the wall at a distance maintains a large rounded rose window, a fine, radioactive stone-like window, a narrow, steep arch window at the bottom and cold daylight through the window hole and the broken dome. the outlook is dark, bright, the light is soft, the air is clear and the shadows are rich in detail. the whole body is dominated by black carbon, cold ash, grey green, dark olive green, low saturated and restrained silver and white light. the damp slabs and shallow water reflect a weak reflection, with real rough barks, edible stone, moss velvet, mud and leaves, rich in material details and natural light. declared, silent, desolate, mysterious, lost temples, grand scales, environmental narratives, film-class maps, physical reality, fine light, a sense of space, clear, but not overly sharp\u3002<\/p>\n

The style test runs down, if the role image in any concept map is particularly right\u3002<\/p>\n

You can just turn it into a third view of the person and save the rest of the person from a zero\u3002<\/p>\n

So the art and role images are linked\u3002<\/p>\n

\"66c05c19j00tlc4q500pbd000pf00edp\"<\/p>\n

\"4435fc79j00tlc4q00kqd000pt00b0p\"<\/p>\n

Asset design: the most time-consuming step of the whole process<\/strong><\/p>\n

When the script and style are finalized, the next step is not to produce video, but to make assets first\u3002<\/p>\n

Assets are the things that recur in the film. There are three main categories - people, scenes, props\u3002<\/p>\n

BECAUSE AI CAN'T REMEMBER THE MIRROR: WHAT THE CHARACTER WEARS, WHAT THE ROOM LOOKS LIKE, WHERE THE TABLE IS. THE FRONT IS NOT FIXED AND EVERY MIRROR IN THE BACK IS LIKELY TO BE REGENERATED AT RANDOM. SO, IN ORDER TO BE STABLE, IT IS NECESSARY TO STABILIZE RECURRENT ASSETS\u3002<\/p>\n

Personal assets<\/p>\n

When the style is set, the next step is the fixed-person three view\u3002<\/p>\n

Because the easy question for a person is to change his face -- different batches of video, changed character<\/p>\n

Regardless of the type of short film, before it is produced, a standard reference map is created for the main players\u3002<\/p>\n

There are four points\u3002<\/p>\n

Point 1: Try not to be background-free\u3002<\/p>\n

\"08996f74j00tlc4q00mod000pw00ayp\"<\/p>\n

Error Demonstration<\/p>\n

\"1a0fcea2j00tlc4qz00hnd000pz00avp\"<\/p>\n

Correct demonstration<\/p>\n

For the first time, the simpler the background: pure colour, complete subject, clean light. The more complex the context, the more disruptive information the model has to address\u3002<\/p>\n

Point 2: Make sure you have a good figure\u3002<\/p>\n

In particular, I would like to ask you clearly:<\/p>\n

SMALL HEAD, CLOSE TO 28, FULL BODY. OTHERWISE IT'S EASY TO HAVE A BIG HEAD, SHORT LEGS, LIKE A GAME OF Q-STYLE CHARACTERS\u3002<\/p>\n

\"67954ca0j00tlc4r9009fd000a200i5p\"<\/p>\n

Error Demonstration<\/p>\n

\"de54e96ej00tlc4ri0006fd000an00ivp\"<\/p>\n

Correct demonstration<\/p>\n

Point 3: People try not to pick up props\u3002<\/p>\n

The power of the picture is greater than the weight of the text\u3002<\/p>\n

In a make-up picture, the character is carrying a props, even if you write \"Don't take anything\" in the text at the time of the video, and the props are still on the screen\u3002<\/p>\n

So let's set the \"man\" down and make the prop alone\u3002<\/p>\n

\"52c48522j00tlc4rv00kmd000pv00awp\"<\/p>\n

Error Demonstration<\/p>\n

\"d28d1a0fj00tlc4s200dxd000py00b4p\"<\/p>\n

Correct demonstration<\/p>\n

Point 4: The figure must be changed to three views\u3002<\/p>\n

This is the easiest to ignore in the whole process, but I will do it now\u3002<\/p>\n

A lot of people produce a good-looking semi-portrait, and they just take it for video\u3002<\/p>\n

THE PROBLEM IS: A POSITIVE PICTURE, AI HAS NO IDEA WHAT THIS GUY LOOKS LIKE BEHIND HIS BACK\u3002<\/p>\n

You turn him around, his clothes change, his face changes, his hair changes\u3002<\/p>\n

\"cce1029ej00tlc4sc00fd000pv00b2p\"<\/p>\n

\"ef485c11j00tlc4sj00kdd000px00axp\"<\/p>\n

This is a three-view structure<\/p>\n

I used it very easily: the left for the face lock, the right for the body, the dress and the back structure\u3002<\/p>\n

This one can be used as a reference in the back, whether it be near, side or back\u3002<\/p>\n

Summary of the introduction:<\/p>\n

I've had a week of these character tips, and the results are as follows\u3002<\/p>\n

\"42c14344j00tlc4st00ipd000pu00dop\"<\/p>\n

\"5e0148b3j00tlc4t300gyd000px00dip\"<\/p>\n

It's not easy to create a summary. Kneel to the audience and ask for their support<\/strong><\/p>\n

Personal tip (9:16)<\/p>\n

\u25cf<\/p>\n

Advanced senses, atmospheres, senses, clarity, fine nature, oc renderings, cg renderings, true hostage feeling, 3D paranoia, real people's five-eye ratio, characters like bjd, muscular, hairy, skin-sensitive texture, real, hairy and rooty, details like five details, hairy face glowing, virtual engine background glitters, warm\/cold hard light (select one). The light on the face is evenly bright. Black and soft, pure black background, make-up photos, front lenses, half body photographs. + your hint<\/p>\n

Three views (16:9\/21:9)<\/p>\n

\u25cf<\/p>\n

Generate a three-dimensional view of the body and a facial close-up. The top left is 1\/3 full of super-large facial features, 2\/3 right, positive view, side view, back view, two or eight body parts ratio, pure background<\/p>\n

Props assets: separate props that will be involved in the movement<\/p>\n

Once the character is identified, there are other items that need to be prepared individually\u3002<\/p>\n

This time, the sword, swords and acoustics were made\u3002<\/p>\n

The acoustic stone appeared to be a small adornment, but it had to be pressed to patrol at the beginning of the forest, and Luen's call for help came through\u3002<\/p>\n

The camera will be specific, and it needs a clear look\u3002<\/p>\n

The use of silver Gothroscopes and ice-blue-haired crystals has been envisaged, with proxies and different angles\u3002<\/p>\n

\"614d2bd5j00tlc4tj00fod000no00ecp\"<\/p>\n

The hint can be written like this\u3002<\/p>\n

\u25cf<\/p>\n

Fantasy short set, with an acoustic stone that can be worn in the collar, a silver Gothic cologne wrapped in ice-blue light crystals, with minor wear on the edge of the metal. Solid background, left side, with clear features, right side, side and back, with the same shape, material and mosaic structure, and back side, with the wearing of a connection. No characters, no hands, no words, no watermarks\u3002<\/p>\n

Before we do the spectroscopy, we find these props from the script, marking the first place they appear\u3002<\/p>\n

There's something in the back\u3002<\/p>\n

VI. Video generation<\/strong><\/p>\n

All the characters, scenes and props are ready to start producing videos\u3002<\/p>\n

Upload assets to LibTV:<\/p>\n

\"03df3cej00tlc4t00ard000lc00j6p\"<\/p>\n

link to corresponding asset, select sd2.5 model, full reference, select the corresponding ratio and number of seconds\u3002<\/p>\n

\"760311bej00tlc4u2006ed000pn00e5p\"<\/p>\n

A simple template for hints<\/p>\n

For example:<\/p>\n

\u25cf<\/p>\n

CG FILM MASS 3D REVEAL SCRIPT FANTASIA, 16:9\u3002
\nTHE OPENING CHARACTER A IS ON THE LEFT SIDE OF THE PICTURE, WITH THE BODY TO THE RIGHT, AND THE CHARACTER B IS ON THE RIGHT SIDE OF THE PICTURE, FACING THE PERSON A, WITH A WOODEN TABLE BETWEEN THEM\u3002
\nCamera 1:2s Midview fixed, character A looks up to character B and says, \"You're here.\"
\nCamera 2:4s cuts the profile of character B, who did not answer and held his right hand slowly\u3002
\nCamera 3:3s character A's face is getting serious and the camera is moving forward slowly. Only white, clothing friction and indoor sound, and no background music\u3002
\nNo subtitles, words, watermarks, music<\/p>\n

But when the video is actually generated, almost all people will have a problem \u2014 different batches of video, and the person's position is easily changed\u3002<\/p>\n

THE LAST ONE IS STILL LEFT OF THE HOME, RIGHT OF THE HOUSE, TO THE NEXT ONE, AND AI MAY BE ABLE TO REPOSITION THEM DIRECTLY\u3002<\/p>\n

BETWEEN BATCHES, AI DOESN'T AUTOMATICALLY REMEMBER THE SPACE RELATIONSHIP IN YOUR LAST VIDEO\u3002<\/p>\n

So every time I start a new video batch, I'm going to rephrase them:<\/p>\n

The location of the person and the person, and the location of the person and the scene\u3002<\/p>\n

So every time a new video batch is opened, two things are re-stated: the location of the person and the person, and the location of the person and the scene\u3002<\/p>\n

This step seems cumbersome, but it is very important for the coherence of the story\u3002<\/p>\n

\"97784e20j00tlc4ud006jd000q400iap\"<\/p>\n

In cases where the position of the person is more complex, a frame of silence can be intercepted from the previous video generated\u3002<\/p>\n

Upload to the full reference of the video in this batch and clearly tell it: \u201cThis figure refers to the person's position and position.\u201d<\/p>\n

SO, AI HAS A VISUAL REFERENCE, AND IT'S EASIER TO CONTINUE THE CHARACTER POSITION IN THE VIDEO, SO THAT THE BACK-TO-BACK LENS IS MORE NATURAL\u3002<\/p>\n

\"6b123371j00tlc4uo00cvd000pl00c7p\"<\/p>\n

When you encounter a penetrating lens, you can also intercept the frame of silence, correct the image and recreate the video separately and replace it with the original one\u3002<\/p>\n

There is no need to reset the entire draw and save points\u3002<\/p>\n

\"0dbd1a36j00tlc4u07ud000ps00erp\"<\/p>\n

VII. Clips fine-tune: Sound is the key to your video success or failure<\/strong><\/p>\n

Once all the shots are produced, they will enter the final editing and later stages\u3002<\/p>\n

The material is integrated into the clippings, which are grouped in sequence, and the lens is adjusted to the rhythm, with subtitles, music, sound of the environment and the necessary rotations\u3002<\/p>\n

HERE'S A LITTLE SUGGESTION: WHEN YOU GENERATE VIDEO, TRY NOT TO LET AI AUTOMATICALLY ADD BACKGROUND MUSIC\u3002<\/p>\n

BECAUSE THE BGMS THAT COME OUT OF DIFFERENT LENSES CAN EASILY BE DIFFERENT, AND IT'S GONNA GET REALLY JUMPY\u3002<\/p>\n

It's better to stay in the editing stage, to listen more coherently and to feel better\u3002<\/p>\n

Sound effects also need to be supplemented at a later stage\u3002<\/p>\n

AI DIRECTLY GENERATED VOICES CAN SOMETIMES BE THINNER THAN THE IMPACT OF IMAGES\u3002<\/p>\n

So this time I'm reprogramming, and I'm repainting, and I'm repainting, and I'm filling, you know, the sound level of action, the environment and people\u3002<\/p>\n

On the image, I also partially modulated some of the footage, and duly added a little light to give light to the high light, so as to give further prominence to the sacred taste of angels and temples\u3002<\/p>\n

\"e695a5ddj00tlc4v9005d000ps00dkp\"<\/p>\n

At the end: Run through, then perfect<\/p>\n

HERE, THE COMPLETE PROCESS OF AN AI VIDEO FROM 0 TO 1 IS CLEAR:<\/p>\n

AN IDEA TO WRITE A SCRIPT, A STYLE TEST, A THREE-VIEW VIEW OF THE PERSON, A SCENE \/ PROPS ASSET, A VIDEO GENERATION, A CLIPPING<\/p>\n

The best thing to do next is really run it\u3002<\/p>\n

The hints will never end, but the first video will have to be made by yourself\u3002<\/p>\n

The first one does not have to pursue perfection and run the process as a whole, more than anything\u3002<\/p>\n

There are many problems, and the only way to do that is to know where the card is; many of the techniques, too, are slowly coming up in a few adjustments\u3002<\/p>\n

Let's start with article 1 and then with articles 2 and 3. Every time we finish, we'll go back and see what steps we take, what steps we take, how faster we'll go next\u3002<\/p>\n

You've done a lot of things, and it's your own way of doing it\u3002<\/p>","protected":false},"excerpt":{"rendered":"

CAN ORDINARY PEOPLE MAKE BIG HOLLYWOOD MOVIES? IF I TOLD YOU, IT'S LOW COST, YOU DON'T HAVE TO GO TO THE THEATER, YOU'RE HOME ALONE, YOU CAN MAKE A BIG MOVIE OF THE RING-STYLE. AND THIS ARTICLE, I'M GOING TO TAKE THIS WHOLE TOOL OUT OF YOUR HEAD. AFTER LEARNING IT, YOU DON'T JUST LOOK AT IT, BUT YOU REALLY KNOW HOW TO MAKE A BIG, HIGH-LEVEL VIDEO THAT BELONGS TO YOU. WHAT STEPS ARE TO BE TAKEN BETWEEN AN IDEA AND A COMPLETE FILM? JUST RECENTLY, I'M DOING AN AI CLIP LIKE \"THE RING.\" TAKE THIS CASE TODAY AND UNTANGLE THE PRODUCTION PROCESS FROM 0 TO 1. FROM HOW THE SCRIPT IS WRITTEN, HOW THE CHARACTERS ARE DETERMINED, HOW THE SCENE IS DONE, HOW THE MIRRORS ARE REMOVED, HOW THE VIDEO IS GENERATED, HOW THE FILM IS FINALLY CUT INTO A FULL FILM, I'LL CLOSE IT<\/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":[3264,956,5321],"collection":[],"class_list":["post-57324","post","type-post","status-publish","format-standard","hentry","category-jiaocheng","category-baike","tag-ai"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/57324","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=57324"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/57324\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=57324"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=57324"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=57324"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=57324"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}