AI VideoHOW DO I GET UP TO 15 SECONDS, AND MORE THAN 15 SECONDS OF IA VIDEO QUICKLY? TODAY WE SHARE THREE OF MY FAST-TRACK METHODS
First 15-second hint:
Foundation: a 15-second film clip of China’s ancient snow mountain war. Main character: General Qin Qin Zhonging, General's horse {Mixed 3}, General's Red Gun {Mixed}, Rebel: an army outside the country }Mixed 5} } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } Images of a strong visual impact. The images have a strong sense of shock and shock. The whole picture has a rhythm and a film sense and produces only sound effects of the environment, not background music. Real people write Chinese epic style. The image prohibits any subtitles and any white fonts; the image appears in text and uses transparent fonts. It's all about the quality, the classic martial war movie style. It's a very cold snowy canyon. The style reference for all of the person's facial details is image one: low angles for the mirror. The camera was attached to the horse's foot and foot, and followed the horse slowly to build up the snow, the snow was crushed by the horse's foot and the ice was splattered to the camera. The camera swayed along the front leg of the horse, passing through the Qin Zhao Ning's red and black armor with a red Zhao Ning rifle and the rope in her hand, and eventually advanced to her nearness. Qin Zhaoning was riding at once, looking at the Canyon. Image two: The camera switches to the sky view. Qin Zhaoning led the army in the narrow snow canyon. The female general was at the forefront, and the team was caught in the middle by steep snow cliffs on both sides. Image III: The aerial lens continues to rise slowly and move across the canyon. In the snow and snow at the top of the cliff, a large number of enemy shadows are gradually emerging. Toggle the spectroscope, the enemy stands from the back of the rock and snow, and some of the enemy forces combine to push the rock at the edge of the cliff, which slowly falls from the wall. The enemy was laid on both sides of the canyon, forming a high-lying siege. Image IV: A soldier who looked up and found enemy forces on the cliffs and the stones that began to roll, turned his face and pointed to the top and shouted: "There is an ambush!" The pebbles struck individual soldiers, and the balconies started to riot and the horses were frenzied. The camera went to the face of Qin Zhao Ning, where she looked up, her eyebrow was tightened and she looked back at her soldiers and was stoned. In the eyes, both anger and perseverance appear。
Method I: Key frame interception
This is an improved way to use pre-sequent videos to generate follow-up videos, and to change the original reference to the whole video to a few key frame pictures。
It's got two advantages, one of which is to compare reference video with raw video, it's to halve the score, and the other is to directly use reference video to produce video, and to generate three consecutive times, the image starts to have some noise, and that way it can solve this phenomenon to a certain extent。
Specific exercise: the key frame for intercepting the pre-sequenced video -- generating the colour card (used in a regular style) -- using the key frame of the card + the presequence video as a reference material -- generating follow-up video
- Key frame capture: Focus on the following:
- Colour card generation (2K paint mass, image 2.16:9)
ANALYSES ALL THE REFERENCE PICTURES I UPLOADED, EXTRACTS THE MOST REPRESENTATIVE OF THE MAIN COLOURS, SUPPORTING COLOURS AND DOTS, AND ORGANIZES THEM INTO A HORIZONTAL PROFESSIONAL DESIGN COLOUR CARD. THE IMAGE FOLLOWS A VERY SIMPLE AND RESTRAINED EDITORIAL DESIGN STYLE: A WARM WHITE-WHITE PAPER BACKGROUND WITH A ROW OF ROWS WITH A SHORT COLOUR ATMOSPHERE DESCRIPTION; 12-13 RECTANGULAR RECTANGLES OF 12-13 EQUAL WIDTH IN A HORIZONTALLY ORGANIZED ORDER, WHICH ACCURATELY REFLECTS THE OVERALL COLOUR OF THE REFERENCE MAP FROM SHALLOW TO DEEP NATURE, AND DOES NOT ADD THE COLOURS NOT FOUND IN THE REFERENCE MAP; THE CORRESPONDING HEX COLOUR VALUES ARE INDICATED AT THE BOTTOM OF EACH LINE, WITH A SHORT CHINESE STANDARD COLOUR NAME AND THE INTENDED USE SCENE. OVERALL, THERE IS A SENSE OF QUALITY IN INDUSTRIAL DESIGN FILES, FILM COLOUR STUDIES, MATERIAL SAMPLES AND ADVANCED BRAND VISUAL MANUALS. FULL WHITE, WELL-SATURATED, LOW SATURATED, DUMB PAPER, SOFT AND NATURAL LIGHT, FACE-TO-FACE, CLEAN, CLEAN AND UNDECORATED
- Follow-up video generation:
-Phrasing:
[Fundamental setting: a 15-second film clip from the ancient China Snow Mountain war. Follow-up action to create a frame image: General Qin Zhonging {Mixed 6}, third view {Mixed}, horseback, hand-held red jangle gun in the picture 反 } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } } Images of a strong visual impact. The images have a strong sense of shock and shock. The whole picture has a rhythm and a film sense and produces only sound effects of the environment, not background music. Real people write Chinese epic style. The image prohibits any subtitles and any white fonts; the image appears in text and uses transparent fonts. It's all about the quality, the classic martial war movie style. It's a very cold snowy canyon. A style reference for all people's facial details
Image one: Panorama, Qin Zhaoning's army was hit by stones, many wounded. Toggle Qin Zhao Ning's face, with a strong eye, and shout, "Foreign army, follow me out of the valley!" Image two: Sideside with Qin Zhaoning leading the cavalry to the valley. Turning to the first perspective, the enemy cavalry appeared slowly in the front wind. Image III: Using a film-level multispectroscopy switch and combination, Qin Zhao Ning turned his head, his eyes turned stronger, he shouted, the horse stormed into the enemy, and several enemy soldiers were shot down. It also entered the back of a turn of the canyon, followed by enemy forces from outside. Image one: On the front, Qin Zhao Ning was running in the canyon alone, followed by enemy forces, two: Switching the hostess' periscope, deforming the background, and the enemy was chasing after the other. Suddenly, the hostess pulls the rope, stops the horse and closes to the low angles of the horseshoe. Then, in the near view of the female master (there was no one behind her), the female master seemed to have seen something and to have a strong breath. _
Caveats:
(a) The screenshot (except for the frame) should not be quoted directly in the reminder box, but only as a reference
In the second video, the scene of the initial scene, as far as possible, is not a reference to the picture in the material (otherwise, there would be some noise in the image)。
Method II: Spectrum integration
This method is a 25-30-second reminder, followed by a repeat of a 15-second video with the same hint, and then split the video into multiple spectroscopys, and then, like a director, choose a useful spectroscopy, based on the action and rhythm of the person in the video, and remix it。
Specific exercise: Generates a 25-30-second long hint -- multiple times the same hint -- each video is split into multiple spectroscopes -- based on the person action/sequence, etc., and completes the combination of the lens
-Phrases generation: (Gemini or GPT)
I'm going to take a 1,000-word hint, about 25 seconds - 30 seconds. Story: The female general came out of the canyon on horseback, followed by a chaser, with an enemy of thousands of troops ahead. Turning back, the general then looked at a very wide ice river next to it, and the general then ran in the direction of the river, and the leader of the hostile forces called for her to be taken and not to be allowed to run, and the general was on her way to combat the enemy (with cavalry infantry) (which required a lot of details) and to kill many of them. The female general ran into the glacier, and the enemy stopped on the bank, afraid to go down, and seemed worried about something. The female general ran fast on the ice and suddenly caught a storm. Refinement: Then my hint
- Video generation
Image 1: In the immediate aftermath of the episode, using a wide-angled view, in the direct front of the female general, the enemy formations in the mountains are present in the blizzards of the snow, and the strangulation of thousands of horses and horses melts into the cold. The image was replaced by that of the female general, whose eyes were cast to the right. The lens is transposed to the spectroscope, and a vast, vast glacier that cannot be seen at the end lies on the snowy wasteland. The snow of geese and fur dances like muddles, and the mountains of snow in the distance rise and fall, and the naked ice surfaces dark and dark blue. Image II: The female general squeezed the rope, and the horse squeaked with a chilling heel, and forced an extremely dangerous drift into the heavy snow, driving towards the dead, wide ice. Image III: In the midst of the enemy, the camera went into a long-focused camera (Telephoto lens close-up), and the leader of the enemy made a deafening noise (as his eyes were suffocated in the cold air with a big mouth and a heavy white haze). He lifted up the heavy machete in his hand with a deep blood cell, the knife shone in the cold under the dim snow, pointed to the female general who fled to the ice in the snow, and ordered her to die. As the leader roars, the picture cuts back to the center. A group of cavalrymen and infantry, straddled by snow from their knees, came out of the flank formations and leaned in the direction of intercepting the female general's path. Hundreds of horses are running at the same time on the fourth scene: the lens cuts into the extreme accelerator, twitching and reeling mirrors, the female general uses the speed, the muscles of the right arm to burst, the long guns in her hand, and the sharp tip of the gun goes right through the wooden shield raised by the first masked infantry and pierces his chest nails without stalling. The female general uses the horrid habit of moving forward with her horse, flipping his wrists very fast and tearing it horizontally! The body of the infantry was picked out like a rag bag. Immediately thereafter, an enemy cavalry on the left waved a machete and hit her head at horse speed. The female general showed the power of the horse and the heart of the waist. Her feet were standing on her horse, and the upper half fell back in an instant, almost in parallel with the horse's back (the classic martial arts “iron bridge” shunned motion). The enemy's machete emptied her nose and the flying snowflake. In an absolute moment of staggering between the two horses, the female general rose up like a spring, and returned to her body, holding a long gun with her hands behind her back, and with her waist he took a very ferocious retort (Backwards space rifle). The gun pointed precisely into the back of the interlocking enemy cavalry, where he was stabbed directly from the horse's back and thrown into a deep snow pile. In eight short seconds, the horse was undivided, and the female general was shot like a dragon and was born in a storm to kill a trail of blood, showing a strong, hard-core, hard-to-skinned sense of epicism. Image five: Low angles with a high hemorrhage with the shot of a black horse through the wall, and the hooves leap out of the heavy snow and set foot on the vast ice. The camera was raised immediately, using a wide-angle Drone shot from the UAV perspective, and the female general was fast on the ice alone. Hundreds of desperate pursuits have been stoked to death by the snow banks on the edge of the glaciers. The enemy, in the middle of the snow, was staring at the fragile ice, which was wide and shrouded in a dark blue light, with no one dared to take another half step forward。
Once the first generation has been completed, the key frame that can be used is intercepted as a reference material for the second generation。
And then the first one is already available, so you can remove the hint of the image and click on it
In the same way, two or three times, a video of several images is generated。
- Smart splits + clips
Use a cut-off lens with a smart split, and then combine the rhythm of the video with each link of the mirror to complete the mix of the video, with a transposition between parts。
Comment: The first two approaches combine to produce more than 15 seconds of video for most of the scenes, and the production of general short/short films is almost sufficient。
Method III: Storyboard method
Principle: Using a 30-second full storyboard, the whole video's image, the style, is fixed. And then, in two minutes, you're gonna create two 15 seconds before and after, and then you're gonna put it together。
SPECIFIC EXERCISE: GENERATE A 30-SECOND PICTURE OF THE STORYBOARD -- THE FIRST 15 SECONDS OF MR.'S VIDEO -- THE KEY FRAME IS INTERCEPTED AND THE NEXT 15 SECONDS OF VIDEO IS GENERATED -- TWO VIDEO JIGGLES -- THE STORYBOARD PICTURE AI
Plugin generation: (for Gemini or)
I'm going to generate a Japanese school-style, pink-can-sbuzz-water-advertisement
-Storyboard Image Generation (GPT-image2)
The whole picture and layout: Generating a very fine, Japanese-style commercial advertising spectroscopy (Storyboard). The image scale is 16:9 and the background uses a warm, gray, rough paper texture with a mild film particle sense. The top banner contains a style label for the project title “Sakura letterhead |Sakura POP bubble water” and for 30 seconds, 16:9 scale, full colour and “Shunji Iwai Style x Cinematic Nostalgia x Pure Sound x Narrative”. Scene 01-08. Each module consists of: top scene numbers and precise time axes to 0.1 seconds (e.g. 0.0-4.0s); two parallel frames (Start Frame and End Frame) to display a dynamic moment; the following is a camera motion description with arrows (photo-language) detailing the terms "Static" , "Slow Track" (Dolly-in), "Rack Focus" and "Handhold Tracking". Visual narrative: Scene 01 (0.0-4.0s) [P] : Camera 1 shows empty classrooms after noon, white curtains blown up by the breeze and light outside the windows produced a slight over-exposure (Halation). Camera two slowly pushes to an old wood table by the window, with visible dust hanging in the air. Scene 02 (4.0-8.0s) [Silent Test]: Once the lens was swayed with a hand, a short-haired girl (in a loose summer sailor's suit) reached the empty desk with her hands and feet. The second camera turned to a close-up, with young girls carefully placing a pink SAKURA POP bubble at the corner of the table, with condensed beads on the surface. Scene 03 (8.0-11.5s) A folded white paper (letter) was pressed at the bottom of the jar, with a slight break in the finger. In camera two, there were footsteps in the hallways, girls turned their heads in shock, they couldn't see their faces in reverse, and only the hair of flying dances. Scene 04 (11.5-15.0s) [light staggered] : a camera shows a classroom to pull the door open and a young man's blurry image appears at the door. Camera two uses the Focus Switch (Rack Focus) and the focus quickly shifts from young people far away to pink bubbles on the nearest desk, where the sun falls right on the cherry. Scene 05(15.0-19.0s) [Coagulated calorie]: The hand of a close-to-earth youth picks up the jar of bubble water, a drop of cold condensed water falls to the tip of the finger and damps the desktop. Camera two shows white paper (letters) on the bottom of the jar being exposed to air, with the edge blown up slightly by the breeze. Scene 06 (19.0-23.0s) [Voice echoing] : Camera one with a young man's long finger to pull the loop open and a fine pink fog comes out of the flames in reverse light. Camera two uses a match clip (Match Cut), which cuts in an instant to the girl hiding behind the corner wall of the corridor, closes her eyes against the wall as if she were listening to the sound of the can open. Scene 07 (23.0-27.0s) [unknown ending]: a cherry flower with a low angle on the camera, a young man's head on top of his head, a bubble water, a twirling throat, looking out the window. Camera two is moving slowly (Pan), and the unopened folded white paper is still lying still. On the desktop, there are water marks. Scene 08 (27.0-30.0s) [Young youth]: The lens cuts back to the corridor, and the young girl has left, leaving empty corridors and moving light. The second image is now white, and the central and fine-looking Japanese font, "SAKURA POP" , is written below with a very small font, "Summer without an exit" , an unconventional commercial button that displays a great whiteness. Bottom Design Code (film level custom): One third of the bottom of the picture is a detailed brand manual. Key Color: contains five colours: Decorated cherry powder (1 TP5TE8D3D8), film blue (1 TP5T8FA8B4), dilution white (1 TP5TF7F9FA), tree shadow green (1 TP5T4A5D54), warm gray wood (1 TP5TC1B6A9) with code. Highlight low contrast and low saturation. Typography: all words are marked with a "building of a purple dynasty" or similar traditional solar liner, with very fine words and extremely wide word spacing (+15%), which displays a fragile sense and poetry. Audio (Music & SE): Mark " Pure environment sound and pronunciation " (absolute backgroundless music/No BGM). The effect is: the wind blows through the curtains, the faint sound from the far, the small breath from the young girl, the silt slide from the door to the classroom, the sound of the can open (scrambling-scrambling), the sound of the delicate bubble burst, and the sound of the school bell from the far. The sound design is required to have a strong sense of space and an ASMR-like white noise. Technical Specifications: Contains 24fps, 1920x1080 delivery criteria and a check box for the Fujifilm film simulation colour LUT (e.g. Classic Negative or Eterna). The style requires that the whole is a very transparent and depressed Japanese-young film style. Using "Natural Light & Soft Focus" natural light and soft focus lens, the image is cold and blue, with warm perigee edges at the height. It is necessary to use Chinese for all scripts, to guide the construction of minor drama conflicts through physical interaction and vision, to follow the extreme white aesthetics, to abandon a strong sense of commercial oppression and to pursue emotional values and film art。
-The first video is generated
Video alerts: (Do not generate BGM, otherwise the next two lines will be difficult to spell) Base-set: based on the information on this spectroscopy storyboard, generate a professional commercial commercial in pink cans of bubble water, filled with a sense of Japanese aesthetics and narratives. Image content generation is 0-15 seconds only. Style setting: Visual tone: modern Japanese youth aesthetics, simulations of the air permutation of Medium Format Film in 120, or digital filmmaker with a pure sense of cleanness of Black Pro-Mist. The images are clean and the particles are very weak. Phototelevision: High-key Backlight and wide-reflective environmental light. A slight spill of high light (Bloom/Halation) with soft cherry pollen and sunlight; and a cold blue (Cyan-blue) in the dark. The day system, with its low contrast and low saturation, is curable. Do not generate background music, only sound. Image taken from: Scene 01 (0.0-4.0s) : Camera 1 shows empty classrooms after noon, white curtains blown up by the breeze and light outside the windows produced a slight over-exposure (Halation). Camera two slowly pushes to an old wood table by the window, with visible dust hanging in the air. Scene 02 (4.0-8.0s) [Silent Test]: Once the lens was swayed with a hand, a short-haired girl (in a loose summer sailor's suit) reached the empty desk with her hands and feet. The second camera turned to a close-up, with young girls carefully placing a pink SAKURA POP bubble at the corner of the table, with condensed beads on the surface. Scene 03 (8.0-11.5s) A folded white paper (letter) was pressed at the bottom of the jar, with a slight break in the finger. In camera two, there were footsteps in the hallways, girls turned their heads in shock, quickly left the scenes, lost their faces in reverse light, and only the hairs of flying dances. Scene 04 (11.5-15.0s) [light staggered] : a camera shows a classroom being pushed and the door being pulled open by a teenager, and a young boy's blurry clip appears at the door. Camera two uses a focus shift (Rack Focus), which rapidly shifts from young people far away to pink bubbles on the nearest desk, where the sun hits on the cherry blossom
-The second video is generated
Finally, two video clips were completed and a 30-second video was completed. Individual shots are not satisfactory, not all of them. Reference is made to the following hints and storyboards and screenshots (with significant savings)
These are the three ways to generate more than 15 seconds of video。