The core elements of the script (roles, scenes, props and special effects) and the spectroscopy were shared in the mandatory part of the script
I'M GOING TO GET AN I.A. SHORT PLAY
The core task of this note is to adopt a list of detached characters, scenes, props, special effectsBean bun, dreamTwo tools, includingFigure three, main scene, prop and special effects, provide precision material for the subsequent generation of spectroscopy images, with the core requirement being “mapping and splitting lists, consistency of styles, meeting specifications” ~
I. Core course objectives
Proficiency in the process of soybean bags, i.e., the production of living images from dreams, and the ability to describe accurate editorial tips in conjunction with the decomposition of scripts, leading to an efficient conversion of “text-required visual drawings”
Note: [If a flying book is used, you can try to use a shortcut to the field, select a suitable model, and automatically generate a hint based on a visual description; if you use a bean bag for a cleavage, you can also generate a description attempt based on a bean bag, adjusted for a direct dialogue if it is not appropriate. ]
Example of flying book operation:
2. Identification of criteria for the generation of character three viewes, scene maps, trophies and utensils to ensure that the drawing details fit the script set without bias or omission
II. CONTEXT CONTEXT AND RESPONSIBILITIES
It is important to follow the previous splitting list in the drawingsKeep the whole style consistent(e.g. phantom, reality, doubt, etc.) Avoiding material fragmentation and subsequent production
The whole style story, suggest a style by which to look at the same kind of short play in the market
(i) Third view of the person
1. Core requirements:Should includeIt's the three angles of the front, side and backPrecision reduction of role haircuts, clothing, deformities, emblematic movements and detailed features (e.g., accessories, clothing trails), co-ordinated, unified
2. Key points:Identify the visual elements of character identity and character, mark differences in emotional state in different settings, and avoid inconsistencies in three view details
3. Step of integrity:
(1) We've sorted people's stories
It's a small one
Jiang Fubao: Visual age: 3 years old, female master, modern society animal, good, smart, independent, resilient, inexhaustible, vendetta. She died as a Chinese doctor after working overtime and overworking her three-year-old, who crossed the village of Jiangjiang in Tianling. The child was loved by his family, which allowed him to feel the warmth of her home, and she decided to live a good life with her own room for rental and a golden finger for hearing animals and plants
(2) Generate corresponding role cards
(3) Additional important information
Identity and the background of the times and the appearance of the person, the style of the dress, for example: it's a farmer, it can't be a fancy dress, it has to be a character
(4) Examples of birth maps (example role design)
(5) The images produced are a three-view view of the person, his expression, his attitude is all on a one-size-size-fits-all image, which is divided by a dream, into three-views of the pure colour background and his face
According to the profile, a corresponding role card is generated:
[Phone Name]
Visual age
Sex
[identity/era background]
[core character keyword]
"Specific Keyword"
"Cress style"
Key Set/Gold Finger
♪ Live graph hint ♪
[3 View Tips]
• Heads:
• Side:
• On the back:
Key additional information:
I. Identity and the background of the times and the appearance of the person, and the style of the dress should be consistent, for example: this is a farmer, it cannot be a fancy dress, it must be the character
2. It is important to be detailed by reference to the following requirements:
Performing role design (Character Design), image style: 3D sensor-style UE5 rendering based on [role card], in turn producing [scale set table] [model set table] [ emoticon set table] [hatting table]:
The role card
– IMAGE STYLE: CG-STYLE CUISINE, CG-BASED ANIMATION
– IDENTITY: FORMER INTERNET NETWORK ANIMAL PROGRAMMER, WHO TRAVELED OVER OVERTIME AND DIED IN THE SERVICE OF THE SANCTUARY OF HEAVEN, USING THE PERFORMANCE, KPI, PPT AS A MASTER OF SCIENCE。
– Characteristic characteristics: seemingly lazy, unstoppable, mouth-to-mouth-to-mouth, master poacher; well-defined, critical moment-to-time, serious, good at dismantling problems, developing processes, and stabilizing emotions. “Impressed by the “dry” “ineffective effort” and the belief that “undoing something stupid is too hard”。
– APPEARANCE – BLACK HAIR, ALWAYS A LITTLE BIT OF A MESS, BUT LONG PERIODS OF FATIGUE. THE UPPER BLACK T-SHIRTS, THE DARK ROBES WITH THE ELECTRIC VIBRATOR ON THE OUTSIDE, THE LOWER JEANS, AND THE FOOT ON THE “SPIRITED SNEAKERS” ARE AT ODDS WITH THE ANCIENT FRIAR。
– MARKER PROPS: THE MAGIC CHANGE IS SIMPLE, AND LOOKS LIKE AN OLD SMARTPHONE THAT POPS UP A PPT, A MIND MAP; A HAND-IN-HAND MARK THAT CAN DRAW FLOW CHARTS, FORMULAE, SHORT-LIVED WORDS AND LIGHT, AND STRAIGHT TO THE ESSENCE。
– Typical gold phrase: “Don't use tactical diligence to cover up your strategic laziness — this is a cuisine, not a brick move.” _Other Organiser
1. Portion Sheet image style: 3D Monument UE5 rendering
– Demonstrating role height comparison, main body ratio structure (head ratio 1:7)。
2. [Model Sheet / Turner] / Three View Image Style: 3D Sensor UE5 Render
– Standard positive view: front, side, back。
3. "Expression Sheet" video style: 3D Monument UE5 rendering
– Nine different facial expressions and emotions showing the role in 3x3 tables。
4. [Pose Sheet / Action Sheet] Imagestyle: 3D Discretion UE5 Render
– Nine different positions, representative dynamics, representational features, representational character, motor characteristics, corresponding to classic gold sentences, displayed in 3x3 tables
On the basis of the hint, you can figure out:
It's not hard to see the direct effects of bean buns, such as the fact that a person's face and pose are not working well, which may be too much of a hint, and it's really difficult to present on a map
So it's better to separate the three-view and the face-portrait
For example, there is a message that directly provides role cards and face positionsSo no one's ever really thought about why the short play needs to produce these, not just positive, overall images of people
That's actually..In order to keep the character in the short play, after all, the video character is dynamic, and the more details are provided in the picture, the less likely the problem will be
I didn't know much about it until I was in the endOne key to generate three viewsThe toolslike oioii and sekoIt's mostly a one-click video platform, which is a direct three-view
So, to make short play videos, you have to have three views, but you didn't know! So it's a 10-hour draw card, and it's numb to the pit
PersonallyI mean, if you want to be stable, it's a bean wrap
(ii) Main scenario
1. Core requirements:Focus on the core scene, the type of retrofit, the overall atmosphere, the core set and the photo properties, the visual level of the core and the secondary areas, and the suitability of the drama
2. Key points:Controls the complexity of the scene, matching the aesthetic and subsequent spectroscopy (not too cumbersome to cause confusion in the spectroscopy), and the light effect needs to serve the scenery environment (e.g., a strange scene with low saturation)
3. Step of integrity:
(1) BASED ON THE SCENE MAP TIPBOARD (THE TIPWORD IS STILL BASED ON THE IMAGE DESCRIPTION IN THE SPECTROSCOPY AND IS EDITED USING THE AI TOOL)
(2) Difficulty: Multiple angles of the scene
THIS MULTI-PERSPECTIVE FIGURE NEEDS TO BE USED AS A BASE MAP BASED ON THE APPROPRIATE BASE MAP DRAWN IN (1), AND THE ISSUE IS LEFT TO AI FIRST
The effect of the dream:
In fact, all kinds of graphics tools can try to do the above, but the results are mixed, so you can just choose your own satisfactory angle, and you can get plenty of pictures over and over again
(iii) Props
1. Core requirements:Core props need to be presented at multiple angles (head, side, state of use), with colours, materials, shapes and details related to role/situation, with secondary prototypes as a clear basis
2. Key points:The size of the prop needs to be coordinated with the role, scene ratio to avoid an imbalance in size, and the detail of the prop directly influences the authenticity of the subsequent spectroscopy
3. Phrasing + Figure
Prompt tip
(a) A three-view combination of 16:9-scale props, based on the structure of the pedagogy reference map (a complete consistency)
Each map contains a standard face-to-face view of the front, side, back, with a clean white base/no shadow, no person/scene/negative elements, and a design sense for the layout profession, with clear design features
(iv) Special effects maps
1. Core requirements:Revert special effects names, trigger scenes, visual styles, colours and dynamic logic, with dynamic orientations (e.g., the spread of magic effects) for static effect maps
2. Key points:Combine short play speed, special effects do not allow for complex redundancies, visual effects match scenes, roles, avoiding noise, priority for recognition
3. Phrasing + Figure
RECOMMENDATION: IN MANY CASES, THE SPECIAL EFFECTS ARE EXPRESSED WITH PRECISION, SO IT IS SUGGESTED THAT SMALL PARTNERS USE PHOTO INVERTS TO ALLOW AI TO GENERATE A TEXT MESSAGE FOR YOU
Tool exercise: bean bag + dream use instructions and attention
(I) BEAN BAG (AI GENERATION+PHRASING TOOL)
1. Core uses
Fits to generate a third view of the person, a simple prop and a graph of the effect, while supporting the optimized hint(Attemption for poor performance of the dream map, adjust accuracy with bean buns)It's good for a little white
2. Operational processes
(1) Phrasing:Based on splitting list, pressCore Elements + Details + Style Requirements + Perspective RequirementsBuild hints
(2) Generating operation:Opens the bean bag, selects the Image Generation function, pastes the hint, sets the size of the picture (mainly for a uniform format) and clicks to generate it
(3) Effect adjustment:If you generate graph detail deviations, add amendments to the original hint and recreate
3. Notes
(1) Precision is required to avoid vague expressions (e.g., “good-looking clothing” instead of “white shirt plus black suit pants with buttons on his sleeve”)The more the details, the more the effects fit
(2)When a third view of a person is generated, the three views and three perspectives (specific face, side, back) are identified, otherwise it is easy to produce a single perspective
(3)We need to avoid adding too many elements at a timeEasier to create muddled and unbalanced maps, with step by step to optimize details
(4) After generationSave original maps in a timely manner, indicating the corresponding role/protocol nameFacilitation of subsequent classification management
(ii) The dream (visualized biographic tool, with emphasis on scenery and atmosphere presentation)
1. Core uses
It's a good way to generate major scenes, complex trophies and special effects with an atmosphere that is better than bean buns with a sense of quality and atmosphereFitting requires material that enhances visual impact
2. Operational processes
(i) Preparation of prior periods:Open a dream, log-in an account, create a local folder of material in advance (by “person/scene/props/effects”) to facilitate rapid archiving after generation
(2) Phrasing:Refinement of bean bean packs, supplementing the atmosphere and light description
(3) Parameter Settings:Select the corresponding model (4.5 4.1 can be tried) and adjust the resolution (suggested maximum resolution to ensure material clarity)
(4) Generate and fine-tune:Click to generate, using a "local redraw" feature correction (e.g., repositioning of the prop in the scene, additional details); or regenerated after optimization of the hint if the overall bias is large
3. Notes
(1) The “outdoor/outdoor” “whole atmosphere” and core scenes need to be identified when the scene is generated to avoid the generation of non-compliant scene elements
(2) A slightly slower rate of dream generation than a bean bag, with complex image generation patiently waiting to avoid abortion
(3) When the local redrawing is done, the precision box selects the area to be adjusted, and the hint is added only for the region
(iii) Summary of adaptation scenarios for the two tools
1. Priority bean packs: three view of the person, simple trophies, phrase optimization, quick first draft
2. Prioritization of the dream: main scenes, complex trophies, special effects with an atmosphere, material that needs to raise image quality
3. Co-use: building basic tips and first drafts with bean buns to optimize details and atmosphere and improve the efficiency and quality of maps
IV. Summary of learning and learning
(i) Core focus
At the heart of the map"Phrasing Precision + Tool Fitness", the precision of the split list directly determines the impact of the map, and it is difficult for vague splits to optimize the hint
Bean buns and dreams each have an advantage in combining efficiency and quality, and white does not need a single tool, with a focus on selecting tools and fine-tuning indicators based on the type of material
In addition, the mapping process needs to be refined and refined with patience in order to generate material that meets subsequent production requirements
(ii) Real-time double-hatting
1. Achievements:Generation of three person viewes, major scenes, core trophies and special effects maps for the script, filed by category, consistent style of material, and basic details of the split list
2. Core issues:Imbalance in the proportion of the three-view side view of some of the characters and redundancy in the details of the complex scenario (following the need to optimize the hints and reduce non-core elements)
3. Optimizing direction:Increases in the accumulation of high-quality warning templates, which are tailored to optimize the representation of the number of people and the details of the scenes; exercises, i.e., dream local redrawing, parameter adjustment techniques, to improve the efficiency of material correction; and pre-emptions of life maps to optimize the details of the split list in advance
Additional matters of concern
1. All drawings generated requireUnique style(b) Avoiding the drawings of people for cartoons and scenes, resulting in the subsequent fragmentation of the spectroscopy
2.Backup in timeAll material (local + cloud) to avoid error and renaming
3. Time frameKeeps the hintSubsequent materials of the same kind can be reused and optimized (there is a possible need for a coherent supplement) to improve productivity
4. If there is a disproportionate deviation between the creation of the graphs and the splitting list, then the splitting list is retrofitted, then the hints are optimized to avoid a reversal of the end
The core of the short play is not based on a single tool, but a combination of "precision split" + high-quality hints + tool fit"
For beginners, it is not necessary to pursue complex operations, to develop a splitting list, then to select a pair of tools based on the type of material, and to fine-tune the hints, so you can generate the material that fits the demand
The third view of the person is refined, the scenery is organized in an atmosphere, the trophies are scaled up, the special effects map is re-adapted, and this logic is used to effectively reduce back-to-work and to provide a solid basis for subsequent lens production and video generation。
WITH THE ACCUMULATION OF LIFE-STUDIES AND THE CONTINUOUS OPTIMIZATION OF TIPS AND TOOL-USE TECHNIQUES, THE QUALITY AND EFFICIENCY OF MATERIAL CAN BE IMPROVED OVER TIME, AND THE PRODUCTION OF SHORT PLAYS IN AI CAN BE IMPROVED ~