Storyboard in words: how to describe video frame by frame
The 'Storyboard in words' tool transforms a video idea into a textual description of frames, camera angles, actions and transitions. It does not generate video, images or sound, but creates structured text that can be used as a basis for shooting, animation or editing. To make the result useful, it is important to fill in the fields correctly and understand the tool's limitations. This article explains how to work with the form, what data to enter and how to check the resulting storyboard.
Open tool: Storyboard planner ↗How the form is structured and which fields to fill in
The tool form includes three main fields: 'Your text or idea', 'Role or characters' and 'Duration or length'. The first field accepts up to 8000 characters, the details field - up to 1000 characters. This is enough to describe a scene, but not the entire script altogether. If the idea is large, break it into several parts and create separate storyboards for each.
The 'Role or characters' field sets whose narration is being told and who acts in the frame. Without it, the tool may create impersonal descriptions. Specify the main heroes and their functions: for example, 'head chef', 'customer and seller', 'pottery master'. This will help the tool describe angles and actions oriented towards characters.
The 'Duration or length' field affects the number of frames. For a clip up to 30 seconds, usually 3-5 frames are enough, for a minute-long video - 6-10. If duration is not specified, the result may be either too short or overloaded. Indicate approximate time so the tool maintains a balance between detail and volume.
Example of filling in: idea - 'Video about making cappuccino in a small coffee shop'; role - 'Barista, 25 years old, works behind the counter'; duration - '30 seconds'.
How to formulate an idea to get specific frames
General formulations like 'video about nature' or 'coffee advertisement' lead to abstract frames: 'landscape', 'cup', 'close-up'. Such a storyboard is useless for shooting. Add specificity: object, action, environment and goal. Instead of 'coffee ad' write: 'A short video about how a barista makes a cappuccino, starting with grinding beans and ending with serving the cup'. The more detailed the idea, the more accurate the frames.
Use active verbs and indicate what is happening in the frame. For example: 'Barista takes portafilter, tamps coffee, turns on espresso machine'. The tool better understands sequences of actions than static descriptions. If you need to show an emotion or atmosphere, add adjectives: 'warm light, cozy atmosphere, steam over cups'.
Check if it is possible to understand what to shoot from the description. If frames repeat or are too abstract, clarify the input. Remember, the tool does not generate video and does not create images - only text.
Input: 'video about nature'. Result: 'Frame 1: forest. Frame 2: river. Frame 3: sky'. Corrected input: 'Video of a morning run in the park: runner, trees, dawn, dew drops'. Result: 'Frame 1: general view of runner on path, camera follows him. Frame 2: close-up of sneakers, slow-motion shooting. Frame 3: panorama of dawn over pond'.
Role of characters: how to avoid impersonal frames
If you skip the 'Role or characters' field, the tool may create frames without acting characters or confuse roles. For example, for a video about cooking it is unclear who is cooking: a chef, a child or a robot. Always fill in this field, specifying the main acting characters and their functions.
For educational video specify: 'Host - experienced chef, shows the process'. For advertising: 'Girl-buyer and seller'. This will help the tool describe actions and angles oriented towards characters. As a result, frames will appear where characters perform actions, not just objects.
If the video has multiple roles, list them briefly. For example: 'Pottery master, 40 years old, works in workshop; his apprentice, 20 years old, observes'. The tool will take both into account and offer frames with their participation.
Input: idea 'video about pottery', role not specified. Result: 'Frame 1: clay on table. Frame 2: hands shape it'. With role 'Pottery master, 40 years old, works in workshop': 'Frame 1: medium shot of master, he sits at pottery wheel. Frame 2: close-up of his hands forming a jug. Frame 3: general view of workshop, master smiles at finished work'.
Duration and number of frames: how to find balance
The 'Duration or length' field directly affects the number of frames and their detail. If you specify '1 minute', the tool may create 5-7 frames. If '10 seconds' - 2-3 frames. Without specifying duration the result may be either too short or overloaded.
For short clips (up to 30 seconds) enough 3-5 frames. For a minute-long video - 6-10 frames. This will help the tool maintain a balance between detail and volume. If there are too many frames for the stated duration, reduce the input or reduce duration. If too few - increase it.
Remember, the tool does not do editing and does not calculate exact timing. It only offers a textual description that you can adapt to real shooting conditions.
Input: idea 'video about baking bread', duration not specified. Result: 12 frames, including kneading, rising, baking, slicing. For a 30-second video this is excessive. With specification '30 seconds' the tool will output 4 frames: 'Frame 1: general view of kitchen, baker kneads dough. Frame 2: close-up of dough rising in form. Frame 3: oven, bread bakes. Frame 4: slicing finished bread, close-up'.
Tool limitations and realistic expectations
'Storyboard in words' works only with text: it describes frames in words. The tool does not generate video, photorealistic images or sound. It does not do editing and does not publish the result. Its output - a text document that can be copied or downloaded in TXT format.
Some users expect a ready-made video or visualization. This is not the case. Visualization - task of other programs. The tool only helps to structure the idea and describe the sequence of frames. If you get a text with frame descriptions, that is the correct result.
Consider volume limitations: first field - up to 8000 characters, details field - up to 1000 characters. The model may make errors, and queue and availability may change. Always check logic and manually edit text if necessary.
User enters idea and expects to see a ready-made video. Instead receives text: 'Frame 1: close-up of hands shaping a cup. Frame 2: medium shot of potter at work'. This is the correct result - a basis for shooting.
How to check and improve the resulting storyboard
After generation carefully read the text. Check if it contains angle descriptions (close-up, general view), actions (what characters do) and transitions (smooth transition, abrupt change). If something is missing, add this to the original data and generate again.
Use the 'Details' field for clarification. For example, if a specific transition is needed, write: 'Transition between frames 2 and 3 - dissolve'. The tool will take this into account. You can also specify desired angles or emotions.
Remember that the result may contain errors. The tool does not guarantee perfect storyboard. Always check logic and manually edit text if necessary. If frames contradict each other or miss important moments, simplify the input.
After generation you see: 'Frame 1: general view. Frame 2: close-up'. Missing actions. Add to input: 'Show how potter takes clay and starts shaping it'. Result will become more specific.
The 'Storyboard in words' tool is an assistant for creating a textual description of a video. To get a useful result, avoid generic formulations, fill in all fields, specify duration and do not expect it to produce a ready-made video. Check the output and edit the input if necessary. By following these tips, you can effectively use the tool for shooting planning.
Open tool: Storyboard planner ↗Interface translations are machine-assisted and may need refinement.