Video by scenes: how to create a script with timing and storyboard
The "Video by scenes" tool turns an idea into a structured script: a sequence of scenes with timing, narrator text and visual descriptions. It does not shoot video or narrate the text, but creates a textual basis that you refine for your task. In this article we will look at how to fill in the form fields, how to formulate requests for different types of videos and how to check the result so that it is suitable for shooting or storyboard.
Open tool: Video script writer ↗How the form is set up and what to provide as input
The form contains three fields: "Your text or idea", "Role or characters" and "Duration or length". In the first field insert a draft, theses or a brief concept outline - up to 8000 characters. The second clarifies whose voice the narration is from and who appears in the frame. The third sets a time reference: "60 seconds", "2 minutes", "90 seconds".
The tool returns only text: a list of scenes with timing, narrator replays and visual row descriptions. Video, images and sound it does not create. The result can be copied or downloaded as TXT. This is important to bear in mind so as not to expect a ready-made video.
The more specific the initial data, the more useful the draft. Specify the topic, audience, key messages and tone. If there are mandatory phrases or facts, add them to the idea field. Avoid contradictions: for example, do not ask for "30 seconds" and "5 minutes" at the same time.
Example request: "Make a script for 2 minutes introducing the district repair cafe. Role: friendly host. Duration: 2 minutes. Include three scenes: entrance, work zone, thank guests".
Example 1: short promo for a local project
Suppose you are promoting a district repair cafe where volunteers help fix household items. A 2-minute video is needed for social media. In the idea field describe the atmosphere, key zones and call to action. In the role field specify "friendly host, off-screen voice". In the duration field - "2 minutes".
The tool will return a scene-by-scene script. For example: scene 1 (0:00-0:20) - general view of the entrance, host welcomes viewers and explains what a repair cafe is. Scene 2 (0:20-1:10) - close-ups of work tables, host tells what items can be repaired. Scene 3 (1:10-1:50) - volunteers and guests interacting, host emphasizes the value of community. Scene 4 (1:50-2:00) - call to visit the website or group.
Check that the timing fits within 2 minutes when narrated. Usually a narrator reads about 130-150 words per minute. If there is more text, shorten the replays or combine scenes. Make sure frame descriptions are realistic for your filming group.
Result fragment: "Scene 2 (0:20-1:10). Frame: close-up of hands screwing a screw. Narrator: Here we repair what is precious to throw away: lamps, chairs, toys".
Example 2: training video for employees
The second script is internal training on warehouse safety procedures. Idea: show three rules for moving boxes. Role: "instructor, calm and clear voice". Duration: "90 seconds". In the idea field list the rules and typical errors.
The tool will propose a structure: scene 1 (0:00-0:15) - introduction and goal; scene 2 (0:15-0:45) - first rule, lifting a box; scene 3 (0:45-1:15) - second rule, moving; scene 4 (1:15-1:30) - third rule and brief summary. For each scene there will be narrator text and frame description.
Such a draft is convenient to use as a basis for a storyboard. Check that there are no ambiguous formulations in the text that could be misinterpreted. If your company has approved instructions, check each rule against them. The tool does not replace an official document, it only helps prepare a script.
Example request: "Script for 90 seconds: three rules for safe box lifting on a warehouse. Role: instructor. Duration: 90 seconds. Add a warning about a typical mistake - leaning back".
Example 3: plot for social advertising
The third example is a social video about separate waste collection. Idea: show the path of a bottle from home to recycling. Role: "storyteller, warm and inspiring tone". Duration: "60 seconds". In the idea field describe key points: kitchen, container, truck, factory.
The result may look like this: scene 1 (0:00-0:10) - bottle on table, narrator asks a question; scene 2 (0:10-0:30) - bottle ends up in a container, narrator explains the benefit; scene 3 (0:30-0:50) - recycling, narrator talks about new things; scene 4 (0:50-1:00) - call to sort waste. Frame descriptions will help the operator understand what to shoot.
Check that the tone matches the declared audience. For social advertising it is important to avoid pressure and accusations. If the result turns out too preachy, rephrase the request, adding "without preaching, with respect for the viewer". Remember that this is only a draft, not a final text.
Fragment: "Scene 3 (0:30-0:50). Frame: conveyor, bottles turn into granules. Narrator: From an old bottle a new jacket or bench can be made".
How to check and edit the result
After generation read the script in full. Check logic: are any important steps missed, are there repetitions? Check timing: read the text aloud with a stopwatch. If the script is longer than the indicated time, shorten the replays or divide it into two parts.
Evaluate frame descriptions. They should be executable: do not require complex special effects if you have no budget. If necessary replace descriptions with simpler ones. Make sure the narrator text matches the role: for children - simple words, for specialists - exact terms.
The tool may make errors: inaccurate facts, poor formulations, timing violations. Always check facts and adapt the text to your tasks. Do not use the script without editing, especially if it contains instructions or important data.
Check: "Read aloud scene 2. If it takes more than 30 seconds, remove one sentence or speed up the pace".
Tool limitations and typical errors
The tool does not create video, images or sound. It does not generate photorealistic pictures and does not narrate the text. Speech can be synthesized separately using eSpeak to WAV, but this is not part of this tool. Vector images are available only in SVG and PNG, but they are not photographs.
Among the limitations: no export to DOCX or PDF, no batch processing, no automatic saving, no integrations and publication. Cannot upload documents, use OCR or voice cloning. Maximum length of the first field - 8000 characters, details fields - 1000 characters, speech - 3000 characters, images - up to 4 MB, browser audio - up to 20 MB or 5 minutes.
A typical error - expecting a ready video. Another - too general a request, for example "make a video". The more specific the idea, role and duration, the more useful the draft. Also do not insert contradictory requirements into the request: for example, "simultaneously 30 seconds and 5 minutes".
Incorrect request: "Make a cool video". Correct: "Script for 60 seconds about the benefits of morning exercise for Instagram. Role: trainer. Duration: 60 seconds".
The "Video by scenes" tool is an assistant for creating a textual basis of a video. It saves time on structuring ideas, but requires your review and editing. Use specific requests, indicate role and duration, check timing and facts. Remember the limitations: the tool does not generate video, sound or photorealistic images. This is only a draft that becomes the basis for your creative work.
Open tool: Video script writer ↗Interface translations are machine-assisted and may need refinement.