✳ IskraAI

Storyboard in words: how to describe a video idea for the storyboard planner

The «Storyboard in words» tool transforms your video idea into a textual description of frame sequence: camera angles, actions and transitions. To make the result useful, it is important to correctly fill in three fields: «Your text or idea», «Role or characters» and «Duration or duration». In this article we will explain how to compose an exact assignment, which details are important and how to evaluate the resulting storyboard.

Open tool: Storyboard planner ↗

What the tool does and its limitations

«Storyboard in words» is a text planner. You describe a video idea, specify the role or characters, and also the desired duration. The tool returns a textual description of frames: what is visible in the frame, how the camera is positioned, what actions occur and what transitions are used between scenes. This is not video generation and does not create images. The tool does not draw photographs and does not create audio. It only helps to structure the intent in a list of scenes.

Important to understand the limitations. The tool does not guarantee that the described frames will be realistic or easy to execute. It does not check the chronometry and does not synchronize actions with sound. The result may contain errors, and the queue and availability may change. Input is limited: the first field - up to 8000 characters, details field - up to 1000 characters. If you use other functions, such as dubbing or subtitles, they have their own limits, but within this task we consider only textual storyboard.

The tool is not intended for creating photorealistic images or video. It does not send audio to AI if you use local transcription. All images that may be generated in other modes are vector compositions, not photographs. Therefore, do not expect «Storyboard in words» to have a visual row - only a textual description.

Example of a poor assignment: «Make a video about a cup». Too general, no roles, no duration. The tool may output an incoherent set of frames.

Field «Your text or idea»: what to write

This is the main field for describing your idea. The more specifically you describe what should happen, the more accurate the storyboard will be. Indicate the topic, key events, setting and general atmosphere. For example, if you want a video about making a ceramic cup, write: «Video shows how a master creates a cup from a piece of clay with their hands: kneading, forming on a potter's wheel, firing, painting and the finished product». Such text gives the tool a clear sequence.

Avoid abstract phrases like «make it beautiful» or «make it interesting». The tool cannot guess your preferences. Instead, describe exactly which actions should be in the frame. If you care about style, indicate it in words: «calm pace», «close-ups of hands», «transitions via darkening». But remember: the tool has no style presets and does not adjust change levels. It just interprets the text.

You can set a structure: beginning, middle, end. For example: «Beginning: empty workshop, close-up of clay. Middle: hands form the cup on the wheel, water splashes. End: finished cup on a shelf, soft light». This will help the tool break the video into logical scenes. If you do not specify a structure, it may propose its own, but it may not match your intent.

Good assignment: «Short video about a master molding a clay cup. Show stages: kneading clay, centering on the potter's wheel, pulling up walls, forming the handle, firing in the kiln, painting and the finished product. Calm pace, focus on hands and material texture.»

Field «Role or characters»: who we show

In this field list the acting persons. These can be people, animals or even items if they play an active role. For example: «potter, his hands, clay cup». If the video has no characters, write «no» or «none». This will help the tool not invent extra heroes. If you want a character to have a certain image, describe it briefly: «elderly man in an apron», «young woman in sportswear».

Do not overload the field. Enough 2-5 characters or roles. If there are too many, the storyboard may become messy. If there is one character, indicate this. If the video is from the first person, write «first-person operator» or «off-screen narrator». This will affect how the tool describes camera angles.

Remember, the tool does not create dialogues and does not script lines. It only describes who is in the frame and what they are doing. If you need lines, you will have to add them separately. Within this task we focus on visual storyboard.

Example: «Roles: potter (man 50 years, apron), his hands, clay cup. No other characters.»

Field «Duration or duration»: how to indicate time

Indicate the desired video duration. This can be approximate time: «30 seconds», «2 minutes», «1 hour». The tool does not create video, but can use this information to suggest the number of scenes and their approximate duration. The more precisely you specify, the better. If you do not know the exact duration, write a range: «from 1 to 2 minutes».

Do not specify too small a duration for a complex idea. For example, if you want to show the entire process of making a cup in 10 seconds, the tool may propose too many frames or, conversely, too few. Better choose a realistic chronometry. If the video is short, describe only key moments. If long, you can split into chapters.

Remember, the tool does not synchronize frames with sound or music. It only gives a textual description. Duration is a guide for structure, not exact chronometry. You can always manually adjust the storyboard.

Example: «Duration: 1 minute 30 seconds. Divide into 6-8 scenes.»

Example of poor and good assignment

A poor assignment is often too general or contradictory. For example: «Make a video about a cup, make it cool and everything. Characters: all. Duration: irrelevant». The tool will not understand what exactly you want. It may output a random set of frames that does not match your intent. Also bad if you write a huge text without structure, with many details but without indicating sequence.

A good assignment contains specificity. Here is an example for our case: «Idea: short video about the process of making a ceramic cup from a piece of clay to the finished product. Show stages: kneading clay, forming on a potter's wheel, creating the handle, firing, painting, final view. Calm pace, focus on hands and material texture. Roles: potter, hands, cup. Duration: 1 minute 30 seconds. Divide into 6-8 scenes.»

Compare: in the poor variant there are no roles, no duration, no structure. In the good - everything is indicated. The tool will be able to propose a logical sequence: close-up of clay, medium shot of the master at the wheel, hand details, transition to kiln, finished product. This does not guarantee an ideal result, but increases the chances of a useful storyboard.

Poor: «Video about a cup, cool, all done». Good: «Video about making a cup from clay. Stages: kneading, forming, firing, painting. Roles: master, hands, cup. Duration: 1 min 30 sec. 6-8 scenes.»

How to check the result and what to do with errors

After pressing the «Generate» button you will get a textual description of the storyboard in a separate panel. You can copy it or download it in TXT format. Check if the storyboard matches your idea. Pay attention to the sequence of scenes: do they change logically? Are camera angles and transitions indicated? If something is missing, you can return to the form and clarify the assignment.

The tool may make errors. For example, invent extra characters or miss an important step. This is normal, as the model is not perfect. If you notice an error, try rephrasing the assignment: add more details in the «Your text or idea» field or clarify the roles. Do not expect the first time to get an ideal storyboard. It may take several attempts.

Remember that the result is not a final script. It is a draft that can be edited manually. You can add, delete or change frames at your discretion. The tool does not save history and has no settings panel, so save the text immediately after generation.

Example output: «Scene 1: close-up of a piece of clay on the table. Camera slowly approaches. Action: master's hands take the clay. Transition: dissolution. Scene 2: medium shot of the master at the potter's wheel. Camera from the side. Action: master centers the clay. Transition: overflow.»

Tool limitations and typical errors

The tool does not generate video, audio or photorealistic images. It does not create dubbing and does not work with subtitles within this task. If you try to use it for creating a ready video, you will be disappointed. This is a planner, not a production studio. Also it does not support export to DOCX or PDF, only TXT. No batch processing, automatic saving or integrations.

Typical error - expecting the tool to come up with creative transitions or non-standard camera angles itself. It follows your description. If you do not specify transitions, it may use standard ones (for example, «cut» or «darkening»). If you want something special, write it in the assignment. But remember, the tool does not guarantee that the suggested transitions will be technically executable.

Another error - overloading the input with details unrelated to visual storyboard. For example, do not write dialogues or describe sound accompaniment. Focus on what is visible in the frame. If you indicate too much text, the tool may get confused. Better split the assignment into several parts if the video is long.

Error: «I want the video to have music and off-screen voice». The tool does not create audio. Correct: «Video without sound, only visual row».

«Storyboard in words» is a useful tool for quickly creating a textual video plan. To get a good result, fill in the fields specifically: describe the idea, list the roles and indicate the duration. Avoid generic phrases and do not expect the tool to do all the work for you. Check the output, adjust the assignment if necessary and remember the limitations: this is only text, without video and audio generation. Use it as a draft for further work.

Open tool: Storyboard planner ↗

Interface translations are machine-assisted and may need refinement.