✳ IskraAI

Storyboard planning with words: how to describe shots, camera angles and transitions

The «Storyboard planning with words» tool transforms your video idea into a textual description of shots: what is visible in the frame, how the camera stands, what happens and how one shot transitions to another. This is not a video generator and not a picture storyboard - you get structured text that can be used immediately for shooting or passing on to a team. In this article we will examine how, by changing the input data, you can get different storyboard variants, compare them and refine the chosen one. We will use a fictional example: a short video about how a ceramic mug appears from clay.

Ouvrir l'outil: Planificateur de storyboard ↗

What exactly the tool does and what data it needs

«Storyboard planning with words» is a text tool. It takes three fields as input: «Your text or idea», «Role or characters» and «Duration or length». The first field is the most important: describe the video idea, key objects and desired mood. The second field helps set whose narration is being told or who is in the frame. The third field limits the duration, so the tool does not stretch the story into ten minutes if you need a 30-second clip.

The tool does not create images, video or sound. It does not generate photorealistic pictures and does not read out text. Its task is to give a coherent description of a shot sequence: angle, action, transition. The result comes in a separate panel, it can be copied or downloaded as a plain text TXT file. There are no style presets, sliders or levels of changes in the interface - you work only with text.

Important to understand the limitations: maximum length of the first field is 8000 characters, details field up to 1000 characters. If you upload an image for inspiration, its size must not exceed 4 MB, but remember: images here are vector compositions, not photorealistic generation. Audio for transcription is processed locally, not sent to AI. This means your voice notes do not go to external servers.

Filling example: input — «Short video about how a master molds a clay mug, from a piece of clay to the finished product. Calm atmosphere, warm light». role — «Master ceramist, close-up of hands». length — «30 seconds».

How to change criteria to get different variants

One and the same plot can be storyboarded differently, if you change the accents in the input data. For example, in the mug example you can focus on the process: then the tool will offer more frames with hands, clay, potter's wheel. If you shift the accent to the result, frames with the finished mug, tea steam, interior will appear. Third option - focus on emotions: close-ups of the master's face, his concentration, a smile at the end.

Change not only text, but also the «Role or characters» field. If you specify «master and his apprentice», frames of interaction, experience transfer will appear. If you specify «hands only», the tool will focus on tactile details. The «Duration» field also affects: for 15 seconds the storyboard will be compact, for 60 seconds - more detailed, with pauses and transitions.

Another technique - change the task formulation. Instead of «show the process» write «show the process through details: water splashes, rotation of the circle, clay texture». The tool does not guess for you, it follows your instructions. The more specific the criteria, the more predictable the result. But do not overload the first field: 8000 characters is a lot, however excess details may confuse.

Variant A (focus on process): «Close-up: hands knead clay. Medium shot: potter's wheel rotates. Close-up: fingers form the walls. Transition: dissolve into frame with finished mug». Variant B (focus on result): «Overall shot: master puts mug on shelf. Close-up: mug with tea, steam rises. Medium shot: master sits to drink tea».

Comparing three variants: what to look for

When you have three text storyboards in hand, compare them by three criteria: completeness, logic of transitions and correspondence to duration. Completeness is how many key moments of the story are covered. In the mug example the process includes preparing clay, molding, drying, firing and finishing. If a variant omits firing, the viewer may not understand how clay turned into a durable object.

Logic of transitions - the second criterion. A good storyboard leads the viewer without jumps. A bad example: close-up of hands, then immediately general shot of the finished mug on the table - unclear how the transition happened. A good example: close-up of hands, then medium shot of potter's wheel, then flow onto the finished mug. Transitions can be direct (cut), smooth (flow) or contrasting (sharp angle change).

The third criterion - duration. If you indicated 30 seconds, and the storyboard has 12 frames at 5 seconds each, that is already 60 seconds. The tool may not indicate exact seconds for each frame, so check if the described sequence fits your time. If not - reduce the number of frames or combine actions.

Comparison: Variant A - 8 frames, covers all stages, but transitions are sharp. Variant B - 6 frames, logical flows, but firing omitted. Variant C - 10 frames, detailed, but does not fit 30 seconds. Choice: take Variant B and add one frame with firing.

Choosing and refining the selected variant

After comparison choose the variant that best suits your task. It is not necessary to take it entirely: you can take the basis from one variant, and good frames - from another. For example, in the mug example basis - Variant B (logical transitions), but we add a frame with firing from Variant A. Thus you get a hybrid, which more accurately reflects your idea.

Refinement can concern not only content, but also formulations. If the storyboard says «camera shows hands», this is too general. Replace with «close-up: hands squeezing clay, fingers leave grooves». Specificity helps the operator or artist understand exactly what to shoot. Also refine angles: «from above», «from the side», «over the shoulder», «macro».

Another step - check for feasibility. If the storyboard has a frame «clay turns into mug with one movement», this may be hard to shoot without editing. Specify: «series of three frames: piece of clay, wheel rotation, finished mug». The tool does not guarantee that all frames are easy to shoot, so final check is on you.

Refined fragment: «Frame 4. Close-up: potter's wheel rotates, master's hands form the mug walls. Transition: flow to frame 5. Frame 5. Medium shot: master puts mug in kiln for firing. Transition: cut».

How to check the result and what limitations exist

Checking the result starts with reading aloud. If you stumble or lose the thread while reading, the storyboard is illogical. Then check if all key objects and actions are mentioned. In the mug example clay, hands, wheel, kiln, finished mug should be present. If something is missing - add manually or regenerate with refined input.

Second stage - check for duration correspondence. Count approximate time for each frame. If the storyboard has 10 frames and you have 30 seconds, average frame - 3 seconds. This is normal for a dynamic video, but for a calm process may be too fast. Reduce the number of frames or increase duration.

Tool limitations: it does not create video, does not read out text, does not generate photorealistic images. It does not automatically save history, does not have batch processing and does not integrate with other services. The model may make mistakes, and queue and availability may change. Always check the text manually. If you need a picture storyboard, this is not the right tool - here only text.

Check: reading aloud - logical. Counting frames: 8 frames for 30 seconds - approximately 3.75 seconds per frame. For a calm video this is acceptable. Checking objects: clay, hands, wheel, kiln, mug - all present.

Practical tips for stable results

To get predictable storyboards, use structure in the first field. Start with one phrase: «Video about how...». Then list key stages via comma. At the end specify desired mood or style: «calm», «dynamic», «close-ups». Such structure helps the tool not go off track.

Do not be afraid to rephrase the task if the first result did not please you. The tool does not remember previous requests, so each time you start with a clean slate. If you want to get another variant, change at least one field: role, duration or accent in description. This will give noticeably different storyboard.

And last: save successful variants in a text file. The tool does not have automatic saving, so copy the result manually. Thus you can return to it later or compare with new generations. Remember, this is only a draft - the final storyboard is always refined by a human.

Tip: if you need 3 variants, make three requests with different accents: «focus on process», «focus on result», «focus on emotions». Save each in a separate TXT file.

«Storyboard planning with words» is a text assistant for planning videos. It does not replace a director or cinematographer, but helps quickly get a shot structure. By changing the input data, you get different variants, compare them for completeness, logic and duration, and then refine the chosen one. The example with a ceramic mug shows how one plot can be storyboarded differently. Remember the limitations: text only, no video and no photorealistic image generation. Check the result manually and adapt it to your tasks.

Ouvrir l'outil: Planificateur de storyboard ↗

Les traductions de l'interface sont assistées par machine et peuvent nécessiter des améliorations.