A written brief can establish the subject, setting, action, and camera direction before any footage exists. Use this page to decide what belongs in the prompt and what you will need to check in the result.
The useful A/B question is what you can supply at the start: a description, a still image, or footage that already exists.
You have a product concept but no approved photograph. Write the setting, subject, action, and mood as separate parts of a brief.
A prompt-first approach lets the team discuss the intended scene without treating an unmade image as an approved reference.
insmind ai video generatorYou already have an approved packshot and want its appearance to remain the visual starting point.
An image-led approach is easier to evaluate when preserving the product's look matters more than inventing a new scene.
insmind image to videoYou have a recorded clip with a soft or noisy appearance, rather than a scene that needs to be invented.
Work on the existing footage instead of asking a prompt to reconstruct its people, timing, and composition.
insmind video enhancerYou need to decide whether a generated clip or a conventional edit better fits a brief with precise timing.
Compare the two workflows before committing to a shot that may need frame-level control.
insmind vs video editorWords communicate intent, not an exact frame. Make the important constraints explicit, then expect to review the interpretation.
Name the subject and what changes during the shot. Replace broad requests such as “make it cinematic” with observable movement, such as steam rising while the camera moves closer.
Specify the setting, lighting, and camera movement independently. A text to video brief can suggest all three, but it cannot supply the fixed composition that a reference image would.
List any critical object details, text, or sequence of actions. A prompt alone does not guarantee that these will remain exact across a generated clip.
Choose the starting material that best protects the part of your idea you cannot afford to lose.
Judge the result against the brief rather than against how polished the first frame looks.
Text to video cannot establish an exact approved layout or product appearance by description alone.
Workaround
If a particular still must guide the shot, begin with that image and compare the result against it.
Words such as “slowly” or “then” do not ensure a precise duration or frame-by-frame sequence.
Workaround
Check the order and pacing of the full clip; use an editor when exact cuts or timing are required.
Small markings, readable text, and repeated object features can differ from the written brief or change during motion.
Workaround
Review the beginning, middle, and end, and do not rely on generated footage as proof of exact product details.
Starting material
Written brief
A description of a scene that does not yet have a fixed appearance
Existing still image
A visible frame with an established subject and composition
Best-defined element
Written brief
Intended action, setting, and mood
Existing still image
The starting look of the subject and scene
Open to interpretation
Written brief
Appearance, layout, and motion
Existing still image
Motion and what appears beyond the still frame
Useful first check
Written brief
Does the clip show the requested subject and action?
Existing still image
Does the clip retain the important parts of the source image?
Weak fit
Written brief
A shot requiring an exact approved frame
Existing still image
A concept without a suitable image to start from
Revision material
Written brief
A clearer description of missing or incorrect details
Existing still image
A better source still or clearer motion direction
Describe one subject, one visible action, and the intended camera movement. Open the destination to explore its current video options, then check any result against your brief; available controls and output terms depend on that destination.
Explore video toolsA written description is the starting material for a prompt-first text to video workflow. Whether the linked destination accepts that input, and which controls it offers, should be checked there before you rely on it for a project.
Name the subject, setting, visible action, and camera movement. Include a few essential appearance details, but avoid burying the main action under a long list of competing instructions.
Do not assume it will. Inspect markings, shape, color, and consistency throughout the clip; if an approved appearance is essential, consider an image-led starting point.
An image-led workflow is a better starting point when you want the still's subject and composition to guide the result. You can still describe the desired motion, but should check how closely the moving scene follows the original image.