Notebooks: cover still I2V after type is locked
Type lock then motion.
Type lock then motion.
A notebook cover is a typesetting job that happens to become a clip. The title, the ruled prompt, the series name have to be the strings on the SKU. Motion is allowed after that still is honest. Motion first is how you ship a pretty six-second lie with a misspelled series on the board.
Ideogram V4 is the still row: text-to-image, posters and logos, accurate text rendering, 6 credits flat per call, 1:1 / 16:9 / 9:16 / 4:3 / 3:4. Lock the cover there. Then Seedance 2.5 Image to Video animates that still: requires the image, up to 720p, four to thirty seconds, native audio with the picture, optional last frame when the move has a destination.
Type is a still job
Notebook brands die on the title. "Gratitude Journal — 90 Days" is a string merch already signed. A video model asked to "show a premium linen notebook with gold foil gratitude" will invent foil, invent a serif, and invent a word. That is not a cover. That is a cousin of the SKU.
Ideogram V4's job on this brief is the cover as a poster. Prompt the type you actually print. Check the still at 100%. If a letter is wrong, regenerate the still. Do not send a bad letter into image-to-video and hope motion hides it. Motion holds the misspelling for thirty seconds.
Aspect is a cover decision. 1:1 for a pack shot, 9:16 for a feed hold, 4:3 if the listing photography is 4:3. The still sets the clip. Six credits is the call, not a per-second cinematic meter. Spend it until the type is the type. Then stop.
Motion is the second object
Once the cover is locked, the generate is only camera and light. Seedance 2.5 I2V takes that still as frame one. Prompt a slow push, a tilt that reveals the elastic, a hold on the foil. Leave the title out of the motion prompt. Re-describing the series name is how the model treats the lettering as optional and paints a new word.
You can add a second image as a closing frame when the shot has a destination — front cover to back, closed to open on a blank page you already designed. A last frame is control, not a default. A single hero still with a camera move is the usual job.
Thirty seconds in one pass is the production reason this family holds a cover: stitching four short clips around a title is how the lettering drifts. 720p is the ceiling on this endpoint. A 2K grade plate is a different model. Identity of the type is the I2V choice.
Seedance I2V is still the still-to-motion job is the router. This page is only the notebook cover.
Do not animate a misspelled cover
The expensive week is generating motion while the title is still a draft. Copy changes. The SKU name gets a hyphen. Every one of those is a new still, then a new I2V. Picture lock on the cover is cheaper than video lock on a wrong word.
Do not ask Seedance to typeset. Scene models sometimes paint glyphs you cannot correct without regenerating the cover. Generate a clean cover. Animate that. Overlay a CTA after lock if the offer is still moving. Native audio on the I2V row is a room, not a voiceover of the blurb.
FAQ
Can I skip Ideogram and prompt the cover inside Seedance?
You can. You should not ship it. Seedance 2.5 I2V requires an image because the still is the contract. Text-to-video a "linen gratitude journal" is how the title becomes a different product.
Do I need a last frame of the open notebook?
Only when you have a second approved still of the inside and the shot is meant to travel there. A single locked cover with a slow push is the usual ad.
Why Ideogram V4 instead of any text-to-image row?
This brief is type. V4's catalog job is high-quality images, posters, and logos with accurate text rendering. A cover is a poster that happens to wrap board. Use the row that is built to spell.
Is 720p enough for a product page loop?
For feed and most PDP clips, yes. For a 2K plate you intend to grade, pick a 2K generate and still start from the locked cover. Resolution is delivery. Type is the still.