Nature centers: exhibit stills; museums already own the gallery page
Lock the diorama or trailhead kiosk you actually built. PixVerse I2V after identity. Gallery identity is /for/museums. Do not generate school groups.
Lock the diorama or trailhead kiosk you actually built. PixVerse I2V after identity. Gallery identity is /for/museums. Do not generate school groups.
A nature-center reel that books a visit is one exhibit you maintain: this marsh diorama, this trailhead kiosk, this bird-skin case the interpreter already stands beside. It is not a museum gallery page wearing a boardwalk. It is not a generated class on a field trip. Photograph the exhibit. Animate that photograph. Leave invented children off the path.
The exhibit you built is the plate
The visitor already knows what a nature center looks like. What they need is this room: the painted marsh, the numbered stop on the trail, the kiosk map that matches the loop they will walk. That is a still you already own if you stood in the hall with a phone after hours.
A paragraph that says "cinematic nature center, children discovering wildlife, golden hour boardwalk" is not a plate. It is a cousin building and a cast you did not clear. The model will invent a second diorama, a third taxidermy mount, a lake the site plan does not have. The volunteer who built the marsh will notice before the caption does.
Keep other rooms out of the prompt. One exhibit per generate. Photograph the real elevation. Approve that jpeg: panel colour, the map that is actually printed, the mount that is actually in the case. Then animate.
Empty is honest. Occupied is a likeness problem. If the only photo includes a person, crop them or shoot again.
AI image to video is the launcher: the upload is frame one, not a style hint. Prompt only the motion. Do not prompt "school group enjoying the new wetland at golden hour." That restages the exhibit as a campaign still you do not have.
PixVerse I2V after identity
Pixverse 5.6 Image to Video requires an image. Durations are 5s, 8s, and 10s. Max output 1080p. Credits 75 on the 1080p step. Catalog audio is true. There is no aspect-ratio field — the still's crop is the frame. 9:16 in the prompt does nothing. Crop the still to the delivery ratio, approve the diorama, then describe only what moves: a light graze across the painted marsh, a slow push on the kiosk map, dust in a sunbeam that was already in the hall.
A text-to-video prompt on this row is a mislabel. You will pay 75 credits to animate a picture you never locked, or the call will fail for missing image_url. Pick the second count that matches the gesture. A 10-second 1080p pass of "families, cinematic, epic" is how 75 credits buy a crowd you did not grant.
Prompt the delta: what moves, how far, what must not change. Negative prompt is a real field. Put "extra people, school group, morphing animals, new background" there.
Audio is on. Write the sound that belongs to the plate — a quiet gallery, no speech, no music. Do not use this row as a talking-head inventor. If a named interpreter must speak, film that person.
Add a text overlay burns one authored line after picture lock: hours, a trail name, "closed Mondays." If the map in the still is already wrong, every second of motion is a longer wrong map. Overlay the hours. Do not ask PixVerse to letter them in flight.
Museums already own the gallery page
/for/museums is one object, one story: collection objects, labelled reconstructions, curator-led explainers, curatorial review of any generated historical depiction. That page can reconstruct, labelled. A nature center is not a reconstruction brief. It is the diorama as it sits and the kiosk as it was built.
Do not twin the museum playbook as this listing. Do not generate a photoreal 1890 wetland and caption it as the marsh hall. If you needed a reconstruction, you are on the museum page, and you still label it on screen.
Gallery identity stays on /for/museums. Exhibit identity stays on the still you photographed in this hall. If the feed wants a human for trust, hire a real named interpreter. Generation stays on the exhibit.
Do not generate school groups
City accounts: the building still is public; residents are not extras is the civic version of the same rule. Nature-center "visitors" are private people wearing an education brief. Minors on a field trip are a hard stop.
TikTok's AI-generated content help forbids adult private figures used without permission, and anyone under 18 is off the table. A labelled synthetic clip of an unnamed class or a child pointing at a mount is already over that line. Calling it a field-trip aesthetic does not create a grant.
What is allowed: the empty hall, the empty trailhead, the diorama with no visitors in frame. What is not: "we used the class only as a style reference." Style reference of a school photo is how you get their heads. Do not "age them up" from a recap to populate the marsh.
Would a regular say "that is our kiosk"? If they would also say "who are those kids," you generated extras. A nature center is an exhibit with a camera. It is not a class photo.
FAQ
Can I skip the exhibit photo and prompt PixVerse for "our nature center"?
No. Pixverse 5.6 Image to Video requires an image. 5s / 8s / 10s, 1080p, 75 credits at the 1080p step, native audio on. A paragraph invents a hall you do not run. Photograph the diorama or kiosk, crop, then animate from image to video.
If we label the video AI-generated, can we add a photoreal school group?
Not as a general rule, and not on TikTok if those faces are private people. Minors are a hard stop. Labelling tells viewers the media is synthetic. It does not create a grant. Nature centers should not be in the replica business.
Is /for/museums the right playbook for a nature-center reel?
It is the wrong playbook for identity. /for/museums is collection objects, labelled reconstruction, curator review. This page is the exhibit you built. Link the museum page so you do not accidentally run a gallery reconstruction on a trailhead kiosk.
Does PixVerse letter trail hours if I put them in the prompt?
Do not count on it. Burn hours and trail names with text overlay after picture lock. PixVerse will animate whatever type is already sharp in the still, and it will happily misspell a new line you asked it to invent.