Observatories: first-last of the slit closed, then open, on the same dome
Two photos of the same civic dome. Flux 3 interpolates the slit. A T2V night sky invents a different mountain.
Two photos of the same civic dome. Flux 3 interpolates the slit. A T2V night sky invents a different mountain.
Observatory content is a bounded motion problem. You already own both stills: shutter closed across the dome, shutter open on that same hemisphere. The ad is the travel. Do not prompt a model to invent the ridge, the building, or a cousin observatory from a paragraph.
Frame A is closed; frame B is open
Photograph the dome you actually operate — same lens height, same mountain, same light family. Closed slit is the first frame. Open slit on that same hemisphere is the last frame. Composite the open state only if the geometry is still this dome. A different ridge is a morph.
First-last frame takes those two stills and generates the motion between them. Ordinary image-to-video knows where it begins and improvises the rest. First-last has to land. The arc of the shot is decided by you.
The two frames have to be plausibly connectable in the time available. Same dome, same camera, same lighting neighbourhood. A wide of the civic building morphing into a macro of the shutter is a warp, not an open. Keep the hemisphere.
A loop (the same still in both slots) is ambient motion on one plate. This offer needs two states. Closed-to-closed is not an observing night.
Shoot both stills of your dome. A catalog jpeg of a famous telescope on another continent is a different mountain the moment you interpolate.
Flux 3 interpolates the slit
Create a first-and-last-frame transition video is the named agent job: two photos in, one transition out. The agent passes them as first and last frame and picks a first-last-capable model. Cost is priced per model, shown before you confirm. This is not the plain “turn one photo into a video” capability.
Flux 3 First Last Frame to Video is the catalog row for that interpolation: start frame and end frame, native audio, 720p or 1080p, durations from 5 seconds through 20. The listed rate is 9 credits per second. Audio is on — write the mix (shutter travel, a quiet pad) or write silence. A silent brief still gets a soundtrack you did not choose.
Prompt the travel, not the objects. The stills already contain the dome and the ridge. Name the mechanism: hold, then the slit walks open; camera locked; no new mountain, no new building. If the endpoints share geometry, the morph reads as an open. If they do not, you get a mushy crossfade you could have made in an editor without a video bill.
Do not use a last frame that is a different observatory. The model will connect them. The connection is the artefact everyone notices — a dome that changes radius mid-travel, a ridge that slides into another range.
A T2V night sky invents a different mountain
Text-to-video of “iconic observatory under the Milky Way, slit opening, cinematic stars” will decide a cousin dome. The mountain will not be yours. The shutter count will not be yours. 1080p of the wrong ridge is still the wrong ridge.
Image-to-video from one closed still will invent the open. Observatory civic film is the open slit on this hemisphere. Pin the open still as the last frame or you are not selling the travel.
The AI video generator is the door for ordinary T2V. Leave it closed for this job. First-last is the row. Two photos. Same dome.
Native audio is on Flux 3 first-last. Write “no speech” if you do not want a voiceover explaining the universe the stills already show. A generated astronomer pointing at a sky the model painted is a second product glued onto the interpolation. Overlay show times after lock, on type you typed. Do not let the interpolate invent a schedule on the dome.
What this is not
It is not a talking host. It is not a planetarium architecture plate dressed as a night of observing. It is not a single-image wiggle that leaves the slit sitting closed for eight seconds of fake starfield.
Do not first-last a closed dome into a wide of a different civic building. Do not use a last frame from a tourist postcard. Do not prompt galaxies into the gap — the stills already contain whatever sky was actually there.
If you needed a line about public nights or a ticket URL, that is overlay after picture lock. Offers change. The dome should not.
Two photos of the same civic dome. Flux 3 interpolates the slit. Stop.
FAQ
Can I animate one photo of the closed dome and hope the slit opens?
You can. The ending is then the model’s guess. Observatory film is the open slit on that hemisphere. Pin the open still as the last frame or you are not selling the travel.
Why Flux 3 instead of a talking-head model?
Because the product is interpolation between two approved stills, not a mouth. Flux 3 first-last is 5–20 seconds, 720p/1080p, native audio, 9 credits per second. A presenter row will talk over a mountain it invented.
Do the two photos have to be shot on the same night?
They have to share geometry. Same dome, same lens family, same lighting neighbourhood. A dusk still into a midnight still is a time-lapse you should prompt as one. A different mountain is a morph.
Is this the same as uploading one image to image-to-video?
No. One image is open-ended motion. Two images is a bounded landing. The agent notes that this capability needs first frame and last frame. Missing the open slit is a different job. A T2V night sky invents a different mountain.