Guides

    Slow cookers: first-last lid on to lid off, insert unchanged

    Two stills of the same pot. Interpolate. Do not let the ceramic color drift.

    Versely Team6 min read

    Two stills of the same pot. Interpolate. Do not let the ceramic color drift.

    Slow-cooker video is a state change on a SKU you already printed: lid on, then lid off, insert still the insert. The glaze, the knob, the stoneware colour — those sit in photographs you already approved. A text-to-video paragraph will mint a cousin crock. A single still will wiggle the lid and never land on the food. Two photographs of the same pot, connected, is the ad.

    Frame A is lid on; frame B is lid off

    Photograph the unit you would ship: square-on or three-quarter, insert seated, glaze true, handles in frame. Lid on. That file is the first frame. It is also the contract. If the ceramic is the wrong red in the still, every second of motion is a false colourway.

    Frame B is the same pot, same insert, lid off — stew, roast, or empty stoneware, shot in the same lighting family. Same camera height. Same counter language if both stills share a surface. The food is the destination of the lift, not a second brand.

    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: lid on, then lid off.

    The two frames have to be connectable in the time available. A macro of the knob morphing into a wide kitchen is a warp, not a lift. A white insert into a different oval is a new SKU. Keep the pot. Keep the glaze. Let the lid travel.

    Do not use a last frame from a different colourway. The model will connect them. The connection is the artefact everyone notices — a cream insert that becomes terracotta, a knob that never existed.

    Interpolate on Flux 3. Do not regenerate the glaze.

    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 a 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 (a quiet lid lift, a counter settle) 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 lid, the insert, and the food. Name the mechanism: hold, then the lid leaves frame; camera locked; no new kitchen; ceramic stays the ceramic. If the endpoints share geometry, the morph reads as a lift. If they do not, you get a mushy crossfade you could have made in an editor without a video bill.

    Duration should match the lift, not a Reels quota. Five or six seconds between two honest stills reads as proof. Twenty seconds invents steam, a second pot, and a glaze in the middle. Inspect the midpoint. If the middle restains the insert, reroll the interpolation. Do not regenerate both stills "to match the motion." The stills are the spec. A long identity hold of the closed cooker never lands on the insert. Lid-on to lid-off needs both stills.

    The insert cannot change colour

    Ceramic colour is a SKU. Stoneware that ships oat cannot finish paprika. If frame A is the oat insert, frame B is the oat insert with the lid gone — not a prettier cousin glaze the model prefers. Pause at the midpoint. If the well has drifted, the interpolation failed even if both endpoints look approved.

    Do not prompt "golden stew, rustic ceramic, farmhouse" on top of the stills. Adjectives argue with pixels. The stills already contain the glaze. Re-describing the pot is how Flux starts drawing a second pot.

    Keep the insert large in frame. Fine speckle is the first thing a generator loses in the middle of a morph. If the interior colour must read at the open, it has to fill the frame at the open. Panel icons and "LOW / HIGH / WARM" are overlay after lock if they must appear, not glyphs Flux draws on a control that was never in the still. Rice-cooker brands have a different failure — a panel the model restypes. This job is lid travel on one insert.

    Nearby kitchens are not this SKU

    Restaurants sell the plated dish as a menu item. Frame B can look like a portion. Frame A cannot become a house crock the kitchen restyped. Slow-cooker CPG is the glaze on the insert you already photographed. Do not twin the venue page as this listing.

    Do not generate a talking chef lifting the lid. If you needed a founder line about the recipe, that is a different generate, cut after the open pot. The lift shot does not have to talk. Ordinary image-to-video from one closed-lid photo will steam and never land. The insert stays. Only the lid leaves.

    The test: pause mid-lift. Is the ceramic still the ceramic? If the well has warmed or cooled a shade, throw the file away. Interpolate again. Do not "fix" the glaze in a new pair of stills unless the original pair was the lie.

    FAQ

    Can I animate one photo of the closed cooker and hope the lid comes off?

    You can. The ending is then the model's guess. The open insert is the offer. Pin lid-off as the last frame or you are not selling the pot.

    Why Flux 3 instead of a talking-head recipe story?

    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 cooker it invented, and it will mint a glaze legal cannot ship.

    Do the two photos have to be shot on the same day?

    They have to share geometry. Same insert, same lighting neighbourhood, same lens family. A studio lid into a night dinner table is a warp. A different colourway is a different SKU.

    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 insert is a different job.