Magnet pickup: lift the steel washer and leave the brass one

Original video brief; method not independently tested. No input image, physical test or generated result is supplied.

1. Task and source boundary

Show a selective pickup with two identifiable parts. One magnet passes from the brass washer toward the steel washer, then lifts with only the steel part attached. The useful ending preserves both originals in different support states: steel on the magnet and brass on the tray.

Goudsmit Magnetics explains that many steels respond strongly to a magnet, while brass is not visibly attracted by an ordinary magnet. It also notes that capture depends on factors including composition, shape, distance, field and adhesion, and that a magnet test is not complete material identification. The relevant explanation and tables were read directly on 23 October 2026 (China Standard Time). These facts do not establish the behavior of your particular parts, magnet or generated clip. Manufacturer materials explanation

Use one small handheld magnet and two separate washers of similar visible outer size: one known ferromagnetic steel part and one known brass part. The pass, capture and short lift below are an authored shot plan. First confirm the real parts and chosen magnet are suitable for it; exclude plated lookalikes, steel-backed brass, adhesive, linked parts, loose filings, electromagnets and release mechanisms. Changing washer color or magnet branding alone would not create this task.

2. Prepare a readable input

Prepare a photograph or accurate drawing you own or have permission to use. Show a fixed nonmagnetic tray, the brass washer on the left and the steel washer on the right, separated with neither touching the other. Both begin flat on the tray with their holes and individual edge details readable. Include one original hand holding the magnet above the brass washer with a visible gap, and enough room for a short level pass to the steel washer followed by an upward lift. The hand touches only the magnet. No manufacturer image or input file is supplied.

Confirm each part's material separately from its appearance, and check with your chosen magnet that the steel part can be picked up at the planned working gap while the brass part does not follow under the intended clean conditions. This preparation has not been performed here. Color is an identity cue, not a material test. Keep a low oblique view that shows both tray contacts, the approaching magnet face, the steel part's departure and its attachment during the lift. A broad head or hand can hide those events; revise the input if needed. Do not set a universal pickup distance from this brief or use a hidden backing, thread or adhesive to create the planned result.

3. Open the current generator

Use the current MiniMax H3 base image-to-video mode. If First Frame is not visible, enable Add end frame, then place your authorized image in First Frame rather than the reference-image area. Leave the optional Last Frame empty. If that mode or first-frame role is unavailable, or the selected mode requires two images, stop and choose a compatible mode. Adapt the part and appearance names to your actual input. Check the complete prompt, current controls and credit estimate before manual submission. The link prepares model and prompt text only; it does not upload an image, enable a frame slot or start generation.

Open prompt draft

4. Complete prompt and sequence

Start from my authorized image of one hand holding a magnet above the left brass washer, with one separate steel washer on the right of the same fixed tray. Preserve both original washers, their holes and edge details. Keep the camera and tray fixed. The magnet begins with a clear gap above the brass washer. Hold briefly; the brass washer remains flat at its original place. Move the same magnet slowly toward the steel washer at the prepared working height without sweeping or touching either part along the tray. As the magnet approaches the steel part, let that same steel washer leave the tray and meet the magnet face in one readable movement. Preserve the washer as a separate rigid part attached to the magnet; do not fuse it into the head. Then raise the magnet only a short distance with the steel washer still attached and hold within the frame. The original brass washer remains on its original tray contact throughout. Keep both parts visible at the ending, exactly one on the magnet and one on the tray. The hand grips only the magnet and never picks up a washer directly. No new parts, swaps, shrinking, melting, hidden cut, floating washer after attachment, field-line graphics, sparks or added lettering. Do not imply a measured magnetic force or verified material test.

Identify the two original washers and their separate tray contacts. Record the real material and input-permission information outside the image; a gray or yellow surface alone cannot establish composition. Neither part starts on the magnet.
Follow the magnet's brief pass above the brass washer and toward the steel washer. Keep the planned working height readable and both parts in frame; touching or pushing a washer across the tray would introduce a different transfer mechanism.
Inspect the same steel washer leaving its original tray position and reaching the magnet face. Track its edge detail and central opening through that transition. If the magnet head hides departure or attachment, record the relation as inconclusive rather than accepting a replacement part revealed later.
During the short lift, check that the steel washer stays attached as a separate part while the brass washer keeps its original tray contact. Finish by accounting for both originals, not merely counting two ring-shaped objects in the last frame.

5. Inspect an actual output

Only after inspecting an actual output, record each criterion, time or frame, visible observation and pass, fail or inconclusive judgment. A steel-looking ring on the magnet at the ending does not establish its origin or the pickup transition. An obscured contact stays inconclusive. The clip can be checked against this authored plan; it cannot independently identify an alloy, measure a field or prove that magnetism caused a real transfer.

Exactly the two original washers begin separately on the tray, with neither attached to the magnet and both identities traceable.
The camera, tray and original brass washer remain fixed while the same magnet passes above the parts without mechanically sweeping them.
The steel washer's departure from its original tray position and arrival at the magnet face form one readable transition, without a hidden swap or teleport.
The steel washer remains a separate rigid ring with its original opening and edge details after attachment; it does not fuse into the magnet.
The subsequent short lift carries the steel washer with the magnet, without a free gap appearing between them or a second pickup.
The brass washer retains its original tray contact and identity throughout; it is not duplicated, replaced or secretly carried out of view.
The ending accounts for exactly one original washer on the magnet and one on the tray; hidden transitions, material identity and magnetic force are not reported as independently verified.

6. Repair one failure at a time

Both washers follow: first recheck the actual materials, backing and clean setup. If the input is suitable, restate the separate attachment and tray-contact roles; do not relabel a visibly replaced part as brass.
The magnet pushes the parts along the tray: restore the visible working gap and remove sweeping or dragging language. Inspect clearance before increasing the movement.
A new steel ring appears on the head: make the original departure and intervening route readable, keeping the same edge detail; an ending count alone cannot repair a missing transfer.
The steel washer merges into the magnet: preserve its ring opening and separate outline during the lift; reduce the lift distance if the junction becomes unreadable.
The steel washer floats below the raised magnet: require the established attachment to persist, and inspect the junction rather than adding field-line effects.
The hand or head covers the pickup: change the authorized input and camera angle before retrying. Do not invent transparency or a cutaway to replace the missing view.

7. Ending, saved result and limits

The same magnet is held a short distance above the tray with the original steel washer visibly attached as a separate part. The original brass washer remains flat in its starting place. Both washers and the hand-held magnet stay in frame.

After your own generation completes, use the normal result-download control to save the video. Keep the actual input and permission record, part-identification and setup notes, exact prompt, settings and timestamped observations. Saving a clip does not validate its pickup transition or material properties.

No H3 generation, physical pickup, field measurement, pull-force calculation or alloy verification was performed. The working gap and capture behavior must come from the actual setup, not a universal value in a prompt. A rendered movement can imitate selective pickup without proving its physical cause. Follow the instructions for the magnet and parts you actually use.

The sieve exercise separates sizes through fixed apertures; this plan uses similar-size parts without a sieve and checks attachment versus retained tray support. The placement exercise releases one gripped object onto a table; here the hand never grips either washer. The existing mechanic-and-robot prompt already uses magnetic theft, but it pulls many tools and ends with loose metal collecting together. This brief adds a known retained counterpart, continuous identity tracking and a material-preparation limit. These are visual review conditions, not a measured separation process.