Check the transfer as a chain of visible physical events, not as two matching poses. The prop should have a plausible grip, continuous support, and a clear handover. Matching the starting hand and finishing hand is insufficient if the object floats, duplicates, or changes ownership in between.
Start with a source action you can actually inspect
Choose a continuous source shot with the hands and prop visible. Establish who initially supports the object, its orientation, and which hand is expected to receive it. A cropped wrist or long obstruction may make accurate reproduction impossible to verify.
In a hypothetical demonstration, a performer lifts a closed box with the right hand and passes it to the left. This example describes a review method, not a reported model result. The box's lid, corners, and orientation provide useful visual anchors.
Separate source uncertainty from generated failure. If the original hides the release, write “release obscured.” Do not reconstruct the missing mechanics from what would normally happen.
Divide the handover into six phases
Use a compact worksheet before watching the generated version:
| Phase | What to observe |
|---|---|
| Approach | Receiving hand moves toward the same object |
| Contact | Fingers meet a plausible surface |
| Grip | Hand shape and object position agree |
| Shared support | Both hands can plausibly support the load |
| Release | Initial hand separates after receiving support |
| Settle | Prop remains stable with the new holder |
Some actions overlap these phases. The table is a viewing aid, not a requirement that every handover include a long shared hold. Record the sequence actually present in the source.
The object may rotate during transfer. Follow a recognizable edge rather than assuming every orientation change is a defect. Look for abrupt changes that the visible movement cannot explain.
Compare action phases, then inspect the transitions
Adobe distinguishes the Source Monitor for individual clips and Program Monitor for the edited sequence. Use an equivalent clip-and-sequence arrangement in your editor to compare the reference with the recreation.
Align the initial contact, then examine what happens before and after it. Merely comparing the same wall-clock timestamp can mislead when one performance begins later.
Look at the supporting fingers, the object's path, and the gap when a hand withdraws. At normal speed, does the handover feel physical? During closer review, is there an interval where neither hand supports the box? Does a second copy appear? Does a wrist merge with the surface and later separate differently?
Classify problems before attempting a repair
A brief obstruction is uncertain evidence. A visible unsupported jump is a confirmed defect. A different but physically coherent grip is a changed interpretation; whether it is acceptable depends on the original requirement.
Trimming can remove an unnecessary gesture at the edge of a shot. Cropping can change which action is shown. Neither proves that an omitted transfer was reproduced correctly. If the handover is the subject of the shot, hiding it defeats the task.
For another attempt, simplify one requirement: use a clearer reference, a larger prop, or less competing body movement. Preserve the earlier brief so you can identify which change was tested.
Deliver the claim that the clip supports
Review the final exported shot again, including every interval retained around contact and release. Record confirmed observations separately from hidden phases.
Accept a faithful transfer only when the required mechanics are visible and plausible throughout the relevant action. If the clip merely suggests a handover, describe it that way. This is a source-to-result motion check within one shot; a separate continuity review is still needed when editing the prop across multiple shots.