COPY. CAST. CREATE.
A useful Hotel Lobby prompt tells the model who to replace and what to preserve. Copy a starting point below, attach your own references in a compatible editor, and change only the details that matter to your cast.
Create with the original templateThese are instructions for video editors that accept reference footage and images. Replace “first photo,” “second photo” and “reference video” with your editor’s attachment labels if required. Our original generator handles its own prompt; custom prompt input and selective replacement are planned features.
Replace the left performer in the reference video with the person in the first photo and the right performer with the person in the second photo. Preserve both identities, the orange studio, suspended microphone, camera framing and performance timing. Keep their clothing consistent with the photos and their movement natural.
Cast the couple from the two supplied portraits as the left and right performers respectively. Preserve each person’s facial identity, hairstyle and outfit. Keep the reference video’s orange backdrop, microphone, choreography and camera position. Maintain realistic proportions and consistent lighting throughout.
Use the first sibling’s portrait for the left performer and the second sibling’s portrait for the right performer. Keep the two identities distinct, including their individual facial features and hairstyles. Preserve the reference scene, original movement and camera framing. Do not blend the faces together.
Create a stylized pet-and-owner performance using the supplied reference video. Place the owner from the first photo on the left and the pet from the second photo on the right. Preserve the owner’s identity and the pet’s recognizable markings. Adapt the pet’s movement coherently while retaining the orange studio, hanging microphone and overall timing.
Replace the two performers with the subjects in the corresponding left and right reference photos. Preserve the visible costumes, faces and silhouettes from those images. Keep the orange studio, suspended microphone and performance choreography unchanged. Maintain consistent costume details through motion and avoid merging the two characters.
Replace only the left performer with the person in the supplied portrait. Preserve that person’s facial identity and visible outfit. Leave the right performer unchanged. Keep the reference video’s studio, microphone, camera position, choreography and timing intact. Apply identity changes only to the selected left subject.
Say which image supplies the left performer and which supplies the right. Without that assignment, a model can mix identities or use the same face for both positions.
Describe the parts that should stay fixed: orange studio, hanging microphone, camera framing, choreography and timing. Refer to the supplied video rather than asking the model to guess a specific performance from its name.
Choose a visible outfit or a simple accessory instruction. Avoid a paragraph of tiny details that compete with identity and motion consistency. If the reference already shows the clothing you want, say to preserve it.
Do not combine incompatible directions such as a locked camera and a moving orbit, or preservation of the source set and an entirely new location. Each contradiction asks the model to decide which part of the request matters more.
Avoid promising yourself that words like “perfect” or “viral” will solve an unclear input. A sharp, unobstructed face and an explicit left/right assignment are more actionable than a long chain of quality adjectives. If the result drifts, simplify the instruction and improve the source portrait.
Some editors use named attachments, others use numbered images, and some treat the video as a separate motion reference. Keep the role assignments intact while changing the attachment syntax. Do not paste model-specific tokens into a tool that does not recognize them.
Audio is a separate capability. Asking a silent video model to preserve music does not make it an audio editor. Confirm the model’s supported workflow, and use a source you are allowed to edit and publish.
Paste them into an editing tool that supports the required images and reference video. The current homepage generator does not expose a custom prompt field; it prepares the original template instructions for you.
No. Upload your two photos, set their positions and choose the template. The prompt library is useful for creative planning or a separate compatible video editing workflow.
No. Attachment syntax, selective editing, animal handling and audio support vary. Treat each prompt as a clear creative brief, then adapt it to the model you use.
YOUR TWO PHOTOS. YOUR MOMENT.
