AI video effects: a photo and role guide
Compare the photos needed for selected AI video effects, from solo portraits to pets and empty rooms. See each scene’s preview, target format and credit cost.
An adult portrait, two separate portraits, a pet photo or an empty-room photo serves a different role in each AI video effect. Check the photo requirements and current availability in the studio. Browse all video effects.
Bring a character from a child’s drawing to life
WaveMira’s Kids Drawing to Life turns one drawing into a clay-style still or a lifelike clay video. The video opens on the full drawing, then reveals the same character moving as a clay figure on a tabletop. The video is set to 5 seconds at 720p with no audio and costs 90 credits.
Use one drawing of one fictional character that you have permission to use. Keep the whole character in frame and crop out any real faces, hands, names or signatures.
See WaveMira’s Kids Drawing to Life effect.
Compare the original drawing with a clay-style AI still, then watch the independently made lifelike clay video preview. The studio lists current modes and credits.
Hallway runner
Closing Door Prank AI centers on a run toward a narrowing doorway, the squeeze through and a slide or reaction at the end. Use one adult photo for the runner. There is no second upload slot for the person holding the door.
Target video: 9s · 720p · 180 credits.
Show the face and enough of the body to guide the runner. A close selfie leaves the model to invent the outfit and body. The moment at the door matters more than a decorative camera move.
Dog Elevator Prank
Make a dog elevator prank video from two photos. In the first photo, show one dog with its face and paws visible. The second photo shows one consenting adult. The dog presses the elevator button and slides a shoe inside after the adult walks in.
How can I make my dog dance by a pool from one photo?
Start with Dog Pool Party AI and choose a photo with one dog, a clear muzzle, visible ears and room around the paws. The 13-second silent video follows a goggle close-up with cross steps, a back-facing dance and a glance back as a warm light enters from the side.
How can I make a 1950s pet shop ad with my own dialogue?
Use WaveMira’s Husband Emporium pet shop ad maker with one clear photo of a dog or cat. Type one or two short lines for the shopkeeper. WaveMira turns that text into AI narration in English or Mandarin and adds it to the finished video.
Choose a photo where the pet’s eyes, ears and fur markings are visible. In WaveMira’s public 15-second sample, the clerk rings a bell and sets down a nameplate, then the cat appears at about six seconds. Play the sample with sound to hear how the narration fits the reveal.
They Say I’m 2 Much AI: Donut Shop Video
They Say I’m 2 Much AI: Donut Shop Video takes one photo of a consenting adult. The vertical video follows a fisheye walk-in and ends on an overhead donut box reveal. The download uses an original beat.
Current recipe: 15s · 480p · 120 credits. Check the studio for availability.
A clear face and visible upper-body clothing help establish the performer. Look for how the performer moves with the box and how the close-up ends the scene.
Seven more scenes: choose the input before the action
- Pet as Santa: One cat or dog only, with face, fur markings and all four paws clearly visible.
15s · 720p · portrait · 280 credits - Bicep Heart: One consenting adult, waist up. Keep both shoulders, elbows and hands visible, with room beside the head.
9s · 720p · portrait · 180 credits - Santa Naughty List: An empty room or fireplace photo only. Do not include people, personal documents or readable addresses.
10s · 720p · landscape · 160 credits - Mandem in Disguise: One consenting adult, with a clear face and upper body. Leave room around both shoulders.
10s · 720p · portrait · 160 credits - Apartment Window Dance: One consenting adult, standing with face and full body visible. Keep hands and feet inside the photo.
10s · 720p · portrait · 160 credits - Cardo Dinner: Two separate consenting adult portraits. Photo 1 is the seated eater; photo 2 is the standing server. Roles stay fixed.
15s · 720p · portrait · 270 credits - Time Freeze: One consenting adult, waist up or full body, with both hands clearly visible.
15s · 720p · landscape · 280 credits
For the Mandem in Disguise flash outfit change, choose one photo of a consenting adult with a clear face, visible shoulders and enough background to establish the room. In one production trial, the room stayed recognizable after the flash and the camera widened from a half-body shot to a full-body shot. Check the effect page for the current reference and studio availability.
Pet as Santa and Bicep Heart each share one topic page with a separate photo mode. An image costs 5 credits at 1K or 10 at 2K. Choosing photo mode does not start or pay for a video. Check the studio quote and current purchase availability before proceeding.
For Cardo Dinner, assign both roles before uploading. For Santa Naughty List, supply the empty room rather than a portrait. The remaining solo adult scenes need permission from the person shown; pet inputs must contain no people.
Scene interpretation and exact performance edits
A photo-led scene is newly generated. A fixed-performance edit uses a specific reference clip to constrain timing, movement and framing. Similar colors and props alone do not prove the exact performance has been reproduced.
Check each studio for the preview source and whether the effect includes sound. A preview shows how a scene is intended to look or move. It does not grant a licence to reuse a third-party track.
Before a video opens
A scene illustration or shot plan is not an accepted generated sample. The current preview state is stated in each studio, and purchases stay closed while the required checks remain unfinished. When enabled, the studio shows the exact photo roles, duration, format and credits before payment and upload.
Review a finished video from beginning to end. Watch the main action, the subject’s identity, hands and limbs, cuts and the ending. A playable file alone does not establish that the scene worked.
If you already have original and translated subtitles, use the SRT Translation Checker to review reading speed and edit the translation without changing its timestamps.