Give the doorway viewpoint a clear face reference
Create a doorbell-camera-style porch scene from one photo, with a small approach, an awkward reaction and a tiny wave. Use the AI Doorbell tool to create the clip, or read the effect introduction for the intended sequence.

The preset creates a silent, 5-second, 720p video for 284 credits. The scene uses a vertical 9:16 frame, regardless of the source photo’s orientation. Check the displayed quote before generating; five seconds is the finished clip length, rather than the processing time.
Step 1: Choose the reference for this sequence
Your photo is an identity reference for a newly generated scene. It does not set the opening video frame, source background or final crop.
The preset builds a new porch viewed from a fixed wide-angle doorbell position. Use a sharp face with visible upper body; the uploaded background and camera angle do not carry into the new scene.
- Use one person with a sharp face and visible upper body.
- Show a hand clearly for the small wave.
- Prefer an unobstructed reference rather than a close phone or object in front of the face.
You do not need a photograph at a real front door. The source provides the person, while the porch and fixed camera position are generated. A medium reference gives the model more upper-body and hand detail than an extreme face crop. Use a photo you have permission to use.
Step 2: Upload and start one generation
Open AI Doorbell, sign in and upload one JPG or PNG up to 20 MB. Wait for the upload to finish and check the preview against the photo guidance above. Confirm your balance covers 284 credits, then select Generate once.
There is no camera-device connection, porch selector, address field or lens-strength control. Processing can take several minutes. If progress stops updating, check My videos and your balance before submitting again. Contact support if the status remains unclear.
Step 3: Review the full action and ending
The visitor takes one small step toward the fixed doorbell-camera viewpoint. A surprised look and small head tilt create the awkward pause. A short wave ends the reaction while the camera remains fixed. These are the intended actions; generated timing, identity and object motion can differ.
Watch the full clip once at normal speed, then replay the contact points and ending:
| Check | What to review |
|---|---|
| Fixed viewpoint | The doorbell-style camera stays fixed while the visitor takes a small step. |
| Intended lens look | The wide-angle perspective is readable without the face becoming unrelated to the source. |
| Reaction | The head tilt follows the approach, rather than jumping between different poses. |
| Tiny wave | One hand waves with coherent fingers and remains connected to the arm. |
If the face deforms, compare it separately from the intended lens curvature. A sharper reference can help identity, but cannot guarantee the generated wide-angle view will remain consistent. Each new generation uses credits, and a different source does not guarantee a corrected result.
Step 4: Download and finish the timing
Use Download video in the studio, or open the completed result in My videos. After downloading, add a short chime before the reaction in your editor. Keep any caption away from the face near the center of the wide-angle view.
Music, captions and trims are editing steps after downloading. If you share the result, identify it as AI-generated so viewers understand that the scene or movement is newly created.
FAQ
Do I need a doorbell camera recording?
No. Upload one person photo. The porch, lens viewpoint and movement are generated.
Does it ring a real doorbell or include camera controls?
No. This tool creates a silent video scene. It does not connect to a camera or provide security-device controls.
Why does the image look wide or curved?
A fixed wide-angle, doorbell-style view is part of the preset. Extra facial distortion can still be a generation error, so compare the face with the source.
Choose a suitable photo and create a AI Doorbell video. The effect introduction has a quick summary of the photo requirements and limitations.