Photo & workflow guide

How to make a Bulin 47 AI video

Start with one portrait, not an edited video. The goal is to put your authorized character into the lead performer's role while retaining the recognizable freestyle scene and timing.

1. Step 1

   Choose the right portrait

   Use one clearly visible adult face in even light. A front-facing or slight three-quarter view is easier to assess than a distant profile. Avoid sunglasses, heavy filters, face coverings and group photos. Use yourself, a consenting adult, or an original adult character you control.

2. Step 2

   Understand which person changes

   The intended replacement is the lead performer, not the person holding the speaker or the surrounding crowd. You should not need to draw masks, edit a first frame or write a complex prompt. Those are implementation details, not preparation tasks for you.

3. Step 3

   Check the output specification

   Our planned output is a 14-second, 720p vertical MP4. The homepage shows a 6-second source excerpt and links to a separate 19-second reference with sound. A reference clip is not a generated result, and its duration does not establish the final output length.

4. Step 4

   Review the entire result

   Watch at normal speed from start to finish, then inspect turns, hands crossing the face and fast movement. Check that the selected identity remains recognizable, the lead follows the source action, other people remain stable and the audio stays aligned. A good first frame alone is not enough.

5. Step 5

   Share with context and permission

   If generation becomes available, only share results you have permission to use. Label the video as AI-edited rather than presenting it as a real performance. Permission to use a portrait does not automatically grant every right to the original recording or music.