Video trend
With eight people in one video, a problem with a single face is easy to miss and easy for that person's friends to spot. Go through the video once with the sound on, then again person by person. These checks are in the order that catches the most, and they tell you when to post and when to go back to a photo.
Pause once on each of the eight turns and ask the person's friends' question: is that really them? Check that each photo landed in the right slot too. If one face is off and the others are fine, replace that photo with a sharper, closer one.
The video is made in three parts and joined near about 11.5 and 23 seconds, while the camera is swinging between people. Watch these moments in slow motion. A small jump in the background or in somebody's clothes can happen there. If it passes unnoticed at full speed, most viewers will not see it.
Singers also appear behind each other. Check the person at the shoulder of each singer, especially around 12 to 16 seconds, and the group at the end. Someone from a different part of the video may look a little different in the background, and that is the most likely place for a mix-up.
The hands to check are Singer 3's heart around 11 seconds and Singer 6's raised hands around 21 to 23 seconds, which come very close to the camera. Count the fingers at full speed and paused. A brief smear passes; an extra finger you can see at full speed does not.
Watch each singer on the chorus. The mouth movement follows the reference performance, and the original audio is restored at the end, so a small drift can happen. Then check that the singing mouths stop once the song gives way to laughter near the end.
Look for grey patches, thin cyan lines or a faint grid on any person, and any logo or text on clothing that was not in the photo. Re-run for a wrong face, a swapped person, visible guide marks or a badly broken hand. Change the photo first if the face was the problem. Testing at the lowest quality is the cheapest way to judge a new set of photos.
Add a photo for each role shown below. All roles need a photo.
Add one clear photo per person, only of people who agreed to appear. A sharp face near the camera with the whole head and hair works best. Put the person nobody expects in the closer slot.
Choose 480p, 720p or 1080p and review the quote. It covers all eight photo preparations and all 37 seconds, so read it before you generate.
Pause on each of the eight turns, then check the joins at about 11.5 and 23 seconds, the hands near the lens and the mouths on the chorus before downloading your portrait MP4.
All eight people you see up close come from your photos. Each person's face, hair, build and outfit come from their own picture, so nobody needs a costume.
About 37 seconds in portrait, with one phone moving from person to person and no cuts to speak of. The room, the camera's path and the order stay the same.
Your video keeps the clip's own audio, including "Baby" by Justin Bieber featuring Ludacris. Nobody records anything, and the mouths follow the reference performance.
The closer stands beside the first singer from the first frame and takes the camera at the end. The preview is the reference clip, not an Oakgen result, and its eight main performers are replaced by your photos.
It is a 37-second portrait video in one long phone take. Seven people lip-sync the same pop song in turn in a crowded room, then the camera lands on an eighth person while everyone laughs. Oakgen's template replaces all eight with people from your photos.
"Baby" by Justin Bieber featuring Ludacris, identified from its lyrics. Neither artist appears in the clip or is connected to this template. We did not find an established name for this exact clip, so Oakgen calls it Baby Lip Sync.
Yes. The source cast is never kept, so the studio will not start until all eight slots have a photo. You can use a photo twice, but that person will appear twice in the same room.
The person nobody can picture doing this: a parent, a boss or the quiet friend. The closer stands beside the first singer from the start and takes the camera for about the last ten seconds.
No. The video keeps the clip's original audio and nobody records anything. The mouths follow the reference performance, so check them closely before posting.
Yes. There is no costume. Each person's outfit comes from their own photo. The template asks for no logos or lettering from the source clothes to carry over.
They are extras and are not cast from your photos. The template asks for anonymous people in the background, but it is an instruction to a model, so check the background before you post.
Only photos of yourself and of people who agreed to be in the video, all eight of them. Do not upload photos of athletes, celebrities or anyone who has not consented, and label AI videos clearly when you share them.
A single model request cannot cover 37 seconds, so the template makes three parts and joins them near 11.5 and 23 seconds while the camera is swinging between people. Check those moments in your video.
The template is written for people lip-syncing. An animal photo is not blocked, but the result is untested and the mouth timing may not look right.
About 37 seconds in portrait 9:16.
No. It is the reference clip the trend comes from, not an Oakgen result. Review every face in your video before sharing.
5,231 credits per 720p video. The quote includes all eight photo preparations and the complete video. Check it before generating.
Plan a story, keep your character recognizable, and make the finished clip your own.
Make the Baby Lip Sync AI trend step by step: choose your eight people and photos, assign the slots, read the quote, check all the faces and post your group video.
What the Baby Lip Sync AI trend is, why a 2010 pop song works in a crowded room, why the last person makes the video and how the template rebuilds it from 8 photos.
Upload 8 photos. Everything else is set.
Oakgen.ai