# How to Make Dr. Doctor and LOBA AI Videos

> Make Dr. Doctor and LOBA AI videos from one photo. Compare the pool and studio scenes, choose the right crop, and check your export before posting.

- URL: https://trendai.video/blog/how-to-make-dr-doctor-and-loba-ai-videos
- Published: 2026-11-17
- Category: TikTok Trends
- Tags: AI music performance, Dr. Doctor AI video, LOBA AI video
- Author: Trendai.video

Dr. Doctor and LOBA both put your face into a 30-second music performance, but they are not two versions of the same shot. Dr. Doctor follows one singer beside an indoor pool in a vertical handheld clip. LOBA cuts between studio angles around a mic in a wide frame. Choose the format before choosing your photo.

## Pick the scene that fits the post

The [Dr. Doctor pool template](https://trendai.video/trends/dr-doctor-pool-performance-video) puts one singer close to the camera beside an indoor pool and rock waterfall. They point and sway. Its live form says **30 seconds, 9:16**. The camera's proximity makes this a face-first joke: a friend should recognize you before they notice the pool.

The [LOBA studio template](https://trendai.video/trends/loba-studio-performance-video) puts its singer at a standing mic under green neon light. They wear headphones and glasses. The shot switches angles across **30 seconds in 16:9**. The mic and room matter here. If you crop the wide frame into a narrow vertical slice without checking every cut, the singer may be visible while the mic or pointing hand disappears. The studio vanishes too.

| Choice | Dr. Doctor | LOBA |
| --- | --- | --- |
| Setting | Indoor pool and rock waterfall | Green-lit recording studio |
| Camera | Close, handheld performance | Several studio angles and cuts |
| Delivered frame | Vertical 9:16 | Horizontal 16:9 |
| Stronger posting angle | A quick face reveal on a vertical feed | A performance bit where the room and mic stay in view |

Both forms ask for **one required photo of the singer**. Neither asks you to upload a pool, a second performer or a recording booth. The source clip supplies the setting and movement; your image supplies the person being recast. The photos shown on those pages are source references, not proof of how your own export will look.

## Use a face the camera can keep

Start with a sharp solo photo in ordinary light. Both singers face the camera, so show both eyes and your hairline. A front-facing picture with an uncovered mouth is a better starting point than a dim party crop where another person's shoulder cuts across your cheek. That is input advice, not a promise that any specific photo will render perfectly.

For Dr. Doctor, look at the face at phone size before you upload. The camera stays close. Glasses and bangs can fill the shot; so can the mouth. Heavy beauty filters or a hand over the lips leave less useful identity detail. You are casting the singer, not prescribing the polka-dot top or the pool. Don't write a caption about wearing that exact outfit until you've seen your own file.

LOBA needs a different check: the same person must remain recognizable after an angle change. Pick a photo with a clear jawline and visible hair rather than a dramatic profile. Watch the nose shape across cuts. The studio reference includes headphones and glasses, but your upload is not an instruction to add an instrument or another singer. One face. One role.

If your post is about a group of friends performing, neither template casts a whole group. The [Desk Concert guide](https://trendai.video/blog/desk-concert-ai-video-guide) covers a different band setup with separate optional roles. Use that when assigning friends to instruments is the joke; use Dr. Doctor or LOBA when one familiar face carrying a solo is enough.

## Make one version and inspect the hard parts

1. Open the [Dr. Doctor form](https://trendai.video/trends/dr-doctor-pool-performance-video) or the [LOBA form](https://trendai.video/trends/loba-studio-performance-video). Add the consenting adult's clear photo to the required singer slot. Check the thumbnail to make sure you did not upload a group shot by mistake.
2. Choose the resolution offered in your session. Both forms currently show 480p and 720p; check the credit amount on the live form before generating rather than treating a price in an article as permanent.
3. Open the result at full size. On Dr. Doctor, watch the pointing hand move toward the lens and check whether the face and mouth stay consistent. On LOBA, pause immediately before and after an angle change: does the same person still read as the singer, and is the mic in the right place?
4. Play the downloaded file with sound on before cutting it for a post. Both forms label their reference clips as having source audio. That describes the source, not a test of your particular downloaded file or permission to publish its song.

If the face warps near the pool camera, try a less filtered, better-lit portrait before spending another generation on the same bad source. If LOBA holds the face in one angle but loses it in the next, switch to a photo with clearer frontal features. Check the entire clip again after a retry. A good first frame can hide a rough cut.

Don't use the source poster as an example of an output made with your photo. We haven't generated a personal test clip for this guide. Your own export is the only honest before-and-after check.

## Cut for the feed you actually use

Dr. Doctor is already vertical. Leave the performer and pointing hand clear of the app's caption and button overlays; a joke placed over the mouth fights the lip movement. Try a caption about your unexpected poolside concert, not a claim that you performed in that real location. Open with the recognizable face, then let the pool reveal itself.

LOBA is wide. If you post it to a vertical feed, preview a 9:16 crop through **every** camera cut before committing to it, because a crop that holds the singer and mic together in the opening may cut one of them off after the next camera move. Some shots put them on different sides. A safer option is to keep the full 16:9 image with space above or below in a vertical canvas, rather than chopping away the very studio that explains the joke. Check the platform preview, not only the editing timeline.

Neither template gives you the rights to someone else's face or the source recording. Get permission from the person in the upload and avoid implying they actually sang the track. Disclose the AI-made recast where the platform requires it. If the export contains music you cannot clear for your post, mute it and use audio you are allowed to publish in that account and region. Then replay the result: the visual pointing or mic gesture still needs to land on the sound you chose.

