Which photo works best for a Burning Bridges AI video

A clear, well-lit photo of one person facing the camera, with the whole face visible works best. Head-and-shoulders or half-body shots give the most natural result. Avoid sunglasses, hands or hair over the face, heavy filters, group photos and small or blurry faces.

On Burning Bridges AI Video, upload a JPG, PNG or WebP up to 10 MB. Everything below explains why these rules matter, so you can judge a photo before you spend a credit on it.

1. Face size. The face should fill a good part of the frame: head-and-shoulders or half body. The AI rebuilds the face from the pixels it has. A face that is a small patch in a wide shot gives it too little detail, and the result looks like a stranger.
2. Angle. Front-facing, looking at or near the camera. A straight-on photo shows both eyes, both sides of the face and the full shape of the jaw. Strong profiles and steep angles from above or below make the model guess the half it cannot see.
3. Light. Soft, even light across the whole face: a window by day, or open shade outside. Hard shadows, strong colored light or flash glare get carried into the result or filled in wrongly. The party scene is dim, but a well-lit photo still works better, because the model matches the face to the room’s light.
4. Expression. Neutral or relaxed. A wide laugh or a squint is an expression the model has to undo, because the clip has its own. A calm face carries over more of what makes you look like you.
5. Background. Plain and uncluttered. The background is replaced by the party room, but a busy one, or another face behind you, can confuse which face to use.

• Group photos. The video has one lead, so the model needs one face. Crop to one person or take a new photo, and make a separate video for each friend who agrees.
• Sunglasses or hands over the face. Use a photo with the whole face visible. The scene has a hand over the eyes, but the model needs to see your eyes first to draw them around it.
• Hair across the eyes or face. Tuck it back, or pick another photo.
• Heavy filters and beauty modes. Use the original, unfiltered photo. Filters reshape the face and smooth the skin, and the AI copies what it sees.
• Small or blurry faces. Old low-resolution photos, screenshots of screenshots and zoomed-in crops from far away lose the detail the model needs. Use a recent, sharp phone photo.
• Masks, wide-brimmed hats that shade the eyes, and face paint. Use a plain photo.
• Pets and objects. Not supported on Burning Bridges AI Video: each video puts one person in the lead role.

1. Face a window, a step or two back, so daylight falls evenly on your face. Turn off overhead lights that throw shadows under the eyes.
2. Use the back camera if someone can take it for you; it is usually sharper. The front camera works too if you hold the phone at eye level at arm’s length.
3. Frame from the chest up, with your face in the upper middle and some space around your head.
4. Look straight at the lens with a relaxed face: mouth closed or a small smile.
5. Turn off portrait mode, filters and beauty effects.
6. Take three or four, zoom in on each, and keep the sharpest one where the eyes are in focus.
7. Upload it as it is. Most phone photos fit under 10 MB; if one does not, export it as a JPG.

The AI keeps your face: its shape, your features and the overall look of you. Everything else comes from the original scene in Drake’s music video. The clothes, the room, the guests, the light and the movement are the scene’s, so you will be wearing the lead’s outfit, not what you wore in the photo.

Some details can drift, even from a good photo:

• Hair can come out shorter, closer to the lead’s in the original scene.
• Skin tone can come out lighter under the scene’s warm, dim light.
• Expression can follow the scene: a smile in your photo may not survive, because the clip has its own expression.

A front-facing, evenly lit photo with a neutral expression keeps these changes small. If a result drifts too far, a different photo usually helps more than trying the same one again. A video that fails to generate uses no credits, but a finished one does, so choose the photo before you start.

Use only your own photo, or one of someone who has agreed to be in the video, and never a photo of a child. Making the video confirms this.

Your photo is used only to make your video and is deleted within 24 hours. It is not used to train AI models, published or sold.

Before you use a friend’s photo, ask them. A ten-second clip is easy to share and hard to take back. Ready? Upload your photo. Want the background first? Read what the Burning Bridges AI trend is.