How to Make the Outstanding Tiny Desk AI Video

Use a well-lit, front-facing photo with one face in sharp focus. Choose one portrait for Solo, or collect five separate portraits for Band. A plain background is fine; you do not need to photograph yourself at a concert or hold a microphone. Avoid group shots, very small faces, heavy beauty filters and sunglasses hiding the eyes. A photo with visible hair and shoulders gives the model more useful detail than a tightly cropped eye-level fragment.

Upload yourself or an adult who has agreed to this use. Photos of minors, celebrities and public figures are not allowed. Do not use the result to mislead people into believing someone really performed, endorsed a product or said something they did not. Your photo supplies the appearance of the new singer, not a sample from which we clone your singing voice.

Both options make the same performance, about 18 seconds long, in a 9:16 vertical frame. Standard uses 480p; HD uses 720p and more credits. HD is a higher output resolution, not a promise that every facial feature or fast hand movement will be perfect. A clear input still matters, and generated motion can contain small visual glitches.

Choose Solo for one replacement role, the lead singer. Choose the five-person Band template for the lead singer, keyboard player, two backing singers and sax player. Use five separate photos in the labelled order so each person gets the intended role. There is no two-person or Halloween template in this first version.

Your selected photo and quality stay in the browser while you sign in or visit the pack page. A completed purchase returns you to the tool, and the saved draft can continue after its credits arrive. Cancelling checkout does not mean a render has been paid for, and you can return to the photo before trying again.

Single is $9.99 for 10 credits, enough for one solo standard video. Triple is $24.99 for 30 credits: three solo standard videos or two full-band standard videos. One HD video uses 25 credits for Solo or 29 for the full band, leaving five or one respectively. These are one-time packs, not a subscription. Check the current pricing page and the credits on the generation button before continuing.

Watch the full clip with sound before sharing. The finished MP4 includes the trend's original recording, while the picture is an AI remake with your face in the lead. Compare the face through head turns and the mouth during the sung phrases. The backing response comes from off camera in this excerpt, so a pause in the lead singer's mouth is not necessarily an error.

Download the MP4 when it is ready. A silent copy is useful when you intend to choose an approved sound inside a social platform. The included soundtrack does not grant unrestricted rights to music, and the AI clip should be labelled accurately. If a generation fails or is interrupted, its credits return automatically; cash purchase refunds follow the separate refund policy and require contacting support.