Tiny Desk AI Template: Seats, Shots and Cuts Explained

The Tiny Desk AI template is not a single downloadable file. What people copy is an arrangement: one desk, one microphone, five people close enough that you can see their hands, and a camera that stays in the room for the whole take. Once you know the arrangement, you can rebuild it with your own cast and your own audio. The song and the concert behind it are covered in Outstanding Tiny Desk AI; the format itself is explained in What is Tiny Desk AI?.

What is the Tiny Desk AI template?

Most versions circulating online take between one and five photos. A single portrait produces a solo performance; five portraits produce the full band. The template supplies the room, the light and the seating, and your photos supply the cast. That is why the same arrangement can carry a stranger's face, a cartoon character or a group of friends and still read as the trend.

Who sits where in the five-seat version?

The five positions most edits use:

  1. Lead singer: centre, at the microphone, shown closest.
  2. Keys: behind and past the singer's shoulder, angled toward the desk.
  3. Bass: standing at the edge of the frame, near the keys so both fit in a wide shot.
  4. Drums: furthest back, because a kit needs room and reads better at a distance.
  5. Featured instrument: saxophone, guitar or anything else the camera should notice once.

Read the arrangement as a plan, not a rule. The seat order matters because it decides who stays visible when the camera settles on somebody.

How many photos do I need?

One clear portrait per person. Cropping five faces out of one group photo rarely works: each face arrives at a different size, the lighting does not match, and the smaller faces lose detail before the frame is even built. Ask each person for a shoulder-up portrait taken in even light, without sunglasses or a hat that hides the eyes, and with nobody else in the frame. Keep the originals and crop copies.

What does the camera do?

The original performance is the reason the trend reads as one continuous room: a wide shot shows how close the musicians are, then the camera settles on whoever is performing. A rebuilt version keeps that rhythm: establish the room, follow the lead, show one instrument in detail, then return wide. If you cut on every beat, viewers lose track of who is playing what.

Can I use the original audio?

The song in the trend is "Outstanding" by The Gap Band, and the concert clip is Charlie Wilson's performance for NPR Music. That recording is not yours to publish. Use a track you have rights to, or audio from the platform's licensed library, and label the result as an AI-generated performance where the platform asks for it. Reusing a broadcast recording is the most common reason these edits get taken down or muted.

How do I keep the cast recognisable?

Keep the same crop and the same light for every portrait. Use the person with the clearest face as the centre performer, and put the busiest photos at the outer seats where detail matters less. Before you export, watch the clip once with the sound off: faces that drift, hands that bend the wrong way and instruments that change shape are easiest to spot without audio.

Do I need a specific app for the template?

The trend spread through apps that pair the arrangement with a photo-to-video model, and several of them publish their own version of it. What those versions share is one photo per performer, a short excerpt of a song and a vertical frame. Choosing between them comes down to how many photos you can add, how long the clip runs, whether there is a watermark, and what you pay.

Try the cast in the studio

Open the Tiny Desk AI studio to check up to five portraits locally. Select the lead performer first, followed by the other musicians. When the photos are ready, sign in to generate a vertical MP4 with a generated soundtrack. See the Tiny Desk AI video guide to review your finished clip and prepare it for sharing.