Make a Rap Music Video

A rap music video is built from two kinds of footage: performance takes, where you rap the whole song to the camera, and b-roll, the extra shots of places, hands, faces and details. You film several full performances in different spots, then cut between them and the b-roll in time with the beat.

You can do this with a phone. What separates a watchable video from a shaky clip is planning: a concept you can say in one sentence, a written shot list, steady framing and the song played out loud on set so your mouth matches the recording.

Budget one afternoon to shoot and one evening to edit for a three-minute song. More locations and more people add time quickly, so a first video should stay small: you, two or three places and one friend holding the phone.

Write the concept as one sentence: "Me walking my neighbourhood at dusk, ending on the roof." If it takes a paragraph, it is too big for a first shoot. Choose two or three locations within walking distance of each other and check the light at the hour you plan to film.

Then write a shot list: a numbered list of every shot you need, with location, framing and which part of the song it covers. Twelve to twenty lines is typical. On the day you tick them off, which stops you going home without the hook filmed. Draft yours before you charge the phone. Add the time of day beside each line. Soft light in the hour after sunrise or before sunset flatters faces, while midday sun puts hard shadows under the eyes.

Decide the shape first. Horizontal suits long-form video sites, vertical suits short-form feeds. Do not mix them in one shoot. Then set these and leave them alone.

Wipe the lens before every location. It sounds minor and it removes the haze that makes phone footage look cheap.

• Rear camera, not the selfie camera: the sensor is larger and sharper
• Resolution 1080p or 4K at 24 or 30 frames per second, the same for every clip
• Tap and hold on your face to lock focus and exposure so brightness does not pump
• Aeroplane mode on, so a call cannot ruin a take
• A tripod, a gimbal, or two hands with elbows tucked against the ribs

Play the finished song from a speaker and rap along at full voice. Miming quietly looks wrong on screen because your neck and jaw do not move the way they do when you project. The camera’s own audio is only a guide for syncing later. It will be replaced by the real track.

At each location, film the entire song three times: one wide shot showing your whole body, one medium from the waist up, one close on the face. Three locations times three framings gives nine complete takes, which is plenty to cut between. Perform every take as if it were the only one. Move between takes as well: walk toward the lens in one, stand still in the next, so the cuts have variety. Ask your friend to keep your eyes in the top third of the frame.

B-roll covers edits and tells the viewer where they are. Shoot ten to fifteen clips of five to ten seconds each: your shoes on the pavement, a street sign, hands counting or writing, a friend laughing, the sky. Hold each shot still for a count of five before and after any movement.

Match some b-roll to lyrics. If bar nine mentions a train, get a train. One or two literal matches per verse feel deliberate. More than that turns into a slideshow. Add a b-roll column to your shot list and note the bar each clip is for.

In any editing app, put the mastered song on the audio track first. Line up each performance take against it by matching the waveform peaks, or by clapping once on camera before every take and aligning the clap. Mute the camera audio once synced.

Cut on the snare, beats 2 and 4, or at bar lines. A calm verse can hold a shot for two to four bars. A hook can cut every bar or every two beats. Change framing with each cut, wide to close, so the jump reads as intended. Export at the resolution you shot in. Before exporting, apply one simple colour adjustment to every clip equally, so shots from different locations look like one video. Watch the result once on a phone before posting, since that is where most people will see it.

Punchline’s AI rap video generator is a different thing from a filmed video, and it helps to be clear about that. You upload your own photo with your face visible and your own audio of 10 to 15 seconds, and it returns a lip-synced MP4 in a few minutes for 20 credits up to 15 seconds.

That makes it useful for a teaser of the hook before release day, or a visual for a snippet when you have no footage yet. It will not replace locations, movement and performance in a full video. If you plan to post a teaser, pick the 15 seconds of the song you would use.