58th Floor-Style Rap Video Generation Workflow and Prompt Design

Two performers, a teal studio and one hanging microphone create a clear visual foundation for a rap duet. Turning that setup into an engaging AI video takes a little more direction: each performer needs a recognizable identity, a distinct turn on the mic and enough screen time for the audience to follow the exchange.

This guide uses AIRAP's colors 58th floor a colors show theme as a practical starting point. The workflow combines two separate photo references with a brief shared opening, individual close-ups and a wider ending. The creative goal is an original performance built around your subjects and topic.

Figure 1. The theme's visual foundation: a shared teal space, two distinct performers and a single suspended microphone. Wedding attire belongs to this example; your own photo references guide the subjects and their clothing.

1. Prepare two separate photo references.

Use one clear individual photo for each performer. Keep faces unobstructed and avoid heavy filters or very distant subjects. Label the first reference A and the second B, then retain those labels throughout the brief.

Assign A to the left stage position and B to the right. These positions describe where they stand in the studio. A close-up can crop the partner out while that partner remains in the same place off-screen. This distinction helps keep the camera instructions consistent with the scene.

2. Choose the theme, format and topic.

For the guided route, open the ai rap video creator, select 58th Floor and upload the two portraits. Choose the available model, duration, aspect ratio and resolution, then review the credit quote before generating.

A ten-second concept provides a useful starting point for planning both solo turns and a wide finale. A landscape frame gives the duo more space; a vertical frame calls for closer attention to the microphone, hands and headroom.

Keep the audio topic focused on one shared moment. For example:

Two friends celebrate launching their first project. Mention late nights, teamwork and the excitement of a small first win. Keep the exchange playful and confident.

The topic gives the duet something specific to express. In the theme workflow, use the audio-topic field for that subject; the detailed production prompt later in this article illustrates the directing logic behind the scene.

3. Plan a sequence the viewer can follow.

Build the performance around four compositions. Establish the relationship, give each lead a solo view, then reveal the full stage. One possible ten-second shot plan is:

Treat these times as directing targets. Camera cuts and vocal handoffs need their own instructions, and a generated clip may interpret their timing differently.

Figure 2. A solo view should give the current lead a distinct composition. The partner can remain on the same stage mark outside the frame.

4. Write instructions that describe visible behavior.

Broad adjectives such as “cinematic” or “energetic” leave many decisions unresolved. A useful prompt identifies what the model should preserve and what should change over time.

Start with identity: A comes from reference one, B from reference two. Lock the shared scene with a seamless teal wall and floor, neutral face lighting and one hanging microphone. Then assign behavior: the lead raps with a small head nod and a clear hand accent while the partner listens and reacts.

Be explicit about solo framing. “Favor A” can still describe a two-shot. “Show only A's face and upper torso; B is completely outside the frame” defines a different image.

The ending also needs a visible destination. Ask for both performers from head to toe, contact with the floor and open teal space around them. A request to “zoom out slightly” can leave the final shot almost as tight as the opening.

Figure 3. The second solo should retain B's identity, outfit and lighting while giving the response its own screen time.

5. Combine the scene, performance and audio into a core prompt.

The following is an illustrative brief for a video model that accepts reference images and detailed shot instructions. Replace [TOPIC] with the subject of your duet. It describes intended behavior rather than guaranteeing exact cuts, lyrics or lip synchronization.

Create a 10-second original English rap duet using two reference images.
Reference 1 defines performer A; reference 2 defines performer B. Preserve
their distinct faces, hairstyles, outfits and body proportions.

SET: A seamless matte teal studio with matching wall and floor, soft neutral
light on both faces and natural floor shadows. Exactly two performers and
one black microphone suspended between them. A stands on the left stage
mark; B stands on the right. Keep their positions and the equipment stable.

PERFORMANCE: The current lead visibly raps with natural head movement and
one relaxed hand accent per phrase. The partner listens, bounces subtly and
reacts at phrase endings. Keep gestures clear and faces unobstructed.

SHOTS:
0.0–1.2s: Brief frontal two-shot. A starts; B listens.
1.2–4.4s: Hard cut to A alone, face and upper torso in frame. B stays off-screen.
4.4–7.6s: Hard cut to B alone, preserving the same studio and stage axis.
7.6–10.0s: Cut wide, then dolly backward to reveal both performers head to toe,
their feet and ample teal floor. B continues the response during the pullback.
Reach the wide composition before the final beat and allow a natural release.

AUDIO: Original English hip-hop about [TOPIC], around 104 BPM, with a light
funk bass groove and crisp drums. Use two distinct original voices. Begin
vocals promptly, give A and B separate lead turns, then add a very short
shared closing tag. Aim for roughly 28–34 words across the whole clip,
leaving room for breaths. Keep the lyrics concise and the delivery natural.

If the model supports a separate negative-prompt field, keep it focused on the scene's likely continuity problems:

Identity swaps, merged faces, duplicate performers, extra people, outfit
changes, extra microphones, handheld microphones, drifting equipment,
distorted hands, floating feet, background color changes, face-morph
transitions, captions, logos, watermark.

The audio direction calls for a new duet. The theme does not supply the original 58th Floor recording, lyrics or artist voices. If you choose silent output, omit the audio instructions and plan any soundtrack separately.

Figure 4. The wide ending gives the sequence a visible change in scale. These illustrations come from AIRAP's existing theme cover and supplied example video; they are not a new generation made with the sample prompt above.

6. Review the result against the brief.

Watch once for identity and continuity: does A remain A, does B remain B, and does the same microphone stay in place? Watch again for coverage: are the solo views distinct, and does the ending reveal both full bodies? Finally, listen for understandable words, a clear exchange and enough breathing room.

Use the observed problem to guide the next revision. A crowded vocal delivery calls for fewer words. A missing solo calls for clearer cropping instructions. If the ending remains waist-up, strengthen the full-body pullback instruction before changing the lighting or color palette.