58TH FLOOR / FIELD GUIDE

How to Make a 58th Floor AI Video From Two Photos

Choose your two performers, assign the left and right roles, and understand the steps from photo pairing to reviewing a studio duet.

Jump to the step-by-step guide

Quick answer

The short version

Start with two separate portraits. A takes the left role; B takes the right role and the solo close-up. Choose a one-time pack if needed, then explicitly start your video. Track that task through generation, soundtrack preparation and download.

Explore the studio

Before you start

  • Two sharp portraits, with one person in each
  • Permission from both adults to use their photos
  • A modern browser on your phone or computer
  • A one-time video pack or an existing available video balance

A takes the left role. B takes the right role and the solo close-up.

Start with the two people, not a new script

Pick the two people you want to see at the microphone. Their portraits can come from different places and different days. A clean, recognizable face matters more than matching backgrounds.

You do not need to film the performance, find a video-editing template or describe the gestures in a prompt. This preset already has its studio scene and choreography. Use a portrait of yourself or an adult who has agreed to this use.

Assign A to the left and B to the close-up

Choose Photo A for the left performer. Choose Photo B for the right performer, who also appears in the solo close-up. Check both previews before continuing.

The assignment follows the character throughout the video, not whichever face happens to be nearest the camera. If you want the other person in the close-up, swap A and B before starting. Selecting photos only prepares your cast. The paid generation starts after you explicitly click Generate with available video entitlement.

Photo uploadIllustration of the video creation workflow.
Choose your two performers

Review your selection before starting a paid take

Choose a one-time pack: $6.99 for one video, $17.99 for three or $24.99 for five. There is no subscription, and unused videos do not expire. Review the final checkout amount before paying.

Continue as a browser guest or use Google sign-in. Guest and account wallets are separate. Your selected photos are kept in a short-lived browser draft during checkout or sign-in. On return, confirm both photos and explicitly click Generate; payment alone never starts a video.

Let the reference lead the performance

The photos define the new performers. The video reference supplies each role’s gestures, facial expressions, gaze, mouth movements and interactions. It also supplies the teal scene, suspended microphone, lighting and camera cuts.

The intended result follows the reference through a two-person shot, a close-up of B, a full-body view and a return to the pair. The model generates new frames, so it may vary facial detail, motion or timing. This is not a promise of a pixel-exact edit.

One-time video packs$6.991 video(s) included

Live prices come from the current checkout configuration; you can check the pack terms before purchasing.

See One-Time Video Pricing ↗
video generationIllustration of the video creation workflow.

Review the whole duet before sharing

Watch the opening, B’s close-up and the return to the wide shot. Make sure the roles have not switched, both people remain recognizable, and hands and clothing stay coherent. Check the full clip rather than judging one frame.

The preset is designed for a 15-second vertical 9:16 MP4. Save a completed result while it is available. If a task fails or playback is unavailable, check that existing task before submitting a new paid generation.

The silent generated video is combined with the prepared soundtrack in your browser. If soundtrack preparation fails, retry that step without purchasing another generation. Private results are retained for 24 hours; download promptly. A confirmed failed generation restores its video entitlement. A connection error is not proof of failure: check the existing task first.

Choose your two performers

FAQ

Frequently Asked Questions

Do I need the original video file?

No. The preset supplies the reference. You choose two portraits and their A/B roles.

Can I use two selfies from different places?

Yes. Use one clear person per photo. The photo backgrounds do not become the studio setting.

Is a new video free to generate?

Public examples and guides are free to view. A new generation uses a paid video entitlement; selecting photos does not charge you.

Keep exploring the studio

The Best Photos for a 58th Floor AI DuetChoose sharp, well-lit portraits for both roles. Learn why framing, visible facial features and Photo B’s close-up matter.What Is the 58th Floor AI Video Format?Understand the two-person teal-studio format, how A and B map to the reference, and what changes when new people take the roles.

58th Floor AI

Choose your two performers

58th Floor AI is an independent creative tool. Use photos with permission and review your result before sharing.

Explore the studio

If payment or delivery needs attention, use the support details and published refund policy.