Keep the roles straight
Put the listener in Headphone wearer and the speaker in Elevator companion. The order follows what the characters do, so you can ignore which side of your original pictures each person is on. Pick references where both faces are visible without headphones or hands covering the eyes.
Read the exchange
Watch the preview until the listener moves the headphones and turns. That movement connects the companion’s comment to the listener’s reaction. The studio takes the timing and audio from this reference, so typing new dialogue is unnecessary.
Review each close view
After generation, pause on the headphone adjustment and then on the companion’s face. Compare each person with their own source photo. If they have similar hair or clothes, use the eyes, brow and jawline to check the assignment.
Reviewed against the EasyVideoAI studio on November 8, 2026. Check the live studio for current availability and the final credit quote.