Only the first three images you upload will be used in the output video. This limitation provides better control over the final result. Use the best three images to generate the best possible video output from the AI

Avoid images where the model's hands are holding objects or in bent positions, as this can affect the natural movement in the video.

Include at least one back view of the model within your first three images to ensure comprehensive angle coverage by the AI when generating the video

Add one close-up shot in your initial three images to capture detailed features effectively. This is also helpful to capture various feature for the product for example graphics on a t-shirt

Select images where the model is in static poses. Avoid action shots like playing, exercising, or running, as these can result in unnatural AI-generated movements.

Use images featuring only the model without additional objects in the frame for optimal results. This reduces the chance of any interference impacting the final video quality.

Avoid demonstrating products with flowing garments like lehengas or skirts, as these may cause unwanted spinning animations and body morphing issues in the final video.