Sometimes you have the right image, but the wrong aspect ratio.
A portrait photo may look great on your phone but leave large empty spaces when used as a YouTube thumbnail, blog cover, presentation slide, or website banner. Cropping it to 16:9 often makes things worse—you may cut off the subject, background, or important details.
AI image expansion gives you another option.
Instead of cropping the original image, you can extend the canvas and let AI generate the missing areas around it. With a simple prompt, you can turn vertical, square, or narrow images into a natural-looking 16:9 image while keeping the main subject unchanged.
How to Extend an Image to 16:9 With XPT
With XPT, you don't need to manually resize the canvas, clone backgrounds, or use generative-fill tools layer by layer.
Upload your image, tell XPT to expand it to 16:9, and describe what should stay unchanged.
Step 1: Upload Your Image and Ask XPT to Extend It
Upload the image you want to expand.
Then describe exactly what you want XPT to do. A useful prompt should include three things:
- The target aspect ratio: 16:9
- The direction or area you want expanded
- The elements that should remain unchanged
For example:
| Use Case | Prompt |
|---|---|
| General image expansion | Extend this image to a 16:9 landscape format. Expand the background naturally on both sides. Keep the main subject, face, clothes, pose, lighting, and original visual style unchanged. |
| Portrait photo | Convert this portrait image into a 16:9 landscape image by extending the scene on the left and right. Keep the person exactly the same and generate a natural continuation of the existing background. |
| Outdoor photo | Extend this image horizontally to 16:9. Continue the sky, landscape, lighting, and surrounding environment naturally. Do not crop or modify the main subject. |
| Product photo | Expand this product image to 16:9 while keeping the product size, shape, color, position, and branding unchanged. Extend only the surrounding background. |
The more clearly you tell XPT what not to change, the easier it is to preserve the original image.
Step 2: Check the Expanded Areas
Once the new image is generated, check the edges that were added by AI.
Pay attention to:
- Repeated objects or patterns
- Distorted walls, furniture, or buildings
- Changes to the person's face or body
- Inconsistent shadows or lighting
- Unnatural transitions between the original and generated areas
If the image already looks natural, you can use it immediately.
If something looks wrong, you don't need to restart from scratch.
Step 3: Refine the Image With Another Prompt
Continue editing the same image with a more specific instruction.
For example: “Keep the person exactly the same. Regenerate only the extended area on the left. Add soft white sheer curtains to the glass area on the left, making them blend naturally with the existing room, lighting, and perspective. Keep the curtains light and slightly translucent, and make sure they do not cover or overlap the person. Preserve everything else unchanged.”
This is especially useful when the original image contains complex backgrounds such as streets, rooms, crowds, buildings, or detailed landscapes.
When Should You Extend an Image Instead of Cropping It?
AI image extension works best when cropping would remove something important.
Common use cases include:
- Turning portrait photos into 16:9 blog covers
- Creating YouTube thumbnails
- Resizing images for website hero sections
- Adapting square social images for presentations
- Turning vertical AI images into desktop wallpapers
- Expanding product photos for banners and ads
- Reformatting existing images for different platforms
Cropping is still useful when there is plenty of unnecessary space around the subject. But when you want to preserve the full original image, extending the canvas usually gives you more flexibility.
Extend Images Without Rebuilding Them From Scratch
You don't need to recreate an image just because its aspect ratio is wrong.
With XPT, you can upload the original image, ask AI to extend it to 16:9, and continue refining the generated areas with prompts. The key is simple: tell the AI what to expand and what must stay unchanged. This lets you reuse the same image across blogs, websites, presentations, thumbnails, and other landscape formats without cutting out the parts you want to keep.
