My original image generation is in 16:9, and when I input another image of a character to put in the original image, with an aspect ratio of 1:1, the final output comes out to be 1:1 instead of the original 16:9.
Hello,
Welcome to the Forum!
To assist you better, I would need to reproduce this issue. Could you please share the reproduction steps and if possible DM me the images you are using?
Hello, could you please confirm which model you are using and share your prompt along with configurations?
The model is Gemini 2.5 flash (Nano banana). Temperature - 1. Top P - 0.95
The prompt is - Add the man in the image. Keep the aspect ratio 16:9.
Hope this helps.
It sounds like you’re encountering an issue with aspect ratio when combining images. When you input an image with a 1:1 aspect ratio into a 16:9 background, the final output will typically match the dimensions of the input image. To maintain the 16:9 aspect ratio, you might consider the following options:
Resize the Character Image:
Scale the 1:1 image to fit within the 16:9 canvas. This might involve cropping or adding padding to ensure it fits without distorting the original image ratio.
Canvas Adjustment:
Create a new 16:9 canvas first and then place the 1:1 character image onto it, adjusting its size and position as needed while keeping the background intact.
Layering:
If you’re using image editing software, layer the 1:1 character image over the 16:9 background to maintain the desired aspect ratio
Could you please share the original input images you provided to the model along with the full request payload?
The screenshot of inputs and responses is helpful, but having the exact input and payload would be of great help.