3D Render

Translate the text in the speech bubbles/titles from [Japanese/English] into [Chinese]. Seamless erase: wipe out the original text and perfectly fill in the background (for example, a white speech bubble or a colored image background).

Keep the translated text fitting naturally within the bubble without being too crowded.

Match the original image's lighting where the background is filled in.

Original image background, with erased areas seamlessly reconstructed.

Style matching: render the translated Chinese text using a casual handwritten-style font matching the original image aesthetic (or bold Impact font for memes).

Adaptation: ensure the text fits naturally within the bubble and is not overcrowded. Advanced play: change the directional light to side-backlight and the atmosphere immediately becomes more story-like.