Just days following the launch of Nano Banana Pro, Google's newest and most sophisticated AI model for image creation and editing, social media platforms are inundated with AI-altered, photorealistic content, further obscuring the distinction between genuine and synthetic media.
Nano Banana Pro, powered by the new Gemini 3 Pro AI, improves on its predecessor with enhanced multimodal reasoning and the ability to generate images featuring clearer, more accurate text in multiple languages.
While Nano Banana Pro has been making waves with a variety of viral use cases, photo restoration has emerged as the standout favorite. Here’s a closer look at the prompts users can try, based on industry-tested best practices. However, keep in mind that an effective prompt typically follows this structure: [Type of original photo], [Action required], [Specific defect removal], [Desired final quality/style]
When utilizing Nano Banana Pro to add color to black and white photographs, it's important to ensure that the image isn't simply overlaid with sepia tones. To prevent this, consider using these prompts:
Prompt 1: “Transform this black and white image by adding color. Use historically accurate and realistic hues for clothing, plants, and buildings. Ensure skin tones are natural and warm. The color palette should be vibrant yet genuine, resulting in a high-definition color photograph.”
Prompt 2: “Restore the photo by correcting extreme color shifts and fading. Neutralize any red or magenta tints, balance skin tones, enhance natural colors, and ensure the overall color palette looks realistic and true to life. Bring back the natural white balance and vivid, true-to-life colors to the setting. Eliminate grain. Utilize crisp, contemporary color grading.”
Apply this prompting method to restore a classic vintage photograph that is blurry, faded, or has minor scratches.
Prompt 1: “Revive this vintage photograph by eliminating scratches, dust, and creases. Restore faded colors, correct yellowing, sharpen details, reduce noise, and enhance textures to produce a high-resolution, crisp, and natural-looking image.”
Prompt 2: “Complete restoration of damaged old photo. Clean up paper texture, eliminate artifacts and blur. Bring out details in clothing and background. Balance exposure and contrast. Pristine, sharp, 4k quality photograph.”
Use this approach when the physical photo is severely compromised, that is, it has tears, water damage, or deep folds.
Prompt 1: “Eliminate deep surface scratches, mold marks, and emulsion damage across the image. Refine the paper grain texture while preserving details. Significantly reduce noise. Achieve a clear, restored image.”
Prompt 2: “Intensive repair of severely damaged photo. Reconstruct missing torn edges seamlessly. Remove deep fold lines across the center. Eliminate water stains and discoloration patches. Rebuild damaged areas while keeping the original subject intact. Clean, reconstructed photograph.”
Sometimes, AI models struggle to modify old faces, which can lead to them looking too smooth or resembling cartoons. In these cases, try using the sample prompt given below:
Prompt 1: “Revitalize an antique family portrait by concentrating on improving facial features, eyes, and skin texture in a realistic manner. Eliminate any motion blur from the faces. Avoid excessive smoothing of the skin. Preserve the genuine vintage ambiance of the background while ensuring the subjects are sharp and intricately detailed..”