What is Image-to-Image?
Image-to-Image (img2img) takes an existing image and transforms it based on your prompt. Unlike text-to-image which creates from scratch, img2img uses your photo as a structural guide — preserving composition while changing style, details, or content.
The Denoise Strength Secret
Denoise strength (0.0-1.0) is the most important setting in img2img. It controls how much the AI changes your image:
- 0.2-0.4: Subtle changes. Enhances details, adjusts colors. Good for: photo enhancement, minor fixes.
- 0.5-0.6: Moderate transformation. Keeps structure, changes style. Good for: artistic filters, style transfer.
- 0.7-0.8: Major transformation. Loosely follows the original. Good for: complete restyling, artistic reinterpretation.
- 0.9-1.0: Almost a new image. Only vague composition preserved. Rarely useful.
Start at 0.5 and adjust. This is the most reliable starting point.
5 Practical Use Cases
- Photo → Painting: Upload a photo, set denoise 0.6, prompt "oil painting, thick brushstrokes, warm colors". Instant art.
- Sketch → Finished Art: Draw a rough sketch, set denoise 0.7, prompt your desired style. The AI fills in the details.
- Fix Bad Composition: Take a decent photo that needs tweaking. Low denoise (0.3) + specific prompt to fix issues.
- Seasonal Variants: Product photo → transform to winter/summer/holiday scene. Denoise 0.5-0.6.
- Style Consistency: Transform multiple photos with the same prompt to create a consistent series.
Common Mistakes
- Denoise too high: The AI ignores your image entirely. Start low, increase gradually.
- Wrong resolution: Use the same aspect ratio as your input image.
- Vague prompts: Be specific about what you want to change, not just "make it better".