Skip to content

Image Enhancement

Merserk edited this page Sep 3, 2026 · 1 revision

Image Enhancement

Use this to: clean up and enlarge photos, screenshots, scans, AI art. Works with one image or a batch.

Steps

  1. Open Image tab.
  2. Drop images in Input image(s). Drag to reorder if batch.
  3. Leave Neural Rendering settings on default for first try.
  4. Choose:
    • Upscaling factorDLAA = same size, cleaner. Quality / Balanced / Performance / Ultra = bigger.
    • Output formatPNG for best quality, JPEG for small files.
    • RenameAuto is fine (adds timestamp, never overwrites).
  5. Press Render image(s).
  6. Right side: preview → Rendered image files → download single or Download as ZIP.
  7. Check Batch results table: Complete vs Failed + reason.

Only first frame is used for animated / multi-page files.

Advanced
  • Lossy quality (1-100, typical 95) affects only JPEG / WebP / AVIF. PNG / TIFF ignore it.
  • JPEG can't keep transparency — it fills with white. Use PNG / WebP / TIFF if you need alpha.
  • RAW, HEIC/HEIF, SVG supported as input. Output is always standard PNG / JPEG / WebP / AVIF / TIFF.
  • EXIF orientation fixed, sRGB kept, metadata preserved where possible. Diagnostic JSON in logs/ explains warnings.
  • Strength sliders: NR Intensity / Local Tone / Local Structure (0-2), Skin Structure (-1 = native). Lower = safer, higher = stronger but can look artificial.
  • Automatic Mask and NR Preset are experimental hints — content-dependent, can flicker. Keep Off / Default unless testing.
  • DLSS Model Preset J/K/L/M forces a model for all modes. Default lets NVIDIA choose — recommended.

Related: Settings ExplainedFAQ & Troubleshooting

Clone this wiki locally