Browse by modality

Enhance

Restore, upscale, and transform existing media.

What you can do

Enhancement models process an existing image or video: they can increase resolution, reduce noise and compression artifacts, restore focus, interpolate frames, or adjust color and dynamic range. The best tool depends on the defect in the source.

Increase image resolution

Upscale a small image for a larger layout while balancing faithful reconstruction against generative detail.

Remove noise and compression

Clean photos, scans, and video affected by grain, low bitrate, block artifacts, or repeated exports.

Restore sharpness and focus

Use a model designed for lens blur, motion blur, soft faces, text, wildlife, or general sharpening.

Improve or transform video

Upscale footage, interpolate frame rate, deinterlace archives, reduce motion blur, colorize, or prepare SDR material for HDR delivery.

Inputs

  • An existing image or video file
  • The target scale, resolution, frame rate, or restoration operation
  • Strength, detail, creativity, and output controls offered by the selected model
  • A prompt only for enhancement workflows that explicitly use text guidance

Outputs

  • An enhanced image or video rather than a replacement for the original file
  • Higher resolution, cleaner detail, restored sharpness, or transformed color and motion
  • Results that depend on the source quality and the selected restoration model

How to choose a model

Diagnose the source before choosing a model. Upscaling, denoising, sharpening, frame interpolation, and creative restoration solve different problems and should not be treated as interchangeable filters.

Reviewed examples

The reviewed before-and-after image and video below make it easier to judge how enhancement changes real Omnipix outputs.

Limitations and trade-offs

  • No enhancer can reliably recover information that is completely absent from the source.
  • Generative upscaling may invent texture, facial features, lettering, or product details.
  • Excessive denoising or sharpening can produce plastic surfaces, halos, ringing, and unstable video detail.
  • Video processing is usually slower and more expensive than image enhancement because every frame must be handled consistently.

All available model families

Black Forest Labs

View details

Topaz Labs

View details

Topaz Color and Lighting

Topaz models for exposure recovery, white-balance correction, and context-aware colorization of monochrome images.

Topaz Frame Interpolation

Topaz Frame Interpolation synthesizes intermediate frames to change frame rate or create slow motion with Aion, Apollo, and Chronos.

Topaz Image Cleanup

Topaz cleanup workflows for background isolation, archival dust and scratch repair, and object-aware reconstruction.

Topaz Image Denoise

Topaz denoising from everyday low-light cleanup to severe noise removal and generative detail recovery.

Topaz Image Sharpen

Topaz sharpening models for lens blur, motion blur, portraits, wildlife, natural texture, and severe focus recovery.

Topaz Image Upscale

Topaz upscaling models specialized for photos, low-resolution sources, CGI, text, transparency, faces, and generative reconstruction.

Topaz Video Denoise

Topaz Nyx models reduce sensor noise and grain, from fast previews and strong cleanup to fidelity-preserving denoise-only processing.

Topaz Video Upscale

Topaz Video Upscale groups restoration, deinterlacing, face recovery, general enhancement, and diffusion-based models for different source defects.

Topaz Video Utilities

Topaz Video Utilities covers three separate operations: SDR-to-HDR conversion, motion deblur, and black-and-white video colorization.