Noise-Aware Sharpen

One-click sharpening that separates detail from noise

This model was formerly known as "Auto Sharpen".

Model Overview

Noise-Aware Sharpen is an intelligent automatic sharpening model that uses AI-driven image analysis to evaluate both image quality and scene content before applying sharpening. Unlike traditional sharpening tools that enhance detail and noise equally, the model automatically separates true image detail from digital noise and compression artifacts before intelligently restoring clarity. The result is sharper textures, facial features, fur, and fine edges while preserving the natural character of the original image.

By avoiding aggressive edge enhancement, Noise-Aware Sharpen reduces halos, oversharpening, and harsh digital artifacts commonly introduced by conventional sharpening filters. This model is especially effective for low-light, high ISO, wildlife, sports, nighttime, or grain-heavy photography where visible grain or noise is already present or intentionally part of the atmosphere of the shot. It performs best on high-resolution and production-quality imagery where preserving realism and fine detail is critical.

Key Capabilities

• One-click intelligent sharpening with no manual tuning required

• Separates image detail from noise before sharpening

• Preserves natural texture and realistic image character

• Reduces halos, oversharpening, and sharpening artifacts

Model Pricing

Output MP	1	4	8	16	24	32	40	50	64	100
Credits	1	1	1	1	1	2	2	3	3	5

Getting Started

Endpoint

• https://api.topazlabs.com/image/v1/sharpen/async

  • model — "Auto Sharpen"

In order for "Auto Sharpen" to be called, all parameters below must be left as null. This is recommended for batch image processing with various image subject types.

Model-Specific Parameters

• preset — 'normal', 'strong'

  • Controls the strength of the sharpening 'normal' for highest fidelity, 'strong' for extra sharpening to be applied and the most visible results. Defaults to 'normal'. Recommended: omit for automatic parameter selection.

• denoise_strength — (optional) decimal between 0 and 1, default null

  • Amount of noise to remove from the image (for noisy input images, for non-noisy omit). Recommended: omit for automatic parameter selection.

• opacity — (optional) decimal between 0 and 1, default null

  • Controls blending with the original, Recommended: omit for automatic parameter selection, decrease opacity on very strong sharpening levels (outside the recommended range).

Resolution Limits

• input — 512 MP

• output — 512 MP

Examples

Click to view

Click to view

Click to view