Super Focus 3

Generative blur recovery with selective refinement

Model Overview

Super Focus 3 is a next-generation model designed to recover detail from severely blurred or out-of-focus images when traditional sharpening methods fall short.

It selectively refines only the blurry regions of an image, intelligently preserving already-sharp details while handling variable blur with precision. This results in more natural reconstruction of detail, edges, and textures, with improved fidelity over Super Focus 2, especially on faces and skin.

Key Capabilities

• Selectively enhances blurred areas while preserving sharp regions, handling variable blur with precision.

• Produces more natural detail and improved realism, particularly on faces and skin.

• Can be applied multiple times for stronger recovery when needed.

• Powered by NeuroStream for local performance across a wide range of hardware.

Model Pricing

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

Getting Started

Endpoint

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

  • model — "Super Focus V3"

Endpoint-Specific Parameters

• denoiseStrength — decimal between 0 and 1, default 0.5

  • Removes noise from the image.

• faceEnhancement — boolean

  • Leverages our face recovery model to enhance faces.

• faceEnhancementStrength — decimal between 0 and 1

  • Sets the strength of the face recovery.

• faceEnhancementCreativity — decimal between 0 and 1

  • Choose realistic or creative face recovery.

• subjectDetection — "All", "Foreground", or "Background"

  • Choose where enhancements are applied.

These parameters should be appended as key-value fields in the form-data request bodies (i.e. alongside the model, outputHeight, and faceEnhancement etc. fields). Any parameters not explicitly provided are automatically configured by our auto-parameter model. Extra parameters provided that are not supported are ignored.

Model-Specific Parameters

• No parameters

Resolution Limits

• input — 256 MP

• output — 256 MP

Examples

Click to view

Click to view