Back
Gigapixel AI
Sharpen AI
DeNoise AI
Video Enhance AI

Gigapixel AI v6.3 - Updated upscaling models, better artifact removal control, and lots of performance/stability improvements

AUTHOR:
Brian Matiash
Published:
January 12, 2023
Time to read:
13 Minutes

Introduction

Gigapixel AI v6.3 introduces a significant update to the Standard model that adds more natural-looking detail across a wider array of upscaled images suffering from blur and noise, and a new High Quality model that delivers improved quality when upscaling images that are already high quality.

You’ll also be able to control how much detail is retained with the new “Fix Compression” slider. And in addition to several important bug fixes and faster model load times, users with NVIDIA and Intel GPUs will experience greater stability when using Gigapixel AI.

Update to the latest version [ macOS | Windows ] of Gigapixel AI to take advantage of these helpful new features.

Updated Standard model - More natural detail and better blur/noise correction

The updated Standard model provides much more natural detail enhancements when upscaling

The updated Standard model is an excellent starting point for most image upscaling needs because of improvements made in two key areas:

  1. The new model architecture, training strategy, and larger data set allow the Standard model to add more detail while upscaling photos. More importantly, the added detail looks more natural compared to the previous Standard model version.
  2. The model can deliver substantial upscaling improvements to a broader set of images. Previously, the Standard model did not perform well while upscaling images suffering from heavy blur and noise.

The updated Standard model delivers excellent results when upscaling images with low-to-high blur and low-to-medium noise.

New “High Quality” model - Improved upscaling for high-quality and AI-edited photos

The new High Quality (HQ) model is designed to upscale images that are already high quality, such as those taken with modern cameras. It is also ideal when upscaling images that have already been processed using other AI models, such as those in DeNoise AI and/or Sharpen AI.

You’d want to use the High Quality model instead of the Standard model when your image already has sufficient detail or quality before upscaling, or if it has already been processed using DeNoise AI and/or Sharpen AI.

That is because the High Quality model has been trained to be less aggressive with texture enhancement while upscaling when compared to other models. As a result, edge details should look more natural, and areas including text will be less distorted.

New “Fix Compression” slider - Greater control over compression artifact removal

The new Fix Compression slider provides additional control over removing compression artifacts and is available when upscaling using either the High Quality (HQ) or Standard models.

The Fix Compression slider also pairs well with the “Suppress Noise” slider: larger compression artifacts, such as blocky areas, can be mitigated using the Fix Artifacts slider, while smaller artifacts, such as noise, can be mitigated using the Suppress Noise slider. Users will also experience greater sensitivity when adjusting the parameters of the other sliders in the app.

Performance and stability improvements - faster model loading and more reliable GPU support

Mac users will experience faster model load times after updating Gigapixel AI because the app now pre-compiles them. We reduced crashes and improved stability for computers with Intel GPUs, ensuring that they are used correctly when selected. We also fixed Face Recovery for users with certain NVIDIA GPUs in addition to numerous other bug fixes and usability improvements found throughout the app.

Try Gigapixel AI for free today

If you haven’t tried Gigapixel AI yet, be sure to download your free trial to experience all of these significant improvements and see what exceptional AI image upscaling looks like.

Tags
No items found.
AUTHOR:
Brian Matiash

Brian Matiash is a South Florida photo educator who serves enthusiast photographers looking to grow their landscape, travel, and wildlife skills. Learn more on his website and YouTube.