Updated 10 February 2025<\/strong>: Topaz Labs has released Gigapixel 8.2<\/a>.<\/p>\n The update adds the option to upscale images based on their longest side, making it possible to batch process images with varying aspect ratios.<\/p>\n It is also now possible to batch upload and batch process images in the cloud, as shown above.<\/p>\n Topaz Labs has also improved performance of the generative AI-based Redefine model, particularly on macOS, although Mac users are still reporting issues in the release thread<\/a>.<\/p>\n The previous update, Gigapixel 8.1<\/a>, added the option to order physical prints of images from within Gigapixel, via a partnership with photo printing firm WhiteWall<\/a>. <\/p>\n Price and system requirements<\/strong> The software is OpenGL-based, and supports AMD, Apple Silicon, Intel and NVIDIA GPUs.<\/p>\n New licences of the standard edition cost $99. The new Gigapixel Pro<\/a> edition for studios is subscription only and costs $499\/year.<\/p>\n Cloud render credits cost $5 for 20 credits. Credit subscriptions start at $9.99\/month for 80 credits per month.<\/p>\n Read a full list of new features in Gigapixel 8.0 to 8.2 in the online release notes<\/a><\/p>\n Have your say on this story by following CG Channel on Facebook<\/a>, Instagram<\/a> and X (formerly Twitter)<\/a>. As well as being able to comment on stories, followers of our social media accounts can see videos we don\u2019t post on the site itself, including making-ofs for the latest VFX movies, animations, games cinematics and motion graphics projects.<\/em><\/p>\n<\/p><\/div>\n
Gigapixel is available as a standalone application for Windows 10+ and macOS 11.0+; and as a plugin for Photoshop and Lightroom Classic. <\/p>\n
![\"\"](\"https:\/\/www.cgchannel.com\/wp-content\/uploads\/2024\/10\/241028_Gigapixel8_f.jpg\"){kind=tracking}
<\/p>\n
Gigapixel 8\u2019s new Redefine model uses generative AI to fill in missing details in images.<\/em><\/p>\n Originally posted on 28 October 2024. Scroll down for details of the Gigapixel 8.2 update<\/a>.<\/strong><\/p>\n Topaz Labs has released Gigapixel 8<\/a>, the latest major version of its AI-trained software for up-resing photos or rendered images<\/a> without softening them or introducing noise.<\/p>\n Version 8.0 adds a new generative AI-based Redefine model for sharpening and stylizing images, and updates the software\u2019s Face Recovery model.<\/p>\n It is also now possible to process images in the cloud, as well as on your local machine, with cloud processing charged on a credit basis.<\/p>\n Machine-learning-trained tool enlarges images without sacrificing detail<\/strong> Its algorithms are trained using a data set of \u201cmillions of images\u201d, with Topaz Labs claiming that it can enlarge images up to 16x before image quality starts to drop visibly.<\/p>\n Although it\u2019s marketed primarily at photographers, it can also be used on rendered imagery, and has had a dedicated AI mode for CG images since version 5.3<\/a>.<\/p>\n As well as upscaling low-res renders to generate high-res images more quickly, users can remove noise from renders by upscaling them, then reducing them to their original size.<\/p>\n Text prompts can be used to guide the output of Gigapixel\u2019s new generative AI-based Redefine model, particularly when using it at high Creativity settings.<\/em><\/p>\n Gigapixel 8.0: new generative AI-based Redefine model<\/strong> The software\u2019s new Redefine model uses generative AI to fill in missing detail when upscaling images to very high resolutions, or repairing damaged images.<\/p>\n It is also \u201csurprisingly good at denoising and sharpening in some scenarios\u201d.<\/p>\n The model comes with a Creativity setting, which determines to what extent the image is stylized: low values result in realistic sharpening, while higher values produce progressively more and more stylized results.<\/p>\n For high Creativity settings, users can also enter text prompts to guide the content of the AI details generated, in a similar way to a conventional generative AI tool.<\/p>\n The Redefine model currently runs on the GPU, and there are some known issues with AMD GPUs and on macOS: you can find more details in the release notes<\/a>.<\/p>\n Updates to Face Recovery<\/strong> Gen 2 also results in \u201ccleaner teeth and eyeglasses\u201d, and performs better when recovering \u201cobscure faces with little-to-no facial attributes\u201d.<\/p>\n Option to render images in the cloud, as well as on a local machine<\/strong> Cloud rendering is available when using the generative AI models<\/a>: both Redefine and Recover, which was introduced in Gigapixel 7.1 for upscaling very low-res images.<\/p>\n
Gigapixel is designed to enlarge digital images while preserving their sharpness, using machine learning techniques to fill in the missing details.<\/p>\n![\"\"](\"https:\/\/www.cgchannel.com\/wp-content\/uploads\/2024\/10\/241024_Gigapixel8_Redefine_TextPrompt.jpg\"){kind=tracking}
<\/p>\n
Topaz Labs describes Gigapixel 8.0 as the \u201cfirst upscaler using generative AI models to create high-quality photorealistic results\u201d.<\/p>\n
Gigapixel\u2019s existing Face Recovery model has been updated, with the Gen 2 model improving performance when upscaling faces seen in profile.<\/p>\n
The other major change in Gigapixel 8 is the option to process images in the cloud as well as on your local machine.<\/p>\n