V3.9 Update Details
New:
· Face Restoration: Introduced a new AI model specifically optimized for low-resolution and legacy portraits. This update delivers higher-fidelity skin textures and facial details (such as eyes and teeth) for more authentic restoration results.
· SDR to HDR Conversion: Now supports intelligently upscaling SDR 8-bit images to 10-bit HDR, delivering a visual output that more closely mirrors the real world. For optimal results, we recommend using this in conjunction with the Strength slider and Color features to fine-tune the dynamic range.
Optimized:
· Import Experience: Now supports Ctrl+C / Ctrl+V keyboard shortcuts for more efficient importing and adding of image assets.
Fixed:
· Resolved minor, intermittent bugs occurring in specific scenarios to ensure reliable software performance during high-load processing tasks.
V3.8 Update Details
Added:
· AI Eraser (experimental): Added AI Eraser to remove unwanted objects, photobombers, background distractions, undesired text, signs, and other elements.
Fixed:
· Resolved minor bugs occurring in specific scenarios.
V3.7 Update Details
Added:
· TIFF & DNG Output: Adds support for exporting images in TIFF and DNG formats, enabling higher bit-depth output and better compatibility with professional editing workflows.
· Restore Color: Adds an option to revert enhanced image to its original color profile, helping maintain color consistency after applying AI models. Includes a flexible slider (0-1) for fine-grained control based on your needs.
Fixed:
· Fixed occasional crashes caused by Intel GPU interference under specific conditions.
· Removed the "Keep the Prompt" feature previously used for AI-generated images.
V3.5 Update Details
Added:
· Strength Control: The new Strength slider lets you control the enhancement intensity and sharpness level of the AI model, helping you find the perfect balance between image clarity and natural appearance.
· Color Correction: Added color correction tools that let users fine-tune temperature, tint, exposure, contrast, saturation, and more to enhance color tone and overall visual appeal.
· RTX 50 TRT Support: Deeply optimized for the latest NVIDIA RTX 50 GPUs and TensorRT, unlocking stronger AI computing power and delivering faster, more efficient image processing performance.
Optimized :
· Faster Model Loading: Improved initial inference speed — saves around 12 seconds when face restoration is enabled (tested on i7-13700K CPU).
Fixed:
· Fixed potential crashes during inference on RTX 50 series GPUs.
· Fixed possible crashes caused by internal computation on certain AMD CPUs.
V3.3 Update Details
New Features:
· Added support for HEIC input format.
Performance Improvements:
· Optimized intelligent noise reduction during the AI upscaling process to enhance image quality.
· Optimized the multi-threaded processing engine, accelerating image preprocessing speed.
Bug Fixes:
· Fixed an issue where DPI metadata for PNG images was not stored correctly.
V3.1 Update Details
Added:
· Added support for 5 languages: Spanish, Italian, Korean, Portuguese, and French.
· Added a Single Export button.
Solved:
· Resolved several known issues.
V3.0 Update Details
Added:
· More-Detail GAN v3, an enhanced version of "More-Detail GAN v2," improving texture stability and generating more accurate details for a smoother, clearer, and more refined images.
· AIGCsmooth v3, a new AI model with improved support for AI-generated images, delivering more clarity, smoother and better quality.
· AI face restoration, featuring Gradient Overlay for smooth transitions and Routine Overlay for efficient processing, delivering improved image quality and more detailed and natural facial features.
· DPI output settings, providing flexible control over export resolution.
Improved:
· Optimized the color accuracy for more precise output images.
· Minimized checkerboard pattern artifacts in upscaling.
· Improved Export Settings, enabling file name customization for better file management.
