0.1.5

[0.1.5] - 2025-01-20

Changed

• Updated pyimagecuda dependency to version 0.1.2
  • Includes automatic EXIF orientation handling for images loaded in Image nodes
  • Photos from mobile devices now display with correct orientation

0.1.4

[0.1.4] - 2026-01-14

Added

• HueNode: Node for hue adjustment in node-based workflows
• VibranceNode: Node for vibrance adjustment in node-based workflows
• ChromaKeyNode: Node for chroma key effect in node-based workflows

0.1.3

[0.1.3] - 2026-01-07

Added

• RotateNode now supports interpolation method selection (Nearest, Bilinear, Bicubic, Lanczos)
• ZoomNode for viewport-based zooming with center offset controls and interpolation support

Changed

• Preview render time now reflects actual GPU processing time (explicit sync enabled)
• Updated to use GLResource.copy_from(sync=True) for accurate timing measurements
• Refactored resize method selection to use centralized get_resize_function() utility

Performance

• Significantly faster resize operations thanks to pyimagecuda 0.1.0:
  • Lanczos resize: 2.7× faster
  • Bicubic resize: 1.1× faster
  • Bilinear resize: 1.09× faster
• Improved performance in ResizeNode, ScaleNode, AspectResizeNode, ImageNode, RotateNode, and ZoomNode

v0.1.1

v0.1.1 - Improved Buffer Management

This release updates PyImageCUDA Studio to use the improved buffer sizing system from PyImageCUDA 0.0.10, providing more flexible memory management and better buffer reuse.

Changes

Updated Buffer System

• Updated to PyImageCUDA 0.0.10 buffer capacity system
• Buffers now use pixel-based capacity instead of fixed dimensions
• Improved buffer reuse across different aspect ratios

Internal Improvements

• Simplified buffer management code in ensure_buffer() and resize_buffers()
• More robust dimension updates using Image.resize()
• Better handling of buffer capacity checks

Technical Details

The buffer management system now tracks total pixel capacity rather than individual width/height limits. This allows the same buffer to be reused for different aspect ratios as long as the total pixel count fits within capacity.

Example:

# Same buffer can handle different aspect ratios
buffer = Image(1920, 1080)  # 2,073,600 pixels capacity

buffer.resize(3840, 540)    # Landscape (same pixel count)
buffer.resize(2160, 960)    # Different aspect ratio
buffer.resize(1080, 1920)   # Portrait

Requirements

• PyImageCUDA >= 0.0.10

Migration

No changes required for existing projects. All .pics files remain fully compatible.

---

v0.1.0

Node-based image processing powered by CUDA - Design visually, execute programmatically