[1.3.66] 2026-03-05

## Core
- Added `writeEXR()` functions for writing single-channel and multi-channel float images to OpenEXR files with lossless ZIP compression, using the tinyexr header-only library.

## Radiation
- Added `writeCameraImageDataEXR()` and `writeDepthImageDataEXR()` methods for exporting camera and depth data to EXR files, preserving full floating-point precision.

## Visualizer
- Fixed bug in colorbar ticks where ticks could extend past the colorbar.

## Radiation
- Added OptiX 8.1 ray tracing backend for NVIDIA systems with driver ≥ 560. This resolves the driver 590+ incompatibility that prevented the OptiX 6.5 backend from working on modern NVIDIA drivers. The backend is selected automatically at build time: driver ≥ 560 uses OptiX 8.1; driver < 560 continues to use OptiX 6.5.

Author: bnbailey-psl

core/CMakeLists.txt

@@ -45,6 +45,8 @@ add_dependencies( png_static zlibstatic )
 include_directories("${CMAKE_BINARY_DIR}/lib/libjpeg-9a" "${CMAKE_CURRENT_SOURCE_DIR}/lib/libjpeg-9a")
 add_subdirectory("${CMAKE_CURRENT_SOURCE_DIR}/lib/libjpeg-9a" "${CMAKE_BINARY_DIR}/lib/libjpeg-9a")
 
+include_directories("${CMAKE_CURRENT_SOURCE_DIR}/lib/tinyexr")
+
 target_link_libraries( helios PRIVATE png_static  jpeg ) #note that zlib is already linked by libpng
 
 # Suppress warnings for third-party libraries

core/include/global.h

@@ -878,6 +878,26 @@ namespace helios {
      */
     void writeJPEG(const std::string &filename, uint width, uint height, const std::vector<unsigned char> &pixel_data);
 
+    //! Write a single-channel float image to an EXR file with lossless ZIP compression
+    /**
+     * \param[in] filename Name of the EXR image file
+     * \param[in] width Image width in pixels
+     * \param[in] height Image height in pixels
+     * \param[in] pixel_data Float values at each pixel (index at pixel_data[row*width+column])
+     * \param[in] channel_name Name of the EXR channel (default "Y" for grayscale; use "Z" for depth)
+     */
+    void writeEXR(const std::string &filename, uint width, uint height, const std::vector<float> &pixel_data, const std::string &channel_name = "Y");
+
+    //! Write a multi-channel float image to an EXR file with lossless ZIP compression
+    /**
+     * \param[in] filename Name of the EXR image file
+     * \param[in] width Image width in pixels
+     * \param[in] height Image height in pixels
+     * \param[in] channel_data Vector of per-channel float data, each of length width*height
+     * \param[in] channel_names Names for each channel (e.g., {"R", "G", "B"}). Channels are sorted alphabetically per EXR convention.
+     */
+    void writeEXR(const std::string &filename, uint width, uint height, const std::vector<std::vector<float>> &channel_data, const std::vector<std::string> &channel_names);
+
     //! Template function to flatten a 2D vector into a 1D vector
     /**
      * \ingroup functions