Universal Image Viewer with HDR Support

Overview

Universal Image Viewer is a versatile tool designed for quick and efficient viewing of images across various formats using OpenImageIO and OpenCL. While it's engineered to handle a wide spectrum of image types, it boasts specialized features for high-dynamic-range (HDR) images.

Features

• Broad Format Support: Opens and displays images in multiple formats such as HDR, EXR, TIFF, PNG, JPG etc.
• Exposure Adjustment
• Gamma Correction
• Real-Time Performance: Optimized for speed, providing a smooth, real-time experience when applying adjustments and browsing through images.

Installation

Follow these instructions to set up the project on your local machine for development and testing purposes.

Prerequisites

• Git
• CMake (3.8 or higher)
• Python (3.9 or higher)
• Homebrew (for macOS users)
• vcpkg (for Windows users)

Windows Setup

1. Install vcpkg: We use vcpkg to manage our C++ dependencies. To install it, clone the vcpkg repository and run the bootstrap script as follows:

git clone https://github.com/Microsoft/vcpkg.git
.\vcpkg\bootstrap-vcpkg.bat

2. Integrate vcpkg with your system: Run the following command:

.\vcpkg\vcpkg integrate install

This makes the installed libraries globally available to your build systems.

3. Install dependencies: Use vcpkg to install the necessary packages:

.\vcpkg\vcpkg install openimageio:x64-windows pybind11:x64-windows opencl:x64-windows

4. Environment Variable: It is recommended to set an environment variable for VCPKG_ROOT:

setx VCPKG_ROOT <path_to_vcpkg_installation>

5. CMake Configuration: While setting up your project with CMake, add the vcpkg entry to your toolchain settings to make sure CMake uses the vcpkg toolchain:

cmake -E make_directory ./build
cmake ./cpp -B ./build -DCMAKE_BUILD_TYPE=Release -DCMAKE_TOOLCHAIN_FILE=${env:VCPKG_ROOT}/vcpkg/scripts/buildsystems/vcpkg.cmake
cmake --build . --config Release

macOS Setup

OpenCL: macOS comes with OpenCL pre-installed, but it's deprecated and will be phased out for Metal. Currently, we are constrained to use OpenCL 1.2 (as opposed to 3.0 on Windows).

Install dependencies: You can use Homebrew to install the necessary packages:

brew install openimageio
brew install pybind11
brew install opencl-clhpp-headers

CMake Configuration for Older macOS Versions: If you're building on an older version of macOS, ensure you set the deployment target for compatibility.

cmake -E make_directory ./build
cmake ./cpp -B ./build -DCMAKE_BUILD_TYPE=Release
cmake --build . --config Release

If you need your application to support older versions of macOS, you can specify the minimum OS version required for your project by adding a flag to your CMake configuration command. For example, to support macOS 11.6 or newer, you would use:

cmake .. -DCMAKE_OSX_DEPLOYMENT_TARGET=11.6

Python setup

For the Python part, you need to install dependencies via pip:

pip install -r requirements.txt

TBC