This is a simple pathtracer written in C++ as part of our B.Tech. CSE Final Year Project. It has multithreading support for rendering the image.
This project is based on Peter Shirley's series on raytracing.
As we can see in the above image, the rendering process is using all available cores on the system and CPU utilization is 100%.
The rendering process for the whole scene took more than 5 hours when rendering it with 10,000 samples per pixel.
NOTE: Users can change main.cpp file and reduce the samples per pixel or image aspect ratio and height/width to render a lower quality image which will speed up the rendering process. This is ideal when one wants to test the renderer themselves.
This is the final image produced by the renderer.
$ cmake -B build
$ cmake --build build
On Windows, you can build either debug (the default) or release (the optimized version). To specify this, use the --config <debug|release> option.
On Linux or OSX, from the terminal, run like this:
$ build/raytracer
On Windows, run like this(Use Powershell):
.\build\Debug\raytracer.exe
or, run the optimized version (if you've built with --config release):
.\build\Release\raytracer.exe
When running the project it will ask the users:
How many threads you want to use for rendering? (Default: Use All CPU Cores If No Input Provided) :
Enter the number of cores you want to use for rendering or leave it blank and press enter to use all the available cores on your system.
After running the above command a new output.ppm file will be generated in the base folder which can be opened with any image viewer on Linux.
On Windows the image can be opened by using any of the following applications:
1. GIMP
2. Nomacs
3. Adobe Photoshop
- ThreadPool.h - For Multithreading Support (This is a simple C++11 Thread Pool implementation by log4cplus hosted on github at log4cplus/ThreadPool).
- stb_image.h - For image manipulation (Hosted on github at nothings/stb).