rendirt

rendirt is a bare-bones C++ software rendering library for triangle meshes. The library is able to load STL models both in binary and ASCII format. In fact, offline rendering of STL model thumbnails is its primary use case. From this point of view, rendirt means render it.

But beware! rendirt also stands for dirty renderer. This thing is as simple as possible, quite inflexible and mostly unoptimized. Clocking in at ~400 LOCs, it does its (very limited) work in reasonable time and that’s all. This is not meant to be an example of state-of-the-art graphics programming. Decent speed is only achieved with compiler optimizations enabled. Still, the debug build manages to render ~400k tris at 800x600 px in less than 1 second and simpler models in less than 100 milliseconds. It becomes about one order of magnitude faster when compiler-optimized. (DISCLAIMER: those are not accurate measures, just average execution times to give an idea).

rendirt is distributed under the open source MIT license.

Example pictures

Image rendered with the depth shader — Depth shader

Image rendered with the position shader — Position shader

Image rendered with the normal shader — Normal shader

Image rendered with the diffuse directional lighting shader — Diffuse directional lighting shader

Usage

This is the bare minimum to obtain an image. For more details and variations look into the examples folder in the repository, or read the API reference.

namespace rd = rendirt;

std::ifstream file("/path/to/file.stl");
if (!file) {
    // Error handling
}

rd::Model model;
rd::Model::Error err = model.loadSTL(file);
if (err != rd::Model::Ok) {
    // Error handling
}

file.close();

std::vector<rd::Color> colorBuffer(800*600);
std::vector<float> depthBuffer(800*600);
rd::Image<rd::Color> image(colorBuffer.data(), 800, 600);
rd::Image<float> depth(depthBuffer.data(), 800, 600);

image.clear(Color(0, 0, 0, 255));
depth.clear(1.0f); // Important!

rd::Camera view(
    { 0.0f, 0.0f, 5.0f },
    { 0.0f, 0.0f, 0.0f },
    { 0.0f, 1.0f, 0.0f });

rd::Projection proj(
    rd::Projection::Perspective,
    glm::half_pi<float>(),
    width, height,
    0.1f, 100.0f);

size_t faceCount = rd::render(
    image, depth, model, proj * view,
    rd::shaders::position(model.boundingBox()));

// Do something with the image