User Tools

Site Tools


groimp-platform:raytracing:raytracer_algo

Standard ray tracer

The Standard Raytracer option produces rather unnatural-looking images, as this algorithm ignores indirect lighting for diffuse materials and always calculates ideal reflections for reflective materials. However, the calculation with this option is quite fast.

Path tracer

PT is a ray tracing method, with the help of which a global illumination simulation can be carried out. Rendering with the Pathtracer option is qualitatively better compared to the standard ray tracer. The images generated here support diffuse reflections very well, but can be noisy. Its options are located in the grey zone with no grouping label near the bottom of the Twilight renderer settings. be noisy. The top field Pathtracing options contains the Number of paths property. This property is only evaluated if the Pathtracer MT has been selected as the ray tracing algorithm, and has a strong effect on the noise in the rendered image. The higher the value, the less noise the image contains, but the longer the calculation will take. The initial value of 25 is considered a good compromise. The Recursion depth property determines how often light in the scene is scene is reflected back and forth during the calculation. The value 0 would mean that all reflections and refractions are excluded. To obtain realistic images, it should be set to 6. The Priority can be used to determine the resource consumption during the calculation can be determined. The high priority option option, all available resources are used for the calculation. The option medium priority option offers the option of loading the operating system to a limited extent. With the low priority option, the calculation is regularly paused, allowing work in the operating system in parallel with the calculation is possible. In the <key>Antialiasing</key> method drop-down list, the user can select whether the image is to be is smoothed using super-sampling or whether no antialiasing is performed. The number of sampling passes can be controlled via the <key>Grid size of super sampling</key> field. It should be noted that the square of the input value will be used by the algorithm.

Photon mapping

The aim of photon mapping is to determine the global global illumination of a scene efficiently and thus to create realistic images in less time. It can be divided into two phases. In the first phase, photons are sent into the scene from a light source. With every interaction with a diffuse surface, an entry is stored in a photon map. This map is then used to calculate the illumination of the scene.

There are four important parameters that influence the quality of an image rendered using photon mapping: <key>Global photon count</key> determines how many photons are to be emitted into the scene. The higher this value, the more precisely the scene is illuminated. For scenes where only the caustic effects are of interest, it can be advantageous to set this value to 0. <key>Caustic photon count</key> defines how many caustic photons are to be emitted into the scene. Caustic photons are only saved in the photon map when the first interaction with a reflective or transparent object takes place. As the name suggests, these photons are responsible for caustic effects in the rendered image. An increase in this value ensures stronger and more detailed visibility of caustic effects. For a scene with only diffuse surfaces, this value can be set to 0 to save computing time. <key>Photon area</key> determines the size of the area that is scanned for photons when rendering the image. The higher this value, the brighter the scene is illuminated. A lower value ensures more precise lighting. Negative values are not possible. <key>Recursion depth</key> determines the recursion depth at which the calculation of the photons is started.

Bidirecitonal path tracer

BTP is a path tracing process in which ray paths are generated by the camera as well as by all light sources in the scene and are linked together. BPT in GroIMP is controlled via 3 parameters. The <key>Max. eye path depth</key> and <key>Max. light path depth</key> parameters can be used to set the maximum number of vertexes on the eye path and light path respectively. A light path depth of 1 can be used to simulate the operation of a normal PT. Conversely, with an eye path depth of 1, only caustic events are displayed. Values smaller than 1 are not permitted. The parameter <key>Heuristic exponent</key> controls the calculation of the weighting factor w_st, reference the Steidelmüller 2008 report for further details.

Radiosity

The radiosity algorithm was developed in 1984 by Goral, Torrance, Greenberg and Battaile. It is a global illumination method that is ideal for diffuse reflecting surfaces. This corresponds well to reality, since diffuse reflections are often encountered in practice, whereas completely reflective surfaces tend to be the exception. The basis for this algorithm is the law of conservation of energy. The aim is to achieve a balance between the radiant energy supplied by light sources and the radiant energy absorbed by all surfaces. This is achieved by calculating the specific illuminance for each surface. There are four important parameters to influence the quality of an image rendered using radiosity: <key>Hemicube size in pixel</key> determines the number of hemicube pixels of which a Hemi-Cube consists. The higher this value, the more precisely the light falling on a triangle is calculated. <key>Hemicube size in world coordinates</key> defines how large a Hemi-Cube is in the coordinate system of the scene. <key>Subdivision threshold</key>: After each calculation step, the color values of adjacent triangles are compared. If the difference is greater than this threshold, the triangles are subdivided for the subsequent calculation step. A lower value improves the color gradations between adjacent triangles, but increases the calculation time and memory consumption. <key>Subdivision depth</key> defines how many calculation steps are to be performed. The higher the value, the more precisely the image is calculated. However, this also increases the calculation time.

Metropolis light transport

MLT is a path tracing process in which randomly distributed initial paths on the image surface are mutated according to special rules until all image pixels are covered and the desired accuracy and mutation rate are achieved. The MLT raytracer in GroIMP is controlled via a separate parameter box in the raytracing options. The user has the option of selecting the mutation strategies to be used via checkboxes. They can also specify the number of initial paths in the <key>Count of seed paths</key> field. The most decisive parameter for the quality of the final image is the <key>Count of mutations per pixel</key> field, which determines the average target number of mutations per image pixel. It is important to note that the renderer tries to achieve this average value for all pixels of the image. For scenes that are mostly empty (e.g. an object floating freely in space without any surrounding elements), the pixels highlighted with an object are frequented correspondingly more frequently by the algorithm. As the bidirectional ray tracer is used in advance, its parameters also determine the length of the paths also determine those of the MLT raytracer.

See more:

Tutorials:

groimp-platform/raytracing/raytracer_algo.txt · Last modified: 2024/06/18 12:32 by thomas