Physical Renderer Options
These settings are available if you check the Use Physical Renderer box in the Detailed view of PhotoRendering Settings for the CineRender Engine.
The Sampler does just that - it samples the Project’s cache from the angle of view of the camera in a pixel raster. The Sampler has to determine which regions have to be sampled again with regard to antialiasing or blur effects (depth of field, motion blur, blurriness effects) in order to achieve the desired render quality.
The samples gathered (i.e. colors), whose number far exceeds the number of pixels, will in the end be turned into pixels colors via an antialiasing filter.
The following generally applies: The more samples taken at critical regions, the better the final result will be (with correspondingly longer render times).
•Adaptive: Depending on the settings defined, an increased sampling will take place at critical or important regions. This mode is best for use in most cases because it offers a good compromise between quality and render speed (comparable with the default renderer’s Best Antialiasing setting).
The Adaptive sampler is a much more intelligent sampler compared to the Fixed sampler, and often renders faster than the Fixed sampler. On the other hand, if you choose a Sampling Quality of Custom, it’s more difficult to control because is calculates the following 4 values and distributes the samples accordingly:
-Shading Area Threshold
Special attention is paid to object edges, to which additional samples are applied to ensure that they are always antialiased optimally. The Adaptive sampler renders slightly less sharp than the Fixed mode.
•Fixed: A definable and unchangeable per-pixel sample number (which will, however, already be increased in blurriness effect regions) will be calculated (comparable to the default renderer’s Best and Min/Max Antialiasing settings).
This sampler subdivides each pixel in a defined number of sub-pixels, according to the value entered for the Sampling Subdivisions value. For example, to make the blurred region of a motion blur effect visually appealing, you should use a relatively high value (e.g., 4 or 5). This means that 16 and 25 samples, respectively, will be calculated per pixel even in superfluous regions.
Increasing the Sampling Subdivisions level by a single value normally doubles the required render time. Here you can estimate render times quite precisely, which on the other hand is quite difficult using the Adaptive sampler.
Render time is valuable. This is why there is an Adaptive mode that recognizes at which locations more sampling must be done.
•Progressive: This sampler can render infinitely (samples will continuously be calculated), during which time the result will continuously improve. The initially grainy image will become increasingly less grainy, i.e. more homogeneous. The longer it renders the better the antialiasing and blur effects that will be calculated.
The longer the render, the better the result
This mode is primarily suited for fast preview renderings because you can very quickly determine the quality of the image as a whole. No other method offers you such a fast rendition of the entire image (with reduced quality at the start). To this end it is recommended that the Sampling Subdivision value (see below) be set to 0.
Progressive Mode/Pass Count/Time Limit
Use these settings to define whether the render should never end (Infinite) or if it should end after a specified number of passes (Pass Count) or render time (Time Limit, in minutes). A pass in this context is defined as a completely new rendered result over the entire surface of the given image.
Note that when using Time Limit you should allow for enough time for at least 1 pass, otherwise it can result in faulty renderings if the pass is stopped halfway through the images.
This selection menu makes your work a little easier by automatically defining parameter combinations, depending on which setting is chosen. Low, Medium, High refer to the quality of the rendered image - the higher the setting the less grain the image will have but the longer it will take to render. Custom will automatically be selected when one of the parameter values is modified manually.
This setting functions somewhat differently depending on if you select Fixed or Adaptive.
•For a fixed number: The absolute number of samples that are calculated per pixel (depending on the subdivision values, additional samples will be calculated accordingly).
•For an adaptive number: A number of samples that are taken from the scene in order to gather information which will be calculated in a subsequent step (by the three shading parameters), i.e. if the values entered here are too low, greater values for the other parameters will not lead to good results.
The render results using the values shown for Sampling Subdivisions
Shading Min/Shading Max/Shading Error Threshold
The Sampling Subdivisions setting is applied and the number of shading samples required per pixel will be determined based on the shading parameters.
“Shading” is defined as the calculation of a color value based on shadows, textures, refractions, reflections, etc. This can be very complex and time-consuming process, depending on the scene.
The Shading (Min) value defines the minimum number of shading samples that will always be calculated (to the power of two, i.e. 1, 2, 4, 8, 16, etc.).
The Shading (Max) value defines the maximum number of shading samples, in correlation with the Shading Error Threshold value, that should be calculated.
The Shading Error Threshold value controls the number of samples that will in fact be calculated. The lower the value, the more the number of samples will lean towards Shading Subdivisions (Max) (however, only in critical regions!).
Subdivisions: Blurriness Max, Shadow Max, AO Max
The following three subdivision settings are the quality settings for certain effects (e.g. Ambient Occlusion) in the Transparency or Reflectance channels of the Surface material. The strength is defined at the surface level.
These values will be taken into consideration when Progressive mode is used (see Sampler). Higher values will result in a slower, more gradual rendering. For preview renders in this mode it is suggested that each of these values be set to 0.
Blurriness Max: If blurriness effects are defined for surfaces (Transparency and/or Reflectance channel), use this value to adjust the quality (=graininess) of the effect. Larger values will result in better quality but will also increase render times accordingly. Render times for a given effect will double each time this value is increased by a factor of 1.
More samples (right) result in better quality in matte reflections and transparencies
Shadow Max: This parameter refers to Shadow Maps (the Physical Renderer simulates Shadow Maps because it cannot calculate them). Larger values result in better quality but also increase render times accordingly.
AO Max: AO is short for Ambient Occlusion. Use this value to adjust the quality of Ambient Occlusion effects. Larger values result in better quality but also increase render times accordingly.
More samples (right) result in improved Ambient Occlusion quality
SSS stands for Subsurface Scattering Subdivision. Larger values will produce higher-quality results with correspondingly longer render times and vice versa.
This setting can be defined for surfaces via the Sampling Subdivision option (see Sampling Subdivisions).