...
Micropolygon Lenth: Controls the tessellation rate for displacement. The default setting of 1 will dice displacements to 1 pixel's size. (As seen from the dicing camera, typically the render camera). Smaller numbers are finer geometric detail at the cost of creating more geometry.
Watertight Dicing: Closes pinholes in diced geometry or attempts to close gaps in displacement. Useful if you have holes in your solid object alpha.
Dicing Camera: Enables the use of a reference camera for dicing. This can be done globally, or restricted to specific objects (e.g. objects with displacements that "sparkle" as the camera moves) by also attaching this as a per-object attribute.
...
Min Hair Width: Sets a minimum width for hairs/RiCurves. Narrower hairs will be clamped to this value and their opacity will be reduced to compensate for the difference in visual weight. This may improve anti-aliasing quality if not too small.
Trace
Trace Auto Bias and Trace Bias: RenderMan will automatically compute a trace bias for displacing ray origins off a hit surface. However, if you wish to turn off the auto bias, you can set the bias manually here. Per-object settings should have Auto Bias turned off. Increase the value slowly until the artifact (self-shadowing, dark lines, etc.) disappears.
Trace World Origin: Move the world origin for increased floating point precision for raytracing. Sometimes when rendering very large scenes, the camera is moved to a position quite far from the origin. This may introduce precision issues depending on your scene units, values exported to the renderer are very large in these cases. Possible values are:
...
Top Left/Bottom Right: Specify the coordinates (x and Y) for a crop window to render. This can be useful when rendering many iterations for debugging or lookdev when only a specific region is under consideration. You may also use this to shift around the rendered region to tile together a very large render (possibly for print) and send the different regions to a different render node. Please note using crop windows with the PxrVCM renderer may generate different results than a full frame since we only generate photons for the region being rendered as an optimization.
IES Profiles
Scene units to meters: While scenes are exported to rib renderers "unitless", IES files specify a unit measure. Use this setting to specify the scene unit to meter conversion factor. For instance, if 1 unit = 1 meter, use 1. If 1 unit = 1cm use 100.
...
Allows users to customize the cache sizes for various assets, including textures, ray-traced geometry (specifically the tessellation cache), and Opacityopacity (as a function of Presence in materials). The values are in megabytes, and the renderer will automatically unload data from the cache when the limits specified here are reached. These values may be conservative based on your machine's available resources. If your render stats show many cache misses (reloading and re-reading) and your overall usage is below your available memory, you may want to increase these values to improve performance, keeping them reasonable and within your installed memory limitbudget.
Statistics
Level: Enable this to output diagnostic information from render jobs. Useful for timing, debugging, and seeing what RenderMan is doing. Higher numbers are more verbosity. Note there can be a performance impact that adds up by requiring the renderer to lookup these strings.
...
This section maps the specified lobes to an output. You will notice in the Light Path Expression documentation that per-lobe outputs rely on these being correct so we recommend leaving them at their default. You may use User Lobes in custom C++ BxDFs.
Example: You will note in the defaults that Diffuse 3 Lobe ID is Subsurface. The corresponding LPE is:
...
There are no D1, S1, or U1 lobes as these are "fallthrough" when something is unspecified. User 2 Lobe are the collective Albedo results (texture data).