RTAO Quality
This page describes parameters under the "Quality" foldout that are specific to RTAO mode.
Resolution
Determines the resolution at which the effect is rendered. This parameter has the greatest impact on performance and visual output.
Full: 100% of pixels are rendered. Achieves maximum quality.
Half: 50% of pixels are rendered in a checkerboarded (non-temporal) pattern. A good balance between performance and quality.
Quarter: 25% of pixels are rendered. The most performant of the three modes.
Adaptive: Splits the screen into 2×2-pixel tiles and checks whether the geometry within each tile is uniform enough, by comparing the depth, normals and velocity of the pixels and occlusion values inside it. Tiles are then categorized into three groups, rendered at 100%, 50%, or 25% resolution. This can be previewed via the Shading Rate Visualization debug mode.
Performance impact: most severe, consider carefully
Ray Count
Specifies the number of rays launched into the scene to evaluate occlusion. This parameter represents the primary tradeoff between noise level and performance for this mode.
Performance impact: severe, use as few as possible
Sample Rebalancing [Adaptive Resolution]
Since Adaptive Resolution depends on frame content and motion, a large portion of the frame may occasionally require full (100%) resolution due to temporal disocclusion or high geometric complexity. In such cases, Sample Rebalancing stochastically reduces the ray count on random pixels to keep the total number of rays per frame as constant as possible, avoiding sudden performance spikes. Enabling this option may degrade overall AO quality in some cases, but provides a smoother framerate.
Performance impact: beneficial
Screen Space Pretrace
Activates screen-space ray marching that precedes hardware ray tracing. Its primary goal is to hide the mismatch between the rasterized scene and its ray-tracing counterpart (RTAS), by providing screen-space occlusion for objects that are not visible to hardware ray tracing for any reason. It can also save some performance if enough screen-space rays find their intersections in screen space.
Performance impact: none
Mask [Screen Space Pretrace]
Excludes objects from casting screen-space AO on a per-layer basis via the Rendering Layer Mask during Screen Space Pretrace. Most useful when you know which objects are certainly not visible to hardware ray tracing and can mark them through this mask, so screen-space rays target only those objects, leaving the rest to regular ray tracing.
Performance impact: none
Samples [Screen Space Pretrace]
Defines how many steps the screen-space ray marching algorithm is allowed to take across the depth buffer to find an intersection. The higher this value, the more defined the occlusion gathered by Pretrace — at the cost of performance.
Performance impact: moderate
Distance [Screen Space Pretrace]
Controls the maximum distance in meters that can be ray-marched in screen space. Samples are distributed along this distance, so the greater the distance, the more samples may be needed. If the Mask is not used, it is recommended to keep this value low to avoid marching long rays in screen space, which can cause view-dependent artifacts.
Performance impact: small
Thickness [Screen Space Pretrace]
Controls the virtual thickness of objects. Since a screen-space algorithm uses the depth buffer to find intersections, it has no way of knowing the true thickness of objects. Instead, an arbitrary parameter is used — the higher this value, the stronger the occlusion.
Performance impact: none
Last updated