The image renderer 102 may employ the casting determiner 108 to adapt the number of rays used to sample particular locations to various scene conditions, thereby reducing the likelihood of such visual artifacts while avoiding an increase to the number of rays used to sample every location in the scene. In particular, the criteria employed by the casting determiner 108 to determine the number of rays may be tailored to certain problematic locations and scene conditions used for a ray tracing effect. For example, FIG. 1 indicates that the ray tracing pass 114 in combination with the ray tracing pass 116 uses a variable number of rays to sample the lighting conditions for different locations with respect to the light source 120. By way of example and not limitation, three rays are used for the location 130, two rays are used for other locations—such as the locations 132 and 134—and only one ray is used for other locations.

The casting determiner 108 may determine for a given location(s) whether to cast one or more additional rays to sample the one or more aspects of the virtual environment (e.g., lighting conditions) and/or how many rays to cast to sample the one or more aspects of the virtual environment for the given location(s). The determinations made by the casting determiner 108 may be based on the results of analysis performed by the sample analyzer 106.

In embodiments, the casting determiner 108 may make such determinations for each location and/or pixel being rendered (and/or selected groups thereof). For example, the sample analyzer 106 may analyze results of the ray tracing pass 114 for each given location(s) and the casting determiner 108 may determine whether or not to cast one or more additional rays and/or how many additional rays to cast for the given location(s) based on the analysis. These rays may be cast in the ray tracing pass 116, and ray-traced samples from the ray tracing pass 116 may be used to update the results of the ray tracing pass 114 (e.g., by combining ray-traced samples for corresponding locations) to enhance the accuracy of the overall sampling of the virtual environment 110.

The sample analyzer 106 may, for example, analyze variables such as visibilities and/or hit distances of ray-traced samples of the light source 120 for related locations (and the location in some embodiments) in the virtual environment 110. More or different variables may be employed in various embodiment, which may depend upon the ray tracing effect being simulated. A visibility for a location, such as the location 130, may indicate whether the location is at least partially visible to the light source 120. A hit distance for a location, such as the location 130, may indicate a distance between the location and an object, such as the occluder 122 or the light source 120.

In various examples, a ray-traced sample determined for a location, such as the location 130, and for a ray, such as the ray 142, may include a visibility (e.g., visibility value) representative of whether the ray interacted with the light source 120. For example, a visibility value for the ray 142 may be a binary value in which a “1” indicates the ray hit the light source 120 and a “0” indicates the ray did not hit the light source 120. The ray-traced sample may also include a hit distance (e.g., hit distance value) representative of a hit distance between the location and the point at which the ray interacted with an object. For example, a hit distance value for the ray 142 may be a value that indicates a magnitude of the hit distance, where a larger value may indicate a greater distance than a smaller value (e.g., on a linear scale).

In various embodiments, visibilities and/or hit distances (and/or other variables) for locations in a region around the location (e.g., centered at the region) may be analyzed (e.g., within an N-by-N kernel centered at the location) to make the determination(s). For example, to make a determination for a given location and/or pixel, the region for that location and/or pixel may be analyzed. FIG. 2A-2D each show representations of different example combinations of visibilities and hit distances for a region and are used to describe how such differences may impact the determinations made by the casting determiner 108 for that region.

In the examples shown, the region includes locations 210, 212, 214, 216, 218, 220, 222, 224, and 226. Each location may, for example, correspond to a respective location in FIG. 1. However, the visibilities and hit distances in FIG. 2A-2D are not intended to necessarily match the depiction of FIG. 1 for particular locations. In embodiments, the sample analyzer 106 may analyze a group of the visibilities and the hit distances that correspond to a group of the locations based on proximities of the locations in the group to a given location(s). The casting determiner 108 may then make a determination for that given location(s). For example, for the location 130, the casting determiner 108 may define an N-by-N kernel centered at the location 130. In this example, the location 130 may correspond to the location 210 in FIGS. 2A-2D, and the N-by-N kernel may be a 3×3 kernel. The sample analyzer 106 then analyze one or more of the visibilities and/or hit distances that correspond to the locations within that region. The casting determiner 108 may use the results of the analysis for that region to determine whether to cast one or more additional rays to sample the one or more aspects of the virtual environment 110 (e.g., lighting conditions) and/or how many rays to cast to sample the one or more aspects of the virtual environment 110 for the location 130.

A similar approach may be used for each given location(s) and/or pixels. For example, the number of regions may, in some embodiments, match the number of pixels being rendered. Further, analysis of the regions may be performed in parallel, such as by a GPU. In some embodiments, the analysis performed by the sample analyzer 106 and/or the determinations made by the casting determiner 108 may at least partially be incorporated into the subsequent ray tracing pass (e.g., in corresponding executions of a dispatch ray function), such as the ray tracing pass 116. As further examples, the analysis performed by the sample analyzer 106 and/or the determinations made by the casting determiner 108 may at least partially be incorporated into one or more shaders between a subsequent ray tracing pass, such as the ray tracing pass 116. For example, a pixel shader and/or compute shader may be used for at least some of the computations.

In some examples, the group of visibilities and/or hit distances may be determined and/or defined using a different approach. Further, the shape of the kernel may not necessarily be square and/or the size of the kernel may be varied. In some examples, the size of the kernel may be determined based at least in part on a resolution of the virtual environment being rendered. For example, a 3-by-3 kernel may be used for a 4K resolution, a larger size, such as 5-by-5, may be used for a lower resolution, such as 640×480, and a smaller kernel may be used for a higher resolution (e.g., 2-by-2). It is also noted that the region may not be contiguous, may not include the given location(s) and/or may not include multiple locations in various embodiments.

Examples of factors used to make the determination(s) for the location 210 of FIGS. 2A-2D include those based on the variability in the visibilities and/or hit distances within the region. Additional examples of factors include those based on differences between one or more of the visibilities and/or hit distances relative to the location 210 and/or the overall region. Further examples of factors include those based on the magnitude of one or more of the hit distances (and/or visibilities in some embodiments).

Referring now to FIG. 2A, FIG. 2A is a diagram illustrating an example of ray-traced samples for a region in a virtual environment that has full visibility with respect to a light source, in accordance with some embodiments of the present disclosure. For example, the white shading of visibilities 200A in FIG. 2A may indicate that each of the locations 210, 212, 214, 216, 218, 220, 222, 224, and 226 has full visibility with respect to the light source 120 in FIG. 1. In this case, the rays cast from those locations may each have hit the light source 120 in the ray tracing pass 114. In this example, the sample analyzer 106 may analyze the visibilities 200A and based on the analysis, the casting determiner 108 may determine to not cast any additional rays for the location 210. Such an analysis may be based on determining the visibilities 200A in the group do not have at least a threshold amount of variability. For example, the sample analyzer 106 may determine whether at least one of the visibility values corresponding to the visibilities 200A is different than at least one other of the visibility values. If this threshold amount of variability is met, the casting determiner 108 may determine to cast at least one additional ray for the location 210. Otherwise, no additional ray may be cast for the location 210 (e.g., regardless of hit distances 202A for the region). In other examples, a more complex analysis and/or different threshold amount of variability may be employed.

Referring now to FIG. 2B, FIG. 2B is a diagram illustrating an example of ray-traced samples for a region in a virtual environment that has no visibility with respect to a light source, in accordance with some embodiments of the present disclosure. For example, the dark shading of visibilities 200B in FIG. 2B may indicate that each of the locations 210, 212, 214, 216, 218, 220, 222, 224, and 226 has no visibility with respect to the light source 120 in FIG. 1. In this case, the rays cast from those locations may each have hit the occluder 122 and/or other object instead of the light source 120 in the ray tracing pass 114. In this example, the sample analyzer 106 may analyze the visibilities 200B and based on the analysis, the casting determiner 108 may also determine to not cast any additional rays for the location 210. As in FIG. 2A, such an analysis may be based on determining the visibilities 200B in the group do not have at least a threshold amount of variability. For example, the sample analyzer 106 may determine whether at least one of the visibility values corresponding to the visibilities 200B is different than at least one other of the visibility values. If this threshold amount of variability is met, the casting determiner 108 may determine to cast at least one additional ray for the location 210. Otherwise, no additional ray may be cast for the location 210 (e.g., regardless of hit distances 202B for the region).

Referring now to FIG. 2C, FIG. 2C is a diagram illustrating an example of ray-traced samples for a region in a virtual environment that has mixed visibility and high hit distances with respect to a light source, in accordance with some embodiments of the present disclosure. For example, the white shading of visibilities 200C in FIG. 2C may indicate that each of the locations 222, 224, and 226 has full visibility with respect to the light source 120 in FIG. 1 and the dark shading of visibilities 200C in FIG. 2C may indicate that each of the locations 210, 212, 214, 216, 218, and 220 has no visibility with respect to the light source 120 in FIG. 1.

In this example, the sample analyzer 106 may analyze the visibilities 200C and based on the analysis, the casting determiner 108 may determine to cast at least one additional ray for the location 210 (e.g., one additional ray). Such an analysis may be based on determining the visibilities 200C in the group do have at least a threshold amount of variability. For example, the sample analyzer 106 may determine whether at least one of the visibility values corresponding to the visibilities 200C is different than at least one other of the visibility values. If this threshold amount of variability is met (as it is in FIG. 2C), the casting determiner 108 may determine to cast at least one additional ray for the location 210. As with FIGS. 2A and 2B, in other examples, a more complex analysis and/or different threshold amount of variability may be employed.

With regard to hit distances, the dark shading of hit distances 202C in FIG. 2C may indicate that each of the locations 210, 212, 214, 216, 218, and 220 has a relatively high hit distance with respect to an occluder. The white shading of the hit distances 202C in FIG. 2C may indicate that each of the locations 222, 224, and 226 has a relatively low hit distance with respect to an occluder (e.g., a minimum hit distance value as those rays did not hit an occluder). For example, the sample analyzer 106 may analyze the hit distances 202C and based on the analysis, the casting determiner 108 may determine not to cast at least one additional ray for the location 210 (e.g., in addition to the ray that will be cast based on the visibility).

Such an analysis may be based on determining the hit distances 202C in the group do not have at least a threshold amount of variability. In some embodiments, the sample analyzer 106 may determine the threshold amount of variability of the hit distances 202C with respect to a hit distance of the location 210. This may involve the sample analyzer 106 comparing the hit distances of the location 210 to the hit distances for the other location(s), such as to determine whether one or more of the location(s) are within a delta (threshold amount) of one another (e.g., sufficiently similar). In some examples, if one or more of the locations are not sufficiently similar (e.g., at least one location exceeds the delta) the casting determiner 108 may cast an additional ray. However, if each of the one or more of the locations are sufficiently similar (e.g., no location exceeds the delta) the casting determiner 108 may not cast an additional ray, as in FIG. 2C. In this example, the similarities may be considered with respect to those rays that also hit an occluder, which may not meet this criteria for an additional ray.

In addition to (or alternatively from) the casting determiner 108 basing determinations on whether the threshold amount of variability of the hit distances 202C with respect to a hit distance of the location 210 is met, the determination(s) may be based at least in part on determining at least one of the hit distances 202C in the group is less than a threshold value. For example, the sample analyzer 106 may analyze the hit distances for each of the locations of the region. If there are no hit distances for the locations that are below the threshold value (indicating a close hit or short hit distance), the casting determiner 108 may determine not to cast an additional ray for the location 210 (e.g., although a ray may still be cast based on a different determination). Otherwise if at least one of the hit distances are below the threshold value, the casting determiner 108 may determine to cast one or more additional rays (e.g., a single additional ray) for the location 210. As most of the hit distances are quite high in FIG. 2C, the casting determiner 108 may determine not to cast an additional ray based on the hit distances 202C.

Referring now to FIG. 2D, FIG. 2D is a diagram illustrating an example of ray-traced samples for a region in a virtual environment that has mixed visibility and includes low hit distances with respect to a light source, in accordance with some embodiments of the present disclosure. For example, the white shading of visibilities 200D in FIG. 2D may indicate that each of the locations 212, 218, 216, and 226 has full visibility with respect to the light source 120 in FIG. 1, and the dark shading of visibilities 200D in FIG. 2D may indicate that each of the locations 210, 214, 220, 222, and 224 has no visibility with respect to the light source 120 in FIG. 1.

As in FIG. 2C, the sample analyzer 106 may analyze the visibilities 200D and based on the analysis, the casting determiner 108 may determine to cast at least one additional ray for the location 210. Such an analysis may be based on determining the visibilities 200D in the group do have at least a threshold amount of variability. For example, the sample analyzer 106 may determine whether at least one of the visibility values corresponding to the visibilities 200D is different than at least one other of the visibility values. If this threshold amount of variability is met (as it is in FIG. 2D), the casting determiner 108 may determine to cast at least one additional ray for the location 210.

With regard to hit distances, dark shading of hit distances 202D in FIG. 2D may indicate larger hit distances than smaller hit distances. For example, the location 224 may have the largest hit distance, followed by the locations 210 and 210, then the locations 214 and 222. The white shading of the hit distances 202D in FIG. 2D may indicate that each of the locations 212, 218, 216, and 226 has a relatively low hit distance with respect to an occluder (e.g., a minimum hit distance value as those rays did not hit an occluder). For example, the sample analyzer 106 may analyze the hit distances 202D and based on the analysis, the casting determiner 108 may determine to cast at least one additional ray for the location 210 (e.g., in addition to the ray that will be cast based on the visibility).

Similar to FIG. 2C, such an analysis may be based on determining the hit distances 202D in the group do have at least a threshold amount of variability. In some embodiments, the sample analyzer 106 may determine the threshold amount of variability of the hit distances 202D with respect to a hit distance of the location 210. This may involve the sample analyzer 106 comparing the hit distances of the location 210 to the hit distances for the other location(s), such as to determine whether one or more of the location(s) are within a delta (threshold amount) of one another (e.g., sufficiently similar). In some examples, if one or more of the locations are not sufficiently similar (e.g., at least one location exceeds the delta) the casting determiner 108 may cast an additional ray. However, if each of the one or more of the locations are sufficiently similar (e.g., no location exceeds the delta) the casting determiner 108 may not cast an additional ray, as in FIG. 2C. In this example, the similarities may be considered with respect to those rays that also hit an occluder, which may meet this criteria for an additional ray.

In addition to (or alternatively from) the casting determiner 108 basing determinations on whether the threshold amount of variability of the hit distances 202D with respect to a hit distance of the location 210 is met, the determination(s) may be based at least in part on determining at least one of the hit distances 202D in the group is less than a threshold value. For example, the sample analyzer 106 may analyze the hit distances for each of the locations of the region. If there are no hit distances for the locations that are below the threshold value (indicating a close hit or short hit distance), the casting determiner 108 may determine not to cast an additional ray for the location 210 (e.g., although a ray may still be cast based on a different determination). Otherwise if at least one of the hit distances are below the threshold value, the casting determiner 108 may determine to cast one or more additional rays (e.g., a single additional ray) for the location 210. In FIG. 2D, the locations 214 and 222 may have hit distances that meet this criteria so that the casting determiner 108 may determine not to cast an additional ray based on the hit distances 202D.

While FIG. 2A-2D are used to describe various examples of how the sample analyzer 106 may analyze groups of visibilities and/or hit distances and how the casting determiner 108 may determine whether to cast one or more additional rays for locations and/or how many additional rays to cast, these approaches may be varied. For example, visibilities and hit distances may be evaluated separately and/or in combination in order for the casting determiner 108 to make one or more determinations. In some examples, criteria for casting a ray based on visibility may be analyzed and if the visibility criteria is not met, hit distance criteria may not be analyzed by the sample analyzer 106, thereby preserving computing resources (e.g., the visibility criteria being met may be one of the criteria for the hit distance criteria to be met). In other examples, hit distance criteria may still be evaluated.

Additionally, in some examples, the number of rays cast for a given location(s) may be a function of the visibility and/or hit distances. For example, the number of rays could scale with visibility variability and/or other factors. As a further example, the number of rays could scale with hit distance variability and/or other factors, such as the hit distances that are below the delta with respect to the location and/or the number of hit distances that are below threshold value. Also, other factors may be considered, such as the type of object hit by rays, temporal information, etc. For example, while the present examples may employ spatial samples for hit distances and/or visibility, temporal hit distances and/or visibility values may be used in addition to or instead of the spatial values.

Further, in the examples of FIG. 2A-2D visibilities have binary values (e.g., hit or miss). This may be a result of one ray being cast per pixel and/or location in the ray tracing pass 114. Where additional rays are cast per pixel and/or location, visibility values may be non-binary. For example, a visibility value for a location may be an average of visibility values for the rays cast from the location. This may impact the way the sample analyzer 106 evaluates variability and/or other factors regarding visibility. In some examples, a hit distance value for a location may be the lowest of hit distance values for the rays cast from the location (the shortest hit distance). This may also impact the way the sample analyzer 106 evaluates variability, threshold values, and/or other factors regarding hit distance.

As described herein, the ray tracer 104 may cast one or more additional rays or no additional rays for one or more locations and/or pixels, as described by the casting determiner 108. These additional rays may be cast over any number of additional ray tracing passes, such as in the ray tracing pass 116. Ray-tracing samples of a location from multiple rays and ray tracing passes may be combined to form an aggregated ray-tracing sample. In some examples, a visibility value for a location may be an average of visibility values for the rays cast from the location over the ray tracing passes. Also, a hit distance value for the location may be the lowest of hit distance values for the rays cast from the location over the ray tracing passes. The image renderer 102 may use the resultant ray-tracing samples to generate one or more images that are representative of the virtual environment 110 and/or one or more portions thereof (e.g., representative of a camera or other view of the virtual environment 110).

Now referring to FIGS. 3-6, each block of methods 300, 400, 500, and 600, and other methods described herein, comprises a computing process that may be performed using any combination of hardware, firmware, and/or software. For instance, various functions may be carried out by a processor executing instructions stored in memory. The methods may also be embodied as computer-usable instructions stored on computer storage media. The methods may be provided by a standalone application, a service or hosted service (standalone or in combination with another hosted service), or a plug-in to another product, to name a few. In addition, the methods are described, by way of example, with respect to the adaptive ray tracing system 100 (FIG. 1). However, these methods may additionally or alternatively be executed by any one system, or any combination of systems, including, but not limited to, those described herein.

FIG. 3 is a flow diagram showing the method 300 for determining whether to cast one or more additional rays to cast to sample lighting conditions for a pixel, in accordance with some embodiments of the present disclosure. The method 300 may apply to any of the examples of FIGS. 2A-2D, or other examples. The method 300, at block B302, includes determining visibilities and hit distances for pixels using rays. For example, the ray tracer 104 of FIG. 1 may determine (e.g., in the ray tracing pass 114) for pixels (e.g., corresponding to at least the locations shown in FIG. 1), visibilities and hit distances with respect to the light source 120 using at least one shadow ray for each pixel to sample lighting conditions of the pixel with respect to the light source 120. The visibilities and the hit distances may have been determined based on sampling lighting conditions in a scene comprising one or more pixels using at least one shadow ray for each pixel of the one or more pixels.

The method 300, at block B304, includes selecting a group of the visibilities and the hit distances based on a proximity from a given pixel(s). For example, the sample analyzer 106 may analyze a group of the visibilities and the hit distances that are associated with a group of the pixels corresponding to the locations of FIGS. 2A-2D based on distances of the pixels in the group from a given pixel (e.g., corresponding to the location 210) of the pixels.

The method 300, at block B306, includes determining to cast one or more additional rays for the given pixel(s). For example, based on the analyzing, the casting determiner 108 may determine to cast one or more additional shadow rays to sample the lighting conditions of the given pixel (e.g., corresponding to the location 210) with respect to the light source 120.

The method 300, at block B308, includes updating lighting conditions of the given pixel using the one or more additional rays. For example, the ray tracer 104 may cast the one or more additional rays in the ray tracing pass 116 and update the lighting conditions for the given pixel based at least in part on the one or more additional rays (e.g., by determining an aggregated ray-traced sample for the pixel).

The method 300, at block B310, includes rendering an image using the lighting conditions. For example, the image renderer 102 may render an image representative of at least a portion of the virtual environment 110 using the lighting conditions of the given pixel (e.g., and similarly for each pixel of a virtual screen).

FIG. 4 is a flow diagram showing the method 400 for determining a quantity of rays to cast to sample lighting conditions for a location in a virtual environment, in accordance with some embodiments of the present disclosure. The method 400 may apply to the examples of any of FIGS. 2A-2D, or other examples. The method 400, at block B402, includes sampling lighting conditions of locations using rays. For example, the ray tracer 104 of FIG. 1 may cast (e.g., in the ray tracing pass 114) rays from locations in the virtual environment 110 (e.g., corresponding to at least the locations shown in FIG. 1) towards the light source 120 in the virtual environment 110 to sample lighting conditions of the locations with respect to the light source 120.

The method 400, at block B404, includes determining ray-traced samples of the locations. For example, the ray tracer 104 may determine based on the rays, ray-traced samples comprising visibilities and hit distances of the locations with respect to the light source 120.

The method 400, at block B406, includes analyzing a group of the ray-traced samples based on proximities of locations to a given locations(s). For example, the sample analyzer 106 may analyze a group of the visibilities and the hit distances that correspond to a group of the locations 210-226 in FIGS. 2A-2D based on proximities of the locations in the group to the location 210.

The method 400, at block B408, includes determining a quantity of ray to cast from the given location(s) to produce one or more ray-traced samples. For example, the casting determiner 108 may, based on the analyzing, determine a quantity of rays to cast from the location 210 toward the light source 120 to produce one or more ray-traced samples for the location 210.

The method 400, at block B410, includes computing lighting conditions of the given location(s) using the one or more ray-traced samples. For example, the ray tracer 104 may cast the one or more additional rays in the ray tracing pass 116 and compute the lighting conditions of the location 210 based at least in part on the one or more additional rays (e.g., by determining an aggregated ray-traced sample for the location 210).

The method 400, at block B412, includes rendering at least a portion of the virtual environment using the lighting conditions of the given location(s). For example, the image renderer 102 may render an image representative of at least a portion of the virtual environment 110 using the lighting conditions of the location 210 (e.g., and similarly for other locations of the virtual environment 110).

FIG. 5 is a flow diagram showing the method 500 for determining whether to cast one or more additional rays to sample lighting conditions for a location in a virtual environment, in accordance with some embodiments of the present disclosure. The method 500 may apply to the examples of any of FIGS. 2A-2D, or other examples. The method 500, at block B502, includes determining visibilities for locations using rays. For example, the ray tracer 104 of FIG. 1 may determine (e.g., in the ray tracing pass 114) for locations (e.g., corresponding to at least the locations shown in FIG. 1), visibilities with respect to the light source 120 using at least one shadow ray for each location to sample lighting conditions of the location with respect to the light source 120.

The method 500, at block B504, includes analyzing a group of the visibilities associated with a region surrounding a given location(s). For example, the sample analyzer 106 may analyze, for the location 210 of the locations, a group of the visibilities associated with a region surrounding the location 210 in the virtual environment 110. This region may correspond to the locations of FIGS. 2A-2D.

The method 500, at block B506, includes determining to cast one or more additional rays for the given location(s). For example, based on the analyzing, the casting determiner 108 may determine to cast one or more additional shadow rays from the location 210 to sample the lighting conditions of the location 210 with respect to the light source 120.

The method 500, at block B508, includes computing lighting conditions of the given location using the one or more additional rays. For example, the ray tracer 104 may cast the one or more additional rays in the ray tracing pass 116 and determine the lighting conditions of the location 210 based at least in part on the one or more additional rays (e.g., by determining an aggregated ray-traced sample for the location 210).

The method 500, at block B510, includes rendering at least a portion of the virtual environment using the lighting conditions of the given location(s). For example, the image renderer 102 may render an image representative of at least a portion of the virtual environment 110 using the lighting conditions of the location 210 (e.g., and similarly for other locations of the virtual environment 110).

FIG. 6 is a flow diagram showing the method 600 including a decision tree for determining a quantity of rays cast to sample lighting conditions for a location in a virtual environment, in accordance with some embodiments of the present disclosure. The method 600 may apply to the examples of any of FIGS. 2A-2D, or other examples. The method 600, at block B602, includes determining ray-traced samples for locations. For example, the ray tracer 104 of FIG. 1 may determine (e.g., in the ray tracing pass 114) for locations (e.g., corresponding to at least the locations shown in FIG. 1), ray-traced samples using at least one shadow ray for each location to sample lighting conditions of the location with respect to the light source 120.

The method 600, at block B604, includes determining, for a given location(s), whether visibility criteria for the location(s) is satisfied. For example, the casting determiner 108 may use the sample analyzer 106 to analyze the visibilities of the locations 210-226 to determine whether visibility criteria for the location 210 is satisfied. If the visibility criteria is not satisfied for the given location(s), the method may proceed to block B606, where the casting determiner 108 determines to not cast any additional rays for the given location(s). FIGS. 2A and 2B may correspond to examples of this scenario. If the visibility criteria is satisfied for the given location(s), the method may proceed to block B608.

The method 600, at block B608, includes determining, for the given location(s), whether hit distance criteria for the location(s) is satisfied. For example, the casting determiner 108 may use the sample analyzer 106 to analyze the hit distances of the locations 210-226 to determine whether hit distance criteria for the location 210 is satisfied. If the hit distance criteria is not satisfied for the given location(s), the method may proceed to block B610, where the casting determiner 108 determines to cast one or more first additional rays for the given location(s). FIG. 2C may correspond to an example of this scenario. If the hit distance criteria is satisfied for the given location(s), the method may proceed to block B612, where the casting determiner 108 determines to cast one or more first additional rays (e.g., one ray) for the given location(s) and one or more second additional rays (e.g., one ray) for the given location(s). FIG. 2D may correspond to an example of this scenario.

FIG. 7 is a block diagram of an example computing device 700 suitable for use in implementing some embodiments of the present disclosure. Computing device 700 may include an interconnect system 702 that directly or indirectly couples at least the following devices: memory 704, one or more central processing units (CPUs) 706, one or more graphics processing units (GPUs) 708, a communication interface 710, input/output (I/O) ports 712, input/output components 714, a power supply 716, and one or more presentation components 718 (e.g., display(s)). The adaptive ray tracing system 100 of FIG. 1 may be implemented on one or more of the GPU(s) 708 and/or the CPU(s) 706 of the computing device 700. Further, various memory described herein may correspond to the memory 704 and/or one or more instantiations of the computing device