You are here: Rendering > Render Item > Render Item: Settings

Render Item: Settings

The Render item's Settings tab allows you to adjust render settings specific to rendering quality such as anti-aliasing.

Render Item- Settings

The following Antialiasing options are available for the Render item:

There are three main settings that control the "quality" of rendered pixels. These include Antialiasing, Refinement Shading Rate, and Refinement Threshold. To create the highest quality imagery, Modo does all of its computations at the sub-pixel level; taking multiple tiny samples, within the space of a single pixel, and then averaging their result. However, rather than computing the entire image this way, over and over (the "brute force" method), Modo uses elegant logic to determine how to best create the image. To balance quality with performance Modo provides methods to determine which areas of the image should receive the greatest number of samples.

Antialiasing - at the top-most level the Antialiasing control dictates the numbers of "samples" taken per pixel. Modo allows you to specify from 1 to 1024 samples per pixel in powers of 2 (1, 2, 4, 8, 16, and so on up to 1024). We can think of these samples as sub-pixel fragments, imagining a pixel as a square that can be divided into smaller squares. When Antialiasing is set to 8, the pixel is broken down into 8 pieces, and each of these pieces is evaluated and then averaged together to give us the final high-quality pixel. Since it would be expensive time-wise to evaluate every single pixel in an image repeatedly we have some additional controls to limit the areas that receive the maximum number of evaluations.

Antialiasing Filter - allows you to choose different antialiasing filters. The Antialiasing Filter determines the pattern to use when evaluating a pixel. The default is Gaussian, which performs very well in most cases, and offers a good balance of performance and quality. Box, Triangle, Catmull-Rom, and Mitchell-Netravali are additional options. Catmull-Rom generally yields slightly "sharper" results than Gaussian, while Mitchell-Netravali offers good results when dealing with problematic moire (created by fine texture patterns).

Environment Shading Rate - specifies the approximate spacing between shading samples, measured in pixels, used for the environment. It also acts as a cap on the rates specified in the various shader items, allowing shading detail to be increased across the entire frame. The main reason you might want to lower this value would be to improve the motion blur of volumetrics.

Refinement Shading Rate - tells the engine how finely to evaluate any pixel that falls within the Refinement Threshold. When a pixel is being evaluated, first, it is rendered through the standard Antialiasing application. Neighboring pixels are then tested for Refinement Threshold and, if a pixel is within the threshold, then the pixel is finished and the engine moves on to the next. If it's outside of the threshold, then the Refinement Shading Rate kicks in, and the pixel is further evaluated. The value itself basically acts as a multiplier of the pixel's size, so a setting of '.5' would split the pixel in half both ways, resulting in 4 additional shading samples calculated. A setting of '.1' would split the pixel 10 times both directions, resulting in 100 additional shading samples calculated.

Refinement Threshold - essentially a contrast tolerance, which is to say this control dictates how much contrast between adjoining samples is acceptable. Set to 100% the Refinement Threshold would do nothing at all, as it would accept brightness steps as high as 255 (the upper bound of a 24 bit image). Set to 0% no pixels would be inside the threshold and every pixel would be refined. The percent amount spans the 0 to 255 brightness steps so a threshold of 50% would accept pixel contrasts of 128 steps, whereas a setting of 10% would only tolerate 25 brightness steps between samples. Just like the refinement shading rate, we see that an increase in quality when set to lower values, but it also increase computation requiring longer render times.

Refine Bucket Borders - due to the "bucket" nature of Modo's render engine, adjacent pixels in neighboring render buckets can't be used to decide whether a border pixel needs further refinement, because they are either not rendered yet, or fully rendered and no longer active. By enabling Refine Bucket Border, Modo assumes all border pixels in a given bucket require refinement.

TIP:  The Antialiasing setting's effect is most apparent on geometric edges. A setting of 8 is generally acceptable and is a good trade off between quality and speed, though in reproducing effects such as Depth of Field or Motion Blur, you need to increase this value, possibly all the way up to 1024 to produce a pleasing result.

If you find your shaded edges, such as cast shadows, refractions, reflections, and procedural- and image-based textures are aliased, then reducing the shading rate value or lowering the refinement threshold helps to increase shaded edge quality.

If you are finding specific areas of an image to be noisy, and it cannot be solved by Antialiasing and refinement options, such as with soft reflections, blurry refractions, soft shadow edges, or subsurface scattering, it may simply be a case where raising the number of samples specific to the effect eliminates the noise. These settings can be found in the associated properties panels. Always keep in mind though, as more calculations are taken into account, the longer an image takes to render.

Field Rendering - for scanline televisions, there were originally two half frames that made up each whole frame that were interlaced on top of each other to make a single image. Enabling the Field Rendering option properly renders two discreet moments in time and interlaces the resulting half-frames into a single image (in essence, doubling the frame rate). This option is useful to compose Modo-generated elements into video that was originally captured this way.

TIP:  When using Field Rendering, it's recommended that you switch to Box antialiasing to eliminate field cross-talking. Also, post effects, such as Bloom, are generated on a whole-image basis, so should only be applied in a compositing package. Modern film and video capture is largely progressive, where each frame represents only a single moment of time, so in most cases Field Rendering is not necessary.

The following Ray Tracing options are available for the Render item:

Reflection Depth - Modo renders images by ray tracing, sort of like shooting tiny arrows out from the cameras lens and seeing what they hit. If a ray hits an object that is reflective, that ray bounces off the object until it hits another. Should that surface also be reflective Modo again bounces that ray until it finds a surface that is not reflective. If the ray doesn't find a surface that isn't reflective, the Reflection Depth threshold stops the ray from continuing on its path, keeping Modo from infinitely tracing rays between reflective surfaces. The default of 8 bounces is fine for most scenes, and is a good compromise between render accuracy (quality) and render time. In some instances, you may find a need to increase this setting to accurately portray all the reflective surfaces in a scene, or find you need to reduce it to shorten render times.

Refraction Depth - like Reflection Depth, the refraction depth is calculated by ray tracing. When a ray enters a refractive surface with some transparency, the ray bends the appropriate amount and travels until it hits its next surface. If that surface is also refractive, again the ray is bent and sent along its path until it hits a non-refractive surface or is stopped by the Refraction Depth threshold from traveling further. The default setting of 8 is often good enough for most scenes, and is a good compromise between render accuracy (quality) and render time. In scenes with many transparent surfaces, you may find a need to increase the setting to accurately portray all your transparent surfaces, or find that you need to reduce it to shorten render times.

Ray Threshold - for some scenes, firing all those refraction and reflection rays can get expensive time-wise. Ray Threshold eliminates those rays that have little to no impact on the final rendered image. As rays are fired, they are given an importance value that can decrease or increase at each bounce, depending on the surface values contribution. Should the rays' importance fall below the Ray Threshold, Modo decides whether to kill the ray or trace it further (up until the total depth value specified). This killing of rays reduces the overall number and depth of rays traced but has little visual impact on the final image. Increasing the Ray Threshold value further eliminates rays, though at the expense of image quality.

Minimum Importance - importance is an estimate of how much a particular shading evaluation contributes to the final color of a pixel. For surfaces directly viewed by the camera, it's basically what fraction of a pixel the shading evaluation represents, which depends on how many antialiasing samples are being shaded together as part of a single evaluation versus the total number of antialiasing samples in the pixel.

If there are 8 AA samples in a pixel and they are all shaded together as a group (which can happen if they all belong to the same surface and the Shading Rate is large), that single shading evaluation entirely determines the pixel color and thus has an importance of 100%. At the other extreme, if each antialiasing sample is shaded separately (which can happen if the Shading Rate is small or zero), then each of the 8 shading evaluations have an importance of 1/8 or 12.5%. This importance value is used to modulate the number of rays used to compute various phenomena, such as soft shadows, Monte Carlo indirect illumination, blurry reflections, and similar. At 100% importance, the full number of rays specified are used. For example if there are 64 blurry reflection rays specified in a material's properties, then 64 rays are fired during the shading evaluation, but in a shading evaluation with 12.5% importance, only eight reflection rays (12.5% of 64) are fired.

The Minimum Importance setting imposes a lower limit on importance when shading surfaces are directly viewed by the camera. This can cause more rays to be used in some cases, reducing artifacts that can be caused by Clamp Colors but can also possibly increasing render time.

Spectral Samples - when rendering any transparent refractive surfaces with Dispersion, the Spectral Samples value determines the number of rays that each dispersion ray is broken into. The higher the number of rays, the smoother the dispersion calculations are, at the expense of longer render times.

HDR Dark


HDR Bright

No Dispersion - 1 Minute

16 Spectral Samples - 3 Minutes

256 Spectral Samples - 41 Minutes

Maximum Radiance - controls the maximum radiance carried by a ray, limiting it to a user-defined value. This can be helpful in reducing noise and preventing "fireflies" (single bright pixels) caused by small but very bright features of a scene, such as tight specular highlights or surfaces very close to point light sources. Setting the Maximum Radiance to '0' (the default value) means that a ray can carry unlimited radiance.

The following Direct Illumination options are available for the Render item:

Light Samples - Modo uses an adaptive light sampling method that takes into account various attributes from direct light sources, such as their brightness, relative position to the shaded surface, and their intensity falloff, among other settings. These values combine to generate an overall importance amount for each light in a given scene, while still producing an unbiased result. Generally speaking, this method of working renders faster and produces higher quality shading, especially in scenes with many light sources, allowing the render engine to focus its sampling on only those lights that are most important. The main benefit being that light (and shadow) quality can be controlled by a single global setting rather than adjusting a Samples value per light, as done in earlier versions, simplifying the overall rendering process.

The Light Samples value represents the total number of samples generated for all lights in a given scene. Shading noise can be decreased or eliminated by increasing this value, in exchange for longer render times. The higher the value, the smoother the shaded result. Common input values are available under the Value Preset pop-up. To disable adaptive light sampling, set this value to '0'. This, in turn, enables the Samples value per light, and works as in previous versions.

In rare cases there may be situations where individual lights don't produce the desired results. In these cases, a light's Importance setting can be increased or decreased depending on the desired outcome, acting as a multiplier to its evaluated importance. This setting is found under each light's Properties panel when selected.

MIS - MIS stands for "Multiple Importance Sampling" and is a way of intelligently sampling direct lighting for both specular and diffuse shading, greatly reducing noise without a significant increase in render time. For most cases, the MIS default setting of Both produces the best results. In rare cases, you may wish to limit the MIS calculations to just the Specular shading or just to Diffuse shading. Setting MIS to None disables Multiple Importance Sampling, reverting to the legacy sampling method. It should be noted that MIS works with all shading models for diffuse shading, but MIS specular shading only works on the Energy Conserving and Physically Based models when Match Specular and Blurry Reflections are enabled for reflection shading.

Shadows - enabled, by default, the Shadows checkbox directs Modo to globally render all shadows cast by direct lights. Direct lights can be any of the light items, such as Distant Lights, Area Lights, Point Lights, Spot Lights, Dome Lights, Cylinder Lights, and Photometric Lights. Shadows cast by luminous polygons when Global Illumination is enabled are not affected by this setting (their shading is part of the indirect illumination). Shadows can also be disabled on a per-light basis in each light's properties panel, by setting shadow type to none.

The following Geometry options are available for the Render item:

The first two settings under the Geometry section are global controllers that relate to Subdivision Surface (SDS) models in a scene. Subdivision Surfaces is a means of producing a smooth, continuous mesh at render time, based on a low resolution proxy model (sometimes called the "cage" or Limit Surface). Generally, when straight polygon modeling, surfaces are made from faces - each its own tiny flat plane that, when shaded together, simulates a smooth surface - but silhouette edges can reveal the faceted nature of the model. The traditional fix for this limitation is to simply add more polygons, but controlling all those real polygons can quickly become unwieldy. For a SDS model, the polygons in the low resolution model are automatically divided and refined recursively in such a way that the resulting model is a smooth organically curving surface.

In Modo, you can toggle any object as an SDS model by simply pressing the Tab key. You can control the level (number of times the models is subdivided) using the Subdivision Level setting found in the Mesh Item properties viewport. Higher values produce smoother models, but generate more geometry that requires more processing time and memory to render. When the Adaptive Subdivision setting is enabled, this function overrides the Mesh Item setting at render time allowing you to adaptively control all the Subdivision Surfaces in a scene globally.

SubDiv Cage

SubDiv Level 1

SubDiv Level 2

SubDiv Level 3

Original 'Cage' Geometry

SubD level of 1

SubD level of 2

SubD level of 3

Adaptive Subdivision - toggles the Adaptive Subdivision function in Modo. When enabled, Modo adaptively tessellates all SDS meshes at render time. Which is to say, depending on the user settings, the mesh is refined repeatedly until all polygonal edges fall within the set subdivision rate.

Subdivision Rate - when Adaptive Subdivision is enabled, the subdivision rate sets the level threshold for all SDS geometry in a scene. Defined as pixels, the default rate is 10, which means that Modo picks a subdivision level for that mesh, such that the length of edges for any group polygons in the largest subdivision patch appear no more than 10 pixels long (most are shorter than that). This is very useful since it means that subdivision levels automatically adapt to each mesh's distance from the camera, the zoom factor, and similar. If you really want to minimize your polygon count, you can set the subdivision rate to a huge number such as 1000. On the other hand, if you never want to see a faceted edge in a render you can reduce this number to 5 or less. Be warned that reducing the subdivision rate comes at the expense of additional system memory use and performance loss.

The next five settings control Modo's Micropolygon Displacement function: a means of adding fine geometric detail to surfaces by way of a texture-based controller. Similar in nature to Subdivision Surfaces, polygonal surfaces are automatically divided and refined recursively producing finer and finer micropolygons, which are then displaced (pushed in or out in the polygons normal direction) based on a grayscale value (and user settings in the Texture Locator item).

This can result in an incredible amount of detail for models that would be difficult or impossible to create otherwise. This is similar to a bump map, but where a bump map is a rendering trick that simulates detail on a surface and often looks imperfect, displacement is actual geometry. Displacement maps can be created in Modo, using any of the painting or sculpting tools, generated in an external application and applied to the object, or simply applied as a procedural texture layer in the Shader Tree. For best results, you should enable SDS for any surfaces with displacement applied. Please refer to the Effect - Texture Item topic for more information on working with displacement maps.

No Displacement

Displacement Applied

SDS Limit Surface

Displacement Map applied

Micropoly Displacement - when an object's surface has displacement applied in the Shader Tree, this checkbox toggles the rendering of micropolygon displacement on a global level.

Displacement Rate - when Micropoly Displacement is enabled, the Displacement Rate, sets the threshold to which the geometry is subdivided into micropolygons. Similar to the Subdivision Rate, micropolygon displacement is adaptive, evaluating every polygon edge separately. The distance from the camera to the center of the edge (the focal length), and the render resolution are used to convert the Displacement Rate (in pixels) into a distance in world space. If the edge is longer than this and also longer than the Minimum Edge Length, it is split in half. This process continues recursively until all edges satisfy those requirements.

Displacement Ratio - when set to 1.0, objects are tessellated a consistent amount across their entire surface, but when displacing large areas, such as a ground plane in an environment, where a good deal of the geometry lays outside of the view of the camera, most of those polygons are going to waste, filling up your system's memory. The Displacement Ratio reduces the number of displaced polygons outside the camera's view. The higher the number, the greater the reduction in polygons. Keep in mind though that reflective and refractive objects still are able to see these areas outside the camera's view, so high settings may begin to introduce artifacts to your image.

Minimum Edge Length - each edge in the mesh is tessellated until polygons satisfy the Displacement Rate setting. Ungoverned, this can go out of control resulting in ridiculous numbers of polygons. The Minimum Edge Length does exactly what the name implies; it sets a minimum length for any polygonal edge. Once an edge has reached that minimum length it can no longer be split. A simple and effective throttle for controlling the amount of polygons in your scene.

Smooth Positions - when a surface is diced into micropolygons, the initial positions of the micropolygon vertices lie on a curved surface based on the original smooth vertex normals. So, if you start with a regular non-SDS sphere, for example, it'll be diced into a smooth sphere, and then the displacement texture is applied to that. You could even use this as an alternative to subdivision surfaces by just applying a displacement texture that is zero everywhere (although the resulting surfaces are different from SDS in that they actually pass through the original vertices).

The ability to turn off position smoothing, which means that all the initial positions of the micropolygon vertices lie exactly on the flat planes of the original polygons. Doing this in the non-subdiv sphere example would cause the diced version to look faceted. The reason this feature was added was that it's necessary to turn off position smoothing when using a displacement map created by object-to-object baking, since the distances in the map are measured based on the original polygons of the low-poly object and not the smoothed polygons.

Displacement as Bump - when this option is enabled, Modo also applies any displacement maps as a bump map, providing a finer detail evaluation that requires lower number of subdivisions to achieve. This can be a great way to reduce the amount of memory necessary to render complex scenes when combined with lower subdivision rates.

Indirect LOD - when this option is enabled, Modo holds an additional lower resolution version of displaced geometry used strictly in calculating indirect lighting, reducing overhead in GI calculation. However, it may end up using slightly more memory as two distinct copies of the same geometry may be held in memory simultaneously.