Performance

3D Accuracy

The accuracy option controls the tessellation of surfaces when in Edit mode.

You can specify a value for Proportional and Fixed. The preview area to the right shows the effect.

Proportional

Tessellation is calculated according to object size.

Fixed

Sets a fixed sag value (from 0.01 to 10) for calculating tessellation on all objects.

Auto-Adaptive

Computes tessellation according to the object's curvature to best fit the context.

Curves' accuracy ratio

Lets you control the ratio of the 3D accuracy you define using the above-detailed 3D Accuracy options.

Optimize meshes for texturing

Changes the mesh triangulation to enhance visualization when applying textures to objects.

2D Accuracy

Options for 2D accuracy are the same as for 3D accuracy, except the Auto-Adaptive mode that is not available.

Visual Quality Management

Lets you override quality parameters defined in the Visual Quality editor for the static and dynamic rendering modes.

Click Change Visual Quality to define your own preset parameters through the preset editor.

Manage User Presets

Lets you create, edit, or delete groups which store user-defined values for the visual quality parameters.

Visual Quality Management - Rasterizer

Anti-aliasing

Lets you specify the number of internal renderings made to create the anti-aliasing effect.

Anti-aliasing smoothes rough edges to give the appearance of higher resolution by taking into account how much an ideal edge overlaps adjacent pixels.

It also avoids rendering artifacts, such as the staircase effect, by improving the visual quality of lines and surface edges (whether they are common or free border edges) in the whole scene. The image is rendered internally with a resolution higher than the one on screen. Geometry is drawn only once and for CPU-limited objects, it has a very positive impact on performance. On the contrary, for GPU-limited objects, the performance gain is minimal.

With this technique, each given pixel on screen is divided into "n" subpixels: the more subpixels, the better the result.

Option	Description
Setting	Click a value from the list to specify the quality level. None: no super sampling. nx where n is the factor value: each pixel is divided into n subpixels. The quality is n times better. MS is for MSAA (Multisample antialiasing). This method draws the scene multiple times and slightly jitters the position of the camera for each version. The resulting image is a blending of all the previously drawn versions. SS is for SSAA (Supersample antialiasing). This method renders the image at a higher resolution, then downsamples it. Sharp edges are smoother with no staircase effect. Note: There is a price to pay in performance when using this method. The higher the value, the better the quality but there is a price to pay in performance if you choose a very high-quality level. 8x is a good value to start with because you can obtain a nice result with a minor impact on performance. In addition, do not use Edges/Lines. Recommendation: Do not use your driver's antialiasing.
Post process	Lets you select the post-process technique you want to apply. None: no post-process is applied. FXAA: applies the FXAA (Fast approximate antialiasing) technique. This post-process shader-based method smoothes out hard edges (even those inside polygons) in all pixels on the screen. TAA: applies the TAA (Temporal antialiasing) technique. This technique uses the data from the previous frames and the current frame to converge to an aliasing-free image. It is basically a smart average between the N previous frames. As a consequence, it provides a better temporal stability for pictures compared to other techniques like MSAA (Multisample antialiasing). You have less flickering while moving or in the HTC Vive™. Important: A ghosting effect may appear when TAA is activated. The TAA result depends on the camera orientation. For example, a nearly horizontal line does not give good results.
FXAA	Lets you define the quality level when the FXAA method is applied. This option is relevant only when FXAA is selected for the Anti-aliasing Post-process option.
Use with SSAO	Activates anti-aliasing on SSAO (Screen Space Ambient Occlusion). All the post-treatment effects are anti-aliased as well: the normal and the Z map are anti-aliased, which means that some post effects are anti-aliased too. Therefore, activating this option may have a negative impact on performance unless you have at least a Fermi-based card: Quadro Q2000, Q4000, and so on. Important: This option is relevant only for ambiences using the SSAO lighting technique (such as Clean Space). See 3DEXPERIENCE Native Apps User's Guide: Native Apps Advanced: Additional Viewing Commands: Applying Predefined Ambiences for more information about these ambiences. You can use this option only with graphic cards supporting FBO, FBO with multisampling, texture with multisampling extension, and OpenGL 3.x. Otherwise, anti-aliasing is not available.

Allow outline view mode

Allows the rendering of outline edges when the viewpoint is animated through available commands, a direct manipulation, or whenever an animation is playing.

Culling & Model

Option	Description
Min obj size (px)	This option is similar to Level of Details and lets you define the size of objects you want to hide or display in your geometry. Even if you do not want to move geometry, it is often useful to remove details you do not need to see. To do so, specify a high value to remove these details. Setting a high value also enables you to move large parts more quickly. On the contrary, setting a low value displays the details. For example, setting 2 means that objects whose size on screen is lower than 2 pixels are static.
Level of details	Defines the display precision for your objects. A LOD (Level of Details) is an approximation of the mesh intended to reduce the number of polygonal objects in modeling. The purpose of the LOD mechanism is to adjust the polygonal representation of an object to the distance of the user. Specify a low value to see all the details, or a high value to remove details. This option is relevant when: You do not need a high level of detail in your geometry. For example, because some portions are obscured by a visible piece of the object, or are far enough away to make the detail meaningless. Representations support LOD. Otherwise, it has no effect. The object is near the observer. It occupies a large number of pixels on the screen and a precise LOD is required to have a nice rendering. The object is far from the observer. Only a few pixels are used and a rough LOD is enough to render the object.
Occlusion	Lets you improve rendering performance by rendering only visible objects. Objects occluded (hidden) by another object are detected and prevented from being rendered, which optimizes memory consumption and CPU usage.
Strategy	Lets you choose the size of the bounding box for occlusion culling. Precise corresponds to the standard bounding box (1.0) Optimal corresponds to a slightly bigger bounding box (1.25) Fast corresponds to a bigger bounding box (1.6). This option gives a faster result but there is a price to pay in performance. Important: Activate Occlusion first, otherwise the Occlusion Strategy list is grayed. Static and Dynamic modes must use the same occlusion strategy. By default, this option is set to Precise.
Force rendering	Activates rendering optimization on scenarios where visual elements are created (for example, skeleton bones on DELMIA manikins, or VR scenarios with interactors).
	Option is cleared	Option is selected
	Important: This option is available in Static mode only. As internal data is reorganized, there might be an impact on performance depending on the number of loaded and displayed data. By default, the option is cleared for all profiles except Immersive in the Virtual Reality list.

Transparency

Tip: For partially transparent objects, the higher the value, the better the result. But for textures with opaque elements, a Fast quality mode might produce a better result than a Balanced one. To help you with your choice, some transparency modes are associated with the presets available in the Static and Dynamic lists. For example, choosing the Ultra preset automatically applies the Order Independent Colored transparency mode.

Transparency Mode	Description
Alpha Blending Without Sort	Produces an effect similar to looking through clear glass. Use this option when you need to view several transparent objects located at different depths of a scene. For example, looking through a car windscreen at other opaque objects inside the car. This effect is impacted according to the value you specify. You need to specify the transparency coefficient on selected objects through the Transparency slider in the Properties dialog box. See 3DEXPERIENCE Native Apps User's Guide: Native Apps Advanced: Editing Objects: Editing Graphic Properties for more information. Regarding polygons, the result might not be as expected because the triangles are blended with the rest of the scene. As it is too costly and may adversely affect performance, transparent polygons are drawn at the end of the draw phase but are not depth-stored. Therefore, when the transparent polygon is blended, the scene might not be fully drawn behind. Important: When working in Shading with Material mode, if you apply the maximum value (that is, 255) to the transparency in the object properties, then nothing is displayed in the work area. This mode is computation-intensive and consequently has an adverse effect on display performance.
Alpha Blending	Similar to Alpha Blending Without Sort, except that objects are sorted according to their distance from the viewer. This gives slightly better results.
Weighted Average	Improves the quality of transparent rendering, and renders mixed opaque/transparent geometry. This rendering technique also improves the visualization quality of edges: edges in the back are less visible than edges in the front, and their colors are blended with the geometry. This technique applies to 3D geometry only and has no impact on other transparent geometries such as the Compass. Warning: Shaders must be supported. GL_EXT_blend_func_separate must be supported. Floating textures must be available on the device. When this option is selected, there is a price to pay in performance because more memory/CPU is used. In full screen mode, approximately 30-40 Mo of textures are allocated. As the scene is drawn many times, the higher the number of transparent faces, the higher the impact on performance. Note: About this mode: Mirroring with alpha blending is rendered in classical mode and therefore, visual artifacts might occur especially for objects with edges. When the geometry contains a high number of layers, the back edges seem to disappear. The reason is that an average on the layer is computed, and the edge color has no significant impact on the computation. This mode does not support MSAA (Multisample Anti-aliasing).
Order Independent	This technique lets you render transparency in your scene by sorting geometry per pixel. It provides the highest transparency quality. Important: Refraction and roughness are not taken into account.
Order Independent Colored	Provides a better real time transparency by taking into account the albedo color, absorption color, attenuation distance, and Fresnel color. Important: Refraction and roughness are not taken into account.

Shadows

Shadow Type	Description
On ground	Simulates the shadows of the objects on the ground, provided that the current ambience supports this effect.
Inter objects	Enables shadow-casting between objects.
From transparents	Enables transparent objects to cast shadows on opaque objects and on the ground but they do not receive any shadows from other objects.
From area lights	Renders soft shadows cast by rectangle, disk, or sphere area lights. You can have up to 4 area lights. This option can be used with the Inter objects option. Important: This shadow type requires a GPU supporting Vulkan ray tracing. The On ground option has no effect on shadows cast by area lights. This option is not available for the Virtual Reality mode. The following capabilities are not supported: Multi-GPU environment Large scale models High-memory pressure Local clipping planes Shadows cast by transparent objects.
Max map size	Defines the maximum resolution at which shadows are rendered. The higher the resolution, the sharper the shadows.
Use bounding box	Uses the bounding box around the selected object (the bounding sphere) instead of the scene bounding box when casting shadows. The scene bounding box casts more precise directional shadows. However, as the box is recomputed based on all triangles in the object whenever the scene is modified, there is a price to pay in performance. The bounding sphere takes less time to recompute but casts less precise shadows.
Filtering to use	Lets you select the method to use when filtering the shadow map. PCF (Percentage-Closer Filtering) calculates the percentage of the surface that is closer to the light. ESM (Exponential Shadow Mapping) enables hard shadow anti-aliasing. PCSS (Percentage-Closer Soft Shadows) uses a variable penumbra to render realistic soft shadows.
Filtering quality	Lets you define the quality level to apply when filtering the shadow map.

Reflections

Option	Description
On ground	Lets you apply a mirror or SSR effect. None: no mirror effect is applied. Mirror: applies a mirror effect on the ground, and hidden objects are also reflected. SSR: SSR (Screen Space Reflections) replaces the mirror effect on the ground, but only visible parts are reflected. This avoids rendering hidden parts when computing reflections and may be useful when working with heavy objects. SSR allows simple reflections on reflective objects from other objects visible on screen. SSR can replace the mirror effect in predefined ambiences.
Inter objects	Lets you display inter-object reflections. None: no reflections are displayed between objects. SSR: displays inter-object reflections using the SSR algorithm.
	None	SSR
	You can use SSR for inter-objects reflections in the following ambiences: White Review Dark Review Outdoor Indoor City Road All ambiences where the mirror effect can be replaced.
SSR quality	Defines the level quality for the Screen Space Reflections (SSR) algorithm, either Low or High. Use the Screen space reflections quality option with the Reflections on ground, or the Reflections between objects option.

Ambient Occlusion

Option	Description
Allow	Displays shaded surfaces by considering light attenuation from nearby actors. The theory is when an object is surrounded by many other objects, it appears darker because less light can reach it.
Mode	Lets you select the type of ambient occlusion. SSAO: SSAO (Screen Space Ambient Occlusion) is a simplified calculation of the global indirect illumination. It produces enhanced realism by taking into account light attenuation caused by occluding objects. With this technique, the scene is equally lit with soft shadows and small surface details are accentuated to give more relief. HBAO: HBAO (Horizon-Based Ambient Occlusion) provides a more accurate ambient occlusion. HQAO: HQAO (High-Quality Ambient Occlusion) provides world-space ambient occlusion to improve perceived quality. When this option is activated, you can use the HQAO cache size option to define memory usage for the cache. Keep in mind that using less memory uses more compute power every frame to compute things that are not stored in the cache. Important: For both SSAO and HBAO: There is s small radius of impact. It is not made to have large smooth shadow on the floor. There is a price to pay in performance. As these techniques are an approximation of ambient occlusion, it might not work in all cases. Some artifacts might appear on screen for edges because there is no data. For HQAO: This option is available only if your graphics card supports the required features (vk_nv_ray_tracing extension). Activating HQAO deactivates Max sampling, Blur mode, and Max blur level. The following capabilities are not supported: multi-GPU environment, large-scale models, high-memory pressure, and local clipping planes.
Max sampling	Lets you define the maximum number of depth samples per pixel for ambient occlusion.
Blur Mode	Lets you select the blur type. None: no blur effect. EDZ: fast blur with edge detection from Z-depth. EDZN: fast blur with edge detection from Z-depth and normal. AEDZN: high-quality blur with accurate edge detection from Z-depth and normal.
Max blur level	Lets you define the maximum shadow sharpness for ambient occlusion.
HQAO cache size	Lets you define memory usage for the cache. For example, using less memory uses more compute power every frame to compute things that are not stored in the cache.

Bloom

Allow Bloom	Applies a glow-like effect on brightly lit objects.
Max quality	Lets you define the maximum quality level applied to the bloom effect.
Allow depth of field	Makes portions of the scene away from the camera focal point appear blurred.
Enable multi-color capping	Lets you use multiple colors for the capping plane when running sectioning commands or clipping tools. To edit the capping plane, right-click in the 3D area. Note: This option is relevant only for geometry created with the 3DEXPERIENCE.

Downsampling factor

Lets you enhance the rendering performance by applying a downsampling factor to the image displayed on-screen.

The image is generated with a lower resolution to have better performance but with a lower visual quality (the image looks slightly blurry).

This option is especially relevant on laptops with a low-end graphics card. In this case, selecting Medium or High provides a good quality/performance ratio.

By default, this option is set to Ultra.

Material

Option	Description
Metallic flakes quality	Lets you tune the simulation quality of flakes (thin and highly reflective particles) at interactive frame rates when a car paint material is applied. The frame rate defines the number of images displayed per second (FPS). The higher the FPS, the better the result. Virtual Reality scenarios, for example, require a very high FPS. Low: no individual flake is visible, only a rough approximation. This value is used by default for the Immersive preset. Medium: some flakes are visible, but not all of them. High: most of the flakes are visible. Ultra: all the simulated flakes are visible (with two different colors).
Closeup flakes mode	Lets you activate or deactivate large flakes on car paint materials.

Depth Buffer Options

Option	Description
Buffering mode	Default means that normal depth buffering is activated. Log means that the logarithmic Z buffer is activated. This provides better depth precision, which is useful for large-scale models, but there is a price to pay in performance.
Offsetting mode	Edge means that the edge rendering technique is activated. When edges are drawn, some artifacts (such as specular aliasing) that might occur on surfaces are corrected. There is a price to pay in performance. Face is the default mode and provides optimal performance.
Flickering reduction	Improves visual quality by reducing visual artifacts (called "Z-fighting" or stitching effect) that may occur with overlapping models. The scene is prerendered in the depth buffer, without rendering any colors. The depth buffer is a simple 2D image containing depth values of an individual pixel on screen. It is used for comparing the current pixel's depth value with the value already present in the buffer, before deciding to overwrite it. Once the depth values have been retrieved, the scene is rendered again and: A bias is added to the computed z-values. Comparisons are made, but nothing is written to the depth buffer. Color values that fail the comparison are overlapped. Visual flickering is thus significantly reduced by overlapping zones. However, there is a price to pay in performance because the frame rate is decreased. The bigger the model, the more significant the decrease.
	Option is cleared	Option is selected
	Important: This option may have no effect on small structures, and overlapping surfaces with high curvature. This option does not work in ray tracing view mode. By default, this option is cleared.

Visual Quality Management - Ray Tracing

Accumulative Settings

Option	Description
Target quality (dB)	Defines the quality of the converged image as peak signal to noise ratio (PSNR). Low values (between 10 and 20) result in noisy images and a quickly finishing convergence and are suitable for preview rendering. High values (between 40 and 70) result in images without apparent noise but with a convergence time that might be very long. Such values are suitable for final frame rendering.
Samples per frame	Defines how many samples per pixel are traced before displaying the first ray tracing image. 1 enables maximum interactivity. Set to values above 1 makes the first image less noisy at the cost of lower interactivity. Recommendation: Use values above 1 only for compute setups with high compute power like multi-GPU and cluster setups. The value of this option varies exponentially (1, 2, 4, 8, and so on).
Max sample luminance	Lets you activate or deactivate the maximum sample luminance option. If you activate this option, some parts of the computed images may be darker than they should be in reality. In return, image noise converges more quickly.
Base value	Defines the clamping value of result pixels. Lower values result in faster convergence but darken bright effects like highlights or emissive objects. This value is used as maximum sample luminance value for sample paths consisting of many ray bounces. For paths including just few ray bounces, or only specular bounces, higher max sample luminance values are used.
Filtering	Lets you activate or deactivate the Gauss anti-aliasing filter.
Strength	Defines the Gauss filter curve used for smoothing. The higher the value, the smoother the result. This option is available only when Filtering is activated.
Denoiser	Converts a noisy image into a smooth image and improves convergence time for a rendering. Click Denoiser, and then use Start iteration option to specify after how many frames the denoiser is enabled. Recommendation: Enable denoising only if the amount of video memory (VRAM) is above 8 GB. The Denoiser option can be activated if your GPU is: a NVIDIA GPU with a compatible NVIDIA driver installed. generation Maxwell or newer (for example, Maxwell, Volta, Turing, Ampere). Note: About the denoiser: There is a kind of artistically painted style effect in the early stages, but this effect is removed when the image has converged more. Enabling the denoiser automatically grays out the Denoiser option in the Visual Quality Manager.
Start iteration	Applies the denoiser after the specified number of iterations.

Global Illumination

Option	Description
Max path depth	Defines the maximum number of ray bounces.
Light Tracing	Optimizes simulation of caustics effects and indirect illumination. Recommendation: Use this option for scenes with small light emitters or very glossy surfaces, and for interior scenes. Clear this option in standard scenes for faster interactivity and convergence.
Fast subsurface scattering	Lets you switch between fast and accurate subsurface scattering simulation (SSS). Fast SSS converges significantly faster but is only an approximation to reality. Recommendation: Use it for concept design and marketing. Accurate SSS is physically-based but converges a lot slower. Recommendation: Use it for concept validation and light visualization.
Ray offset (mm)	Specifies the offset from a ray hit point at which new rays start.

Downsampling factor

Lets you enhance the rendering performance by applying a downsampling factor to the image displayed on screen.

The image is generated with a lower resolution to have better performance but with a lower visual quality (the image looks slightly blurry).

Miscellaneous

Enable isoparametrics generation

Displays the topological elements defined as being isoparametrics.

Enable two side lighting for faces and surfaces only

Helps you visualize the two sides (that is, front side and back side) of a face or surface by lighting them.

Enable back face culling

For stand-alone faces

Avoids redisplay of back sides of faces or surfaces.

For faces belonging to solids

Avoids redisplay of back sides of faces or surfaces belonging to solids.

Halo when displaying the geometry with hidden lines removed

Enabled

Displays a halo around intersecting edges to create a perspective effect.

Picking

Enables you to pick elements, that is, select elements when pointing them.

Window size for picking

Lets you specify the size (in pixels) of the panel to be used for picking elements in normal mode.

Enable accurate picking

Helps you select more precisely elements that are very close to each other when displaying the geometry. .

Video Memory Management

Memory Unloading

Unloads memory when 90% of the GPU is occupied. The existing in-use memory is packed into a compact space.

Mesh Compression

Reduces memory consumption on the GPU (around 30%) when importing models.