LOD (Level Of Detail) uses camera-to-object distance to switch among contained child levels. (Contained nodes are now called ‘children’ rather than ‘level’, for consistent naming among all GroupingNodeType nodes.) LOD range values go from near to far (as child geometry gets simpler for better performance). For n range values, you must have n+1 children levels! Only currently selected children level is rendered, but all levels continue to send/receive events.

The LOD node belongs to the Navigation component and requires at least level 2, its default container field is children. It is available since VRML 2.0 and from X3D version 3.0 or higher.


+ X3DNode
  + X3DChildNode
    + X3DGroupingNode
      + LOD


SFNode [in, out] metadata NULL [X3DMetadataObject]

Information about this node can be contained in a MetadataBoolean, MetadataDouble, MetadataFloat, MetadataInteger, MetadataString or MetadataSet node.


SFBool [ ] forceTransitions FALSE

Whether to perform every range-based transition, regardless of browser optimizations that might otherwise occur.

SFVec3f [ ] center 0 0 0 (-∞,∞)

Viewpoint distance-measurement offset from origin of local coordinate system, used for LOD node distance calculations.

MFFloat [ ] range [ ] [0,∞) or -1

Specifies ideal distances at which to switch between levels. The range field is a floating-point array of camera-to-object distance transitions for each child level, where range values go from near to far. For n range values, you must have n+1 child levels!


  • Can add <[WorldInfo](/x_ite/components/core/worldinfo/) info='null node'/> as a nonrendering, invisible final (or initial or intermediate) child node that also documents the LOD switch-over rationale.
  • Not setting range values indicates that level switching can be optimized automatically based on performance.

SFInt32 [out] level_changed

Output event that reports current level of LOD children whenever switching occurs.


  • LOD level index counting starts at zero. Thus level_changed value of -1 means no choice, 0 means initial child, 1 means second child, etc.


  • It is an error to define this transient outputOnly field in an X3D file, instead only use it a source for ROUTE events.

SFBool [in, out] visible TRUE

Whether or not renderable content within this node is visually displayed.


  • The visible field has no effect on animation behaviors, event passing or other non-visual characteristics.
  • Content must be visible to be collidable and to be pickable.

SFBool [in, out] bboxDisplay FALSE

Whether to display bounding box for associated geometry, aligned with world coordinates.


  • The bounding box is displayed regardless of whether contained content is visible.

SFVec3f [ ] bboxSize -1 -1 -1 [0,∞) or −1 −1 −1

Bounding box size is usually omitted, and can easily be calculated automatically by an X3D player at scene-loading time with minimal computational cost. Bounding box size can also be defined as an optional authoring hint that suggests an optimization or constraint.


SFVec3f [ ] bboxCenter 0 0 0 (-∞,∞)

Bounding box center accompanies bboxSize and provides an optional hint for bounding box position offset from origin of local coordinate system.


MFNode [in] addChildren

Input field addChildren.

MFNode [in] removeChildren

Input field removeChildren.

MFNode [in, out] children [ ] [X3DChildNode]

Grouping nodes contain an ordered list of children nodes.


  • Each grouping node defines a coordinate space for its children, relative to the coordinate space of its parent node. Thus transformations accumulate down the scene graph hierarchy.
  • InputOnly MFNode addChildren field can append new X3DChildNode nodes via a ROUTE connection, duplicate input nodes (i.e. matching DEF, USE values) are ignored.
  • InputOnly MFNode removeChildren field can remove nodes from the children list, unrecognized input nodes (i.e. nonmatching DEF, USE values) are ignored.
  • X3D Architecture 10.2.1 Grouping and children node types



  • Can add <WorldInfo info=’null node’/> as a nonrendering, invisible final (or initial or intermediate) child node that also documents the LOD switch-over rationale.
  • Insert a Shape node before adding geometry or Appearance.
  • GeoViewpoint OrthoViewpoint and Viewpoint share the same binding stack, so no more than one of these nodes can be bound and active at a given time.
  • Security mechanisms such as encryption and authentication can be applied to high levels of detail, allowing authors to protect intellectual property at high resolution for authorized users while still rendering simple unrestricted models for other users.
  • Contained nodes must have type X3DChildNode, such as Group or Transform or Shape.
  • Apply containerField='shape' if parent node is CADFace.
  • ConformanceNist X3D Examples Archive


See Also

This post is licensed under CC BY 4.0 by the author.