About Loading Data and Performances

Loaded Data Based on User Viewpoint

With 3D Navigate, the loading of data is dynamic and depends on the user's viewpoint. When the viewpoint changes, it requests data from the CloudView index using the viewpoint parameters as input and the root expands progressively. For example, when you load large data, 3D Navigate only displays the objects that are bigger than the pixel culling and that you can see from your viewpoint. If you change the perspective, objects that you could not see now appear.

CloudView expands the structure to find relevant shapes, depending on the viewpoint given as input. The structure is refined taking the intersection of object bounding boxes and the viewpoint. If an object is visible (partially or completely in the viewpoint), then it expands to verify its children. This is reiterated until a shape or a child assembly is found outside the viewpoint.

Loading Boxes

Thanks to the progressive loading, you can dynamically load large amounts of data into the 3D viewer. The convergence of a viewpoint might take tens of seconds, and it is sometimes difficult to see where the geometry is loading or missing. Loading boxes indicate where the loading of objects is not complete yet. They are drawn for all leaf objects of the model in memory that have not been loaded. Once the loading boxes appear, they can either:

• Progressively disappear as the geometry is loading, until viewpoint convergence ("Loading complete" displayed in the status bar).
• Stay on the screen if the memory limit is reached ("Loading stopped" displayed in the status bar). You can then navigate to a more precise viewpoint that allows the loading to complete.

Performances

Progressive expansion based on viewpoint position enables the navigation of large amounts of data with a constant frame rate. Thanks to the use of wrap and progressive expansion, feedback is faster and the 3D appears within a few seconds.

Moreover, there are boundaries to avoid out of memory fails. When there is no more memory, then the viewpoint completion stops. A warning message suggests zooming in to focus on a smaller set of data, thus enabling the viewpoint completion.