The Point Cloud object lets you create precise, three-dimensional models from real-world references by importing large datasets captured from reality as point clouds: sets of data points in the form of 3D objects. 3D modelers can view point clouds in true color in the viewports, interactively adjust the extent of the cloud displayed, and create new geometry in context by snapping to point-cloud vertices. Support for the RCP and RCS file formats enables you to take advantage of a connected reality-capture workflow with other Autodesk solutions: Autodesk® ReCap™ Studio, AutoCAD®, Autodesk® Revit®, and Autodesk® Inventor® software.
Point-cloud objects are made up of closely packed, self-illuminated individual points. Each point has its own color, so one way to think of a point cloud is as a 3D bitmap. Point clouds do not contain any other type of geometry such as edges or faces, so they do not change appearance with different viewport rendering modes, or respond to scene lighting by default. Also, the Point Cloud object currently does not support conversion of point clouds into geometry. You can, however, snap to points, which makes it relatively easy to create geometry that conforms to the shape of the point cloud.
The main benefit of using the Point Cloud object is that you can incorporate complex, realistic 3D objects in your scenes without having to model them explicitly. The consequent drawback is that these objects are not directly editable using traditional 3D-modeling techniques.
When you add a point-cloud object to the scene, 3ds Max automatically applies an Autodesk Point Cloud material to it. In general you don't need to adjust this material, but if necessary it provides controls for color intensity, ambient occlusion, and shadow reception.
Because of the nature of the default material, point-cloud objects do not, by default, respond to scene lighting. That is, they appear with the lighting under which they were originally captured. To enable the object to respond to scene lighting, follow this procedure:
To enable point-cloud objects to respond to scene lighting:
The following illustration shows the result of doing this: Map #1 (Gradient Ramp) has been copied from the Display rollout on the right to the Ramp Shader parameter on the left.
This gives the point-cloud object a look similar to that obtained with the default material, with the added benefit of enabling the object to respond to scene lighting. Note, however, that any shadows incorporated in the original capture remain.
To create a point-cloud object:
This opens a file dialog.
3ds Max displays the point cloud data in the viewport. If it's not readily apparent, you can try a couple of things:
This changes the magnification and viewpoint in all viewports to closely fit the point cloud data.
To use the Limit Box to hide parts of the point-cloud object:
In some cases, you might want to prevent parts of a point-cloud object from appearing in the scene. Examples of this include hiding undesirable artifacts of the reality-capture process, or elements that are irrelevant to the current project. The Limit Box sub-object level provides a box-shaped volume that allows only portions of the point cloud inside (or outside) it to be visible. You can move the entire volume or a single side. You can also rotate and scale it.
When the limit box is enabled, only portions of the point cloud inside its volume are visible.
This causes the limit box to appear in the viewports as a yellow, translucent bounding box encompassing the point-cloud object.
To select the entire box, first select one side, then Ctrl+click to select another side. You can now move the box in any direction. Only portions of the point-cloud object that remain inside the box are visible.
To create geometry from a point-cloud object:
You can use a point-cloud object as a template for creating standard geometry such as editable poly objects. You create the polygon mesh one point at a time, snapping to vertices in the point-cloud object. You can then use standard modeling techniques such as extrusion and sub-object transforms to customize the object.
The Create tool enables creating polygons one at a time by clicking their corners.
By default, the dialog opens to the Snaps panel with the Standard category active and the Grid Points option enabled.
Now it simply remains to create polygons on the surface of the point-cloud object.
The polygon appears and the Create tool remains active.
You can use other polygon-editing tools such as vertex welding to combine the polygons into a single object. You might also want to delete the original object's geometry from step 2, at any sub-object level, leaving just the newly created polygons.
To change the default grayscale gradient map, click the (Gradient Ramp) button below the drop-down list. This opens the Material/Map Browser, from which you can apply a different map, or choose the current map (under Scene Materials) to edit. You can apply any map, but only the Gradient Ramp map is meaningful in this context.
Top view of a point cloud of a terrain, with elevation ramp showing the relative heights of different areas
Red is the lowest elevation and blue is the highest.
To change the default grayscale gradient map, click the (Gradient Ramp) button below the drop-down list. This opens the Material/Map Browser, where you can apply a different map, or choose the current map (under Scene Materials) to edit. You can apply any map, but only the Gradient Ramp map is meaningful in this context.
The number of points currently visible and the total number of points in the object appear in the read-only fields at the bottom of the Level Of Detail Setting group: Displaying and Out Of, respectively.
The following illustration shows the results of using the extreme ends of the scale at different distances with a point-cloud object containing about 444,000 points.
Point cloud object viewed from far away (top) and close up (bottom):
1: Performance (13,600 points visible)
2: Quality (444,000 points visible)
3: Performance (161,000 points visible)
4: Quality (384,800 points visible)
The following illustration depicts a point-cloud model at the lowest Level Of Detail setting (Performance) with three different As Pixel values. Note that the lower As Pixel values permit seeing "through" the closest surface to the underlying surfaces.
1: As Pixel = 0.5
2: As Pixel = 1.0
3: As Pixel = 4.0
The Point Cloud object provides two ways of hiding parts of the point-cloud data: Limit Box and Display Volumes. If you want to hide all but a section of the data that fits inside or outside a single box-shaped volume, the Limit Box method is easier because it doesn't require external objects. But if you want to hide multiple, non-contiguous parts of the data, it might be necessary to use the Display Volume feature with the supported geometry primitives: Box, Sphere, and Plane.
This tool supports only the Box primitive, the Sphere primitive, and the Plane primitive. With the Plane primitive, points qualify as being inside the display volume if they're behind the plane; that is, on the non-rendering side.
Use the Add and Remove controls to manage the list of display-volume objects. To stop adding, right-click in a viewport or click Add again.
1: Point-cloud object with box primitive placed as display volume
2: Box used as display volume with Invert off (the default setting), then hidden. Only points inside the box are visible.
3: Box used as display volume with Invert on, then hidden. Only points outside the box are visible.