Houdini 20.5 Nodes Geometry nodes

Remesh to Camera geometry node

Subdivide input geometry to generate polygons smaller than a camera measurement.

On this page
Since 20.0

Overview

This node repeatedly subdivides the input geometry until it is small enough relative to a Pixel or Dicing Quality measurement based on the specified Camera. It can be used to generate geometry for rendering as an alternative to render-time dicing. For example, if the geometry is to be deformed by VEX or another set of Houdini operations that are difficult to express in Karma or another renderer.

Note

This node uses the TriDivide SOP to perform an approximate, smoothing subdivision, so it works best on flat grids or geometry that is intended to be smooth on render.

Parameters

Group

The primitive group to be subdivided.

Camera

The camera used for measuring the final size of the output polygons.

Tip

For a Solaris camera, you can use a LOP Import Camera OBJ node to first import the camera from LOPs, then specify that node as the Camera.

Measure

The camera-relative measurement criterion that output polygons must meet.

Pixels

Each output polygon is smaller than this area of pixels when projected to the specified Camera.

Karma Dicing

Approximates the criteria used when specifying Dicing Quality to the Karma renderer, as measured by the specified Camera.

Min Area

The minimum area of pixels covered by the output geometry when Measure is set to Pixels.

Dicing Quality

The approximate Dicing Quality setting when Measure is set to Karma Dicing. A value of 1 means that approximately 1 polygon will be created per pixel. The effect of changing this parameter is to increase or decrease the amount of subdivision by a factor of Dicing Quality squared. If you only require more detail at the horizon, consider increasing Z Importance instead.

Geometry Z Measure

The space in which the geometry’s z-coordinate should be projected before measuring if it meets the Min Area or Dicing Quality criterion. The x and y coordinates are always in pixel space, but the z coordinate along the camera direction can be in a different space.

Camera Space

The z coordinate is mapped to camera space, and its value will range 0 to (camera_far - camera_near). This mapping works well for most small to medium-scale scenes, and retains good detail at the horizon. However, for very large camera distances, it can lead to excessive dicing and memory usage as the geometry becomes very stretched at the horizon.

Pixel Space

The z coordinate is mapped to pixel space, and its value will range from 0 to max(xres, yres). This mapping gives more consistent dicing across widely varying scales, but might require increasing Z Importance for long camera views to keep sufficient horizon detail.

Z Importance

A scale on the mapping specified in the Geometry Z Measure parameter. Increasing this value will increase detail at distance along the horizon, generally with lower memory usage than increasing the overall Dicing Quality.

Resolution

The resolution of the image to be rendered by the Camera. By default these parameters use an expression to read the resolution directly from the camera object.

Screen Window Size

The size of the screen window. Increasing this value will extend the subdivision to the offscreen area, which is especially useful if the resulting geometry might be displaced horizontally into view.

Geometry nodes