Point Cloud I/O adds point cloud import and export to Blender across the E57, PLY, LAS, LAZ, PCD, XYZ and PTS formats. Imported and exported clouds use Blender's optimised PointCloud geometry rather than a mesh fallback, and the extension handles million-point datasets.
The PTS support reads and writes the Leica Cyclone text format. On import it auto-detects the column layout from row width — 3 columns (x y z), 4 (plus intensity), 6 (plus rgb), or 7 (intensity plus rgb, the Leica canonical layout). Intensity is normalised to 0–1, and RGB auto-detects between 0–255 and 0–1. PTS export writes a count header followed by x y z rows with optional intensity and colour, controlled by independent Write Intensity and Write Colors tickboxes; intensity is scaled back to the Leica integer range and RGB written as uint8.
LAS and LAZ handling covers the ASPRS LiDAR formats, where each point stores position, 16-bit intensity, return-number metadata and an ASPRS classification code. LAZ is the LASzip-compressed equivalent of LAS. The PTS path is a pure-Python plus numpy implementation with no new wheel dependencies.