Hi @artec186​ , recently I created a workspace including a process to add gradation color to point cloud, would share my experience here.

Step 1: Slice the source point cloud by the number of color gradation (e.g. 100)

(1) PointCloudPropertyExtractor: Retrieve zmin and zmax of the source point cloud.

(2) AttributeManager, AttributeCreator or ExpressionEvaluator: Calculate the height of each slice: slice_height = (zmax - zmin)/100

(3) Cloner: Create 100 copies of the point cloud. Name copy number attribute as "index", for example.

(4) AttributeManager or AttributeCreator: Calculate lower limit ant upper limit of each slice.

lower = zmin + slice_height x index

upper = zmin + slice_heigth x (inde + 1)

(5) PointCloudFileter: Filter the point cloud by:

lower <= z && z < upper

Step 2: Create gradation color (R, G, values

(1) RGBGradationCalculator (from FME Hub): Calculate 100 color values (Format of Destination Color: Comma-Separated Decimal RGB24). The result will be stored into "_color{}" list.

(2) ListExploder: Explode the "_color{}" into individual "_color".

(3) AttributeSplitter: Split the "_color" into 3 elements (R, G, and store them into a list, named "_rgb{}" for example.

Step 3: Set color to the point cloud slices and combine them into one

(1) FeatureMerger: Add "_rgb{}" to point cloud slices, by joining by their index.

(2) PointCloudExpressionEvaluator: Set color to point cloud slices.

Component | Data Type | Expression

color_red | UInt8 | _rgb{0}

color_green | UInt8 | _rgb{1}

color_blue | UInt8 | _rgb{2}

(3) PointCloudCombiner: Combine the slices into a single point cloud.

See the attached workspace example to learn more.

Note: If size of the source point cloud be very large, it would be better to create a custom transformer which performs creating slices iteratively with the Loop functionality, in order to reduce memory usage while translation.

Hope this helps.