To understand why this is a breakthrough, you must understand the "waviness vs. roughness" problem. For decades, engineers used 2RC filters to separate a surface's roughness from its waviness. Imagine driving a car over a road that has potholes (roughness) and rolling hills (waviness). Traditional filters acted like stiff suspension—they worked, but they introduced phase distortion . This means the filtered output signal was shifted relative to the input. In manufacturing, this led to rounded-off peaks and inaccurate edge detection. The 2.1 GDPS Solution The Gaussian filter defined in ISO 16610-21 (colloquially known as 2.1 GDPS) changed the game. Unlike 2RC filters, the Gaussian filter has zero phase shift . It uses a weighted average based on the normal distribution (the Gaussian bell curve) to remove waviness while preserving the true shape of primary surface features.
Reality: The linear Gaussian filter fails at sharp edges (the Gibbs phenomenon) and for extremely short profiles. For those, you would need Part 2.2 (Spline filters) or morphological filters. 2.1 gdps
Reality: It is a mathematical standard. Version 2.1 defines the specific constants used in the Gaussian weighting function. Using a "generic Gaussian filter" is not compliant unless it matches the ISO 16610-21 transfer function. To understand why this is a breakthrough, you