> For the complete documentation index, see [llms.txt](https://docs.sevenpens.com/devnotes/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.sevenpens.com/devnotes/processing-pen-data.md).

# Processing pen data

What happens to pen data between reading it from an API and drawing it. None of this is Windows-specific — it applies to any platform, and the reference implementations here happen to be C#.

Two stages, on the two halves of a pen sample:

* [**Pressure Quantization**](/devnotes/processing-pen-data/pressure-quantization.md) — coarsening pressure to a fixed number of levels, to model what a lower-resolution pen feels like. The rounding rule matters more than the level count.
* [**Position Smoothing**](/devnotes/processing-pen-data/position-smoothing.md) — interpolation and smoothing of the path, which are different operations addressing different defects. Includes what Krita does, and a measurement showing that on quantized input, mild smoothing is *more* accurate than the raw data.

Both stages are worth keeping distinct from the input layer. A defect that looks like it belongs here is often a coordinate problem upstream — see [Framework Coordinate Conversion](/devnotes/pen-input-on-windows/pipeline/5-desktop-to-canvas.md) before reaching for a filter.