Skip to content

Adjustments to N colors and spectra #354

Description

@dylanbeaudette

Follow-up to #332, where N chip color and spectra were determined via direct measurement. While the colors were plausible, they all appear too "light" on most LCD screens. So much so that organic soil material and darker soils like CAPAY, ZOOK, and other using N colors seemed "off".

Since the Munsell -- sRGB conversion was mostly for qualitative visualization, it didn't matter much. Now that CIELAB color coordinates are included in the LUT, implying some level of colorimetric accuracy, there needs to be a little more explanation.

Image

The N chips in aqp were far too dark from 2018--2025, having been estimated vs. measured.

Here is N 2/ over time:

Image

TODO:

  • more detailed description of the Munsell LUT creation process, probably in a vignette
  • more details in manual pages
  • optional argument to for N-chips / spectra "as measured" vs. "looks right"
  • does this have any effect on estimateSoilColor() model development (I doubt it)

Decide:

  • CIELAB accuracy vs. on-screen representation for N chips / spectra

Metadata

Metadata

Labels

No labels
No labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions