A web-based application for baseline correction, smoothing, processing, visualization, export, and optional eLabFTW upload of Raman spectra.

In general, the workflow and processing logic are similar to the original raman-tl.py. For many practical questions, the original documentation is still a useful reference. The main differences are summarized below.

• Graphical user interface
  Processing and display settings can be adjusted interactively in the browser instead of using command-line options.

• Direct HORIBA LabSpec file support
  The app can read:

  • HORIBA LabSpec .l6s
  • HORIBA LabSpec .xml
  • plain text files (wavenumber intensity)

• Metadata display
  Available metadata are shown directly in the app and included in export products where appropriate.

• Additional baseline correction method: SNIP
  In addition to arPLS, the app also supports SNIP baseline correction.

• Interactive Plotly visualization
  Spectra can be inspected interactively in the browser:

  • zoom and pan
  • show or hide spectra by clicking legend entries
  • isolate one spectrum by double-clicking a legend entry

• Structured export packages
  Instead of writing output files directly into the working directory, the app creates ZIP-based export packages for:

  • the active spectrum
  • selected multi-spectra views
  • the full session

• eLabFTW upload
  Export products can be uploaded directly to an existing eLabFTW experiment as attachments.

• Session-based workflow
  The app is designed for interactive work:

  • upload spectra
  • inspect and compare them
  • adjust processing settings
  • export or upload results

1. Upload one or more spectra
2. Select the active spectrum and optional overlay spectra
3. Adjust:
   • baseline correction
   • smoothing
   • peak detection
   • spectral range
4. Inspect:
   • Single View
   • Overlay Spectra
   • Normalized Overlay
   • Stacked Spectra
5. Export results as ZIP packages
6. Optionally upload packages to eLabFTW

• Processing settings are global within a session
  Baseline correction, smoothing, peak picking, and spectral range settings currently apply to all spectra in the session.

• Use moderate numbers of spectra for comparison
  It is possible to load and compare more spectra, but for practical use it is recommended to work with no more than about 5–6 spectra at a time, especially for overlay, normalized overlay, and stacked views.

• Manual intensity scaling and x-shifting are useful for visual comparison and alignment, but they can affect interpretation and should be used with care.
• Export and eLabFTW upload both use a two-step workflow:
  1. create the package
  2. download or upload the generated files

Sung-June Baek, Aaron Park, Young-Jin Ahna, Jaebum Choo
Baseline correction using asymmetrically reweighted penalized least squares smoothing
Analyst 2015, 140, 250–257
DOI: https://doi.org/10.1039/C4AN01061B

C. G. Ryan, E. Clayton, W. L. Griffin, S. Sie, D. R. Cousens
SNIP, a statistics-sensitive background treatment for the quantitative analysis of PIXE spectra in geoscience applications
Nuclear Instruments and Methods in Physics Research Section B 1988, 34, 396–402
DOI: https://doi.org/10.1016/0168-583X(88)90063-8

Paul H. C. Eilers
A perfect smoother
Anal. Chem. 2003, 75, 3631–3636
DOI: https://doi.org/10.1021/ac034173t

based on:

E. T. Whittaker
On a new method of gradutation
Proceedings of the Edinburgh Mathematical Society 1922, 41, 63–75
DOI: https://doi.org/10.1017/S0013091500077853