Skip to content

Background subtraction

S.-H. Dan Shim edited this page Mar 13, 2025 · 17 revisions

X-ray patterns measured through a diamond-anvil cell often have a very intense background. PeakPo provides a background subtraction function to enhance the visibility of weak peaks.

Background subtraction is a powerful tool, but it should be used carefully. For this reason, the default setting for the Bg sub checkbox in the Plot Control box in the Main tab is unchecked.

Background Parameters

The parameters used for background subtraction are located in the Bkgn tab. If you hover the mouse pointer over any widget for more than 1 second, tooltips (balloon help) will appear. The numbers in parentheses next to the labels indicate the recommended values:

  • N Points = 20
  • N Order = 10
  • N Iteration = 20

Presentation of Background Fitting Results

When the Bg sub checkbox is unchecked, the raw base pattern is plotted alongside a dashed line representing the background fit performed by PeakPo. You should carefully inspect whether the background subtraction removes any intensities from diffraction peaks or introduces artifacts in the subtracted pattern. Unusual background fitting behavior is common, particularly at the edges of the selected range (ROI – Region of Interest).

Recommendations:

When starting a new dataset, experiment with different background parameter sets. The most critical parameters to optimize for your diffraction pattern are ROI Min and ROI Max. Once you achieve satisfactory results, you can use the same background parameters for the rest of your dataset, as long as the background and 2-theta range remain consistent.

Tips:

For better results, set the ROI Min and ROI Max values in a region without peaks. This will help avoid any strange artifacts near the edges of the ROI.

Notes:

If your Min or Max values for the 2-theta range are outside the range of your data, PeakPo will attempt to fit the background across the entire 2-theta range. Be sure that your Min and Max values are within the 2-theta range of your data to avoid this issue.

Clone this wiki locally