Compensation and transforms

Check the matrix source

The workbench can use an embedded spillover matrix, import a .mtx file, or compute a matrix from single-stain control samples through the compensation wizard. Check which matrix is active and whether compensation is enabled before interpreting fluorescence plots.

Match the matrix channels to the sample parameters. Do not reuse a matrix solely because two files have similar marker labels; verify the acquisition panel and controls for the experiment.

Review compensation

1. Open the Compensation panel and inspect the Matrix or Heatmap view. Check the channel labels and matrix values.
2. If computing a matrix, add the required control samples to the workspace and use the wizard to assign them.
3. Use Before/After to compare uncompensated and compensated signals. Use QC Grid to inspect channel pairs and Spread QC to review spread diagnostics.
4. Review affected populations after changing a matrix. Save the analysis when the settings and plots have been checked.

Choose display transforms

Use the workbench transform controls to set the display for each channel. Inspect the values around zero and the range of the population of interest. Keep transform choices consistent when comparing the same parameters across samples.

Changing an axis display is not a substitute for checking the spillover matrix. If a plot looks unexpected, inspect both settings separately before moving gates.

Spectral workflows

Spectral unmixing has a separate workflow in Spectral & high-dim. Use its reference and detector configuration controls when working with full-spectrum data. Confirm the channel mapping and inspect its diagnostics before proceeding to clustering or population comparisons.

Record the correction context

1. Identify the source matrix or reference controls and the samples to which they apply.
2. Record the channel mapping and whether the displayed data has compensation enabled.
3. Keep display transforms with the saved analysis, including any per-channel changes.
4. Inspect representative samples again after changing the matrix or transforms.
5. Recheck gates and downstream population measurements before exporting updated results.

Avoid an ambiguous processing history

Determine whether event values were already corrected before import. If that is unclear, return to the original acquisition export and processing record before enabling another correction. Keep original and derived files distinguishable.

For an unexpected channel pair, first verify the sample, channel assignment, active matrix, and transform. Record which setting changed when comparing the before and after views; otherwise a display change can be mistaken for a correction change.