Compensation is one of those concepts that researchers often accept without fully understanding, which is fine until something goes wrong. When a compensation error produces a population that appears to shift with a marker it should not correlate with, the source of the problem is not always obvious. This article explains what compensation does, how the matrix is calculated, and what the most common errors look like in practice.

What compensation corrects for

Every fluorochrome emits light across a broader spectrum than the detector channel it is assigned to. Some of that light lands in adjacent channels and registers as signal, even though no fluorochrome was assigned to those channels. Compensation subtracts this spillover mathematically, using a correction factor derived from single-colour controls. Without compensation, a cell that is strongly positive for PE will appear falsely positive in the PE-Cy5 channel, even if no PE-Cy5 antibody was used.

How the compensation matrix is calculated

For each fluorochrome in the panel, a single-colour control is acquired. The instrument measures how much signal from that fluorochrome appears in every other detector channel. These values form the spillover matrix. The compensation matrix is the mathematical inverse of the spillover matrix. FlowJo and most acquisition software calculate this automatically, but the quality of the result depends entirely on the quality of the single-colour controls.

What makes a good single-colour control

The positive population in a single-colour control must be at least as bright as the brightest positive population in the full panel. If your single-colour control is dimmer than your experimental positive, the compensation will be insufficient and you will see residual spillover in the corrected data. Compensation beads are convenient for surface markers because they are consistently bright. For intracellular markers or tandem dyes, cells stained with the same antibody and protocol are more reliable.

Common compensation errors and what they look like

Under-compensation leaves a positive population that tilts upward in the spillover channel, producing a diagonal smear rather than a clean vertical or horizontal population. Over-compensation pushes the negative population below zero in the spillover channel, which appears as a population with negative values. Both errors are visible in biaxial scatter plots and are correctable if the single-colour controls are still available. If they are not, the data may need to be re-acquired.

When to recalculate compensation

Compensation should be recalculated whenever the instrument configuration changes, when a new lot of tandem dye is introduced, or when the staining protocol changes in a way that affects fluorochrome brightness. It is not a one-time calculation that transfers between experiments. We recalculate the compensation matrix for every run at Nex Lab Flow Zone, using fresh single-colour controls acquired on the same day as the experimental samples.

If you have existing FCS files where you suspect a compensation error, our data re-analysis service can review the original spillover matrix and apply a corrected compensation. Details are on our services page.