Lifetime heterogeneity for two-component discrete, continuous, and experimental systems. Panel *A* shows a two-component system with discrete lifetime components (*dashed lines*) of 2.0 and 3.0 ns, respectively, and a system with two continuous Gaussian lifetime distributions of different width centered around the same lifetimes (continuous lines, *CV* = 50% and *CV* = 25% for the 2- and 3-ns component, respectively). Panels *B*, *D*, and *F* show the standard deviation of the distributions obtained by mixing the two components with a relative fraction of 0% (only the longer component) to 100% in steps of 10%. The standard deviation of the true lifetime distribution (*M*, *gray solid line*), the standard deviations as computed by LiMA in its complete (, *dashed line*) and approximate form (a; ○) are shown. The difference of the phase- and modulation-lifetime estimations is represented by diamonds, and represents a frequently used heterogeneity estimator. Panels *C* and *E* show the phase- and modulation-lifetime estimations (*dashed lines*, *τ*_{m} is always greater than *τ*_{φ}) and their average (⋄). The true average lifetime (*gray solid line*) is in good agreement with the values computed with LiMA (○). Panels *E* and *F* show the experimental results from the lifetime determination of a mixture of two different EYFP mutants exhibiting lifetimes of ∼2- and ∼3-ns and widths (*CV*) of ∼50 and ∼20%, respectively. Dashed lines represent *τ*_{φ}, *τ*_{m} (*E*) and the standard deviation (*F*) for the different relative fractions that were predicted from measurements of the two unmixed EYFP mutant components. The lifetime and standard deviation calculated by LiMA are in good agreement with the predicted values.

## PubMed Commons