Discrete orthogonal polynomial deconvolution (DOPD) is an accurate and
robust method for signal recovery under certain conditions. However,
practical applications have been limited by unpredictable instability.
In this correspondence, the origins of instability, noise effects on
the inverse solution, and appropriate selection of the basis set for D
OPD are addressed.