This article [Opt. Eng. 50(1), 016001 (2011)] was originally published on 6 January 2011 with errors in Eqs. (51) to (57). They are corrected below.

The approximation written in Eq. (51) should have included an ${\omega }^2$ dependence inside the exponential:

The expression for $A$ in Eq. (54) should have been written using the limits

The negative sign in Eq. (56) should have been inside the exponent:

A factor of $2\pi$ was missing from Eq. (57):

The use of $V(Q,X)$ was based on the approximation of Eq. (51) where the primary $\omega$ dependence was quadratic. The function was evaluated at the $-3\mathrm{dB}$ point of the $M_S(\omega )$ curve that yielded $V>1$ under certain conditions. This was incorrectly interpreted as super-resolution behavior. Further research indicates the approximation expression improves performance at low to moderate frequencies but overestimates responses at high frequencies at moderate turbulence levels. In general $V$ is a moderate function of $\omega$, $V(Q,X,\omega )$, that falls below unity at high frequencies $\omega$ in such a way that diffraction limits are maintained. Enhanced responses ($V>1$) often appear for $Q>1$ and $\omega \approx 1/3$.