We report here a technique for processing microcalorimeter data that offers improved live-time over conventional optimal filtering techniques without loss of spectral resolution. Separate filters optimized for pulse amplitude and pulse arrival time (constructed in the usual way from the averaged signal and noise spectral densities) are applied to the entire pixel data stream. Pulses in the resulting filtered streams are then simultaneously fit as the sum of scaled and shifted copies of an isolated filtered pulse template. Analysis using calibration data from the University of Wisconsin/Goddard Space Flight Center X-ray Quantum Calorimeter (XQC) sounding rocket payload demonstrates the ability of this technique to recover pulses separated by as little as the rise-time of the detectors without observable spectral broadening.