A simple new estimator is proposed for direction finding applications which extends conventional phase-only interferometry to incorporate both calibrated phase and amplitude response data from antenna arrays. This is done by appropriately weighting the square of the baseline phase differences with the antenna gains. The incorporation of amplitude data generally provides significant performance improvement over phase-only interferometry with only a modest increase in computational complexity. Furthermore, this performance improvement increases with increasing additive noise and with increasing deviation of the antenna array response from an ideal geometric array response. As such, the new estimator strikes a nice compromise between phase-only interferometry and maximum likelihood estimation (MLE), this latter yielding nearly-optimal performance but at significant computational expense. Performance results are derived analytically for sufficiently high signal-to-noise ratio (SNR) or sample count, and are also demonstrated using antenna array responses simulated with the numerical electromagnetic code (NEC). The simulation results corroborate the analysis and clearly demonstrate that significant variance reduction in direction of arrival (DOA) estimation error can be achieved with the new estimator.<<ETX>>