This paper describes efforts to develop a small in-flow phased microphone array as a reference acoustic sensor and quantify its level-measurement accuracy with new in-flow calibration sources. The devices can be applied to a wide variety of in-and out-of-flow measurement applications and test facilities. Recent tests of the array and two calibration sources in the NASA Anechoic Chamber and Army 7-by 10-ft Wind Tunnel at NASA Ames Research Center demonstrated that the array was functional over a frequency range of 2 to 100 kHz for emission angles of 92˚, 108˚, and 120˚, and Mach numbers of 0, 0.15, and 0.25. The array and calibration sources were then used to assess the relative effects of array installation on background noise and level measurement accuracy by locating the array in the flow (strut-and wall-mounted) as well as outside the flow behind a porous wall screen and a free shear layer. The wind tunnel array measurements were compared with anechoic chamber measurements using discrete microphones at the same source distance and emission angles. At 92˚, the noise floor and attenuation spectra were comparable (within 5 dB) for each installation from 2 to 30 kHz. At higher angles, the array installation behind the Kevlar screen had the lowest noise floor at M = 0.15 and 0.25. The array installation 10" (maximum available distance) behind a free shear layer had higher background noise that is typically reduced to a lower level by moving the array back further from the flow. For emission angles greater than 92˚, both the Kevlar and free shear layer installations exhibited stronger attenuations in measured level, beginning at lower frequencies than either the wallor strut mount configurations. Additional array and calibration source units of the same design are being tested at NASA's Glenn Research Center and Langley Research Center in their aeroacoustic research facilities in an effort to improve measurement accuracy and repeatability.