Shure's official recommendation for maximum mounting height for both the MXA920 and MXA901 ceiling arrays is 12' AFF. The microphone array in the MXA920 is nearly double the size of the MXA901, so a delta in audio quality between the two as mounting height increases makes intuitive sense.
These recordings aim to evaluate the relative difference in change in audio quality as a function of mounting height. Two heights are tested, 9' and 13' AFF. In addition, each height and microphone is tested with both a very quite backgorund noise level and a more realistic, approximately NC 35 background level.
The common area at Richard Dean Associate's Newburyport office features a vaulted glass skylight ceiling reaching 14' at its peak. Despite the glass and hard surfaces, the complex shape of the room and thin carpet keep RT a reasonable 0.4-0.5s. Additional areas openly adjoin to the central area pictured.
24' x 15'
22' x 18' & 24’ x 12’
7'4", 9', and Up to 14''
None
Painted Gypsum, glass
Glass Ceiling in Main Area, Mineral Fiber Tile in Joined Spaces
Thin commercial carpet
Typical office furniture and chairs
RT = 0.43s Broadband (T30)
Baseline: 26.6 dBA, NC20 | With Fan: 42.5 dBA , ~NC35
Small overall volume and complex geometry tame potential for high RT due to predominance of hard surfaces
None
Studio Six Digital: AudioTools RTA and Impulse Response modules Run on M1 iPad Pro
Studio Six Digital iPrecision Mic: ANSI Type 1 small diaphragm, omnidirectional condenser measurement microphone
Larson Davis CAL200
Captured via 18" balloon pop
For the recordings indicated, a small, generic fan provided a typical level of ambient, broadband noise.
Generic fan used in recordings indicated
NA
NA
NA
Recorded, dry, female speech was played back through a small loudspeaker at normal seated height and normal speech level exactly 10' from the center axis of the microphone. The loudspeaker was chosen to approximate the sound directivity of a human head. The distance was both a normal talker location in the room and at a typical boundary of recommended coverage area for the array in its automatic tracking mode.
"Harvard Sentences" Script From "IEEE Recommended Practice for Speech Quality Measurements." IEEE Transactions on Audio and Electroacoustics, Vol. 17, Issue 3, 225-246, 1969. APPENDIX C 1965 Revised List of Phonetically Balanced Sentences (Harvard Sentences). DOI 10.1109/IEEESTD.1969.7405210 (IEEE standard 297-1969) and 10.1109/TAU.1969.1162058
G. E. Henter, T. Merritt, M. Shannon, C. Mayo, and S. King, “Measuring the perceptual effects of modelling assumptions in speech synthesis using stimuli constructed from repeated natural speech,” in Proc. Interspeech, 2014. Silent audio truncated
VLC Media player to USB audio interface
5", coaxial 2-way passive "cube" loudspeaker, low-Z
48" AFF 10' o/c of the medial axis of the mic array Airmed toward center of table, no tilt
15-watt class-D, tabletop amplifier, low-Z
60 dBA @ 1m, 1s ("slow") averaging
Recordings were captured using the Shure Microflex Advanced MXA920 or MXA901 ceiling microphones as indicated. Configuration was selected as detailed below to offer a realistic and largely "default" set of circumstances.
9' AFF and 13' AFF
Single, Dynamic Auto-Coverage area, 20' x 20'
High-pass filter: 125 Hz, MXA9xx onboard Intellimix DSP, undefined slope
None
Default (MXA9xx onboard Intellimix DSP)
On (MXA9xx onboard Intellimix DSP)
To reflect likely real-world configuration, for baseline, no-noise-added recordings, NR was set to "Low". For the fan-noise-added recordings, NR was set to "Medium".
MXA9xx Dante out to Dante Virtual Soundcard on PC. Audio recorded in multitrack digital audio workstation (DAW)
Both the MXA901 and MXA 920 were recorded at 9' and 13' AFF with the room in its quietest state.
Noise Reduction in set to "Low".
Both the MXA901 and MXA 920 were recorded at 9' and 13' AFF with a generic fan adding broadband noise to the room, bringing it to approximately NC35.
Noise Reduction in set to "Medium".
