Further Sound System Development at Church of Our Lady, London, N1

Implications on Intelligibility of Changing Interior Surface Treatment
Following the recent removal of the church carpet it had been reported to us by Fr. Michael Daley that there had been some comments by members of the congregation that speech intelligibility had deteriorated causing the congregation difficulty in understanding the Mass and that there was thought to be some variation in speech quality depending upon the microphone in use with the Altar microphone probably offering the worst performance.

Our Method of Approach
In order to attempt to isolate the source of the problem our procedure was to inject into one of the microphone input sockets a speech intelligibility test signal and then to measure the quality of that signal using a calibrated speech intelligibility meter at various points within the congregation seating area. Our findings were readings which varied from 0.3 to 0.51 all of which represent very poor levels of intelligibility and where one would anticipate that listeners would find understanding speech very difficult. The tests were then repeated using each loudspeaker in isolation having first switched off the  remaining five loudspeakers: as we would anticipate that, where the problem is acoustic, that the speech intelligibility readings would improve when only one loudspeaker is active. It is noteworthy that the readings still only improved to an average of 0.51 compared to our readings of as high as
0.59 even with all loudspeakers operating when the carpet had been in place.

The apparent variation in intelligibility depending upon the microphone in use can be explained by the more directional qualities of the gooseneck and radio microphones (with the associated consistency of user) whilst the very broad pick up pattern of the altar microphone situated directly under the concave ceiling of the Sanctuary would place it at a severe disadvantage due to the undesirable acoustic reflections received from the surrounding surfaces which would, in turn, be injected into the sound system simultaneously with the required speech.

With a building interior such as Our Lady & St. Joseph where all of the surfaces are acoustically highly reflective and where there is, in addition, the considerable disadvantage of a concave ceiling, achieving acceptable speech intelligibility will always be difficult.Whilst the system electronics were working correctly and offering a ‘clean signal path’ we concluded
that the loudspeakers, which were installed around 30 years ago, were now incapable of delivering the required intelligibility in such a difficult environment and so replacement was the only solution. Fortunately, the church had recently benefited from a full re-wire of loudspeaker cabling and so much important preparation had already been completed. Our objective at the conclusion of these works would be to to recover the earlier level of  intelligibility of 0.59 (as measured with a carpet in place) and preferable achieve a reading of 0.65 or better.

Six new loudspeakers were specified and whilst being of a tradition column type appearance making them architecturally acceptable, the modern design incorporated a high degree of passive beam steering enabling us to aim the loudspeakers onto the ears of listeners and to some extent away from reflective surfaces.More power is required for these loudspeakers to operate correctly and so it was decided to take full advantage of the new individual cables feeding each loudspeaker and to install higher power but  more electrically efficient Class D amplifiers, one for each loudspeaker with a total output of 250W each. For improved cooling and overall system conditioning; the entire system was built into a new lockable equipment cabinet. We can confirm that our objective has, indeed, been achieved with speech intelligibility being restored and slightly improved to readings on average of 0.62 within this very difficult environment.