EQ Credit: Acoustic Performance - Sound System Portion

I corresponded with our audio designers (Fred/Richard) and here is a summary of their comments. Since I submitted my comment already, I was not able to add their thoughts to the public comment. So this is the only place I thought to go. (Fred Schafer's comments) Sound Systems: 1. Large conference rooms seating 50 or more people should also have a distance between the talker and the farthest listener as part of the criteria. In a theater the rule of thumb is 50ft. It may be beneficial to include that a sound system should be considered if the farthest listener is >50ft from the person talking. 2. I agree with Richard’s comment (below) with the addition that these representative points should be defined. STI+0.60 and CIS=0.77 are at the upper end of the “Fair” intelligibility rating. This may be acceptable for a mass notification system or fire alarm system. However, I would like to see an audio system for voice and other programming have at least an STI of 0.75 or a CIS of 0.88 which are at the top of the “Good” range as their minimum. NFPA-72_2010 calls for measurement points on 40ft centers. Using this criteria in typical church (40x90) we could get away with only two measurements. A number of measurements per SF would potentially be a better metric. 3. I also agree with Richard (below) regarding the issue of distortion. The sound system, at its maximum output level should not exhibit a distortion level greater than 1% and at a point 3 to 6 dB above that point a distortion level in excess of 5%. While a minimum performance sound level is nice (70dBA) it would seem that this is referenced from some other document (such as NFPA-72) where they are referencing paging or voice notification levels. The sound coverage at the 2K octave band should also be noted as being measured with a slow meter rather than fast meter setting. This will discriminate against the narrow band filtering that would be caused by the interference pattern between the HF elements of two speakers covering the same area. Masking Systems: 1. Items 1 & 2 are reasonable. 2. Item 3 should have a spectrum reference either from the ANSI masking system document or some other generally accepted reference. Richard's Comments: STI of 0.6 or better corresponds to "good" or "excellent", meaning 95% or better sentence intelligibility (87% or better word intelligibility). Appropriate. Sound level spec should include some sort of peak factor, and a distortion criterion at the specified level, such as "capable of providing average levels of 70 dBA using speech signals, with harmonic distortion not to exceed 5% on peaks of 80 dBA." Otherwise, this could be interpreted as permitting a system to only provide pink noise at a 70 dBA clipping point. Radio Slack would love it! "Minimum" standards have a way of morphing into design-center goals, so one has to be careful. A tolerance of + 3 dB is pretty tight for what will likely wind up being low-budget systems, but then they're only looking at the 2 kHz octave band -- the one most critical to intelligibility. (Just a comment). No comments on the masking criteria.