Audio, Business

Playing the Numbers Game: Who is Responsible When it all goes Wrong?

Last month, I noted that some of the current STIPA meters do not meet the current (or even recent) standards and can give incorrect and misleading results. However, these discrepancies almost pale into insignificance when compared to some loudspeaker manufacturers’ data sheets, download information and claimed performances. Not only is it truly impossible to compare some types of product, but the data itself can be erroneous and misleading.

At times, I would go so far as to say the data can be completely worthless or even dangerous, if the equipment is to be used in an emergency sound system. As I pointed out in December’s “Sound Advice,” a classic example of this relates to digitally steered line array loudspeakers. Some manufacturers provide useful, worthwhile data, while others would be better off filing their info under fiction rather than fact. I still find it amazing that the pro audio industry acts in such a facile and juvenile way. When confronted about their data claims and format, most of the manufacturers will say that they need to do this because their competitors do, and they need to keep up. Perhaps such loudspeakers should be rated in terms of dBM (dB Marketing) or dBF (dB Fiction).

Not only does this not help the industry, but it could be dangerous because, at some point before too long, someone is going to get sued. In fact, given the litigious nature of the US, that this hasn’t already happened quite surprises me. If ever there was a case for standardization, it has to be for the measurement and rating of active line array loudspeakers.

Currently, manufacturers rate what the achievable SPL is in a variety of non-comparable ways, including “peak” SPL (often with no signal or bandwidth defined); program SPL (whatever that means, because there is no formal international standard that defines this term); pink noise (unweighted) SPL; pink noise A-weighted SPL; high-frequency, 3-octave average SPL and band-limited, high-frequency SPL, to name but a few. Interestingly, considering that most of these products are installed to provide speech reinforcement or announcements, no one specifies the actual speech level that you are likely to get; that would be far too useful and comparable!

So what happens when a consultant specifies such a speaker, or an integrator installs one (or many), and it doesn’t deliver the expected/required performance? Who is responsible? Or, to put that another way: Who gets sued?

As a consultant, I have a duty of care to my client and need to apply due diligence when selecting and specifying a product. But what if the data for the product is misleading or incorrect? With traditional loudspeakers, it is relatively easy to spot the bogus or inflated claims and downgrade the numbers accordingly. Equally, on a large or life-safety-based project, I will normally obtain a sample unit and give it a quick check to ensure that it produces what I would expect and is in the right ballpark. However, with large loudspeakers and line arrays much over a meter in length, this, of course, becomes appreciably more difficult to do.

“Ah-ha!” you might think: Why not use a reputable computer modeling program and model the setup or even a single unit because, surely, such programs have strict database requirements and this should give you a sensible answer? In theory, this solution, indeed, would be a good idea and can certainly be insightful…in theory! But what if the loudspeaker data provided is wrong or not measured/provided in accordance with the program’s prescribed method? How would you know?

Some manufacturers are very good in this respect and tell you that the data that such and such a program will produce is what you would measure in practice with a sound level meter and pink noise source. Others keep quiet about what you would then measure. Some seem to ignore the rules and provide inflated data that lead to higher predicted SPLs than reality. Again, it is difficult to tell if the modeling data is correct. Should the modeling program suppliers check every data file before allowing it to be registered and used?

Last year alone, in my practice, we spent more than four person-weeks of time just checking supplied manufacturers’ loudspeaker modeling data and information. We found a significant number of errors and discrepancies in what we had been given or downloaded from manufacturers’ websites. Some problems were not easy to spot, such as the one manufacturer that flagrantly ignored the program convention and provided data that was 3dB higher than it should have been; one had a 15dB error in the data, and yet another still hadn’t corrected the erroneous data that we had pointed out two years previously. The errors can produce a range of effects, varying from incorrectly predicted SPL to incorrect STI predictions. In other words, the audibility and intelligibility of the associated sound system would not have been as anticipated or predicted.

Now, I am the last person to want red tape and lengthy bureaucratic procedures, but the last few years have taught me that perhaps it would be a good idea if all loudspeakers were tested to a universal standard. Not only is a “level playing field” created this way, but the data produced is meaningful and actually pertinent to designing a sound system!

A few years ago, the EN54 series of standards was introduced in Europe for sound system equipment that could form part of a Fire Alarm or, essentially, a Voice Alarm/life safety system. The standard not only sets down a rigorous set of tests that have to be performed on the equipment, but also requires that the testing be carried out by an independent, third-party, accredited test house. Furthermore, the manufacturer’s premises, manufacturing procedures and documentation are scrutinized and audited before certification is given. Some of the detailed requirements do seem a little over the top to me, but at least you can have some confidence in the data that accompanies certified products.

Sadly, there is not as yet an EN54 standard for active loudspeakers and line arrays. However, hopefully, when there is, it may stop some of the inflated claims that are currently being passed off as reality. You may feel that I am being too harsh on the pro audio industry. I would disagree: There are too many well documented cases where the performance of the PA system has been shown to be a contributing factor in an emergency situation that turned into a disaster with consequent loss of life. How many marketing dBs is a life worth? Do we really want to find out?

Sound & Communications: May 2020 Digital Edition
Previous ArticleNext Article

Send this to friend