Skip to content

Worship Music in Church: SPL Levels ≠ It’s Too Loud

Working as a professional audio engineer since 2002 and spending quite a bit of that time in a variety of churches and houses of worship, I am consistently encountering the issue of worship music and the associated politics of volume levels and various facilities setting “SPL Limits” and guidelines for the audio engineer on how “loud” or “quiet” (depending on your perspective) the mix of the worship music should be. Furthermore, the individuals who complain that “it is too loud” in fact may have just reasons for complaint, but may not necessarily know in technical terms how to explain what the problem is beyond “it’s too loud”. Throughout the past few years, and more-so recently, I have begun to realize how little the number the $50 SPL meters display can really paint the picture of why people submit complaints. Allow me to explain.

SPL meters measure the sound pressure levels of the loudest frequency within it’s weighting curve. Therefore, a SPL meter alone cannot tell you if the 95db it is showing is feedback at 1k Hz, a CD playing, or a band on stage. The measurement is purely a scientific reading displaying the SPL at a brief moment in time (whether or not the meter is set for FAST/SLOW response), yet all of us would perceive 1k feedback or even a CD playing at 95db as “louder” than a band on stage at 95db, however a SPL meter will tell us that they are all the same volume…

A SPL meter cannot take into account human emotion. It does not reflect if the content you are listening to is in line with your musical style preferences, if the mix of the instruments and vocals is even and cohesive, or if the moments prior to the point of measurement is a slow building energetic movement of music and emotion of the band & room that is leading up to a high climatic point of the musical arrangement…it only reflects a scientific number.

What am I getting at? Organizations who impose fixed limits of volume based off of a momentary SPL reading is not a proper measurement of loudness used to control noise complaints because:

  1. SPL readings do not reflect the quality of the mix. A band mixed at 95db with the vocals buried will get far more loudness complaints as compared to a band mixed at 95db with the vocals very present.
  2. SPL readings do not reflect personal taste. Many individuals are more likely to be sensitive to loudness if the content of the musical styles are more high energy, urban in nature, or all around not in line with the tastes of the individual. For example, we take an older audience and submit them to a pipe organ driven worship environment at 95db vs a R&B style worship environment at 95db. We all know which style the older audience will perceive as louder.
  3. SPL meters are rarely calibrated. How long have we all had the $50 SPL meters and not once calibrated them? I know I have seen the prosumer meters register lower SPL levels as the battery nears death. If the measurement of SPL is a significant priority, how can we revolve so much politics of volume around a $50 piece of equipment?
  4. SPL readings do not reflect system alignment. If an individual is sitting in a place within the venue where the PA system is not properly aligned/tuned or where they are hearing multiple reflections as a result of poor acoustical treatment, the listeners will be subjected to a distracting balance of the frequency spectrum different from those in other parts of the venue. The resulting hard to describe comb filtering is summed up as “it’s too loud”. However this issue will not be reflected on a SPL meter at FOH.

Ultimately, momentary SPL meters do not represent the SPL level of the whole musical experience.

How should loudness be measured then you might ask? SPL meters are not the enemy, they are simply a tool and it is wise to understand how the tool works in order to use it efficiently. Professional SPL meters have the ability to measure a time weighted average, (aka TWA or Leq/LAT), and some meters even log these for historical reference. In other words, the SPL level the meter is displaying is an average SPL level over the course of a defined amount of time (5 minutes, 10 mintes, etc.). This measurement paints a clearer picture of how loud the environment is over a period of time rather than an instantaneous moment. Setting fixed “not to exceed” PEAK levels based off of momentary SPL level readings handcuffs engineers from expressing their artistic influence by using acoustical energy and the mix to work with the band in creating high emotion/energetic dynamic moments. Instead of “not to exceed” PEAK levels, it is better to establish Leq levels over X number of minutes which allows for dynamic moments to still occur within a worship set.  That time-averaged value becomes the “target” SPL level, and then if desired a higher “not to exceed” momentary PEAK level could still be established that both values are taken into account when determining volume.

Others ask which weighting curve should be applied when measuring SPL. The answer is dependent on what you are wanting to measure. Therefore, it is better to have an understanding of how each weighting curve differs from each other. If you are wanting to have a comparison of your SPL measurements to OSHA’s standards, then you will want to measure Leq A-Weighting over time as all of their regulations are based on. The only reference to PEAK in OSHA’s standards is that no one should be exposed to 140db PEAK, otherwise all of their SPL regulations are based on exposure over time.

Ultimately though, the message to understand is that loudness is more complicated than a simple SPL measurement. Setting defined “PEAK” SPL Levels for worship are not as effective of a definition as compared to a time-weighted-average SPL level. Content, mix, and sound system alignment/acoustics can all have contributing factors into motivating individuals to complain about volume.

Find this useful? Take just a moment and give a $1. Thanks!
Digg This
Reddit This
Stumble Now!
Buzz This
Vote on DZone
Share on Facebook
Bookmark this on Delicious
Kick It on DotNetKicks.com
Shout it
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)
  • Nick

    I think the loudness war that CDs have been going through over the past 15 years also applies to this blog. When you set an arbitrary limit to the amount of level to pass through, there is a lot of leeway in what is perceived as loud and not loud (even though they’re the same level) purely based on instrumentation, compression and efficient use of the available frequency spectrum. 

    Over compressed music tends to wear your ears out over time, whereas uncompressed is more comfortable to listen to at a higher volume. When we set dB limits it forces the engineer to use other methods like over-compression to achieve that emotion

    [Reply]