In this month’s tech column, we unpack the concept of reverberation, and the measurement known as RT60. Commonly bandied about by audio professionals, these terms are frequently misapplied or misunderstood. Often when we say reverberation, we mean discrete echoes. Leading into the discussion of reverberation, we will review some general acoustics concepts, and then apply those concepts to the dictionary definition of reverberation. Then we’ll talk practically about applying equalization, and the effect of volume on our perception of reverberation.
Acoustics Concepts
#1: Sound is a wave. An easy, but inaccurate, way to think of sound is like a billiard ball that rolls from one point to another and then bounces off the bumper to head in another direction. However, sound waves expand in space, and when they hit a surface, they scatter off that surface in a multitude of directions, like ripples in a pond. If the frequency of sound is low enough, it will scatter completely around an object, like the object hardly existed.
#2: Acoustic behaviors are caused by room geometry, dimensions, materials and air itself. While this may appear self-evident, it remains important to internalize the root cause of room “sound.” Sound waves move in space and time (i.e. propagate) from the instrument or sound system to the various surfaces in the room, where those surfaces perform some mixture of reflecting, diffracting, and absorbing sound. The interaction of these effects will determine the overall character of the room acoustics.
#3: Acoustics are also influenced by the physical properties of air. Sound requires a medium to propagate in, a way to travel from the sound system to the architecture. That medium is a fluid known as air, which is composed mostly of nitrogen and oxygen. The temperature and humidity of air can have profound effects on how sound travels to the audience and about the room.
#4: Sound covers a huge range of wavelengths. Wavelength is a measure of how far a periodic signal, like sound, travels in space before the wave repeats itself. In the range of human hearing, the shortest wavelengths are on the order of one inch, and the longest sixty feet! This huge range of wavelengths is the genesis of most issues in professional audio. It is hard to provide the same degree of control over both very short and very long wavelengths.
#5: Sound waves interact with room surfaces in a frequency-dependent manner. This is a direct consequence of #3. Most professional sound systems, even small “speakers on sticks” provide good directional control at short wavelengths (i.e., high frequencies). A useful rule of thumb is that directional control is achieved for frequencies above four kilohertz (4,000 Hz or 4 kHz). Above 4 kHz, the sound system’s physical aiming and nominal coverage pattern is representative of where the sound energy will end up heading into the room.
#6: Room surfaces interact with sound in a frequency-dependent manner. A cousin of #2, when sound hits a surface, the surface will generally not return the sound to the air in a uniform manner at all frequencies. It would be convenient if room surfaces all absorbed sound the same way at every frequency, but this isn’t the case. Whether brick or drywall, painted or unpainted, wood floor or carpet, all materials change how the sound arriving from the speakers is returned back to the room.
#7: Room modes always exist, but they rarely behave like the simple textbook examples. Just as physics dictates that pressing down on different guitar frets produces various tones, so too does physical room geometry set up defined modes for all rooms. Of course the model for a vibrating string is one dimensional, but rooms are the much more complicated three-dimensional system. Another practical difference between textbooks and physical rooms is that modes are rarely as severe in volume in real rooms. This is because the boundaries (i.e., walls) of real rooms are not as rigid as those in a textbook model.
#8: Human hearing includes both direct and reflected sound. From birth, our ears are forced to make sense of a world full of both direct and reflected sounds. Our brains fuse a mixture of the direct sound from the speakers and the reflected sounds from the room. This core human behavior is behind why we need consider effects of the room. Generally our ears consider more direct sound at higher frequencies, and more reflected sound in the lower registers.
Reverberation Defined
The section above discusses a variety of acoustic behaviors without ever mentioning the term reverberation. So what exactly then is reverberation? Plainly stated, reverberation is the steady decay of a dense collection of reflections off the surfaces of a space. This dense collection of reflections is termed the reverberant sound field, and often exhibits exponential decay. RT60 is the measure of how long it takes the reverberant sound field to decay 60 dB from its initial volume. See Fig. 1. And RT60 values will vary by frequency, typically being much longer at in the lower octaves, as shown in Fig. 2.
The dense, smooth decay of reverberation stands in contrast to the discrete echoes that wreak havoc on room acoustics. True reverberation, as long as the decay time is reasonable, rarely causes trouble for the mixer or the audience. Instead, it is discrete reflections, usually shortly (<100 milliseconds) after the direct sound that will be distracting to the band, audience or mixer. Now extremely long decay times can be problematic, especially when they are prominent in specific range of frequencies. In other words, if the decay time at 100 Hz is dramatically longer than the decay time at 1,000 Hz, the results can be quite distracting to the audience.
The framework for modern analysis of reverberation was laid down in the 1960’s by Manfred Schroeder of Bell Labs. Schroeder examined at the behavior of reverberation by characterizing the total amount of energy deposited into a space during the decay of the reverberant field. Capturing this detail about energy is called integration, and to this day, Schroeder Integration is used to characterize reverberation.
Schroeder also developed the concept of the Schroeder Frequency. Above this frequency, the allowed modes in a space are dense enough to be considered reverberation, and below this frequency, the room modes are too sparse in space and frequency to be considered reverberation. The problem with room modes at low frequencies becomes that there are not enough of them, and thus they cause noticeable spatial variations in the sound. This is analogous to how discrete reflections cause problems to the listener, while the dense reflections in reverberation typically do not. Generally the larger the room, the lower the Schroder frequency, and the more even the room mode distribution will be.
Equalization and Reverberation
For the FOH mixer, the acoustics are generally outside of one’s control. And once the sound system is put in place, and the audience arrives, placement is fixed for the gig. This leaves the processing tools within the console. Typically this means reaching for an equalizer. While there is no such thing as “equalizing the room,” tonal shaping of sources can be surprisingly effective in influencing the overall perceived balance of the sound system and room acoustics. This is especially true if the reverberation decays differently at different frequencies. The following are four general principles for applying equalization.
EQ Fact #1: Equalization cannot solve spatially dependent issues. If the loudspeakers either over- or under-cover the audience, equalization cannot mitigate how the loudspeakers are aimed. Similarly, equalization cannot change the distribution of modes in the room.
EQ Fact #2: Equalization can be used to balance the combined direct and reflected sound in a room. When the combined output of the room and loudspeaker produces too much energy in a specific range of frequencies, equalization can be effectively used to bring the overall sound level of that range back into balance. This trick works because our brains fuse the direct and reflected sound together as an overall impression of the sound.
EQ Fact #3: Equalization at high frequencies can often ignore the room. At high frequencies (e.g., above 4 kHz) one can often shape the equalization almost without considering the room. This is a combination of three effects. First, most surfaces are absorbing at higher frequencies. Second, the directional coverage of the loudspeakers is well controlled. Third, human hearing more directly considers the direct response at high frequencies ignoring most of the late reflections.
EQ Fact #4: Equalization decisions in the lows and midrange should consider for both the room and the source. Because so much of the sound from the loudspeaker spills around the sides and back of the speakers, the aggregate response of source and room must be considered together.
At the end of the day, equalization is the most common approach to tackle problems that would generally be best solved in another manner (e.g., better acoustics). Despite this reality, equalization can be a powerful tool to balance the overall response of the sound system for the audience. As the direct sound is dominant at high frequencies, the equalization at high frequencies goes a long way towards one’s impression of the signal source. This effect can be further magnified as the humidity changes. Wet air absorbs high frequencies much less than dry, so a room that sounded fine empty at sound check can grow very bright and unpleasant when full of hot, sweaty audience members.
Further, our brains don’t expect a great deal of high frequency reverberation, so equalization is often applied to effects processing to make the behavior seem more natural. By removing the very high frequencies (and the very low frequencies) from a reverb signal, the resulting character of the effects processor will be more natural. Indeed many reverb processors apply this type of equalization as part of their presets.
Conclusion
Reverberation (usually) isn’t the enemy. The dense decay of a space gives it character and richness. Similarly, using reverb effects on a mix adds density and character to the sources. Our brains expect instruments and voice to reverberate in a space! Reverberation goes awry when reflections aren’t dense enough, decay very slowly or have character that changes dramatically with frequency.
A Word About Volume
It is commonly understood in pro sound that one of the best ways to get through a show in a room with very long decay times is to reduce the overall volume. Why does this work? I have my own suspicions about why this technique is effective. First, the surfaces in rooms don’t respond in a uniform manner with increasing volume. As the volume gets louder, walls, floors, equipment, etc. often starts to create other unwanted buzzes and vibrations. By exciting these materials less, their natural internal damping tends to suppress these unwanted behaviors more effectively.
A second effect, and possibly the more powerful one, is that human hearing is not linear with volume. In particular human perception of low frequency volume changes dramatically based on the overall mix volume level. Thus, the expected decay of reverberation changes dramatically based on the playback level. It is my suspicion that this perception shift would be functionally similar to changing the reverberation decay time of the room in a frequency dependent way. The combination of these two effects may indicate why a quieter show tends to be a better show in a challenging space.
Take-Home Tip
A common mistake in the world of acoustic treatment is applying a comparatively thin absorptive material to the wall surfaces in a space to provide additional absorption. Because this absorber is thin, it is more effective in the midrange and high frequencies (such as at 9 kHz, where the wavelength is approximately 1.5 inches) than at low frequencies, where a 20 Hz wave is more than 50 feet long. In such cases, the net result is adding too much midrange absorption in the space, which further highlights the low frequency absorption problems that are not addressed. Instead the space would have been better served with bass trapping techniques.
