|
|
|
By
Ethan Winer It's common knowledge that a small room used for playing or recording music doesn't have true reverb. Rather, a typical furnished room has a series of individual reflections that die fairly quickly. In order to obtain true reverb, sound waves must reflect off multiple surfaces, rather than only one or two bounces before being absorbed by a carpet or soft couch. However, an empty room can have real reverberation, even if it's fairly small. I first began to question the conventional wisdom that small rooms can't have real reverb when I noticed how smooth and even the reverb is in my two-car garage. Hand claps sustain without obvious flutter echo, and whistling different notes gives the same uniform extended decay. To my ears this space gives real reverb even though it's nowhere near the size of an auditorium. The live reverb chambers in many famous recording studios are similar in size to my garage or even smaller, and nobody would argue that those rooms don't produce real reverb! After this came up in an audio forum I decided to measure the reverb times in my garage. I also recorded myself yelling and clapping so people can hear it to judge for themselves if it sounds like real reverb: reverb.mp3 |
SIDEBAR: DEFINING THE TERMS RT60: Reverberation Time, abbreviated RT60, is defined as how long it takes for sound in a room to decay by 60 dB after the source stops. But unless the room is very quiet, the noise floor will likely prevent measuring the full duration. In other words, once sound has decayed by 60 dB it's too soft to measure accurately because the room's own background noise dominates. Sound decays linearly, so acousticians often measure how long it takes to decay by only 30 dB, then double that to get RT60. While RT60 is a useful metric, for rooms where music is played it's even more useful to know the decay times in each octave or third-octave band. Modern room measuring software can display RT60 in third octaves, including the freeware Room EQ Wizard program I used to create the RT60 graph that accompanies this article. Q = DIRECTIVITY: The Q parameter is usually associated with filters to define their bandwidth, or range of frequencies affected. But Q is also used to state the directivity of a sound source. A Q of 1 means the source radiation is perfectly omnidirectional, emitting in all directions equally. A human voice has a Q of about 2 because most of the sound goes forward, and less emits toward the rear behind your head. When speaking through a megaphone the Q can be as high as 15 or 20. Taking into account typical horizontal and vertical dispersion, most loudspeakers fall somewhere between these extremes. So a Q of 3.5 was used for the critical distance calculations in this article. This Directivity article gives a more detailed explanation. |
My garage is 24.5 feet long
by 22 feet wide by 8 feet high as shown in the drawing at left. The photos show the setup
with a Dell laptop computer and Mackie HR 624 loudspeaker in one corner, and a DPA 4090
omni microphone in the opposite corner. Here's the reverb decay time (RT60) measured with
that setup:
CRITICAL DISTANCE? Some people claim that you can't have true reverb unless you're beyond the critical distance in the room, abbreviated Dc. This is defined as the distance from the sound source at which the direct and reverberant sounds are the same volume. Pop music recordings usually place the microphones close to the source to get mostly direct sound. But in a reverb chamber you need to be much farther away so the reverb dominates. This formula shows one method for calculating Dc based on the volume of the room (in meters) and the measured RT60:
This simplified formula assumes that the sound source is omnidirectional, which is not the case for either a person singing or speaking, or for a loudspeaker. The formula below includes a "Q" parameter to take directivity into account, and uses feet instead of meters. See the sidebar at left for a more detailed explanation of RT60 and sound source directivity. Dc = 0.03121 * SQRT((Volume * Q) / RT60) The volume in my garage is its length times width times height: 24.5 * 22 * 8 = 4300 cubic feet (rounded for simplicity) We'll use a Q value of 3.5 because that's about what you get from a typical box loudspeaker at midrange frequencies, and we'll use an average RT60 of 1.5 based on my measurements shown in the graph above: Dc = 0.03121 * SQRT((4300 * 3.5) / 1.5) So even 3.1 feet away from the sound source in my garage is enough to be beyond the critical distance. So much for claims that critical distance prevents true reverb from existing in a small room! Understand that our concern here is mainly with midrange frequencies between about 300 and 3,000 Hz. When reverb is used as an effect, low frequencies are routinely removed to avoid a muddy sound. Further, in rooms the size of my garage and the live chambers in pro studios, low frequency decay times are selective and related to the room's modes. It's also important to recognize that reverb calculations are only an approximation. Especially in a small room, and doubly so if the room contains any absorbing materials. In that case the placement of the absorbers - or couches or carpet - affects the RT60 measured, and also causes the measurements to vary around the room. Therefore, the calculations in this article are by necessity approximations because few rooms are totally reverberant. But even allowing for these minor variations, it's clear that a small empty room with reflective surfaces can have true reverb. DON'T TRY THIS AT HOME KIDS Even though small rooms can have measurable reverb when empty, understand that RT60 measurements are not valid for home size listening rooms that are furnished. In particular, don't believe an online calculator that claims to tell you how many acoustic panels are needed for a bedroom studio based on measured reverb times. Special thanks to Amir Majidimehr for his technical expertise and valuable advice writing this article. Added June 19, 2016: Amir did a more detailed analysis of this concept and posted the results HERE in his forum. |
|
Entire contents
of this site Copyright © 2004- by RealTraps, LLC. All rights reserved. |
)
)
)
