Small Room Acoustics
Last updated on 7/08/10
(simplified, expanded, errors corrected)
The most commonly overlooked aspects of a recording studio is the acoustics. Acoustics is the behavior of sound. Sound refers to waves of compression/decompression travelling through any medium. Without good acoustics, it is almost impossible to get a good recording, mix or master. Of course the most important factor is the source of the sound, meaning the musicians and their instruments. However, a bad room can destroy the sound of a guitarist's $3,000 amp no matter how good the microphone is. A bad room will make a $5,000 drum set sound played by the world's greatest percussionist sound like an amateur playing on a $200 S-mart set. After all, the sound must filter through the room to get to the microphones and there's no way to get rid of the room sound once it is there. If an engineer cannot properly perceive the sound coming out of the monitors, they will make poor decisions about that sound. In order to have a stellar sound, you must have a stellar room. What exactly makes a good room though? Some aspects of acoustics are purely subjective and up to personal taste. Other aspects are seemingly universal. There are so many factors involved in the quality of a room, there are literally volumes of books covering the science and art behind acoustics. Many times, there's more art than science to such texts. Never-the-less, some concepts are fairly simple to grasp. Hopefully, a vague idea can be understood enough to allow one to chose their room and treatments effectively enough to be usable for their intended purpose. At the very least, one can use the knowledge to chose somebody to handle those details on their behalf. Regardless, acoustics are one of the largest contributors to both live and recorded sound. This is important because not only can a bad room influence the recorded sound, playback in a poor room can also mislead a listener as to the sound quality that was actually recorded.
Reflections are probably the most important aspects of acoustics. Reflections, of course, are sound waves
bouncing off of objects. They are a part of everyday life
and help us comprehend our location, the direction and distance of
objects etc. Note that the angle of reflection will always be the
same as the angle of attack just like light to a mirror. Since no surface is perfect, there will
always be some blurring of the reflection.
Sound travels at certain speeds in given media, such as 1,130 feet per
second through air at sea level, room temperature. Since the shortest
distance between two points is a straight line, a listener will hear the
original source of a sound before the listener hears a reflection of the same
sound. The reason for this is because the reflected sound must travel
towards a boundary first, then towards the listener while the original sound
travels more or less straight to the listener. Reflections heard
along with the original source sound are the cause of echoes in a
natural environment, though echoes can also be simulated by other
means. Reflections can cause a lot
of harm to sound quality for many reasons but they can also be a good
thing. They can obscure the true
source of a sound under certain conditions and reduce intelligibility. The gap between a sound source and its
reflection is often called a delay.
But as previously stated, they can also cause a lot of problems. Longer echoes are generally a less
offensive than shorter echoes.
Echoes can cause phase interference. Phase refers specifically to the state of a
wave, either at its peak of compression, trough or anywhere in between If
a peak and a trough pass through the same space at the same time, they cancel
each other (like positive and negative charges). When multiple peaks
interact, they reinforce each other just like multiple troughs will. In
turn, some sounds get reinforced while others get cancelled. This effect varies throughout the
spectrum regularly. When similar complex sounds in different phases
interact, the effect is called comb filtering and is almost always undesirable
as it obscures detail and harms intelligibility. A flanger effect relies on a
continuously varied short delay to cause phase cancellation. In this case, it is a desired special
effect. Generally speaking though,
delays of less than 5ms are particularly harmful and if they cannot be avoided,
one should at least take care to minimize the effect by careful placement of
objects in relation to boundaries and microphones.
In the natural world without walls or ceilings, the First Significant Reflection will always come from the ground. The FSR is the first reflection within 6dB of the loudest reflection in an ambient field. The reason this is important is humans subconsciously use the FSR to determine distance from an object. When an object is very close, such as a person speaking to a listener from 2' away, the initial sound will take about 2ms to reach the listener while the FSR will be about 11ms. The FSR is effectively 9ms to the listener because the difference between direct sound and FSR is what is actually heard. If the speaker is 10' away, the FSR will be about 5ms from the listener's perspective. So the further away the listener is, the shorter the Initial Signal Delay gap, or the time difference between the direct sound and FSR is. In a situation where the FSR is very short, it obscures detail due to comb filtering and other phase issues. The subconscious assumption to a listener is that the sound is more distant even if it's actually very close. This applies to recording situations as well. So in an ideal situation, the FSR will be greater than 15ms because comb filtering and apparent distance is less severe. These time delays between direct sound and reflected sound can be calculated using Pythagorean Theorem for geometric distances. A reflection 10ms-40ms (20 optimally) after the initial sound can actually help one localize the source of a sound. In fact, echoes within this range can be up to 10dB louder than the original source without obscuring it. This is called the Haas effect and many studios actually have specular (meaning, mirror-like) reflectors called Haas kickers in place to deliberately cause this effect. The Haas effect is best used in moderation though and is most effective when at least 10dB BELOW the direct sound in level. This is because even within the range for Haas effect, comb filtering can be a problem. On another note, a parabolic dish can be used to reflect sound into a microphone for the sake of intensifying it just as satellite dishes focus radio waves from long distances into antennae. In such a case, the reflection in a parabolic microphone system is a very good thing.
Having several open microphones in a given environment can have the same effect as a reflection. For instance, if more than one instrument is being recorded at a time, there will likely be multiple mics in the general area. The sound from one instrument will not only be received by its intended microphone but by any other nearby microphone as well. Multiple in-use microphones can produce results similar to reflections except the sound takes a direct path to both points instead of actually being reflected. There is still a delay because the microphones are different distances from the sound source. A mic 5 feet behind another will add about a 5ms delay. Conversely, a microphone in front of a sound source with a wall 5 feet behind it will cause a 10ms delay as the sound must bounce off the wall, then back to the mic. Based on the rule of thumb for sound traveling at 1,130F/S, the distance between microphones in relation to the sound source alone will determine the delay between signals. So a microphone 1 foot away from a guitar cabinet and another microphone 5 feet away will add a delay of approximately 4ms between mics. Sometimes it is impossible to avoid delayed signals between microphones but the effect CAN be reduced by carefully regulating the distance between sound sources and microphones. It may be necessary to get a mic closer to its intended sound source to allow the level of that source to be reduced and thus reduce bleed into another mic. One can also use baffles to reduce the direct sound from one source accidentally being received by an unintended receptor. In case of baffles, it is important to angle them to avoid reflecting a sound into a nearby microphone as this can be more damaging than bleed into another microphone. Microphones can also be angled so that their point of maximum sound rejection is pointed toward the unwanted sound source. Multiple sources for a given sound can also have a similar effect to multiple mics for a single source. On a side note, having many microphones in different parts of a room or worse yet, in different rooms, may be subconsciously disturbing to a listener. This is because one who hears a recording made in this fashion is hearing ambience of different environments at the same time which does not occur in nature. It is for this reason that the most natural sounding recordings tend to be recorded using a single stereo pair of microphones in a fixed position of a room. Also, this is why using multiple artificial reverberators is generally not a good idea unless they are blended together in some fashion.
The most important contributor to the sound of a room is the size and shape of the room itself. When a room has parallel walls (almost all rooms), sound gets caught bouncing back and forth between those walls. Some sound waves are cancelled by their own reflections while others are reinforced. This interference changes according to the position of the sound source and the listener or microphones. Having a room where all would-be 90 degree angles are offset by 10-20 degrees will drastically reduce the redundant reflections between walls and thereby almost eliminating the would-be standing waves. It is not recommended to have angles of less than 70 degrees because long wavelengths tend to be reinforced in corners. Waves that are long have broader ranges of reinforcement and corners act somewhat as traps for the fluctuating air pressure. In general, the more obtuse the angle, the less problematic sound waves tend to be. There is an entire industry built around sound absorption just for corners but this will be covered at a later time. To further explain standing waves, a sound wave traveling from a speaker past the listener and bouncing off the back wall will come back to the listener. This potentially increases the level at that particular frequency by 3 Decibels (AKA dB). Decibels are a logarithmic function of sound energy where 3dB equals a factor of 2. That is, a sound level increased 3dB is doubled in power. If the phases are opposite, though, the intersecting sound waves can almost completely cancel each other. This is of course an ideal which is physically impossible because after all no reflection of energy is 100% efficient but the effect can be very close. The standing wave will also reflect off of the front wall and back to the listener causing the overall sound to be further reinforced (or cancelled). This back and forth motion will reach a point of diminishing return of course due to absorption and scattering of the sound but a standing wave can still be over 9dB louder than free-field waves in extreme cases.
Since the wavelength of the standing wave has a direct correlation to the distance the sound travels within the room, it is commonly called a room mode (or node). One can determine room modes by dividing the speed of sound by the room dimension in question. This will show which frequencies in that environment will be standing waves. If the speed of sound is 1,130 feet per second and your room is 30 feet long, the equation will be 1130/30=37.667. 37.667Hz is the fundamental room mode but multiples of that frequency are also modes such as 75.334Hz. Total cancellation is a more serious consequence of room modes than reinforcement because the information is missing rather than merely exaggerated. What's worse is what's reinforced in one position of a room may be cancelled in even a slightly different one. This is only looking at one room dimension. Standing waves can occur between any of the three dimensions of a given rectangular room. For this reason, it is quite common for studio control rooms to be bilaterally symmetrical pentagons, hexagons or octagons where two or more walls are longer than others. This avoids both parallel walls and acute angles. The front wall, which usually contains the studio window, is often one of the smallest and angled slightly toward the listening position at the top. The ceilings are quite commonly angled as well to prevent standing waves between the floor and ceiling. Generally speaking, the front wall containing the window will be the lowest point for the ceiling and it gently slopes upward toward the back walls. This helps avoid sound from the monitors being reflected to the listener. Rectangular studios are acceptable under the pretense that the ratio of length to width to height are appropriate and that the room is large enough to allow standing waves to diffuse naturally over a short enough period of time. Redundant room dimensions or evenly divisible room dimension ratios will compound standing waves so this is why room shape is almost as important as room size.
Refraction is yet another phenomenon of acoustics. Sound can be bent with changes of density in air just like changes of density cause light to bend when passing through a lens. Sound does not always travel at the same speed. Things like humidity, air pressure and temperature greatly effect the speed of sound. Sound travels faster through dense material faster than less dense material. When high humidity meets low humidity, cold air meets hot etc., the direction of sound gets skewed due to the change in speed because the density of the material changes. A room with a cold floor and a hot ceiling will cause sound to bend as it travels through the room. While this can be a very large issue outdoors, it's not normally a problem in studios. In the morning when the ground is getting hot but the air is still fairly cool, sound tends to bend upward. The P.A. system at the famous race car track in Indianapolis actually has the speakers pointed below the audience because most races are in the morning. Thousands of dollars in power and audio equipment are saved by relying on this fact. In the evening, it's best to have the speakers set fairly high or pointed slightly over the audience's heads. The cooling ground with warm surrounding air will cause the sound to bend downward into the audience. One of the famous sermons from Jesus was made at the bottom of a hill while the audience was on higher ground instead of the expected other way. Jesus was addressing a very large audience that would otherwise be unable to hear him speak if he had been on high ground or if it were evening. Since this was in the morning, the audience was in the optimal position to hear him speak, especially since the hill helped capture the speech and block out extraneous noise.
Resonance can be a huge and often overlooked problem. Even if great care is taken to avoid all kinds of modes, the various materials and even the air of the room itself can vibrate in coincidence with another sound source. Resonance is a frequency at which a specific object vibrates naturally. Tap a wine glass and hear the resonance of the crystal. It goes beyond that though, if a sound is present that happens to contain the resonant frequency of a nearby object, the object can amplify it that tone. The classic cliche' of an opera singer shattering a glass with their voice is no joke. If a source sound is loud enough and close enough to the resonant frequency of another object, the object can actually destroy itself. Never-the-less, an object subject to resonance can cause sonic artifacts similar to standing waves in a room. Acoustic instruments rely on resonance to amplify the low level of their own sound. The body of an acoustic guitar amplifies the sound of the strings through resonance of the wood and air chambers. Walls, while not normally very good sound boards, can resonate as well. Putting sound damping material on these types of objects should do well to reduce resonance. It can be a big problem for monitors to transmit vibration through the objects on which they sit causing them to resonate as well. Isolation is the answer in this case. If the monitors are on a large surface, they should be moved and placed on speaker stands which are not as apt to resonate. Even at that, it may be necessary to have sound absorbing material between speaker stands and the monitors they support. It can be a good idea to put instrument amplifiers on sound damping material as well because the floor can add to resonant problems. All rooms have specific volume and the air in them is elastic in nature, making it ideal for resonance. A cavity that contains a specific volume of air intended to vibrate is called a Helmholtz resonator. A good example of a Helmholtz resonator is an ocarina or blowing across the mouth of a bottle. A room can also act as a Helmholtz resonator. Though it's almost impossible to completely get rid of resonance, just like reflections, this problem can be tamed by treatment. Strangely, a resonator covered with a thin membrane can be added to a room to actually REDUCE such issues. This works by displacing the problem frequency into the resonant cavity. Materials good for reducing resonance are generally flexible, very dense, but resistant to movement such as rubber or once again, compressed fiberglass.
Transmission is a great issue for isolation purposes. Though beyond the scope of this article, a lot of money goes into reducing transmission of sound through walls in studio situations. All objects transmit sound but some are better at it than others. Isolation booths are often placed on small pieces of rubber which does not transmit sound very well for that reason. Rubber is also often placed between wall surfaces such as gypsum board and their supporting studs to reduce transmission through the studs themselves. Minimal numbers of screws or nails are used as they transmit high frequencies very well. Air is a great insulator so by using as few studs as possible and using very broad studs such as 2x6 instead of the usual 2x4, isolating one room from another is much easier. In fact, many recording rooms are actually smaller rooms placed inside larger rooms. The larger room is generally a large rectangular prism shape while the room inside it is suspended (usually at an angle to the surrounding room to reduce standing waves and resonance between walls) with cables or placed on thick rubber feet similar to isolation booths. In general, thermal insulation material is also good sound insulation and electrical insulation. An exception to this is the expanding foam that many people use for insulation of houses. The reason is, even though it's a highly porous material, it dries hard and is thus subject to resonance itself at a multitude of frequencies. Hard objects generally transmit energy better than soft objects. Soft, dense objects tend to absorb vibrations because they can compress and stretch with air while density adds to their ability to catch vibrations instead of simply letting the vibrations pass through them.
Reverberation (reverb for short) is defined as multiple random echoes which are so dense, individual echoes cannot be distinguished. Early reflections are the first reflections to be heard by the listener before the reverberant field forms. These are followed by the diffused reverb tail. The reverb time refers to how long it takes for the ambient field to decay by 60dB from the initial sound and is noted as Rt60. Rt60 for a room is a function of the room's spacial volume. As the volume doubles, so does the reverb time. Now, the actual Rt60 of a room is not only a factor of volume but also the reflection coefficients (the ability to reflect sound) of the materials used in the room. Therefore, the development of reverb is a parabolic function of a given room's properties depending on not just size, but also shape and reflectivity. Small rooms by definition cannot have a true reverberant field because the reflections never become completely dispersed by the time the echoes have sufficiently decayed. However, this is a flaw in the definition of a small room because as the properties of the room change, so does the diffusion of echoes within that room. A room that is very small in size could have a reverberant field if it is reflective enough and shaped in such a way as to avoid flutter echoes, or echoes which occur at rapid predictable intervals (usually due to close, parallel walls). Flutter echoes can become fully diffused eventually. Since a reverberant field is timed by how long it takes the sound to decay 60dB from its original level, flutter echoes that become diffused after they decay to 61dB lower than the initial level do not quality as reverb. On the other hand, if they become diffused at -59dB, this could be considered reverberant, but probably not for any practical purposes. Go into a normal room in a house such as a living room or bedroom and clap your hands crisply once. Most likely you will hear your clap sustained as a fluttering or robotic sound in such a room. That is a flutter echo. Now repeat the same process in a gymnasium or a concert hall and you will hear a smooth ambient sound. This is reverb. Hear the difference between a small room with a flutter problem verses a diffused reverberant field. Both samples were taken from the same room treated differently. The conundrum becomes apparent because acoustician definitions state that small rooms do not have reverb but large rooms do. One has to consider the reverb example a large room while the flutter example must be considered a small room, even though they're both from the same room. A large room can also be treated with sound absorptive materials in such a way that only a few echoes can be heard before they decay completely. Definitions aside, flutters are almost universally considered problems while reverb is almost universally considered to be good.
Room dimension ratios play heavily into the overall quality of sound obtainable in a given space.. It is imperative to avoid redundant dimension and multiples of redundant dimensions in an acoustic environment. It is logical to assume that a room 12' x 8' x 10' will be far better than a room that is say 10' x 10' x 10' because standing waves would otherwise be compounded. A room with equivalent multiples such as 10' x 20' x 8' will also be a problem because 20 is evenly divisible by 10 as well as other common factors. Therefore, it will have similar problems with standing waves as a 10' x 10' x 8' room. However, it will not be as bad in the a 10' x 20' room is not as bad because only half the standing waves along the 10' dimension will match up with the 20' dimension (not to mention overall size helps). One way to find out if your room has acceptable ratios is to measure the room in all its dimensions then try to divide each dimension by the others. So if you have a room that is 28' by 23' by 8', follow the process where 28/23=1.22, 28/8=3.5 and 23/8=2.88. The overall ratio of this room is 3.5:2.88:1. Since none of these ratios are close to being even factors, this room will be acceptable. Certain room ratios sound better than others even if they may all be mathematically acceptable. The reason for this is due to the natural overtones of sound. A single note played on an instrument actually contains many tones at a time. Even though the fundamental frequency of such a note may not create a standing wave, its overtones might. Not only that, a chord (three or more notes played simultaneously) may contain several wavelengths which coincide with modes of room dimensions. Therefore some more complicated theory is needed. Ancient Greeks found that rooms with the ratios of 2.62: 1.62:1 sounded universally good. Though this ratio is not perfect in all cases, it is good in general. In the case of our 28 x 23 x 8 room, there are only two coincidental modes throughout most of the audible spectrum which is quite good. You can calculate redundant modes by plotting out all possible modes and then finding which ones coincide. A faster trick is to simply look for common factors among the dimensions. There are many resources that will plot them for you though. One is the Room Mode Calculator at http://www.realtraps.com/modecalc.htm. This will only calculate axial room modes. That is, it will only chart modes between parallel pairs of walls. There are also tangential and oblique modes which are modes caused by the interaction between several dimensions. These modes tend to be less intrusive than axial modes but are important never-the-less. Since there are no known resources for calculating all three mode types together and charting them visually, one must hand chart them. A good resource is http://www.marktaw.com/recording/Acoustics/RoomModeStandingWaveCalcu.html for calculating all three types of modes.
Treating the Problems
All that having been stated, naturally bad sounding rooms can be treated to give acceptable results using a combination of diffusers and absorbers. Learning to properly set up these materials can take a lifetime but a few simple principles can be followed to gain a basic understanding. Absorbers, like one canguess, are intended to absorb sound and reduce echoes within a room. Diffusers scatter echoes rather than absorbing them so that instead of having a single large echo off of a given surface, there will hopefully be many small echoes travelling in several directions. Of course all absorbers and diffusers work on different principles and thus behave in different ways. Due to the long wavelengths, low frequencies are particularly difficult to absorb or diffuse so typical absorptive materials are not generally effective for low frequencies. Common acoustic treatments such as Auralex or Sonex are VERY good absorbers at high frequencies (short wavelengths). Low frequencies tend to just pass right through them and bounce off the boundary to which they are attached unscathed. In order to effectively damp low frequencies, the absorber/diffuser must be dense enough that it impedes the motion of the free field sound. It also must be at least 1/4th the depth of the desired effected wavelength. Waves are measured in degrees where a full wavelength is 360 degrees. All waves are at a complete standstill when reaching a boundary because the wave compresses the air in front of a boundary. The air then acts as a spring board sending the wave away from that boundary. A wall will thus always be 0 degrees in relation to any waves hitting it. The most absorption occurs between 45 degrees and 135 degrees of a wave since 90 degrees is the peak of the wave's energy (1/4 wavelength). The wave has a trough at 270 degrees, which is also oddly another peak of energy, but this is less relevant to the subject. In order to sufficiently absorb the longest wavelengths perceptible to human hearing, the absorber must theoretically be at least 14' thick. The formula is stated as 1,130 (feet per second) /20Hz= wavelength of 56.5'. 1/4th of 56.5' would then by 14.125. The results in practical function are not totally true as one may suspect because extremely low frequencies (below 30Hz) tend to simply pass through many wall materials, including gypsum, and are thus less of a factor.
A sufficiently dense, solid but semiflexible object placed a certain distance away from the wall can act as an absorber for specific wavelengths. This is called a diaphragmatic absorber. The idea is waves from the initial sound cause the material to move. The sound also passes through the material so it is reflected by the wall behind it. The sound bouncing off the wall travels opposite to the diaphragm which is still moving primarily in coincidence with the direct sound. The opposing forces cause the reflected sound to be greatly reduced in power. An absorber for a specific frequency is calculated as follows: 170 divided by the square root of Md. M is the density in pounds per square foot of the material used and d is the depth of air space in inches between the material and the wall. So if you have a piece of plywood that weighs 2lb per square foot and space it 3 inches from the nearest wall, the absorption frequency will be 69.4Hz. The absorption only reduces the amplitude (power at a given point) of the reflected wave but it does have the advantage of making room modes broader and less extreme. If one were to use a porous material that is both dense and flexible, much broader ranges of wavelengths can be absorbed. A very common material and one of the best, is semi-rigid fiberglass. Fiberglass products such as Owens-Corning 703 is about 3lb per cubic foot, is fairly solid and is sold in large sheets making it ideal for mounting on stands, suspended from the ceiling or using small spacers to stand it away from walls. Keeping absorptive materials away from the wall allows said material to be more effective at absorbing low frequencies because the motion of waves is always nulled at the wall. Since it's a porous material, the diaphragmatic absorption equation is fudged by 4 inches (or whatever the thickness of the material is). All material between the front surface and back surface factors into the equation. Also, due to the porous and dense nature, high frequencies are absorbed very effectively as well. The idea, though, is to make the absorbers work as evenly across the spectrum as possible. Typical acoustic foam is more effective at absorbing high frequencies but not as effective at soaking up low frequencies. So even though typical acoustic foam may have an absorption coefficient of 1.2 (1 means 100% of sound is absorbed), that pertains primarily to high frequencies while below 250Hz or so is often less than .2 (20% is absorbed). With semi-rigid fiberglass, .9 for high frequencies is typical but .8 is common for low frequencies. Beware that when acoustic tests are done, they only account for the wall surface area covered by the absorptive material which is why absorption coefficients (also called Noise Reduction Coefficient or NRC) above 1 can be seen. The testing facilities fail to factor in back, side, top and bottom surfaces of the absorptive material itself and thus has a huge error in NRC calculation. Never-the-less, 4 inch O-C 703 mounted spaced and slanted away from walls will absorb across the audio spectrum quite evenly. The slant makes for a varied distance from the wall in order to increase the width of it absorption range. Please note, it is best to not kill an entire wall but to place absorption in key, but scattered places around the room. Treating small, scattered areas rather than large areas can increase the efficiency of the absorption up to threefold. This is because sound travelling through a space has more likelihood of hitting absorption rather than bare wall. More surface area of the absorptive material is exposed as well this way which also improves efficiency.
With diffusion, similar principles apply as to absorption. However, the idea is to scatter sound, not absorb it. There are likewise, many ways to do this, some are more effective than others. The most simple way is a convex curved surface. However again, it must be sufficiently dense otherwise long wavelengths pass right through it and bounce of the wall behind it. There's several commercially available diffusers on the market from many different manufacturers. Some of them rely more on curved surfaces to spread any reflection throughout the room while some rely on Phase Reflective Diffusion, which means the object has many small parallel reflective surfaces to create reflections of different timing. This equates to many small reflections rather than a single, hard echo from the untreated surface. There is a newer kind of diffuser in the last few decades called a Reflective-Phase Grading, which uses multiple reflective surfaces divided by thin walls to create wells of varying depths. This creates small resonant pockets of air that compress and radiate widely throughout the room. The most effective versions of these are Quadratic Residue Diffusers and Primitive Root Diffusers. These are all based on the same basic principles but the math behind determining the depths of the wells is different. It should be noted that Auralex seems to prefers the Golden Ratio for making diffusers. The Greek's theoretical ideal acoustic ratio of 2.62:1.62:1 is applied to the depth of the surface of the diffuser in addition to other surface variations. While this is not known to be the best, it is inexpensive to manufacture and when treated further to increase density (they're made of thin hollow plastic), can be very effective for a significantly lower cost than RPG diffusers. Another factor is that most RPG type diffusers only scatter sound in two dimensions while Golden Ratio diffusers such as the Auralex T'fusor scatters in all three dimensions. It should be noted that PRDs can also come as a simple, less expensive 3D diffuser called a Skyline Diffuser though is still quite costly at several hundred dollars USD for four square feet. There is also something called a Binary Diffuser which is essentially a flat board with holes in it to allow some wavelengths to pass through it while reflecting others. The surface behind the diffuser then reflects the remaining sound. This is a cost effective, low profile way to diffuse sound but is not nearly as effective as the QRD.
Another commonly overlooked aspect of acoustics is diffraction. This is a strange phenomenon where sound passing an object can seem to radiate from the edges of that object. Yagi antennae used for UHF TV signals have many elements, but only one pair of elements is actually electrically fed to the receiver. The rest of the elements use diffraction to redirect stray waves into the receiver's elements. The iris in a camera lens used to control exposure causes diffraction of light as a side effect which softens the image. The size of an object influences the effect of diffraction proportionately to the wavelength. Small objects effect shorter wavelengths while larger objects mostly effect longer wavelengths. When sound waves pass through gaps of otherwise occupied space, the sound will appear to originate from those gaps. In the case of multiple objects in a room between monitors and a listener, the waves radiating off of the edges can interfere with both each other and the direct sound. In a control room, the listener needs the purest direct sound possible from the monitors. It's best to avoid having objects between the monitors and the listener. Even the desk sitting below head level can cause issues. Video monitors should be placed between the audio monitors, slightly further away from the listener to avoid diffractive interference. The phenomenon is not always bad though. A large grid made of some solid, nonresonant material placed in a room can spread sound without the sound ever reflecting directly off of anything. The room can seem more spacious with less interaction between areas of the room at the same time. So in a recording room, diffraction can be very useful. Shotgun microphones use diffraction to cancel out sound coming from the sides, thus making it more directional. The diffraction is caused by a long tube with slits cut in it (usually along with an additional grid-like screen). Loudspeakers can be made highly directional with such gradient tubes also.
The animation here shows sound waves hitting a wall with holes in it and the diffracted waves emerging from the other side.
When treating a room, some people try to simply kill all reflections using absorption. Even if it were possible to remove all reflections, it would leave a very unnatural sounding room which would not translate well to a real world environment. Pure diffusion without absorption can also create many problems as well. Simply scattering sound can lead to a cluttered environment. Clarity of sound is more difficult to achieve because of the high reflectivity and diffusion of the sound field, especially for a mixing environment. Most studio control rooms are based on the LEDE method of controlling sound. LEDE means Live End, Dead End. Between the front wall where the monitors are and the listen position primarily has absorption while from the listening position to the back of the room is left live. The LEDE idea originally came from the notion that the listening position should be completely devoid of early reflections so that the early reflections coming through the mics in the recording room can be heard. However, it is impossible to completely remove all reflections, especially in extremely long wavelengths. A better idea is to extend the ISD gap for the FSR to 10ms or greater. The amplitude of early reflections should be at least 10dB-15dB below the direct sound as a good compromise over the theoretical LEDE technique. Most rooms incorporate combinations of many styles of treatment. Broad band absorbers are put in all corners to reduce bass resonance while spot treatments are placed on the ceiling and on the side walls between the monitors and listening position. More spot treatments are often added to the wall directly behind the monitors. This creates a Reflection Free Zone of sorts where the listening position is relatively low in direct, early reflections. Diffusion is used to prevent any hard echoes from reaching the listener directly and hopefully create more of a reverberant field than flutter echo. It has been said, though, that if the boundary in question is within 10' of the listener (be it human or microphone), then absorption should be used instead of diffusion as diffusers need the space to scatter sound sufficiently. In a recording situation, there is no dedicated listening position so a mixture of diffusion and absorption with gaps in between treatments is ideal. Some very good rooms may simply have diffusion. Spot treatment is also a great idea if there are designated areas for certain applications such as putting absorption on the ceiling over a drum area.
A great trick for finding places to spot treat in a control room is to have a friend move around the room with a mirror. Somebody should sit in the listening position and see if the monitors can be seen in the mirror. For instance, if they place the mirror on the mixing console and the monitors can be seen in it, this will be the point of a significant reflection and should be treated. Of course one cannot put acoustic tile on their mixing board so instead, the monitor and board positions should be adjusted so that the reflection from the board stays below neck level. The object is to redirect the reflection in a direction that will not allow it to be heard. The ceiling is especially important because it is usually the single largest, closest boundary to the listener aside from the console. Don't neglect the fact that a reflection from the ceiling will also reflect off of the console, making the ceiling that much more important to treat. In recording, treatment may be used between microphones and the sound source, though a reflection may also be desired as an effect. In the case of a small amplifier where the speaker center is 1' off the ground and the microphone is 3 inches from the speaker at the same height, the ISD will be 2ms. So either very heavy absorption needs to be placed there or the amplifier and mic should be raised off the ground to reduce comb filtering. In recording, it is usually desirable to have the ISD gap to be greater than 5ms for maximum clarity. Please note that the listening position of a control room should cover a range of motion. In a small mixing room, it will be difficult to get a decent ISD gap because object in the room must be so close together. However, since a small room will likely have fewer listeners in it, an RFZ of four square feet may be satisfactory while a large mix room should go for an RFZ of more like twenty five square feet or more to hold more people.
While this may seem like a complex and somewhat exhaustive essay on the basics of acoustics, learning acoustics takes a lifetime. Even at that, many properties of sound are yet to be understood, such as why two seemingly identical rooms can sound completely different. Thus, this is not a complete theory of sound and there will be some inaccuracies in the aforementioned information. As always, one is welcome to E-mail the author to make suggestions for improvement. In the mean time, some helpful hints....
http://www.ethanwiner.com/acoustics.html for info on rigid fiberglass
http://www.saecollege.de/reference_material/pages/Coefficient%20Chart.htm for many other absorption coefficients.
http://en.wikipedia.org/wiki/Acoustics for general acoustics information.
To E-mail, just type what you see in the image below into another window. Sorry for the inconvenience but it prevents Spam bots.