Room Acoustics Basics
Ever been watching TV and no matter how loud you made the volume, it was still hard to hear? Ever walk down a long, narrow hall and clap your hands to hear a strange twang echo? Well, those are just some of the effects that rooms have on sounds and yes, they play a role in your recordings. So, it is important to first consider the acoustic quality of a room before you start recording your awesome, will-change-the-world songs. Once the tracks are recorded, it is very difficult or near impossible to undo the effects of the room in the mixing and mastering stages. So, deal with this before you start recording.
There are volumes and volumes of information written on the subject of acoustics and most of us don’t have time or energy to absorb that. The focus here is to give you enough information to understand what makes a bad room for recording and how to identify problems before the record button gets pressed. Ways to fix room acoustic problems without spending a fortune are covered in another lens, “Finding and Fixing Room Acoustic Problems,” but first, a handle on room acoustics will be a valuable tool in getting high-quality recordings. Plus, knowing a little about room acoustics will help you pick a room that is already a good choice or needs little adjustment for recording. You just have to know what to listen for in a room.
There are essentially three acoustic properties that dominate the quality of a room, 1) room modes, 2) reverberation and absorption levels, and 3) reflection, combing and diffusion properties. There are certainly other acoustic properties but you’re not going to earn your PhD in acoustics here. Let’s just try to keep it simple so we can do what we really want to do, get a quality recording of some great music.
I. Room Modes and Resonance
Ever play a particular note on your instrument and something in the room vibrated as a result? Well, that note hit the “resonance” of that object. What does resonance mean? If you take a piece of string and tie one end to something and then shake the string until you get the string to vibrate with a single arc, you are at the point where the least amount of energy is required to shake the string. This is why when you pluck an instrument string, it vibrates at a specific rate or frequency (i.e. the tuned note). The energy easily transfers back and forth from the bridge to the nut at this rate. Everything faster or slower just dies out quicker. Take the tie piece of string and try to change the rate at which you shake it and see how erratic it becomes. Well, rooms also have resonances and just like the energy in the guitar string that transfers back and forth from the bridge to the nut, sound in a room bounces back and forth from wall to wall. So at the resonant frequency, the sound will take much longer to die out compared to other sounds not at the wall-to-wall resonant frequency. There are other types of resonances created from more than just two walls (known as mode types), but the wall-to-wall modes almost always dominant the room.
Why should you care about room resonance? Well, say you are recording a song that is predominantly in the Key of A and the room resonates at a frequency that is near E-flat. After a recording of the song in the room, there is now this strange flat sound in the recording you just can’t put your finger on and so, the room has ruined the recording. Typically, if you can’t hear people talking on TV no matter how loud you make the volume, it’s because of the room resonance is covering the sound. The same thing will happen to your recordings if the room has heavy resonances.
The way to tell how many modes are in a room is based on the number of parallel surfaces. For example, in a rectangular room there are three modes: front wall to back wall, left side wall to right side wall, and ceiling to floor. Again, there are other modes but these are typically much less of a nuisance compared to the wall-to-wall type.
The frequency at which they resonate at can be determined by the distance (D) between each wall using the formula,
F = 565/D (resonant frequency of wall-to-wall room mode)
Depending on where you place a microphone, a room resonant can make a played pitch (or component of the pitch) louder OR softer. In any case, it “colors” the sound and distorts the intended recording.
Remember the example with the E-flat resonance? A room that has two walls about 14 ½ feet apart will produce a resonate frequency of about 39 Hz, which is a very low pitch E-flat.
Are there any good distances from wall to wall? No. Your enemy is parallel surfaces and in particular, rooms that are square or cubical. What if I just put a lot of stuff in the room? Won’t that help? Not necessarily; particularly if you just randomly do this. Rugs, drapes and furniture do not provide much mode damping (they do affect reverb and imaging, but not modes). There are ways to dampen mode frequencies (e.g. bass traps, wall wedges) but this will be covered in a separate article, “Finding and Fixing Room Acoustic Problems.”
Here is the Take-Away for this section:
- Avoid recording in rooms with lots of parallel walls (especially square and cubical)
- Irregular room shapes will provide much better recording quality
- If you can’t avoid parallel walls (like the ceiling to floor), then at least know what the approximate resonant frequencies are so you can use that information to fix or avoid these modes
Just because you found a room that doesn’t have many parallel walls, doesn’t mean the room is recording-friendly. There are still two other factors to consider, reverberation and diffusion.
II. Reverberation and Absorption
What is reverberation? Sound bounces off of all surfaces whether it is a full reflection or a partial one. However, unlike direct echoes or resonances, these reflections hit other surfaces and still other surfaces until the sound dies out. This is known as reverberation and summation of the reflections result in a different coloring of the sound based on the size and shape of the room. This is why the reverberation in a church sounds different than in a concert hall. In fact a room with no reverberation at all is not real pleasing to listen to music in, but a room with excessive reverberation will make the music sound blurry and lack real clarity. So, the goal is to find the right balance of reverberation. However, if there were a situation where you had to choose a room with too much reverb versus a room with very little reverb, choose the room with very little reverb (i.e. a very “dry” room). Why? Thanks to the development of software technology and process power, artificial reverb can be added to tracks after recording. This is particularly appealing since different types and levels of reverb can be auditioned for the same track. Beats packing up the equipment and moving to another room just to see if you like the reverb better. However, sometimes we find a room that has such a nice reverb quality to it and you want to capture it with your recording. If not, the goal will be to make the room a dry as possible and ideas for getting a room dry are covered in the article, “Finding and Fixing Room Acoustic Problems.”
So how do you know a room has too much reverb for recording? Good reverb to the ear typically does not translate the same for multi-track recording; always go for a little less, particularly if the music is a faster tempo and more rhythmic. Nevertheless, the first thing to consider is the overall spatial volume of the room. Based on typical professional recording studio rooms and other sources, rooms no smaller than 3,000 cu-ft should be used for recording. This is mostly because any smaller, the resonant modes become too difficult to reduce using absorption and other methods. Not to mention, it’s going to be pretty cramped (e.g. 3,000 cu-ft = 8-ft ceiling x 20-ft front-to-back x 18-ft,9-in side-to-side). For recording, a maximum of 30,000 cu-ft should be the limit; this would be a typical basement with a 10-ft ceiling, 75-ft long by 40-ft wide. Any bigger and it will get more difficult to reduce the reverberation without professional grade treatments. Once you have a room that is a good size for recording, you have to figure out if the reverb is too much for a clear, professional sounding recording.
There are two answers to the question, “How much reverb is too much for a room that you will record music?” The first answer is the more professional approach that requires you to set up a microphone and a speaker for recording a “white noise” that you turn off to have a recording of the decay of this sound. The time it takes for the reverb to drop 60 dB (decibels) is known as the RT60 time of the room. If you go to some of the musical instrument and audio equipment websites, they sell audio analyzers that measure the RT60 of a room. However, they are not cheap and there is a quick and dirty way to do it if you have recording software that displays tracks in a dB scale (covered in Section IV). So, good reverb RT60 levels for recording music should be in the range of 0.5 seconds to 0.75 seconds. The second answer is the less sophisticated approach and that is to just make the room as dry as you can. Then once the tracks are recorded you can add reverb digitally in the mixing. Making a room dry requires a little understanding of absorption and damping.
As much as some materials reflect sound, there are other materials that absorb sound. The problem is that most absorptive materials do not absorb sound the same at different frequencies. For example, curtains might absorb mid-range sounds (200 Hz to 1000 Hz) well but do not impact the lower range much. So, although you reduced the reverb in the mid-range the room now sounds boomier. The bottom line is that thick is better. Fluffy pillows and drapes with a lot of folds will provide better sound absorption in the lower ranges. Carpets are good for mid to high frequency ranges (1 kHz and up). Room modes are a different animal and things like bass traps and wall wedges will be needed but the same idea of thicker-is-better still applies. Overall, try to have a balance of materials for all the frequency ranges. A little adjustment of how many items and where they are in the room will make a big difference in how dry the room will get. If the room starts to sound like you’re in a tin can, the high range needs more absorption and if the room gets too boomy, you need thicker, fluffy items to reduce the lower range. More on this will be covered in the article, “Finding and Fixing Room Acoustic Problems.”
The last topic I want to mention related to room reverberation is the sense of spatial placement. If a listener is standing in one spot in the room and the instrument is played at several different locations, the location change is audible to listener and NOT just because of the volume different between each ear of the listener. The reverberation build up will be different based on where the instrument is placed. One of the key aspects of this is the arrival gap. This gap between the direct sound of the instrument and when the reverb arrives creates a sense of space. Thus, sounds that have different arrival gaps and reverbs give the listener a better sense of spatial placement of the instruments in the room. So, when recording in the room (particularly rooms that are not very dry), record the different instruments as far apart from each other as possible to give them better separation in the mix-down of the tracks. Even though in a live performance the instruments are closer, placement of them further apart when recording will make the total mixed down recording feel much bigger and more live. For much drier rooms, this is not as important but this same effect can be achieved with digital reverberation if you know how to manage the settings to emulate the feeling of spatial placement.
Here is the Take-Away for this section:
- Rooms with less reverb are better for recording than ones with more, even if the reverb has a good sound to it.
- Rooms between 3,000 cu-ft (e.g. larger living room) and 30,000 cu-ft (e.g. large unfinished basement) are the most suitable sizes for recording.
- Thicker and fluffier materials for lower frequency absorption and carpets, rugs and flat curtains for higher frequency absorption.
- Record each instrument at different locations to achieve a better sense of separation on the mixed down track (further apart the better).
So, at this point you know that irregular shape room with some materials to absorb sound are the best candidates for recording. However, there is one more consideration that is related to echoes know as imaging that can ruin a recording even in a mode-free, low-reverb room.
III. Reflections, Combing and Diffusion
Although reverberations can be considered a kind of sound reflection it is really a full blend of many, many reflections in a decaying pattern. A direct reflection is a single reproduction of the direct sound and in cases where the reflection is coming from a far enough away location, it is called an echo. In very close scenarios, the effect is not audibly heard as a second copy of the original sound. One result is a blurring of the source and second is a flanging or combing of the sound. It’s that twangy sound you get when you clap your hands in a narrow, hard-wall hallway. In recording, this can happen if your instrument is too close to a hard flat surface that sends a reflected image of your instrument to the same microphone you are recording. This can also happen when you record two tracks at the same time and each track has the same instrument recorded but with a slight delay between them. After mixing them down together, the flanging will occur.
The solution to preventing combing to occur in your recordings is the THREE-TO-ONE rule. If your instrument is 1 foot away from the microphone you are recording, then every other microphone and flat, reflective surface must be at least 3 feet or more away to prevent the combing filtering effect. For example, if you are in a room with an 8-foot dry-wall ceiling and you have a vocal microphone that is 5 feet above the floor, the furthest distance your singer can be away from the mic is 1 foot before the potential for flanging will occur in the recording. Another example is overhead mics for drums. Say you had two microphones for hanging over the kit for recording and the distance from the left crash cymbal to the mic directly over that is 1 foot. Then the other mic over the right crash cymbal has to be more than 3 feet away from the left crash cymbal. Plus, a flat dry-wall ceiling has to be more than a foot above the microphone. The drawing at the beginning of the section illustrates this. I’m not going to lie to you, drums are very hard to record. So it may take you a little while to find just the right placement and number of the mics and location of the kit in a room. However, sometimes placement and separation of the mics are not enough or the room constraints limit your options.
Any flat surface (even ones with a lot of absorption) can generate a reflection and therefore a potential for combing (or flanging). So, one way to reduce the amount of reflections that could create problems like combing, is to diffuse the reflected sound by changing the surface from flat to something that would better scatter the sound. Something as simple as a piece of wood or stiff cardboard bowed outward from the wall would provide good diffusion of the sound. Diffusion also helps improve the quality of the reverberation as well. Another idea is to have things on or against flat surfaces with random angles and shapes to force reflected sound to disperse in every direction instead of the same direction as a flat surface would. If caves had more absorption to them, they would make great locations to record since the cave walls are so filled with random shapes, sizes and curves.
Here is the Take-Away for this section:
- The THREE-TO-ONE Rule: for every 1 foot your instrument is way from the recording microphone, all other microphones and flat surfaces should be at lest three feet away from that microphone.
- Rid the room of as many large flat surfaces as possible by adding things against them to scatter the sound in random directions or even just an outward-bowed piece of wood or cardboard will make a big difference.
By applying these simple ideas to selecting and setting up a room for recording, you will notice a tremendous difference in the quality of your recordings. The more professional sounding your recordings are the more listeners will view you as a serious musician. You don’t want somebody saying, “This sounds like you recorded it in your garage.” A great song deserves a great recording and it all starts will the right room.
If you want to know a little more about the room you want to record in, the next section covers some simple ideas for taking acoustic measurements to evaluate the room and how well you did in treating it to sound better.
IV. Quick, Cheap and Dirty Acoustic Measurements
This section is a little more advanced in terms of being able to take advantage of the digital recording software to do your own acoustic measurements. If you have software packages like Cakewalk SONAR, Steinberg Cubase, DigiDesign ProTools, or other package that you can record and mix audio tracks, then you have the ability to do some basic acoustic measurements and analysis. Everyday, I’m seeing more and more plug-ins (e.g. VST, DXi) for these types of room measurements. I have even seen a number of “free” spectrum analyzers as plug-in modules for most recording software programs. The point is that more and more people are beginning to realize the value in making sure the room you record in gives you the best results and a little understanding of room acoustics will help you get the most out of these emerging tools. Nevertheless, with just your recording software you can make a few simple checks about the room before you start recording.
Find Those Modes
The first types of measurement that we will try to get with a quick and dirty technique are the room mode resonances. To do this you, you will need:
- A “white noise” wave file (>5 seconds long) to play on your audio player.
- A good microphone that has a low-frequency response (e.g. big condenser type used for vocals or bass drums).
- An audio spectrum analyzer plug-in for your recording software. Particularly one that lets you change the frequency range so you can set it below 1,000 Hz.
This measurement procedure is more for before-and-after testing to see if you were able to reduce the room modes. However, in the least you should be able to locate the modes in the frequency spectrum. In case you are not sure what a “frequency spectrum” is, this is a plot of all the frequencies that are being picked up by the recording microphone. If you record a single note, you will find that it is not just that pitch but many other tones and frequencies that appear in the plot. If you record a single snare drum hit, you will see a very wide spread of frequencies in the plot. White noise shows up on the frequency plot as a complete flat horizontal line of frequencies. So, any effects of the room such as the modes will show up as peaks in this horizontal line, which is why we use white noise to find the room modes.
First Step: you need to estimate where the biggest room modes are in the frequency spectrum by finding the largest parallel surfaces in the room. If the room was rectangular, it would be the ceiling-to-floor, front wall-to-back wall, and left side wall-to-right side wall distances (in feet). Using these measured distances, apply the formula:
Frequency = 565/Distance
for each measured distance between large parallel surfaces.
Second Step: Place a larger speaker (e.g. woofer-type) connected to your sound system in the middle of the room and a large condenser-type microphone (e.g. vocal or kick drum mic) about three feet away from one of the corners in the room (modes are strongest in the corners). Play the white noise sound file through the larger speaker and record this using the large condenser-type microphone you placed in the corner. Make sure you note the exact positions of the speaker and microphone for later reference.
Third Step: Feed the recorded sound file into your frequency analyzer and zoom in on where you calculated the room mode frequencies. Note: some frequency analyzer programs (or plug-ins) don’t let you change the range to compute the spectrum of frequencies and this may make it difficult to see the mode frequency bumps in the plot (i.e. the low-resolution at low frequency has “smoothed” out the modes in it’s calculation of the plot). Thus, find a frequency analyzer program (or plug-in) that lets you set the range so you can get better resolution down at the lower frequencies (10’s of hertz) where the modes are likely to be. FYI: a drop in the peaks of only 3 dB means the volume has been cut in half and a drop of 6 dB means the volume is now only 1/4th the original.
Fourth Step: Add bass traps, corner fills and other mode reducing materials to cut the modes and repeat the measurements in the previous step. Compare the before and after spectrum to see how much reduction you have been able to achieve.
Let me just reiterate that the drier the room (i.e. less reverb) the better for recording and this is mostly due to the advances in computer technology and digital emulation that makes it easier to create very realistic sounding artificial reverberation. This assumes the user of the reverb application knows what they are doing (see the Mixing lens in this series). However, you may still want to see how dry the room is before you start setting up your gear and bringing everybody in for the recording. So, this section shows you a quick little trick for getting a rough estimate on the reverb level of the room. I also want to again point out that there is a growing number of programs and plug-in for this and it might be of some value to you to do a quick internet search for your own benefit.
To do the quick and dirty reverb level check, you’ll need set of drum sticks or some balloons (yes, I did say balloons) and a microphone. Given the option of balloons or stick, pick the balloons. Set the microphone up somewhere near the center of the room and start recording when the room is very quiet. As you are recording, pop a balloon with a pin or bang the sticks together once. Do this every 5 seconds to have a series of pops/bangs to look at in your recording software.
In your recording software, there should be a way to set the scale in the wave view window to be in decibels (dB). Once you have the scale set, zoom in on one of the pops in the wave view window.
In the dB view, the pop (or stick snap) will not completely drop out to zero; the remaining wave data is the background noise of the room and electronics (i.e. the noise floor). The RT60 is the time it takes for the pop to decay 60 dB, but in most practical situations the decay disappears into the noise floor before it decays the full 60 dB. So, you need to draw two horizontal lines, one from the start of the pop waveform and a second 60 dB below that that line (e.g. if the first was at -10 dB then the second would be at -70 dB). Using a straight edge, try to line it up with the peaks of the drop off and the intersection of the second horizontal line with the straight edge will approximately indicate the RT60 time. In the picture above, the RT60 time is about 0.33 seconds; this is a dry room and good to record in. Keep in mind, you are just trying to get a rough idea of the RT60 time; there are more accurate methods but we’re trying to keep it simple and cheap.