Problems with room resonances

By mic on 9/5/2006

You probably already know that your room affects the quality of your home recording. Sound is airborne waves which are physical in nature and are thus affected by physical structures. Sound waves in rooms are affected by walls, floors, ceilings, curtains, furniture, televisions, light fixtures, and other. In short, anything that obstructs the path of the sound wave will alter it and will ultimately change the sound that our ears perceive. The end effect of this may be desired or undesired. Typically, since the human ear is used to hearing sound after such room resonances, those may make the sound appear more "natural" and "lively". This is why recording engineers artificially introduce room resonances in sound recordings through reverb units. Without them the recording will sound "dry" and unnatural. So why exactly are our room resonances problematic?

The problem

Our problem is that our small apartment recording studio does not produce the resonances that we want. We spend a lot of time trying to get rid of unwanted resonances and a lot of our recording sessions have degenerated into acoustic experiments to make our recording studio quieter. This is a time consuming exercise and usually not a very productive one.

To explain why our apartment does what it does is a complex task and so we will start from the very beginning: Any surface reflects sound. Not every surface does that in the same fashion though. First, sound reflected from some surfaces has larger amplitude than sound coming back from other surfaces. In an empty room with hard surfaces, for example, you may hear more of the room reflections than in the same room with carpeted walls, floor, and ceiling. Second, different surfaces tend to produce reflections with different "frequency content". In other words, you can have one surface that reflects better mid-range frequencies while some other surface may reflect more of the higher frequencies. Third, surfaces have different shapes and forms, sending sound in different directions. Fourth, surfaces may be varying distances away. Since sound takes time to travel, the farther a surface is, the more time it takes for the sound to come back reflected.

Imagine the common home recording studio set up in a small apartment. The room in which you record will likely have floors that are different from the walls, a ceiling that still different from the walls and floors, and windows with blinds or curtains. The room may be filled with various objects. Any of these objects will reflect the sound and all will do so in a different way. Each of these objects will reflect not only the original sound, but also the sound already reflected by some other surface. The end result of what you hear is a complex combination of the original sound and a very, very large number of reflected waves, all with different amplitude, different frequency content, and, last but not least, coming in at different times and from different directions.

The end result of what you hear in a room is extremely "complex". Room reverb differs from an echo. If you shout in the Grand Canyon you will hear an echo coming back in several distinct repetitions. Those would be easy to describe as they are distinct. When you shout in a small room, distinct repetitions do not exist. Reflections in a small room come back fast and in multitude. The original sound and the reflections are combined in a continuous "blob" of a sound. We cannot describe sound reflections independently as you can do for an echo. Instead, we can describe the whole "blob" of reverberated sound. Here are some of the properties then that we can use to describe reverb: 1) overall length: In a room which is larger and reflects more of the sound resonances may take a while to die out, which will produce a reverb with larger length. In a small "dead" room reverb will have shorter length; 2) coloration: Some rooms may accentuate the higher or lower frequencies; 3) diffusion: In empty rooms it may be possible to hear some reflections more distinctly, whereas in other rooms there will be a lot more reflections producing a smoother reverb; and 4) early reflections: It may be possible to hear the first reflections more distinctly than the rest of the reverb and those may have their own properties as per 1, 2, and 3 above.

The small apartment room

My recording studio is a medium-size empty room with empty plaster walls, a mixture of wood and carpeted floors, and very, very large windows, covered by plastic blinds. Resonances in this apartment are loud and bright. They create problems and there are no cheap ways to get rid of them. We have lots of issues, for example, with how our vocals sound. My frequent complaint is that the room robs our lead singer of his lower pitches making his voice sound thin on the recording. We could have the singer sing quietly, but that would just rob the song of its dynamics. At the end of the day we have mostly failed to correct for room resonances and have resorted to recording dry and introducing reverb in the recording artificially later. The apartment has affected most of our recording techniques: we changed microphone and instrument placement, moved furniture, and tried to dampen the ambient sound of the room (including building vocal booths from blankets, mattresses, and PVC pipes).

Instrument placement

Our general strategy is to place the instrument in a room away from too many boundaries or near some major absorbing element (i.e., in the center of the room facing some couch or tapestry). This greatly reduces the first order reflections. We avoided corners and metal shelving, which made the sound louder and brighter. We also avoided large glass structures such as sliding doors and stayed away from other hard large physical barriers. If strong, the early reflections from those could rob the voice of tone and impact, especially when the singer needed to really open up. In a small room early reflections come soon, are still distinct, and are loud. Loud singing just does not work near a wall or door.

Wall coverings

We had trouble finding materials that dampen the sound effectively. There are commercially available products but those are too expensive and we possess no desire to remodel a rental apartment. Thus, we moved furniture, propped up cushions, and even tried to build forts out of pillows. None of these produced the right sound. In an act of total desperation we sang under a blanket. It sounds stupid and looks even dumber, but produced some good results. The blanket cut down on early reflections, cost nothing (but our pride), and was a temporary alteration.

Here are my two cents on commercially available products for room treatment. Most of these do work as advertised (some others are complete rip-offs). I actually own some acoustic treatment products for my stereo listening room, and we found that those helped out quite a bit. If you can truly afford such stuff purchase it. I like what I have, but I would never advise anyone to get such items, because they are just pricy. I own a Room Tunes mini-pack – a few small "pillows" located at various corners and on various walls of my apartment. There are many different treatments though. I can try to categorize them but that is well beyond the scope of this post.

Mike selection and placement

I am almost afraid of writing these paragraphs as recording engineers probably will not like what I write. There are a lot of good mikes out there but, for the purposes of recording in a small apartment, they may prove useless. We spent a large amount of time trying a Rode NT large condenser mike with a cardioid pick-up pattern. We moved this microphone about the room and took it to different apartments. The fact of the matter is that none of the recordings of this mike have or will make it in any of our albums. This mike just picks up too much.

We resorted to using a Shure SM57. This is a small directional microphone designed primarily for instrument recording. It is also relatively cheap. It works well for certain drums (e.g., snares) and for acoustic guitars. For us, it also seems to work well for vocals. Since the SM57 is directional it generally picks up most of what is directly in front of it and very little from what comes from the back or the sides. This helps reduce resonances. In our experience, the SM57 also somewhat accentuates mid-range frequencies and so it gives a certain mid-rangy coloration to our vocals. The mid and lower tones of our vocals tend to get lost with certain mikes, but not as much with the SM57, especially if we close-mike singing. We usually end up between a foot and half-a-foot from the mike (making pop screens important). Mike placement is important, but the specific mike placement described here is the aftermath of our choice of mike. The problem with mike placement is that one can write books about it. For now, it suffices to say that, as with selecting your mike, you should experiment, and you should do so independently for different instruments and different vocals. Finally, we found the SM57 very consistent, which means that if we wanted to add thickness by overdubbing vocals we only need to remember the mike placement to get vocals that match the previously recorded vocals in tone.

Before we make an ad for the SM57 out of this article, we should mention that we tend to record the same two vocals and the same acoustic guitars. Perhaps for other instruments or for other persons' vocals other microphones will work better. The SM57 does not record a shaker very well, for example, and will probably not do justice to a fiddle (not that we have tried). Also, we only get a very poppy kick drum out of it.

In summary

Room reverberation is a very complex problem, which requires lots of experimentation and depends a lot on personal taste. We hope that we can compile more articles in the future with more detailed technical explanations on how reverb works, including artificially created reverb with reverb units and software, on mike placement, and on mike selection. For now, we remember to experiment. It is always difficult to remove unwanted sound qualities from already recorded sound. It is much better to spend a lot of time trying to get the right sound before doing the actual recording.

authors: mic & isosceles


Copyright 2006 by Kaliopa Publishing, LLC