[SoundStage!]Max dB with Doug Blackburn
Back Issue Article
February 1999

Bass

This month I’m going to drop on you most of the "facts of bass" I’ve managed to collect over the years from manufacturers, books and personal experience. Some of the things you read here may contradict what you’ve read in other magazines. I wish it were just as simple as telling you the information you’ll see here is right and the other stuff you’ve read is wrong. Unfortunately, I can’t prove to you what I’ll tell you is true unless you read the same books, experience the same products, and speak in earnest with the same manufacturers as I have. One thing I can promise you, I am not going to make up anything you read here. It’s all as true as I can make it based on my study and experience. Some of the information will be brief factoids, other subjects will take some detailed explanation. The article isn’t going to read as an organic whole due to the nature of the subject. So be ready to jump around a bit as I discuss some of the realities and misconceptions about bass and how to get it in your system.

The nature of bass

As far as equipment goes, good bass is expensive. There is no quick and inexpensive way to get good bass. Musically, the bass octaves are undoubtedly the most expensive octaves in terms of equipment costs. The bass octaves are: deep bass -- 20Hz to 40Hz; bass -- 40Hz to 80Hz; and upper bass -- 80Hz to 160Hz. These may not match up with the octaves on an organ or piano keyboard, but you can arbitrarily make an octave be whatever you want it to be for purposes of discussion. Pick any number and double it and you have an octave for purposes of discussion. The octaves on musical instruments are based on musical notes which may have frequencies like 42.3Hz, unnecessary complication for purposes of this month’s discussion of bass.

Sound travels though air as waves. The length of the wave gets longer as the frequency decreases. How long are the wavelengths? Hard question to answer. The length of the wave of any particular audio frequency varies with air density/pressure -- just as the speed of sound varies with atmospheric conditions and altitude. In fact, it is because the speed of sound varies significantly over time that the wavelengths of audio frequencies in air varies. Let’s say the speed of sound at your altitude and current atmospheric pressure happens to be 1150 feet per second. To calculate the length of a 100Hz sound wave, you want the answer to come out in feet per cycle, so make 1150 the numerator and the divisor will be 100(Hz). You get 11.5 feet for the length of a 100Hz sound wave. Tomorrow the speed of sound for you might be 1100 feet per second, which will make 100Hz sound waves 11 feet long. Does this make a difference to the sound of your system? You bet it does! Ever wonder why you like the sound of your system more on some days than on other days? You think it is related to temperature? Sure, it could be. Think it is related to humidity? Yes again, that could do it. And electricity? Yes, another source of time-to-time differences. But I’ll bet not all that many audiophiles have considered how day-to-day changes in the speed of sound affects the sound of their systems. Wavelengths change and the acoustics of the room will change a bit with the wavelengths. I’ll even go so far as to speculate that setting up a system to sound good at 5000 feet elevation is different than setting up a system to sound good at 500 feet because of the significant change in the speed of sound between those two elevations -- assuming the same room and speakers in both locations, of course.

So we know 100Hz sound waves are going to be in the range of 11 to 12 feet long or so. How about 50Hz? Double the length -- 22 to 24 feet. This means that if one dimension of your room is 23 feet, on some days you’ll be able to contain a fully developed 50Hz wave within the room and on other days you won’t! Pretty interesting, eh? I’m not saying there will be a huge qualitative or quantitative difference in the sound of 50Hz in that room on different days, but it is inevitable that some sonic difference will arise from the change in wavelengths from day to day.

Twenty-five Hertz waves in air are twice as long as 50Hz -- 44 to 48 feet. And 20Hz, the lowest frequency we’re supposed to be able to hear, would clock in around 55 to 59 feet long. Do you need a room with at least one 60-foot dimension in it to hear a real 20Hz in the room? No. Your ears actually pick up sound in a different way, reacting to the compressions and rarefactions that happen in the air as the sound propagates though the room. Twenty Hertz creates 20 compressions and rarefactions per second and your ear will pick that up even if you are listening to headphones that respond to frequencies that low. Otherwise your ear canal would have to be 60 feet long -- we would look rather odd if our heads were 60 feet wide.

If you can still hear 20Hz even though your room is too small to contain a 20Hz sound wave, why does it matter that the 20Hz sound wave is almost 60 feet long? For two reasons. The wavelength defines the size of objects which are capable of acting on a 20Hz sound wave in a meaningful way (like 5dB or 10dB or more). Those objects really need to be at least 1/2 the length of the frequency they are trying to control in order to have more than a very small effect on the bass frequency in question. This has to do with the physics of waves. What this means is that you can’t make a passive device the size of a softball or shoebox that would profoundly and directly affect a bass frequency. The softball-size or shoebox-size device may have some collateral room-tuning effect that could be interpreted as having some control over the bass frequency. But the degree of the control will be quite a bit below the 5dB, 10dB or more that a purpose-designed and sufficiently large device will have. This is an important concept. Smaller items like Tube Traps, Pressure Zone Controllers, on-wall tuning panels or other products that touch a wall or floor or ceiling in the right way at the right position can alter the resonant property of the wall or other surface in a way that may have some smaller (yet still perhaps useful) effect on low frequencies. But to act on a low frequency (for purposes of this discussion, let’s assume 100Hz and lower) in a profound way (5dB, 10dB or more), that strongly amplifies or damps the wave, you need devices that are 1/2 the wavelength or larger. Around six feet for 100Hz and around 28 feet for 20Hz. This physical law is the reason horn-loaded loudspeakers have to be physically huge to reproduce bass and deep bass octaves; a horn would have to be at least 1/2 the length of the lowest frequency in order to reproduce that frequency. A horn good down to 20Hz would have to be nearly 30 feet long. Likewise, a bass trap that would have a significant impact on a 20Hz sound wave would have to be almost 30 feet long. But there would still be some compromise. Making the horn or bass trap the full length of the sound wave would further increase control over the bass frequency.

The second reason that the length of bass sound waves is important is that below a certain frequency (room dependent), where you place the subwoofer in the room really makes little difference in bass balance within the room. This is very different from what you’ve read in other publications for years. Loudspeaker placement is still important because the loudspeaker produces frequencies that are much higher with shorter wavelengths than subwoofers. So loudspeaker placement is not really governed by this "rule." But with subwoofers -- let’s say your room is 13 feet wide and 17 feet long with an eight-foot-high ceiling. If the speed of sound is 1150 feet per second, the width of your room can support wavelengths down to 88.4Hz, the length can support wavelengths down to 67.6Hz and the height can support wavelengths down to 143.75Hz. If your subwoofer crossover point is 60Hz, almost everything coming out of the subwoofer will be larger in wavelength than any dimension in your room. You can put the subs in the center of the room, or just about anywhere in the room with little or no qualitative difference in the sound of the bass. One thing though, as the subwoofer approaches a two-sided or three-sided corner, the wall/floor/ceiling surfaces will "horn load" the subwoofer and the bass may sound a bit louder near a two-surface or three-surface corner. But the quality won’t change much and the balance won’t change much. The higher the crossover frequency for the subwoofer, the more concern there may be for where the subwoofer is located. So you can use the walls and corners of the room to "lift" the subwoofers’ bass energy in the room a little if you need it.

One subwoofer or two?

Generally, the smaller the main loudspeakers, the higher the crossover point needs to be. The higher the crossover point, the more you need two subwoofers. Any time a significant portion of the subwoofer’s operating range goes over 100Hz, you really want to seriously consider two subwoofers. Use this rule of thumb to decide if you really need two subwoofers for imaging reasons: Double the subwoofer’s crossover point, and if that number is higher than 100Hz, an argument can be made that two subwoofers will be better for stereo imaging than a single subwoofer. If the doubled crossover point number is higher than 100Hz and you can only deal with having one subwoofer for space or budget reasons, that single subwoofer should unquestionably be placed in the center of the two main loudspeakers for imaging purposes. Placing a single subwoofer off to one side when it will be reproducing any sounds above 100Hz will pull the image to the side. Not many subwoofers crossover as low as 50Hz or lower. This implies that not many subwoofers are going to perform optimally when you only use one. You can use one subwoofer, just know that there will be a compromise as compared to using two subs.

Home theater is an exception. In a dedicated home theater, reasonably full-range main speakers and a single subwoofer can be quite adequate if the subwoofer is connected to the LFE (low-frequency effects) channel of a surround processor. This setup will do nothing for music reproduction though. If your goal is integrating a music system and a home-theater system, read on.

Here are the "rules." Feel free to break them for any number of legitimate reasons, but understand that by doing so, a compromise of some sort and some magnitude will be introduced. You may find the consequence inaudible or inconsequential. That doesn’t mean it is unimportant or incorrect; it just means that in your system, it wasn’t a significant factor.

dB’s rules for great bass in a combined music and home-theater system

  • The best way to get great deep bass in a home stereo system where the loudspeakers are, say, under approximately $10,000 per pair is to use two subwoofers with relatively full-range main speakers. The deep bass is produced outside the enclosure of the main speakers, helping to keep the sound of the main speakers unmuddied by having a lot of bass energy in the enclosure. When you hit approximately $10,000 per pair or so, the enclosures can be complex enough and exotic enough to be able to deal effectively with having deep bass frequencies generated within the same speaker enclosure. This isn’t the case automatically in speakers over $10,000 of course. Much depends on the skill of the designer. There may be a few rare exceptions under $10,000 where the loudspeaker can make great deep base in the same enclosure, but these would be very special loudspeakers indeed.

  • Getting the blend between the main speakers and the subwoofers right is hard. To do it right, the bass of the main speakers must be flat to one full octave below the subwoofer crossover point. If the Crossover is 60Hz, the main speakers have to be flat to 30Hz! This means big main speakers. If the crossover point is 80Hz (quite a common crossover point), the main speakers need to be flat to 40Hz, which still probably means at least three-way speakers with a decent 8" to 10" woofer. Minimonitors or average bookshelf speakers are not going to be a good match for a subwoofer with an 80Hz crossover. If the crossover point for the subwoofer is 100Hz, the main speakers need to be flat to 50Hz, still a pretty serious speaker! Small two-way monitors are going to be flat to maybe 70Hz, with 80Hz more likely. This means the subwoofer crossover has to be up at 140Hz to 160Hz. This is very, very high for a subwoofer crossover point. In fact, it is so high that you really need 10" or smaller drivers in the subwoofer to get decent sound up that high -- the subwoofer crossed over at 140Hz to 160Hz will still need to be an excellent performer one full octave above the crossover point. In this example, that means the subwoofer has to be a good performer up to 280Hz to 320Hz, way up into the vocal frequency range. A 12" or larger single-driver subwoofer is going to be overwhelmed by the need to perform at frequencies that high. You just won’t get good integration. In fact, there are few, if any, subwoofers that you can buy at any cost that sound good from 320Hz down to 20Hz. There are certainly none at lower price points that people with smaller, affordable speakers will be out looking for.

  • LFE subwoofers make lousy music subwoofers. Great music subwoofers can do movie soundtrack bass pretty well, but music subs cannot and should not be connected to the LFE output of the surround processor. The best music subwoofers will utilize some kind of crossover. The best LFE subwoofers are crossoverless and may be "peaked" to provide a more exciting sound within the bass frequency range most prominent in movie effects (30Hz to 50Hz). The best music subwoofers will be linear in response all the way down to 20Hz. A best-of-both-worlds subwoofer would have a crossover plus a peaking adjustment, like adjustable Q (resonant frequency). This would permit setting the subwoofer flat for music, then be able to "peak" it for movies.

  • The best type of subwoofer crossover for music is one that rolls off bass into the main amplifier. This relieves the main speakers and amp from having to produce as much bass. There will be clarity gains from eliminating this bass energy from the speaker enclosure and from relieving the power supply in the amplifier from some of its most demanding work.

  • The best "drive" connection for music subwoofers is for them to be connected to the output terminals (usually binding posts) of the main amp. This seems wrong on the surface, but there are advantages. The main loudspeaker impedance will be 4 to 8 ohms while the input impedance to the subwoofer amp will be 500,000 ohms or higher. Virtually no power will come to the subwoofer from the main amplifier. Power goes to the low impedance loudspeakers instead. The subwoofer will only be "looking at" the output of the main amplifier to know what to reproduce. This means if the main amplifier is a soft-sounding tube amplifier, the subwoofer will create complimentary soft bass so that there is no big discontinuity in sound quality between the main amp and speakers and the subwoofer. If the main amp is a Krell, the subwoofer will produce strong, powerful, controlled bass that matches what the Krell is delivering to the main loudspeakers. This kind of connection means the wires used to connect the subwoofer amp to the main amp are not very critical.

  • While passive crossovers from lower midrange frequencies on up can be made today which are actually better than any active crossover in the same frequency range, bass frequencies are still quite difficult to deal with for passive-crossover design. When selecting your subwoofers, the crossover is a critical consideration that will affect overall sound and integration to a significant degree.

  • Most ideal setup: a pair of subs connected to main speakers with surround processor set to "large front speakers" with an LFE subwoofer of identical type/sound quality running from the surround processor LFE output. Next best option: a pair of subs connected to main loudspeakers, "large" main speakers selected in surround setup and no LFE subwoofer selected in setup. If the surround processor is properly designed, the LFE bass will be redirected to the "large" main speakers. This pair or subs should ideally include the ability to adjust the Q (or some other parameter) to permit "peaking" the subwoofer to get more exciting movie sound while retaining linear response for music.

  • All-out home theater setup: subwoofers on the surround-channel speakers. Dolby Digital[TM] and DTS[TM] are both full range capable and some movie soundtracks do indeed send some pretty deep bass to the surrounds. Music-type subs will work the best with the surround speakers, especially if they are adjustable for "peaking" should you want to use it. These subs would only come into play for movie soundtracks or for the specialized Dolby Digital or DTS music CDs.

  • Over-the-top home-theater setup: subwoofer on the center-channel speaker. Here, too, a music-type adjustable sub would be the best choice.

Special home-theater considerations

LFE subwoofers should not have a crossover per se. They should have a natural and/or cabinet-assisted roll-off on the top end, not an electronic filter to roll off the top-end response. The LFE output of a surround processor limits the frequencies delivered to the subwoofer. So not having a crossover means fewer parts in the signal path and better sound.

Some surround processors don’t deal with bass correctly. If you aren’t using an LFE subwoofer, the LFE bass disappears from some surround decoders. If you are using full-range main speakers or main speakers with subwoofers, you may not be getting all the bass on the soundtrack if the LFE bass is not redistributed at the proper level to the main left and right channels. There’s no easy way to identify this in advance. Perhaps the Video Essentials DVD includes an LFE-only track that would easily reveal whether a decoder was properly redirecting LFE bass to the main speakers when there was no LFE sub connected. I have not yet seen the Video Essentials DVD, so I can’t confirm that it has this capability.

Five-point-one-channel processors add additional crossover "poles" when you select "small" loudspeakers for any of the speakers in the system, and this may actually make the speakers sound worse than calling them "large" speakers so that the processor does not add another crossover "pole." This is true for main, center and surround speakers. Unless they are VERY small speakers, you may get better sound by calling the speakers "large" even if the speakers are only modest in size.

Subwoofers designed primarily for home theater will typically suck for music. Don’t spend a lot of time and money trying to force a home-theater subwoofer to sound good playing music. You just won’t be able to make it happen. You’ll get bass and you may even find it reasonably entertaining. But an LFE sub will rarely sound as good as a "music sub" when playing music.

That’s about it for bass this month -- hope you found the information useful. See you next month.

...Doug Blackburn
db@soundstage.com

 

[SoundStage!]All Contents
Copyright 1999 SoundStage!
All Rights Reserved