Bass Integration Guide – Part 2
This article is the second of a three-part series on bass integration, covering all major aspects of accurate bass reproduction in domestic listening rooms. In Part 1, certain performance targets were presented along with instructions on how to measure the current bass performance. In almost all instances, the targets won’t be met and the shortfall will be significant. Acoustic treatment is introduced in this article as the first step towards improving bass performance, because it deals with the related acoustic problems at their source.
Is acoustic treatment really necessary?
Acoustic treatment is essential if accurate bass reproduction is the goal. It’s often argued that the use of EQ or multiple subwoofers can replace acoustic treatment. In most cases, those will be required in addition to treatment. This guide is written for those who want the best possible performance. Imagine going to an exclusive restaurant and ordering only dessert. One who wants to enjoy the full experience doesn’t weigh up the pros and cons of dessert versus an appetizer.
It is very easy to recognize an overly live room with excessive reflections and flutter echo, but in the bass range, acoustic problems are more difficult to identify by ear. Our auditory processing does not easily distinguish between the bass source and the effects of the room itself. Often a subwoofer will be blamed for poor performance, when in reality the problems heard are actually coming from the room. As a result, the need for treatment is frequently overlooked.
Bass traps are more commonly seen in exclusive high-end dedicated listening rooms that have been professionally designed. Most of the equipment in these exotic rooms is beyond reach for those not so well-heeled, but the bass performance of these systems is one aspect that the average audio enthusiast can obtain without an excessive price tag.
In Part 1, I mentioned how the room dominates the sound below about 200 Hz in domestic rooms. The primary cause of this is room modes, which cause large swings in bass response as well as modal ringing where the energy of some frequencies take much longer to decay. The result is seen in both frequency and time domain measurements. Subjectively, the bass may sound boomy or absent and in some cases both may occur at the same time. Common solutions include plugging ports, turning down subwoofers, or excluding them completely. This does not fix the underlying acoustic problems, but merely hides them. Many audiophiles learn to accept less bass, but often don’t realize the extent to which music is diminished in the process. This guide is intended to help audio enthusiasts rediscover bass.
Understanding your measurements
Here is one example of a room without treatment:
The main thing to recognize in this waterfall plot is the variation in decay rates. You can see that certain frequencies are ringing with a high-Q tail. This is highlighted as a red line. You can see that either side of this point, the bass decays much quicker. This is an indication that the room is not providing enough bass absorption. The highlighted 180 Hz peak is also clearly visible in the 2D decay plot in Figure 2.
The areas shaded magenta highlight problem areas where the decay exceeds the target. You can see that, in this room, additional absorption across the entire range is needed. We are less concerned with the decay below 40 Hz, and extreme solutions to achieve effective absorption down this low aren’t warranted.
For more information on these plots:
Recall that in Part 1, performance targets were mentioned. Of particular interest is the decay target, indicated by the white line in Figure 2. The requirement is for 20 dB of decay at 150 ms above 40 Hz where a 300 ms window time has been chosen.
Figure 3 shows the way in which the room envelope and bass traps work together to provide bass damping. The room envelope can be considered the enclosing boundaries that are seen by bass sound waves. If listening occurs in a larger open-plan space, then the envelope may be much larger than the area designated as the listening room.
Consider an acoustic sound wave, which is reflected, absorbed and transmitted when it hits a room boundary. A heavy concrete envelope will tend to reduce transmission and absorption, resulting in a large proportion of the energy being reflected back into the room. This causes substantial problems and a room like this requires extra attention. A more typical room construction that includes light timber framing and plasterboard (drywall) lining performs better due to the reduced reflection and the increased absorption and transmission. Bass traps are shown in red. These add further bass absorption and ensure that this absorption is spread over the required bandwidth.
Please note: the diagram in Figure 3 is highly simplified to illustrate the effects of absorption, reflection and transmission in a room. In reality the sound waves in the bass range travel in all directions simultaneously.
To read more about the importance of room damping:
Narrow and broadband absorption
A room envelope typically provides narrowband absorption. It acts as a resonant system, and the absorption and transmission varies greatly with frequency. This is why virtually all rooms require added bass traps to ensure broadband absorption. Helmholtz resonators and membrane traps are narrowband devices. If they are the only type of bass traps used, then it’s important to ensure that they cover 40 – 200 Hz. This means having traps that are tuned at different points to cover the range when combined.
Resistive porous traps provide broadband absorption. Their upper limit of effectiveness is determined by the properties of the material, and the depth of the device determines the lower limit. Where possible, broadband traps should be used.
The appeal of membrane traps is that they can be much less obtrusive. Ideally they should be used to add to broadband traps, but in some cases membrane traps may be a more practical solution.
Enclosing boundaries that are heavy and stiff – such as concrete or masonry – cause bass problems that are substantially worse than typical listening rooms. Difficult rooms require more careful attention. If your room is lightly constructed, you can move on to step two. External brick veneer walls don’t pose a problem. A concrete slab requires no attention unless the ceiling is also concrete, in which case the ceiling should be addressed as shown in Figure 4. (Images from The Soundproofing Company.)
It should be noted that sound isolation is a different requirement and is not covered in this article. Contrary to what one might expect, sound isolation at low frequencies suffers a little when a solid wall is modified as shown due to resonances in the air space. This is called the triple leaf effect.
This compromise is necessary where bass quality is the primary focus. The sound isolation can be improved further, but the expense is significant and the construction involved beyond the scope of this article. For more information, see:
The anatomy of a bass trap
A broadband trap is the simplest type. This is the principal trap recommended in this guide. It consists of a porous yet dense material such as rigid fiberglass, typically 100 – 150mm in depth. The fiberglass is the only material needed for it to work acoustically, but a frame is normally included and the material is wrapped in fabric to contain the glass fibers. Many traps have a membrane that reflects the midrange while allowing bass to pass through, where extra mid and high frequency absorption aren’t desired. Without this layer, resistive traps can cause a room to sound dead. Broadband traps are most effective when the distance to the boundary is one-quarter wavelength of the frequency in question, but they are effective at lower frequencies than this would suggest. Broadband traps can also be called resistive traps.
Membrane traps are not quite as simple and there are many variations. Most designs include a box with a membrane on the front and a thin panel of rigid fiberglass inside, which increases the bandwidth. Some are completely sealed, while others have a perforated membrane.
While broadband traps are simple to design and construct, membrane traps are more difficult and many of the details are critical. Factors such as the precise sizing of the holes (if used) or the thickness of the sides of the box can have a substantial impact. Before constructing such a trap, it is recommended to read the BBC research papers referenced at the end of this article.
Foam wedges marketed as bass traps should be avoided because their performance is poor. I have measured the equivalent of many such products and the results are posted on my blog:
Both broadband and membrane traps work best in corners, but their placement requirements are a little different. Broadband traps require spacing out from the boundary to work best.
Membrane traps work best close to the boundary surface, and they also work well in vertical corners, bulkheads and behind furniture. The bulkhead trap could also be placed on the diagonal, which can be expected to yield a similar result with less material. The bulkhead shown is similar to those used in many home theatres, often without performing a bass trap function. A rectangular section trap such as this could also be used for cable runs and lighting. Some care is needed from a safety point of view where lighting and traps are combined.
How much do you need?
While you certainly can kill a room with too much mid and high frequency absorption, this is unlikely to occur in the bass range. Aesthetics, practical limits and cost are limiting factors. As a general rule, it’s best to provide as much bass trapping as you can accommodate without giving up too much real estate.
1. Start with four broadband bass traps in vertical corners:
- 600 – 1200mm wide on the diagonal
- 100 – 200mm thick
- Rigid fiberglass with a density of 48 – 96 kg/m3
These traps can be expected to provide an adequate level of broadband absorption in most rooms. Difficult rooms will require more treatment but even rooms with good performance prior to treatment will benefit from further treatment.
2. Add broadband bulkhead traps:
- 300 – 900 mm in both width and depth
- 100 – 200 mm thick
- Rigid fiberglass with a density of 48 – 96 kg/m3
Rigid fibreglass is the standard material for broadband traps, but there are other materials that can also be used. Rockwool is a suitable alternative, along with high density insulation products with similar properties to fibreglass. The absorption coefficients in the 100 – 200 Hz range should provide an indication of performance.
3. Add membrane traps in other available corner areas:
- In the gap between furniture and the wall
- Along a wall where it meets the floor or ceiling
- Along one vertical corner surface where a door prevents the use of a broadband trap
Practical limits may dictate that broadband traps are replaced with membrane traps. In most rooms, it won’t be necessary to use all of the above traps. Most would be well served with four large vertical corner traps, but if they are made small (600mm wide) then additional traps may be needed. If uncertain, a waterfall plot can provide some feedback. When the decay is very even and ringing is well damped, further traps offer no real improvement.
Deciding when you have enough traps
The decay targets outlined in Part 1 provide some guidance. A waterfall plot is perhaps the most intuitive in terms of deciding when adequate absorption has been provided. The key thing to look for is how well modal ringing has been damped. As absorption increases, the in-room waterfall plot starts to look like a nearfield plot where all frequencies decay at the same rate.
Figure 11 illustrates with measurements before and after the addition of three vertical corner traps. This is a respectable improvement, with the two most prominent peaks around 44 and 75 Hz with serious ringing having been significantly damped.
While the ringing at 75 Hz has been completely damped, the 44 Hz mode still has some ringing, as highlighted in Figure 12. You can also see that the sharp tails from 100 – 300 Hz now decay at a more even rate. Some dips above 100 Hz are also introduced, but those were not caused by the traps and can safely be disregarded. The microphone had to be shifted and repositioned to put one of the traps in place. A slight shift in microphone position can have significant effects.
While the waterfall gives a more intuitive picture, the 2D decay plot allows us to more accurately assess how well damped the modes are with respect to our decay targets. Figure 13 shows the decay plot with the three vertical traps – a good result in terms of bass damping, but we are still short of the target. The two modes that were identified in the waterfall could use some more damping. Above 100 Hz we can also see some ringing remains, but if we increase the broadband absorption below 100 Hz, the ringing higher up should also be resolved. There are some significant dips seen here, but at this stage we aren’t concerned with those, because steps later in the process will address them.
Using larger bass traps will effect an even greater improvement. In Figure 14 you can see that, with larger traps, the ringing at 44 and 60 Hz has been well damped so that it no longer stands out. The decay is quite consistent and the waterfall indicates that the traps are working well. Don’t expect to meet any targets at this stage of the process. The key thing to look for is the consistency of the decay. As the damping is increased, the waterfall will start to look smoother.
You may notice that there is significant roll-off below 30 Hz. This is due to different EQ and bass management settings and isn’t related to the larger traps that were used. Likewise, the boosted bottom end in Figure 12 is not related to the influence of the traps at all.
If bigger traps are not an option, then an alternative would be to use some membrane traps to target the 40 – 65 Hz region. To improve the region above 100 Hz, additional broadband traps could be installed. Given that they would only need to be effective to 100 Hz, they could be smaller in size (bulkhead traps, for example).
The audibility of modal ringing
It should be noted that there is some debate regarding the audibility of modal ringing below 100 Hz. There are those who would argue that the attention paid to reducing decay below 100 Hz isn’t warranted. This line of thinking can lead to the omission of acoustic treatment, focusing instead on the other aspects that will be covered in Part 3 of this series. My experience indicates that the improvement in damping modal ringing acoustically is dramatic and that the same result can’t be achieved by other means. The improvement is so dramatic that I have noticed it while in another room with the door shut.
Further reading on bass traps
The BBC has published the results of considerable research into the design of membrane traps for use in their studios. This paper documents a modular membrane trap that works in the 50 – 100 Hz range:
This paper covers an absorber effective above 100 Hz:
The research shown in this report shows how a membrane trap can have some unwanted and unexpected effects:
- BBC RD 1980-12 – An investigation into the mechanism of sound absorption in a low frequency modular absorber (PDF file)
This paper shows a design intended to be switched on and off. You probably don’t need that aspect, but it’s useful to have another design to consider:
The advantage in building the BBC traps is that their performance has been measured. Other designs are available, but many of them don’t have published measurements, so results will be unpredictable. Even a quick skim read through these papers will indicate why membrane traps aren’t my first recommendation. The details are critical and any variation from a known design can lead to poor results.
The next step is to determine the optimum placement of speakers, the number of subs required and their best positions. All bass sources are then aligned in the time domain and crossover settings are chosen. EQ is applied to achieve maximum smoothness and the final response is adjusted to achieve the desired target curve. These steps will be covered in Part 3.