Bass Integration Guide – Part 1
Exceptional bass is nearly always the exception in high-end audio. You can buy great midrange, but truly accurate bass reproduction requires a room-specific solution. The investment and effort involved is considerable and that is why exceptional bass is so rare.
This article is the first of a three-part series. The aim in part one is to establish the need for the strategies that are covered in parts two and three. Part two covers acoustic treatment. Part three completes the integration process with speaker placement, phase alignment and EQ.
Completing part one will help you to understand your current level of bass performance. Parts two and three will then transform the bass performance. Complete every step in this guide and the result is expected to rival the best bass you have ever heard.
Throughout this article you will find links to my blog, which expands on particular points.
Why accurate bass is difficult
Achieving accurate bass reproduction in a domestic room is difficult because the room itself dominates the sound. Even a good room has many problems that require attention. We can easily identify a poor sounding room at higher frequencies, but in the bass range we are unable to make that distinction due to the limits of our auditory processing. As a result, we may blame the subwoofer for poor sound when in fact it is the room itself that is the problem.
The primary problem is room modes, with their associated frequency and time domain aberrations. In-room measurements reveal large swings in response that can be as much as +/- 20 dB. Some frequencies will decay at a much slower rate. This is referred to as modal ringing.
Whether you have a good or bad room, the same strategies are required.
Why you can’t trust your ears
A common view among audiophiles is that you can get accurate bass by tuning by ear, without using measurements. Once the nature of room-related acoustic problems are understood, the limitations of this approach become clear. While our ears are the final judge of sound quality, they are highly ineffective when it comes to the kind of specific data that is required to achieve our outcome. Our ears may tell us that there are bass problems, but we then need to use measurements to find and solve those problems. Once that is done, you can then evaluate the result subjectively.
What’s in this guide
The complete Bass Integration Guide outlines a 5-step process to getting great bass:
- Start with at least one high quality sub.
- Measure the existing bass performance of your room.
- Install acoustic treatment.
- Determine speaker/sub placement and the need for more subs.
- Apply EQ and complete integration.
In this first part, steps one and two are covered. While I’m tempted to try to convince you that both bass traps and EQ are essential, this guide is about finding out for yourself. By taking measurements of your own room, you can determine for yourself what you need.
This guide will show you how to meet certain performance targets. I have developed these targets after many years of experimentation, measurement and evaluation.
- Bass extension: Adequate low-frequency extension for your typical source material (this could range from 14 to 40 Hz).
- Headroom: At least 3 dB headroom at your maximum desired output level at all frequencies. This means cone excursion should not exceed 70% of the driver Xmax.
- Target curve: The combined system bass response including all subwoofers and the mains should match the target curve shown in Figure 1 below. The curve is determined by three variables that need to be chosen based on user preferences. The curve shown is just one possibility, but there are many variations where each variable is adjusted within the ranges shown. The midrange level rises at F1 until it reaches F2, below which the curve is flat until the low frequency roll off point. The boost amplitude and both knee points (F1 and F2) are to be chosen from the ranges shown in Figure 1.
- Frequency response: Flat response in every significant seat +/- 3 dB relative to the chosen target curve with third octave smoothing applied.
- Spectral decay: Decay rate of 20 dB in the first 150 ms from 40 – 300 Hz. (Note: The time window setting for any waterfall or decay plot used to determine this must be set to 300 ms.)
The target curve is flexible to allow for different user preferences. You should determine a target curve that works in your system, factoring in the level at which you typically listen and typical source material. Some would dispute the validity of using a target curve. Common concerns are addressed on my blog:
It should be noted that not every curve possible within the suggested ranges will work. For example, if you were to add 8 dB of boost and start shelving up at 150 Hz (F1) to a knee at 80 Hz (F2), there would be too much lower midrange resulting in a muddy sound.
You will need a calibrated measurement microphone along with a suitable preamp and stand. The cost of these items is around AUD $250 and the software is both free and easy to use.
Dayton EMM-6 is the recommended mic. It is almost identical to the Behringer mic shown above, but it’s cheaper and has better unit to unit consistency. Either mic can be purchased pre-calibrated directly from Cross Spectrum labs (Dayton EMM-6, Behringer ECM8000). You can also get an existing mic calibrated. Pre-calibrated mics are supplied with a calibration file that is unique to each mic. Purchased elsewhere, both mics come with a generic calibration file, but they can be out by as much as 10 dB.
This mic requires a preamp that provides phantom power. A popular choice is a Behringer mixer such as the Xenyx 502, an inexpensive unit. USB options are also popular, but they can suffer from compatibility issues. Units with an analogue output don’t have this problem, and the Behringer mixer is a cheaper option. EQ is included as well and can’t be defeated, so the controls should be set to their central position. Regardless of the choice of preamp, any deviations from a flat response can be compensated for in the calibration process when setting up the software.
Introducing the Room EQ Wizard
Room EQ Wizard (REW), is the program that we will use in this guide. It is very powerful and easy to use. Thanks to the generosity of its creator, it is also free. Here are some useful links to help you get started with REW:
- Download from Home Theater Shack
- Initial setup Instructions on my blog
- Check the online help files
- Visit the REW forum for more help
Step 1. Start with a high quality sub and mains with decent bass capability
You will need at least one high quality sub, and mains that have decent bass capability extending below your intended sub crossover point. A sub is required for independent placement flexibility and level control – mains should be placed with a primary focus on achieving the best imaging and this will often dictate a position that does not perform well in the bass range. Subs also allow the ability to set the level independently without overloading the mains.
You may need to add a further two subs to smooth out the room response, but whether this is required will be determined later.
At low frequencies, all of the subs will add to the output. Above approximately 40 Hz, they will tend to smooth out the room response without adding significantly more output than a single sub. You may choose to have one larger subwoofer placed for maximum extension and room gain with smaller subwoofers placed for maximum smoothness. Alternatively you may choose multiple identical subwoofers.
Consider your output demands
Estimating your output requirements and how they can be met can prove to be a challenge. The first challenge is choosing a suitable target. After completing the integration process and room problems have been tamed, you may choose a higher than expected bass level, one that would previously have sounded boomy.
The second challenge is that simulations are often inaccurate in estimating both room modes and room gain. That introduces a great deal of uncertainty.
Overlapping mains and subs
The conventional approach is to cross the mains over to a sub in the 80 – 120 Hz region. For example, bass content is filtered from the mains with a high pass filter when they are set to “small” in your AV receiver settings.
The approach recommended in part two of this series loads the room from multiple positions. Each position is carefully chosen so that when all bass sources are combined, the result is smoother than any of the individual speakers. Often it is beneficial to allow the mains and subs to overlap. That way, the mains and a single sub load the room from three positions. In the ideal three-sub setup you will have five bass sources.
In the example in Figure 6 below, the best location for the sub (red line) has a dip at 70 Hz. The mains (black line) don’t have the same dip so they are allowed to run down to about 60 Hz to eliminate the dip. With this arrangement all dips below 100 Hz are eliminated. Below 100 Hz, EQ is required only to shape the response to match the target curve. A conventional crossover would fail to eliminate the dips regardless of the number of subwoofers. No possible subwoofer locations were able to remove the dip.
Step 2. Measure the existing bass performance of your room
In this step, you will establish the performance of what you have now.
First, take a nearfield measurement of your subwoofer, as illustrated in Figure 7 below. If there is a port, you will need to take a separate measurement with the mic near the port as well. Do the same for your mains, including woofers and ports.
Due to the close mic position, the room has no influence on the response measured, resulting in the smooth response shown in Figure 8, free from room effects.
Generate a waterfall plot of each measurement. As shown in the example in Figure 9, the waterfall plot confirms the absence of room modes in the response. We are seeing the response of the sub without the room.
The decay plot in Figure 10 shows the same information as a waterfall, but it is easier to identify the precise drop in SPL at particular time periods. A waterfall more clearly shows modal ringing caused by the room. The decay plot shows a rapid level of decay without room modes. In the first 150 ms, we see mostly 20 – 30 dB of decay. This sets the limit of what we can achieve – as you will soon see, the speaker itself is certainly not the limiting factor in a real room.
Please note: The white lines have been added to assist in identifying the decay but REW does not create them.
For more information on these plots:
Now place the mic in the listening position so that it points downwards, with the end where your head would be. Run a series of measurement sweeps in REW, for each of the following:
- Left speaker only
- Right speaker only
- Left and right speaker together
- Each subwoofer individually
- Left and right speaker and all subs together
If you have more than one significant seat, repeat the measurements for each position.
Now you can analyze your measurements by creating waterfall and decay plots for each measurement sweep. Note that is important to set the time window for both the waterfall and decay plots to 300 ms, otherwise the stated targets won’t make sense. The screenshot below illustrates the correct settings.
In the waterfall plot in Figure 12, you can see a great deal of modal ringing, causing extended decay times. One particular mode around 180 Hz is marked. The peak in response is minor, but it decays much more slowly than the region below. If not for the influence of the room, we would see the rapid and even decay shown in Figure 9. The large variation in decay time shows the need for acoustic treatment. The decay rate is as slow as 30 dB in 300 ms.
The decay plot in Figure 13 shows the slow decay even more clearly, with the response shown in red, and the response after 150 ms in purple. The white line shows the target decay of 20 dB within a 150 ms time window. You can see that there is substantial modal ringing at many points, which cause the bass to sound exaggerated and boomy. The dips are less obnoxious in character but are still a problem.
Making sense of the results
Now you have a lot of measurements to analyse. You will find it helpful to view some of the response plots overlaid. Click on the All SPL tab and you can choose which plots you would like to overlay. You can also adjust their relative levels – click the controls icon above the top right corner of the charts.
- Frequency response. Compare your results to the performance targets. Apply third octave smoothing to the measurement of your mains and subs combined. Does the response fit the +/- 3 dB range up to 300 Hz?
- Waterfall. Now view the waterfall plot. In a well-tamed room the decay is even without any obvious modal ringing.
- Decay plot. Deselect all but T=0 and T=150 ms. Using the cross lines you will be able to estimate the decay at this point and compare it to the 20 dB target.
Save your measurement file with a descriptive name. It will prove useful later for the purposes of comparison.
Introduction to parts two and three
Nearly every system will fail all three tests above prior to undertaking a process such as described in the rest of this guide. In part two, we will mitigate the primary cause of most bass problems by treating the room. We will then use further measurements to better optimize the placement of speakers and subs. We will also determine if more subs will prove beneficial. In part three, more advanced techniques to complete the integration will be covered.
The concept of overlapping the mains with multiple subs in order to smooth the room response has been suggested by Dr Earl Geddes (www.gedlee.com) on a number of online forums. See for example the thread Multiple Small Subs – Geddes Approach on diyaudio.com.
The house curve is explained by Wayne Pflughaupt in House Curve: What it is, why you need it, how to do it on hometheatershack.com.