It’s really important to get the low end right at live events. Research has shown that 3/4 of what people consider to be high quality sound comes from the low frequency content.

Subwoofers are a big part of that low frequency content, supporting and extending its capabilities. However, subwoofers also require careful setup and alignment to ensure optimal performance.

If you’ve ever had trouble getting your low end right, then you might want to read this article. It will explain why subwoofers need to be aligned properly and how to do it.

What is subwoofer time alignment?

Subwoofer time alignment is the compensation for arrival time differences between sources at the listening position. The difference in arrival times may be caused by a physical distance offset or an electronic delay. It is not frequency dependent.

The journey of sound from transmitter to receiver is not instantaneous. If two sources are separated by any distance then their sound arrivals will also be separated. This is the common situation with mains in the air and subs on the ground. From the listeners perspective the subwoofer is closer and must therefore be delayed (or physically moved) to be time aligned with the main.

Does high frequency sound travel faster than low frequency sound?

In short, no.

The speed of sound in an ideal gas depends only on its temperature and composition. The speed has a weak dependence on frequency and pressure in ordinary air, deviating slightly from ideal behavior.

Wikipedia

To compare the speed of sound a different frequencies using the Rasmussen Report PL11b at 20ºC, 50% RH, 101.325kPa:

20Hz: 343.987m/s

20kHz: 344.1206m/s

From 20-20,000Hz the speed of sound changes by only 0.1336m/s.

What causes subwoofer time misalignment?

Subwoofer time misalignment can be caused by acoustic or electric latency. Acoustic latency occurs when two sources do not have matched distances. Electric latency happens upstream in the signal chain, often in a digital signal processor (DSP).

Unless the receiver sits equidistant from both sources, some amount of acoustic latency will always occur. Ignoring any boundary effects, imagine a situation where the entire audience stands at 1.6m height. With a subwoofer on the ground and a main speaker at 3.2m height there is no difference in distance from each speaker to the audience. Everywhere else, there is.

Electrical latency can occur anywhere in the signal chain, but often occurs when one source is processed separately and differently than the other. If two matching copies of a signal are sent to the main and sub then there is no latency, but if the signal for the sub is processed independently through ten plugins then there will be a difference in latency.

How much latency or misalignment is too much?

When a main is stacked on top of a sub we don’t usually worry about the acoustic latency. When the sends for main and sub are split in a DSP we also don’t usually worry about the electrical latency.

Why is that?

Acoustical latency

The wavelengths of low frequencies are relatively large and require a big change for misalignment to bother us. For the purposes of this article I will define a significant misalignment as anything beyond 60º or 17% of a cycle because it will produce a reduction in summation of 0.5dB.

How far apart do our speakers need to be to create a 60º misalignment?

The operating frequency range of a Meyer Sound 750-LFC is 35-125Hz. The highest frequency has the shortest wavelength and therefore the greatest risk. The wavelength of 125Hz is 2.75m, about the height of a male ostrich. 17% of 2.75m is 0.46m, about the length of your forearm.

If we return our example of a seated audience with a sub on the ground then the main would need to raise up to 5.15m to be 0.46m farther away than the subwoofer from the mic position.

Don’t try to generalize this example into a rule. You could just as easily put the sub in the air with the main, but 0.46m behind it to create the same misalignment or change the microphone position.

It is difficult to generalize, unfortunately, because the relationship between source and audience will always be different. However, I can see how it is helpful to translate alignment to distance. This is why the SubAligner app includes maximum distance offset in the Limits pop-up.

The opportunity here is that after you have performed an alignment for a single location then you can move out from that location in any direction while observing the change in distance offset to find the edges of the area of positive summation (aka the coupling zone, aka <120º).

Electrical latency

Matched electrical latency is maintained by splitting the send to main and sub at the last moment necessary. This doesn’t mean you can’t mix to subs on a group in your console if you prefer, just make sure that the sends to main and sub are coming out of the console with the exact same latency. You can verify this with an audio analyzer. 

Time alignment vs Phase alignment

Subwoofer time alignment can be confused with subwoofer phase alignment because the two are interconnected. Time offset causes phase offset, but phase offset doesn’t necessarily cause time offset.

In most cases the timing is set to “align” two (or more) signal sources so as to create the most transparent transition between them. The process of selecting that time value can be driven by time or phase, hence the relevant terms are “time alignment” and “phase alignment.” These are related but different concepts and have specific applications. It’s important to know which form to use to get your answers for a given application.

prosoundweb.com

Time alignment connotes a synchronicity of sources, e.g., they both arrive at 14 milliseconds (ms). Phase alignment connotes an agreement on the position in the phase cycle, e.g., they each arrive with a phase vector value of 90 degrees.

prosoundweb.com

We have already seen how acoustic and electronic latency can affect time alignment. Let’s look closer at what can affect phase alignment.

What is subwoofer phase alignment?

Phase alignment is the process of matching phase at a frequency and location.

If a sine wave is generated starting at the 0º position of its cycle and then fed into a subwoofer, will it come out at 0º?

That will only tell us the story at one frequency, though. How can we look at the story of the entire operating range?

What does sound look like before it goes into a subwoofer?

This video compares the input and output of a microphone cable passing sine waves at 41, 73, and 130Hz with an oscilloscope. Traveling at the speed of light the mic cable appears to create no time offset.

I could insert a video comparing the input and output of a microphone cable with an impulse response, but without anything in line, they look the same. I added a 1ms delay to put the IR in the middle of the graph.

This image shows the transfer function of a microphone cable with a magnitude and phase graph. The magnitude and phase trace are effectively flat. Exactly what we want from from a cable.

What does sound look like when it comes out of a subwoofer?

This video compares the input and output of a subwoofer passing sine waves at 41, 73, and 130Hz with an oscilloscope. I have removed any latency so that we can focus on phase shift created by the sub.

This video compares the input and output of a subwoofer with an impulse response (IR). The IR seems to get stretched out as the amount of phase shift changes over frequency. This is the normal behavior of a transducer who’s group delay, and therefore phase shift, is variable and unable to reproduce every frequency at the same time through the operating range.

This video compares the input and output of a subwoofer with a magnitude and phase graph. Unlike most full-range speakers, the phase response of a sub never flattens out. It’s a moving target.

Do all subwoofers have the same phase response?

A subwoofer’s response will change with its mechanical and electrical design. Matching drivers in different boxes may have quite different responses. Even the same combination of driver and box might have a small contrast in response because a typical manufacturing tolerance is ±5dB.

For this reason it is important to avoid making assumptions based on a manufacturer’s spec sheet, but instead measure the final product and prove it to yourself.

Does the phase response of a subwoofer change with level?

A cold subwoofer operating within its nominal range should maintain a steady phase response against any change in level. But, as a sub approaches maximum SPL or begins to heat up, its response may become non-linear. This behavior will vary from subwoofer to subwoofer so it’s important to avoid driving two different subwoofers with the same channel.

Unfortunately, I don’t know a rule of thumb to guide you, but it would make sense to compare the response of a subwoofer when it’s cold to when it is hot. When I worked at Amex in Slovakia and we were setting up a new system, Igor would punish it outside playing loud music for a few hours and listen to it afterwards.

Of course you can measure this change with your audio analyzer, but another fun test is to push on the driver with your hand when it’s cold to feel how rigid it is. Run it at a maximum level for two hours. Push on it again. Feel how it has become less rigid (increased compliance).

Here is a graph from Heat Dissipation and Power Compression in Loudspeakers from Douglas J. Button showing a sample loudspeaker before and after full-term power compression. The solid line is the one with more heat and a worse quality rating.

Does the phase response of a subwoofer change over distance through the air?

…allow me to remind you that the loudspeaker’s phase response, within its intended coverage, typically doesn’t change over distance, unless you actually did something to the loudspeaker that invokes actual phase shift, i.e., applying filters of some sort which you should be able to rule out!

merlijnvanveen.nl

Here is the magnitude and phase response of a subwoofer measured at 1m and 100m. The only thing that has changed is the level due to the inverse square law.

Room interaction however, will make it appear like the loudspeaker’s phase response is changing over distance because the room makes the traces go FUBAR.

merlijnvanveen.nl

Here’s what that above measurement looks like if I enable four boundaries. At 100m the reflections have transformed the phase trace (blue).

Where is the acoustic center of a subwoofer?

Why does distance offset not correspond exactly with phase offset?

All other things being equal, the distance offset measured from your microphone to your subwoofer may not exactly correspond to the measured phase offset in your audio analyzer. This is due to an interesting acoustical phenomenon documented by John Vanderkooy.

As a useful general rule, for a loudspeaker in a cabinet, the acoustic centre will lie in front of the diaphragm by a distance approximately equal to the radius of the driver.

J.  Vanderkooy, “The Low-Frequency Acoustic Center:  Measurement, Theory, and Application,” Paper 7992, (2010 May.)

This fact becomes important when estimating delay times for subwoofer arrays where a small distance in the wrong direction could compromise the results. It may also be important if you are attempting to estimate subwoofer phase delay from far away without prior access to its native response.

What is the subwoofer crossover frequency?

A subwoofer’s recommended crossover frequency may exist on its spec sheet, but when it comes to subwoofer alignment in the field we must look beyond a single frequency to the entire crossover region affected by the alignment. To make an exaggerated theoretical example, imagine if you turn the subwoofer up by 100dB. The crossover region where they interact will also move up.

The crossover region is commonly found where magnitude relationships are within 10dB because you have the highest risk of cancellation and the highest reward of summation. To find this region in your audio analyzer, insert a 10dB offset and find the magnitude intersection. Some audio analyzers offer other tools like cursors.

What causes subwoofer phase misalignment?

The most common reason for subwoofer phase misalignment is user error. This may seem like a bold or aggressive claim, but manufacturers have historically placed their responsibility on their customers.

There are many subwoofers in the world and only a small number of them have detailed instructions on phase alignment within a narrow set of limitations. The rest require the user to discover an optimal alignment for themselves. This is further complicated by the fact that reflections can make measurement and listening tests misleading or impossible when performed under typical field conditions.

We saw above that what comes out of a subwoofer is not what goes in due to system latency and phase shift. Some products take this fact into account and are specifically designed to work together and are phase aligned when equidistant, therefore only requiring compensation for any distance offset. Other products are designed to work together, but are not phase aligned when equidistant. The third, and most common, scenario is that sound engineers like me and you end up combining products from different generations, families, and manufacturers that were never designed to work together.

I should pause here for a moment to say that I’m not passing judgment or point a finger. I don’t have enough aware of all conditions to say why things are this way, just that the complications exist. And honestly, I enjoy the puzzle. See any of the video on my YouTube channel from the past couple of years for evidence. 🙂

What are the consequences of subwoofer phase misalignment?

Let’s ask Nexo.

Consequences of badly aligned systems
Mis-aligned systems have less efficiency: i.e. for the same SPL you will be obliged to drive the system harder, causing displacement & temperature protection at lower SPL than a properly aligned system. The sound quality will decrease. The reliability will decrease as the system is driven harder to achieve the same levels. In certain situations you may even need more speakers to do the same job.

NXAMP4x1 User Manual v3.1

Do subwoofers need time alignment?

Yes, subwoofers need time alignment any time there is a distance offset creating acoustic latency. They also need phase alignment in any event when they are being combined with another source that is not already phase aligned when equidistant.

Do not assume that your main and sub are phase aligned when equidistant just because they came from the same manufacturer. You have a 33% chance of creating cancellation instead of summation.

How do you time and phase align a subwoofer?

Although there seem to be many methods, I have only ever found one that works reliably and has all three unobtainable characteristics: fast, cheap, and good. It may sound like I’m about to go into some wild conspiracy theory you’ve never heard of, but the method I use is also recommended by L-Acoustics, d&b audiotechnik, RCF, and Coda Audio (and probably more). It involves two steps: first in the phase and then in the time domain.

1. Create a relative equidistant alignment preset using filters, delay, polarity, etc. (this is the fun part)
2. Modify that preset in the field using the speaker’s absolute distance offset by adjusting the output delay time or physical placement.

The method goes by various names, but I’ll give Merlijn van Veen the credit for the Relative Absolute Method since he introduced the idea to me. I then packaged the idea into an app called SubAligner. It not only includes alignments for many major brands, but a total of 39,183 possible combinations between different brands.

How do you verify subwoofer alignment?

How do you know if you’ve done it correctly?

A listening test should reveal higher SPL and a tighter response around the crossover region. SubAligner offers a black and red pulse to focus your ears in the right area.

An audio analyzer should show matching phase response between each speaker and expected summation in the magnitude response through the crossover frequency range. Appropriately filtered IR peaks should be aligned.

All of these methods should work, but can be ruined by reflections. In these worst case scenarios, I still rely on the Relative Absolute Method because I’d rather use something I know to be true than try to speculate on what might be true. I have written more about this in Don’t Align Your Subwoofer to a Room Reflection and Can you remove reflections from live measurements for more accurate alignments?.

Have you tried this method? What were your results?

Acknoledgements

I want to thank Francisco Monteiro for the feedback and patience with my many questions and misunderstandings.