One of the biggest obstacles to achieving pro-quality results in a home studio is its acoustics. Most of us work in spaces that are not purpose-built as studios and are unlikely to have been professionally treated. The acoustical anomalies of untreated rooms distort the frequency response, making it hard to create mixes that translate accurately to other speaker systems.
One answer, of course, is to get your studio treated. Unfortunately, that’s not only expensive, but it often requires physical alterations to the space that may not be practical — especially if you don’t own the property. Fortunately, there is a software-based-solution in the quest for flat-frequency response: Sonarworks Reference 4 is a suite of two applications and a plug-in that correct the sound of your studio monitors at your listening position. To do that, it uses convolution-based equalization governed by a calibration profile that you create through a comprehensive but easy-to-execute measurement process.
The system also offers headphone calibration through a sizeable collection of custom calibration profiles, including a range of popular makes and models. Whether you’re using the speaker or headphone correction, the end result conforms to Sonarworks SR standard, which is designed to provide neutral (e.g. flat) response. You’re also offered post-measurement frequency customization options if you’re going for a target sound that is not entirely flat.
The developers hope that Sonarworks SR will be adopted industry-wide to provide a single listening standard that allows music to translate easily across all listening systems. Currently, Sonarworks Reference 4 only supports stereo output, so you can’t use it for surround mixing.
In this Masterclass, I’ll explain how the system works and how to get it up and running.
WHAT YOU GET
Sonarworks Reference 4 Studio edition ($249, Mac/Win) includes the Reference 4 Measure and Reference 4 Systemwide applications, plus the Reference 4 DAW plug-in. Sonarworks also makes True-Fi, a consumer version for headphones only, but we are going to focus on the Studio edition.
The Reference 4 Measure application guides you through the measurement of the sound of your speakers at your studio’s listening position, and then creates a custom calibration profile that conforms the sound at that spot to Sonarworks SR. You need a measurement mic when using Reference 4 Measure, and Sonarworks offers its XREF 20 mic along with the software in a bundle called Reference 4 Studio Edition with Microphone ($299). You can also use another Sonarworks measurement mic or a third-party model.
The Reference 4 Systemwide application and the Reference 4 plug-in are what you use to apply Sonarworks SR to your computer audio or your DAW, and from both, you can access your custom-measured calibration curves and the factory preset headphone curves. Reference 4 Systemwide uses a virtual audio driver that lets you conveniently apply Sonarworks SR across all your computer audio applications. If you’re using the plug-in, you insert it directly on the master bus of your DAW or other audio application.
When you’re mixing and want to check something on headphones, you can switch Reference 4 Systemwide or the Reference 4 plug-in — whichever you’re using at the time — to the calibration curve for your headphones and get a similar frequency response to what you’re getting on the speakers. That level of consistency is useful not only when switching between headphones and monitors in the studio, but if you sometimes mix on headphones while traveling, and then finish projects in your studio or vice versa.
Sonarworks chose not to implement any 3D spatial effects in its headphone calibration: They are not trying to make the headphones sound like speakers in the way that, say, Waves NX does. Instead, they focused on frequency response and phase to make the response as similar between speakers and headphones as possible.
A MEASURED RESPONSE
Testing your studio’s acoustics with the Reference 4 Measure application is a relatively simple process that takes about 20 minutes to complete for each pair of speakers. Starting with the recently released version 4.1, Sonarworks has made the testing process more comprehensive, yet more user-friendly than before.
The software walks you through the various measurement steps simply and quickly. You don’t need to measure your room dimensions or do any calculations in advance; it’s all handled by the software. All you have to do is hold the measurement mic and follow the instructions.
In one of its early screens, the software helps you make sure you’ve correctly configured your system. It provides you with a handy checklist that is new to version 4.1 (see Fig. 1). The list prompts you to turn on the phantom power to the measurement mic, set the sample rate to 44.1 kHz, and make sure that your input and output devices are from the same audio interface. It also reminds you to make sure your mic input isn’t routed to the output going to your speakers.
Once you’ve checked off the items on the list, the software asks you to enter the microphone ID# of your Sonarworks measurement mic, which allows it to know which of its models you’re using. If you have a different brand of measurement mic, Sonarworks recommends that you load a calibration file for it, which is a text document that should be available from its manufacturer.
The application doesn’t leave anything to chance. It even offers the option of viewing tutorial screens that give you detailed instructions before each significant step (see Fig. 2).
Next, Reference 4 Measure walks you through adjusting the mic input gain (on your interface) to the optimum level for the measurements. Before doing so, it allows you to set the Countdown time, which is the number of seconds (during this step and for subsequent ones) after you press the button labeled Start Measuring, before it actually starts to measure data. I recommend setting it to 5 seconds so that you have enough time to get positioned correctly before the software begins measuring.
For the input calibration step, once you hit Start Measuring and the countdown ends, the software produces a “whoop whoop” noise used for the level setting. The software prompts you to turn the input gain up or down until it’s in the correct range.
I found it helpful to rest my free hand on the input level control of the interface channel I was using, so I didn’t have to feel around for it if it needed to be adjusted, or turn my head to find where it was. You want to keep your eyes on the screen so you can see when you get the level in the right range.
The next step in the measurement process allows the software to calculate the distance between your speakers. You’re prompted to place the mic so that it’s almost touching the center of the woofer for the left speaker first, followed by the right, as tones get played to help the software determine the distances.
Once that process finishes, which only takes a few seconds on each speaker, you’re shown what those measurements are. If something seems wrong, you redo the test. One of the many nice touches of the Reference 4 Measure software is that for most steps you get a Back button that allows you to repeat a step if you think you did something incorrectly.
Next, you’re asked to place the mic at ear level — at your listening spot — so that it can measure the distance from there to each speaker. Knowing the location of your listening spot is essential because Reference 4 Measure will base its calibration curve, which is determined by the primary measurement process (and is coming up next), on where your listening spot and speakers are.
By default, Reference 4 measure assumes your listening spot is centered between your two speakers. If it’s not, you can uncheck the My Listening Spot Is Centered checkbox, and the software will determine where it is based on the response from the mic. So, if for some reason you don’t sit centered between your speakers, the software will compensate.
TESTING 1, 2, 37
The final and most important part of the testing regimen is the measurement of the sound of your monitors at your listening spot. When you press Start Measuring, the software shows a diagram of your listening position, including an icon representing the real-time location of the mic. For the first measurement point, a large colored-in circle appears. A loud clicking sound plays as you move the mic until the screen shows it to be inside the circle. Then the software tells you to hold your position, and you hear what sounds like a laser from a sci-fi movie play a couple of times, until the software indicates that it has completed its measurement for that spot.
Then another circle appears, representing the next measurement point. You move the mic into it and the software captures it, after which the next circle appears. A total of 37 different points are measured during this process, which takes about 10 minutes to complete (see Fig. 3). You can pause it in the middle if you need to, and you can also go back to the previous measurement if you feel like there was a problem; for example, if your phone rang or your dog barked while data was being captured.
For these measurements and all the other steps in the process, Sonarworks recommends that you hold the measurement mic with your arm outstretched. That way, your body will cause fewer reflections and, as a result, the measurements will be more accurate.
Once all 37 positions have been captured, the software tells you the process is complete and shows you the correction curve it created in the form of a graph that compares it to their uncompensated state (see Fig. 4). It prompts you to save the curve as a Speaker Profile.
The first time I used Reference 4 Measure, I put the mic on a stand for all the measurements, figuring that it would allow me to get more precise results, because the height would stay the same for every point. I assumed that if I were hand-holding the mic, there would be more variations.
I mentioned this to the folks at Sonarworks, who told me that it is not necessary to use a stand, because there is a margin of error built into the measurement process to compensate. The calculations average out the fluctuations of hand movement over the course of all 37 positions, so it doesn’t have an appreciable impact on the results. And it is easier to move the mic to the next location when it is held by hand. (It was undoubtedly much more awkward to move the mic into the correct spot when I used the stand.)
With the measurement finished and your calibration profile created, you are ready to apply the correction to your system. Whether you do so using the Reference 4 Systemwide application or the Reference 4 plug-in depends on a number of factors. In many ways the experience is similar with both the application and the plug-in, because both offer the same control set, and their GUIs look identical. Both also offer two operating modes, Zero Latency and Linear Phase, although it’s not quite that simple, as you’ll see shortly.
Once you launch the Reference 4 Systemwide application (see Fig. 5), it stays on in the background and applies your chosen calibration curve through a virtual audio driver that shows up in your application’s audio preferences or your computer’s sound preferences as a stereo output. An “SW” icon appears on your menu bar. From the icon, you can choose a speaker profile or open the full GUI.
If you’re running multiple audio applications, the Systemwide software may be the more convenient choice because you can switch between applications and the correction will be applied globally, no matter if you’re listening to your DAW, audio editor or Spotify (although for DAWs and other audio production apps, you might have to select Sonarworks Systemwide in the audio preferences.)
There is a trade-off for that convenience, however, because the application’s virtual software driver adds 55 ms of latency; interestingly, that is when it’s set to Zero Latency mode. (Sonarworks acknowledges that the nomenclature is confusing and says it will change it in an upcoming release.) Although 55 ms sounds like a lot of delay (especially if you were tracking), I didn’t find it to be distracting when mixing, although some people might! If I were writing automation with a fader or some other realtime activity that required precise timing, the delay could be problematic.
If you don’t want to deal with latency, the Reference 4 Plug-in, which you insert on the master bus of your DAW, is the better option. It’s Zero Latency setting is accurately named. You do get 50 ms of latency in Linear Phase mode, but, as I mentioned earlier, it’s not diffcult to deal with when mixing. Linear Phase mode creates no phase shift plus it is more accurate, and it is certainly worth using, at least toward the end of a mix session.
The plug-in is also your only option if you want Reference 4 correction when tracking, because the Systemwide app’s virtual driver has no inputs. Even if it did, the latency would make tracking impossible.
For me, there’s one big downside to using the plug-in. Because you put it on the master bus, you have to remember to bypass it before bouncing. The Reference 4 correction is intended for monitoring only and would skew the frequency response if you printed your audio with its correction. It doesn’t seem like a big deal to remember to bypass it, but I found that it was easy to forget in the heat of a mixing session.
Mainly for that reason, I found it more convenient to use Systemwide when mixing because it performs its calibration outside of your DAW’s signal path, so bypassing is unnecessary. I’d almost rather put up with the latency of the application, than face a situation where I just rendered a final mix along with alternates and then discovered that I’d left the correction on and have to do it again.
Sonarworks plans to add a feature down the road that warns you to turn off the correction when it senses you’re going to bounce a mix. That would be welcome.
Other than choosing the correct calibration file, you can use the default settings of either the Reference 4 Systemwide application or the Reference 4 plug-in and get good results. But if you want to customize the behavior and response, some user adjustments are available in the GUIs of both.
The Safe Headroom switch (which is on by default) keeps the output low enough to prevent clipping, but can be turned off. if you want more level. There are also limiting options available that allow you to restrict how much boosting the software can do in particular frequency ranges during the correction process. If you monitor at loud levels and are worried that the EQ boosts of the calibration curve might blow a speaker, you can set the limiter for more aggressive protection.
You also get a Mono button, which is handy to have on your output when mixing so you can check mono compatibility. Another calibration option is called Bass Boost and Tilt. When enabled, it gives you a 2 dB bass boost and presents you with adjustment controls for the bass level. It also features a Tilt slider, which operates like a tilt EQ, boosting bass and cutting highs simultaneously, or vice versa, depending on where you set it.
Both of these controls are there to allow you to tailor the Sonar SR calibration. I asked Sonarworks about this and was told that some people are used to mixing in bass-heavy rooms, or situations where they’re accustomed to hearing highs or lows accentuated, and they don’t want to eliminate those sonic characteristics from their monitoring entirely.
Another option, Predefined Target Curves, has two choices. The first is X-Curve, a theater sound target curve that attenuates both the bottom and top end. The other, B&K, simulates the response of B&K hi-fi speakers from 1974. In Sonarworks software before version 4, you could find simulations of a range of studio monitors, but the company pulled those out due to trademark issues and hopes to be able to return them in a future version.
Interestingly, the GUIs also have a Wet/Dry control. I asked Sonarworks about why this would be useful, and they told me that it is a legacy feature from early versions of the software. Back then, Sonarworks had yet to develop its current stellar reputation. People using the software didn’t always trust the results entirely and wanted to be able to dial in the calibration only partially.
WORKING WITH SR
Up until now, I’ve mostly been talking about the features in Sonarworks Reference 4 and how to use them. But the critical question is: how well does it work? Based on my experience, very well.
With Reference 4 correction applied, the frequency response of my two sets of studio monitors (PreSonus R80 and Genelec 1029A) is now quite consistent. That consistency extends to my Audio-Technica ATH-M70X headphones as well. Naturally, there’s a big difference between the experience of listening to speakers as opposed to headphones, but it feels similar from a frequency response standpoint.
I did quite a bit of mixing while monitoring with Reference 4, using either Systemwide or the plug-in, and I found that my mixes translated better to other systems than ever before. My untreated studio tends to attenuate bass, so before using Reference 4, I had to consciously undermix the bass to compensate. With Reference 4, the bass levels in my mixes translate more accurately.
I’ve found the Sonarworks system to be a valuable tool providing a flat mixing environment that would cost a lot more to achieve by way of acoustical treatment.
If studio acoustics are an issue in your studio, give Sonarworks Reference 4 a try. It could be the easy, and relatively inexpensive, solution that proves perfect for you.
Marty Cutler is the author of The New Electronic Guitarist, published by Hal Leonard.