Sonarworks HD Reference Review: Sonarworks have delivered a superb package that will transform the sound of your speakers, removing the guesswork involved with mixing audio that will translate well to other systems. A game changer.
A relatively new name on the scene, Sonarworks were formed in 2011. Based in Latvia, the outfit specialise in what they call ‘accurate sound’. Through their expertise in psychoacoustics and digital signal processing, the team engineer high-end software solutions that seek to provide true reference sound through headphones or speakers. This review concentrates on their speaker calibration system, so for our review of the headphone calibration click HERE to be taken to the relevant page.
Sonarworks claim their speaker calibration software delivers ‘accurate sound in 15 minutes’ for any room. Naturally, this is a hugely attractive proposition when one considers the cost, expense, and space requirements for physically treating a room – not to mention the uncertainty that the results will be good when doing so. Furthermore, with many project studios operating in very tight or dual-mode spaces, the physical treatment approach may not even be a viable option.
The software approach is ostensibly a simple one – record a known reference sound (usually a full sweep across the frequency spectrum) playing out of the speakers through a flat response microphone (this will give us the signal, plus the sound of the room). Next, compare that signal to the original known reference sound (i.e. the signal, without the sound of the room). The difference in the curve between the two is, effectively, the sound of ‘the room’.
We are being overly simplistic here of course, but in essence – that is the process. This exact opposing curve can be used thereafter as the last processor in your DAW’s signal path to correct the sound of your speakers. The result is ‘true’ reference sound through the removal of the room in real time from your audio, as it plays.
The Sonarwarks system comprises two parts – the measurement software (a standalone application), and a plugin for utilisation within your DAW. The measurement software is employed in the first instance to analyse your room. The calibration curve produced by the measurement software is then imported into the plugin and sits at the end of your DAW’s signal path (i.e. the very last plugin on the master bus).
There are two separate versions of the software – the standard edition, and the HD edition. Although Sonarworks haven’t nailed down the full feature set for the HD edition at the time of writing, what we know so far (and this is subject to change) is that the non-HD version can only be used with the Sonarworks microphone. Additionally, the HD version allows more than 24 measurement points to be taken, and allows for the analysis of different ‘sweet spots’ on the fly (for example if you want to switch the accurate sound from your mix position, to the clients position on the sofa at the back of the room).
In the package you get a single licence of the room measurement software, and two licences of the plugin (which comes in AU, AAX Native, RTAS and VST flavours).
System requirements are either Windows (XP, Vista, 7, 8 – 32-bit and 64-bit), or Mac OSX 10.7 (or later, 32-bit and 64-bit). A minimum of 2Gb RAM and 1024×768 display are recommended, and a USB port or internet connection for installation and licensing.
Additional requirements for operation are a measurement microphone with a flat frequency response (included in the Sonarworks bundles or available separately) and, importantly, any audio interface with phantom power (for XLR microphones).
Calibration begins with the measurement software that is provided on a USB stick. At the time of shipping the software provided is up to date, however the developers do seem very active with new releases and it is therefore recommended you ignore the version supplied and instead head straight to the Sonarworks website to download the latest version.
Installation is simple, and the package weighs in at just under 90Mb. Once installed the activation is carried out either online (through the entry of your email address and activation key), or offline (through a supplied licence file).
At this point I feel it is worth mentioning the test environment used for the review. The effective space is a typical small ‘box room’ project studio that measures 2m wide, and 2.4m long. The main monitors that will be calibrated are Adam F5s and a pair of M-Audio (owned by Avid) BX8a which are actually very bottom-heavy and boomy in the current space. It’s important to note that the room being analysed has absolutely no treatment of any kind, and suffers from the usual issues one would expect in such a small space – compromised definition in the highs, boominess and boxiness in the lows and lower mids, and a general ‘smearing’ across the frequency spectrum (which is particularly pronounced in the 100 Hz to 500 Hz range). Stereo imaging is poor and because of this I have typically not completed final mixes in this room because of the aforementioned qualities which, in other words, makes this the perfect environment to see what the software is really capable of!
Firing up the software, the first thing you need to do is select your audio input/output (I/O) options, your ‘latency detection mode’, and microphone type. Because of the nature of how the measuring process works, accurate knowledge of the latency in the system is crucial. The reason for this is that Sonarworks use a number of proprietary technologies to make calculations that determine the measurement area size automatically, and crucially where in the room that the measurement microphone is placed. Dubbed the ‘Automatic Microphone Positioning System’ (AMPS) this is all very clever stuff and operates through the playback of various short noise impulses and brief frequency sweeps to pinpoint exactly where in the room the microphone is. Whilst the basic method of latency detection requires no additional configuration, if your interface is capable of sending the same audio to another output simultaneously, or you have a good quality Y-cable, I recommend choosing the ‘loopback’ option which is more accurate, and can measure larger spaces.
Choosing the microphone type is a critical step – and it is pointless continuing the process unless this is configured correctly. To save a lot of confusion and limit the possibility of taking inaccurate measurements, I strongly recommend purchasing one of the reference grade measurement microphones directly from Sonarworks (available bundled with the software), as they can provide you with a unique calibration file, that is solely aligned to your specific microphone. Regardless of the measurement microphone purchased, Sonarworks analyse the microphones in their lab one-by-one, measure the frequency response, and then generate a unique calibration file for each. Following this you simply import the file they provide for your microphone (determined by its serial number) into the software to ensure you are working as close to reference as possible. Other microphones can be used, but there is so much variation between different makes and models that you will be working ‘in the dark’ to some extent. Unless you can guarantee that your microphone has a flat frequency response, or the manufacturer is able to supply you with a calibration file, then it’s best to play safe and contact Sonarworks directly.
The early stages of the measurement process involve checks of your microphone level, speaker level and, if you are using loopback, the frequency response of your audio interface is analysed and then factored into the measurement process.
Once the preliminary measurements are completed you then move on to the first room measurement which is taken from the sweet spot between your speakers. The software then produces an initial Amplitude Frequency Response (AFR) for your room which shows the various peaks and dips across the frequency spectrum. As you can see in the diagram below, the 10dB peak above 100Hz in this particular room is typical, and quite an eye, or should that be ear?, opener! If it turns out that the room is amplifying the lower frequencies produced by your monitors you’ll know that any mixes that you have previously completed will not translate well. Another interesting thing to note from the adjacent diagram is the ‘dip’ at around 18KHz on the left channel, with a corresponding (smaller) ‘peak’ on the right channel. It turns out that the room being measured unfortunately has a number of shiny reflective surfaces on the right hand wall which is probably the contributor to why the right side of the room is appearing brighter as higher frequencies are being reflected by them.
Following on from the above, the next stage involves an analysis of your ‘listening spot’ (i.e. where you will sit when listening which may not necessarily be in the sweet spot). This presents interesting possibilities in the HD version in that the listening spot can be changed on the fly and, if there is more than one person in the room, the sound can be tailored to work for their individual location.
Once the ‘listening spot’ has been analysed we can move on to the final measurements, which involves moving around the room holding the mic and pointing it at the phantom centre of your stereo image in order to allow the system to take several acoustic snapshots. During this process the speakers first emit a series of clicks to determine the microphone location, and then a series of frequency sweeps are emitted at various volumes. The more snapshots, the more accurate the model – in the HD version of the system, 70 measurements are a good target recommendation which actually takes only 5 minutes or so to complete.
As the system knows where you are for each measurement (via the AMPS system), a detailed analysis of the AFR at several points in the room, along with an understanding of where exactly each analysis point is within the room feeds into the calibration engine to provide an enormously detailed model of the acoustic characteristics of the space.
The diagram shows the finished measurement overview with the locations of each analysis point. Once the process is complete, it is time to export your calibration curve ready for import into the plugin.
The plugin should be placed as the last processor in the signal path – this will generally be on your master bus, at the very end of the chain (including after any metering). This is important to keep in mind as the plugin is not designed to modify your audio in any way, it is purely there to aid the listening process – after all it is a plugin for your speakers and not for your music. As such, it should be disabled before you do any export, and only enabled for listening during the time in which you are listening in the same room, on the same monitors that it was calibrated for.
At its highest quality setting the plugin introduces a fair bit of latency – on my system this was in excess of 60 ms, which is certainly too much to be able to track with. Conveniently the plugin has a simple slider which decreases latency (and quality) at the same time, with a visual readout of exactly how many milliseconds of latency are introduced for each step.
Another welcome touch is the dry/wet rotary knob, which graduates between original sound and an ‘accurate’ sound. This is an excellent feature for slowly adjusting to the new reference sound over time, especially if you have become accustomed to the issues in your room and previously relied on a ‘sixth sense’ to compensate your final mixes.
If we move over to the ‘custom’ section now, this allows for you to custom craft the frequency tilting, which can add a slight bass or treble boost (as required). If you are one of the people that prefers to mix with a slight boost/reduction at the high or low ends, this is very welcome feature. Alongside this, the ‘simulate’ section further pushes the flexibility of the software to mimic other speaker types through your own monitors (including those famous white-coned workhorses; the Yamaha NS-10s amongst others).
Finally, the plugin sports a handy mono monitoring switch (which removes the need for a separate plugin in my case), a listening spot selector, and a bypass button (that maintains levels when both active/inactive – perfect for comparing the sound with/without room at the same amplitude).
There are so many advantages to using this system – whether you work in a small space, or in a well-treated, high end professional studio. Being able to compare the sound of the corrected and uncorrected audio so easily is very enlightening in that what you may have thought was a detailed, flat sound from your nice shiny studio monitors can be revealed to be the polar opposite.
In the case of my project studio, and my main nearfields (Adam F5s), the system has worked wonders. I was quite aware the room had issues, but over time had adapted to working around them (for example the understanding that there was an appreciable peak in the 100 Hz to 200 Hz range, that would need to be compensated for in the mix). Mixes, when they had to be completed in the room, would need to be checked on multiple systems to see how they translated. (I’m sure I’m not the only one with a mountain of scribbled on CD-Rs and a notepad in the car!). This software revealed the exact nature of the issue, across the whole frequency spectrum, homed in on it, and fixed it.
And the sound now? One word – incredible. The first time I tried it, I played song after song through my monitors, and just listened to and enjoyed this new lease of life the whole system had been given. All of that low-mid smearing had vanished, the high end was wonderfully open and clear, the low end beautifully focussed and tight. The whole spectrum was presented in such stunning definition, with excellent width. A separate profile for my lower end M-Audio BX8a monitors, showed that they too responded excellently to the process – the bottom end reigned in tightly, and hugely increased definition in the mids and highs. Having always considered these units unsuitable for serious mixing applications, they are now very useful indeed (a prospect not thought possible in such a small room before).
The ability to simulate other speakers is a welcome bonus – the Yamaha NS-10 profile in particular. Whilst this is simply an equalisation clone, the results are excellent, and genuinely believable. Close your eyes and you would swear those white cones were right in front of you – very cool indeed.
Downsides? Besides a couple of teething issues with configuring loopback correctly, and wishing there was more in the way of help files and video tutorials, my only grievance is that the sound is so good it is a pain to have to disable it when tracking and writing parts! However, this could soon be addressed with Sonarworks making mention of a forthcoming ‘zero latency’ version which would mean the plugin could have a permanent home on the mix bus from tracking, right through to mastering.
If you are in any way reliant on hearing reference sound from your speakers, and are even slightly unsure that you are getting the best possible sound from your room – this is simply a must have. I’d go further to say that if you are in the market for new monitors, then I would factor in the cost of this software (and microphone), because there is no doubt the sound would be significantly better than by simply buying more expensive monitors.
The idea behind what we have here is in itself, not new, but the technologies Sonarworks have developed to model, measure and calibrate the space are in a different league to anything that has gone before – and the result is, as they say, a ‘game changer’. Having used the system for some time, I can’t imagine that I would ever mix or master anything again without it. Grab yourself the microphone through Sonarworks, and take the 21 day trial for a spin, and I’m certain you will feel the same, whatever kind of room you work in.
Now that I have measured my project studio room and calibrated it with this software, I can’t imagine ever working without it. An incredibly vivid, detailed soundstage is now present, stereo imaging is excellent, and every aspect of the sound is delivered with wonderful focus and clarity. Once the sound of the room is effectively ‘removed’, mixing and mastering audio that translates well to other systems is a breeze.