Steven Slate Audio VSX

Can headphones and modelling software really replace a professional mix room?

The Steven Slate Audio VSX system caused something of a stir when it was first launched. The idea that they would bring an end to the issue of monitoring translation by promising “Perfect Sounding Speakers. Now In Headphones” seemed for many, me included, to be somewhat over‑ambitious. However, the initial brouhaha has died down a little now, and a new version of the VSX software has recently been released, so it seemed time for SOS to take a look. Along with introducing some extra room models, the new release also previews an HRTF feature in two of the room models that is slated to be included globally in a future release.

Bundle Up

VSX comprises a pair of headphones and a software suite including AAX, AU and VST plug‑ins, along with a newly introduced ‘Systemwide’ application that enables VSX to do its magic outside of a DAW, processing any audio that the host computer is playing. But what exactly is the VSX magic?

VSX is a monitoring modelling system that aims to render, through its headphones, the sound that would be experienced in a variety of different locations or from alternative hardware: Steven Slate’s mix studio, a car interior, some alternative headphones and a smartphone, for example. In the case of modelled studios (and club or car spaces), the principle is that VSX creates a binaural analogue of the environment and its audio hardware, allowing one to hear how a mix will sound in those locations when played through the monitors or speakers present there. As an extension of the technology, VSX also offers non‑binaural models of some other generic headphones. VSX comes in two editions, Essential and Platinum, the latter providing access to an expanded list of modelled environments and headphones — 20 in total (with more apparently on the way), compared with seven for the Essential edition. Modelled environments can also be purchased individually, so if, for example, you can’t possibly mix without knowing how your work will sound inside a Tesla, that model can be added to an Essential VSX edition.

Can Do

Before getting into how well the VSX binaural modelling works, I’ll look at the VSX headphones themselves. They are somewhat generic in appearance and style but nonetheless of high manufacturing and finish quality, with a nicely padded headband and generously dimensioned oval ear pads. The connection cable attaches via a 3.5mm jack to the left earcup, and the headband and ear pads are covered with faux leather. They are unremittingly black other than the grey VSX logo. As well as lacking colour, they also lack any of the luxury aesthetic touches that seem in recent years to have become de rigueur for ‘audio wearables’, but to my mind that’s a positive; headphones intended for professional use are tools for a job, not fashion items.

The VSX headphones are of closed‑back, circumaural design and feature a high‑tech beryllium‑coated diaphragm and a low‑frequency porting arrangement called APS (Acoustically Ported Subsonics) that is described as a “sophisticated internal tuning vent system and patent‑pending bass coupling for optimal low‑end performance”. I’m not entirely sure what that actually means in real‑world audio engineering terms, but it appears to describe a headphone driver loading technique that at low frequencies allows the headphones to operate as a kind of hybrid between closed and semi‑open. In any case, and owing to the circumaural design, low‑frequency performance will depend to some extent on the quality of the air seal around the ears. The more generous the ear pads, the more consistent that seal is likely to be, and the VSX ear pads are generous.

In use, the VSX headphones are comfortable, with about average weight and ear‑pad pressure. Comfort in use is important, because if the VSX proposition really does result in “Perfect Sounding Speakers. Now In Headphones”, chances are you’re going to be wearing them for extended periods.

Plugging In

Used conventionally, without the VSX app running, the VSX headphones perform reasonably well, with a wide bandwidth, accurately rendered and extended bass, and a relatively neutral tonal balance (although not quite as neutral as my mid‑price circumaural headphone reference, the AKG K371). At the same time, they sound to me a little lacking in the upper midrange, and display what I’d describe as a slightly nasal character. That said, all headphones, and especially closed‑back types, display an aural signature to some degree, and the VSX are generally typical of the breed.

Such use is probably the exception rather than the rule, because with the VSX Systemwide app running it’s likely that most of the time the headphones will benefit from the EQ that the app applies. And with VSX Systemwide running (but in bypass mode so that no room modelling is applied), the subjective tonal balance of the headphones does indeed change: the upper midrange balance is restored and the slight nasal quality is effectively suppressed. They sound much closer to my K371 reference, and actually pretty good.

Diagram 1: The frequency response of the Slate VSX headphones without any EQ (red trace), and that of an AKG K371 (blue), as measured on a Neumann KU100 dummy head.

To illustrate the comparison between the VSX headphones (both before and after the VSX EQ) and the AKG K371s, I pressed my sound designer wife’s Neumann KU100 binaural head into measuring duties and fired up FuzzMeasure. While the KU100 isn’t a calibrated headphone measuring rig, it’s perfectly able to illustrate headphone comparisons. So, Diagram 1 illustrates a frequency response comparison between the un‑EQ’d VSX (red) and my AKG K371 (blue). The most significant differences are that the K371 has more low bass (the K371 is said to comply with the ‘Harman Curve’ headphone response target, which includes a significant bass lift), a flatter midrange and more energy in the 2‑4 kHz region. These characteristics confirm what I heard. I wouldn’t read much into the deeper dips of the curves above 4kHz, because the high‑frequency region is very much affected by how different headphone ear pads interact with the pinnae, and the generic pinna shape of the KU100 will appear to work better with some headphones than others. Our ears average things very effectively at high frequencies, too.

Diagram 2: Comparing the VSX response with and without the Systemwide EQ correction (red and green traces, respectively).

Diagram 2 shows a comparison between the VSX headphones without EQ, and their response when driven via the Systemwide app. Again, the comparison reflects things much as I heard them. With EQ applied, the VSX headphone response flattens nicely, and it sounds that way too.

So, considered in passive mode I’d put the VSX headphones in the ‘competitive if unremarkable’ category, and when the Systemwide EQ is added, I’d bump that up to ‘really pretty good’. However, the VSX system as a whole isn’t just about the headphones, but whether the complete VSX package means an end to mix translation worries.

Software

Rather than use the VSX as a master bus plug‑in, I began by firing up the Systemwide app and listening to some familiar material while experimenting with different room/speaker/headphone models. My VSX review sample included the full Platinum set of rooms so I had a lot to play with.

The VSX plug‑in and the Systemwide app look the same. In the bottom‑left corner are a bypass button and level control. Having a bypass function separate from that likely to be found in a DAW plug‑in instance is important, because it enables the VSX binaural room modelling to be switched off without loss of the headphone EQ. Remember, though, that if you’re swapping between VSX headphone and loudspeaker monitoring you will need to bypass the VSX plug‑in completely, otherwise your monitor will potentially have VSX headphone EQ applied. In this scenario it probably makes sense to have two monitor mix buses set up in the DAW — one for VSX and one for speaker monitoring and mix bouncing.

To the right of the bypass switch is a Depth knob. Depth adjusts the intensity of the binaural modelling, and in doing so, effectively changes the apparent listening position, with more depth making it seem as though the listener is further from the monitors. The Depth control is deleted for VXS non‑binaural models such as headphones. To the right of the Depth knob are ‘push‑button’ switches that provide options depending on the room model selected. A mix room model, for example, might have two or three pairs of monitors (typically nearfield, mid‑field, and far‑field), and these switches enable their selection. Finally, on the right is an output level control.

Next up on the VSX are five switches that enable room model favourites to be quickly recalled, an option switch that inserts a two‑second ‘palette cleanser’ silence when switching between room models, and an option to insert a five‑band EQ in the output. Before I get on to the room models themselves, right at the top of the display are options for Ear Profile and, in two room models, HRTF. The Ear Profile options are designed to accommodate the natural variation in human ear‑canal diameter. This is significant because ear‑canal diameter influences subjective tonal balance. Smaller ear canals potentially result in slightly emphasised upper midrange. Finally, on the two new room models, VSX offers two HRTF options, previewing a feature scheduled to be released more widely in the next version of VSX. HRTF stands for Head Related Transfer Function, and describes the effect of different head sizes, shapes and ear positions on binaural hearing. We all have a different HRTF, so any binaural recording or reproduction process relies on the use of an average (this is partly why binaural audio works better for some people than others). Providing HRTF options means that VSX binaural encoding can be adjusted to suit different listeners.

The main part of the VSX screen is taken up by the room model browser, with its pictorial representation of the available rooms and their monitoring. Selecting a room is simply a matter of clicking on its icon. Once a room is selected, the VSX headphone feed straight away reflects the modelled audio of the room and monitoring. And that all brings me on to how well it works...

Close your eyes and it’s really not difficult to imagine being in a studio hot seat.

In The Room

I have to admit to being a little sceptical before I began using VSX, but now, with some experience of how it behaves I’ve come to believe it works rather well. The binaural room models genuinely do a pretty convincing job of moving the listening perspective to another place with an alternative set of monitors. Close your eyes and it’s really not difficult to imagine being in a studio hot seat. I’ve always had a slight problem with the kind of modelled alternative monitor technology of, for example, Sonarworks or ARC room optimisation, because it really isn’t possible accurately or fully to turn one monitor into another simply by manipulation in the frequency domain (without any control of monitor directivity and a whole host of other factors). The VSX approach is much more successful because the full signature of the alternative monitors and room are encoded in the binaural analogue.

Listening to a range of familiar material and mixes with VSX engaged I found it completely fascinating to hear how things sounded in different rooms. For the most part I was reasonably relieved that music I’d mixed in my own room mostly survived the experience of being played in, say, Steven Slate’s room, although there were a couple of “What was I thinking?” moments. The car, boombox and smartphone models too seemed both intuitively convincing and useful also.

I said earlier that I was initially sceptical of VSX, and the reasons for that are twofold. Firstly, how do we know that the VSX room models are accurately representative? They sound convincing, but without access to those specific rooms it’s impossible to be certain. And secondly, as with speaker modelling in room‑optimisation apps, I can’t help but wonder if somehow it’s cheating. Traditionally, mix skills are acquired the hard way; by making mistakes and learning from them. And through those experiences, an understanding develops of how to assemble the components of a good mix — one that both does the music creative justice and will translate to other playback systems, with one’s own room and monitoring. By offering an easy mouse‑click visit to multiple different playback environments, I wonder if VSX is to some extent undermining that learning process. Use VSX and your work will probably translate well, but perhaps you won’t have done all the heavy lifting of understanding how and why...

VSX is really convincing, though, so perhaps I’m just old and grumpy. Maybe I should stop worrying about the philosophy, and just get on with enjoying Steven Slate & Co’s impressive achievement.  

Information

Essentials Edition £325, Platinum Edition £499. Prices include VAT.

SX Pro +44 (0)800 6522 320.

sxpro.co.uk

stevenslateaudio.com

Essentials Edition $299, Platinum Edition $499.

stevenslateaudio.com