DSPeaker HeaDSPeaker Headphone Virtual Surround System
Most surround sound receivers have a headphone jack. When using that jack, most of them apply some type of faux-surround sound processing. Dolby has their Virtual Headphone mode and Audyssey calls theirs Personal Surround. Interestingly, I don't see those being used very often. As I browse receiver models from Denon, Onkyo, Marantz, Sony, etc, I mostly see proprietary solutions, likely chosen to save money on licensing fees. Regardless of which method is used, I don't find any of them to be all that convincing. They all aim to expand the presentation to sound more speaker-like, but just end up sounding like a mess of reverb. Most of the time I'd rather deactivate all processing and watch a movie in plain old stereo.
PC gamers are better off for multiple reasons. For starters, a sound card is more likely to have dedicated DSP-based hardware - Dolby Virtual Headphone shows up a lot (in the Asus cards for example), and there are others like Creative's CMSS-3D which are well regarded. Both of these sound more convincing to me as compared to the typical Denon and Sony solutions. Creative was also heavily involved with the OpenAL API which was used in several dozen games across multiple platforms. I think as PC and console gaming continues to become more advanced, we'll see further implementations of this type of thing - some good and some very bad.
Over the years, there have been some rather significant attempts at simulating surround through headphones via dedicated hardware boxes. Astro makes various models aimed towards gamers, many of which include dedicated headphones with a mic. These are not intended for "serious" headphone enthusiasts and are priced and marketed accordingly. JVC had the DH1 and a few variations, but those were incredibly rare and have been discontinued for a while. Years ago, AKG had their Hearo 999 which was favorably reviewed by Jude at HeadFi. It went for $1200 and used a proprietary wireless headphone, though regular headphones could be plugged in as well.
These days, if one desires a high quality hardware-based system for simulated headphone surround sound across a wide range of usage types, there appears to be three options currently in production. The first is also the oldest - the Headzone Home from beyerdynamic. With an MSRP approaching $2000 the Headzone is essentially the beyerdynamic A1 headphone amp with an added DSP for surround simulation (and optional head tracking), plus a bundled DT880 headphone. I've heard the Headzone on several occasions and it was generally effective but never really amazed meand for that price I expect to be blown away. Part of the issue is the amplifier section based on the A1 which in my opinion is a spectacularly overpriced piece of hardware. But if this thing sold for a much lower price I'd have no problem recommending it.
The next option, the best option if one can afford it, is the Smyth Realiser. The Realiser has been around for several years garnishing rave reviews and leaving rabid fans in its wake. Rather than merely simulating surround sound in general, it actually imitates a specific speaker setup - be it stereo or a 7.1 system. This can be any setup the user can get their hands on to measure, which opens up a world of possibilities. By all accounts the Realiser is supremely impressive but the price is somewhat prohibitive$2910 for the base system or $3760 with the recommended Stax earspeaker option.
The third option is the focus of this reviewthe DSPeaker HeaDSPeaker Home (approximately $500). This unit is similar to the Headzone in terms of basic layoutit's got a box that connects to your source and handles all processing, along with a sensor which mounts on top of your headphones and tracks movement. But there are some differences, the core of which is the processing engine used to create the effect.
DSPeaker is very closely associated with VLSI Solutions. VLSI designs and manufactures integrated circuits, some (but not all) of which end up in audio gear. They specify a broad spectrum of devices which use their chips, from baby monitors to HiFi applications and everything in between. DSPeaker is a spin-off company made up of VLSI employees and using VLSI components for audiophile related gear. An early product of theirs was the Anti-Mode 8033 favorably reviewed HERE and their latest, the Anti-Mode 2.0 Dual Core has received quite a bit of industry praise. Both of those models focus on room correction, which isn't all that different from surround simulation when you think about it - applying DSP to a signal to make it fit your specific application, thereby (hopefully) improving the listening experience.
The DSP unit is a very compact little box. It has a sort of "lip" on the front section that, when combined with the optional spring loaded mechanism, helps it balance on top of your display. If I was using this permanently I'd want to use something like the 3M Command strips to secure the DSP box without any permanent residue.
The box has most of the important bits on the rear panel: USB, Toslink in and out, headphone jack, sensor connector, and a 3.5mm line-in. It draws power over USB, so the included AC adapter must connect to the USB port if Toslink is used as a source. Note the absence of any RCA outputsexternal headphone amps need not apply. The front panel is a simple affair with a signal emitter for the head tracker and three sets of LED status indicators. Installation is easy; for computer use, just set the box on the monitor, connect via USB, install the included software, and that's it. For living room use I simply connected via optical out from my Blu-ray player. My player will output audio concurrently over HDMI and optical, and I think most others behave similarly, so your main speaker-based surround system can remain undisturbed. There's a Toslink output for passing audio through the box to a surround receiver but that strikes me as a very "pre-HDMI" type feature, and doesn't really seem necessary these days.
Inside the box, the heart of the unit is a VLSI VS2000a. Aside from handling D/A conversion, this versatile chip manages incoming USB signals, takes care of DSP duties, and even has a built in headphone driver (which I don't believe is being used in this particular application). There's also a multi-purpose Micronas chip on board which handles the Toslink input, provides surround sound decoding, and probably some other functions that I'm missing. I also see an Altera programmable logic device, and some other odds and ends too. In short, there's lots of processing power on board. The actual integrated headphone amp is opamp based and sports a .1 ohm output impedance. DSPeaker recommends using headphones rated at 250 ohms or less in order to ensure satisfying volume levels, but that also depends on sensitivity. In my experience 300 ohm Sennheisers worked just fine.
Aside from the Box, the other main component is the sensor used for head tracking. It attaches to the top of most headphonesanything with a solid frame or pad should be fine, but some AKG and Audio Technica Models could have trouble due to their split designs. Rather unfortunately, a hard-wired connector is required, running from the head tracker to the DSP box. This becomes unwieldy when combined with the headphone cable, which of course also must connect to the DSP box. And remember, the DSP box mounts on top of the display, so you'll want the cables routed on the side. I had to use an extension cable for most all of my headphones, in order to reach my couch. A wireless solution would be ideal, but I imagine it would increase both price and complexity of the system.
The other important bit is the credit card sized remote. Since the DSP box has no buttons, the remote is essential for operation. It handles volume adjustment and input switching as well as setup functions. Not the prettiest remote you'll ever see but it gets the job done. It even allows quick selection of separate user profiles for two different users, which I thought was a nice touch.
The heaDSPeaker system actually consists of two different processes: head tracking and HRTF-based surround simulation. Both work together to make the result more convincing. It is possible to use the system without head tracking, and it sounds decent in a pinch, but ultimately the head tracking aspect makes it that much more lifelike.
HRTF is tricky to simulate because every set of ears is physically different. DSPeaker tackles that problem by providing multiple ear/head models. 5 default options are available straight from the remote control, and a 6th button is programmable via software. The user can choose from a pool of 45 different options, which means there should be something for everyone.
VLSI Solution has a patent for their head tracking process. The entire patent application is somewhat confusing (as those tend to be) and even the abstract is a bit convoluted. Essentially, it boils down to the main box communicating with the head tracker via ultrasonic frequencies. The tracker has 2 receivers; one on each side. As the head tracker rotates from left to right, the sensors each receive slightly different signals based on their different locations. The system then calculates phase differences between each signal and thus determines how to process the sound. Sounds simple enough but it's really not. DSPeaker talks more about their process here. It's pretty impressive how the head trackera rather simple little thing in appearancecan be used to calculate spatial locations in such an accurate fashion. But then again, the human body does amazing things with just a pair of eyes or a pair of ears, and routinely localizes sights and sounds in a three dimensional landscape. Mother Nature's been doing this stuff for a long time.