Spatial Hearing, Surround Sound & A Lot Of Voodoo

What Is Spatial Hearing?

A true reproduction of the frequency spectrum is only part of what we're looking for. We also need to assess how well individual sound sources are resolved, and how easily they may be spatially located. This, in turn, represents the playback's precision, in which spatial hearing plays an important role.

The human body has two ears. Between them, the head serves as an acoustic barrier. But how do we actually perceive spatial information relayed through acoustic signals, and what enables us to attribute acoustic events to a specific source in space? This is based on two factors: the respective time differences (meaning when a specific sound reaches both ears) and the difference in intensity (that is, the difference in sound pressure level).

Usable information about the spatial position of a sound source due to differences in intensity and delay time can only be recognized and processed by the ears and brain if the sound has distinct characteristics (sudden occurrence, spectrum, pressure level, etc.). For example, it is almost impossible to spatially separate sources of background noise, such as in the woods or in a city, when the observer is surrounded by a large number of similar sources. Faster and higher-contrast changes in tonal characteristics make it easier to localize the distinct source of a sound.

Transmission Delay Differences

Transmission delay refers to the time difference between the moment of a sound event and the time needed for that noise's sound wave to reach both ears. If the source is not directly in front of the observer (with a deviation of at least 3° or more), logically, the sound reaches the ear closer to the source before reaching the other one. The resulting time delay depends on the difference in distance the sound wave needs to travel to reach the respective ear. Our ears are capable of detecting time differences as small as 10 to 30 microseconds!

Difference In Intensity

A difference in intensity (or loudness) occurs when the wavelength of a sound wave compared to the size of the head is small enough to allow for reflections to occur, with the wave breaking at the head as an obstacle. As the picture shows, this scenario creates a so-called sound shadow on the opposite side. However, it occurs only for frequencies above approximately 2 kHz and grows with rising frequency. For the longer wavelengths of lower-pitched sounds, the head is not really an obstacle.

Localization Of Acoustic Events

If an acoustic event occurs outside the head (for example, generated by loudspeakers), we are confronted with a process called localization. Interpreting data provided by the ears enables our brains to locate the event's origin spatially.

Incidentally, the head is always in motion to provide more precise spatial localization, since turning, raising, lowering, and tilting facilitates localization across all three planes (X, Y, and Z). In this case, and only in this case, the result would be true three-dimensional sound. But this cannot be achieved using a normal speaker setup, which positions all sound sources at more or less the same height.

Special Properties Of Headphones

When you put headphones over your ears, perception of the stimulus always occurs right at the head! And when sound waves produced by the headphones are in sync, the sound's source is always perceived to be in the middle of the head.

Lateralisation is the sound source's apparent movement from the middle of the head to one side. This headphone-specific perception of a supposed sound source "wandering" is either due to a time difference (signals are played back with an offset in time) or intensity difference.

Ears are almost never completely identical. So, with a different sensitivity for both ears, a lateralisation towards the better ear may also occur without a difference in stimulus! This is why an exact adjustment of sound balance is always the first step in optimizing the listening experience.

Surround Sound With Headphones

However, lateralisation is much more! The phase shift, or the waveform's offset as soon as it reaches the ear, may also result in a perceived change in alleged position of the sound source. This is because the auricle itself is of great importance for locating the source of acoustic events; it simultaneously serves as both a sound receptor and filter by linearly distorting impacting sound waves in different ways, depending on the direction of and distance from the source.

Each person's ear cups are unique. Hence, everybody perceives sound a little differently. The shape of the auricle alone influences how a sound wave bounces off the exterior, how it enters the ear canal, and how it is transported to the eardrum. Even hair on the auricle plays a role in this process!

What does all of this have to do with headphones? True three-dimensional hearing requires movement of the head along all axes, which isn't technically possible with tightly seated binaural headphones.

Spatial representation of a sound source's actual position along the Z-axis, either in front or behind the head, or on the Y-axis, below or above the head, are impossible to convey with just one-dimensional headphones!

Even if the software works with phase shifting, manipulates the interaural time difference and intensity, and modifies the frequency spectrum so that sounds behind the head are a little deeper or duller, it simply cannot create a real spatial sound experience. After all, it's using experiences collected from real life sound events to compensate, and thus, from time to time, may be open to manipulation.

What Is The Benefit Of Several Drivers Per Earpiece?

Headphones with several drivers installed at different angles can help with two-dimensional representation because, depending on the auditory sensation and hearing experience, sound that reaches the auricle at different angles can give rise to something like a spatial sound sensation.

The downside of such systems, however, is the uncertainty inherent to multi-source sound creation. The various drivers may coincidentally influence each other in disadvantageous ways, whether that's phase shifting or sound cancellation, due to their close proximity to each other. Enjoying music is not really possible with such a configuration. Even linear reproduction across a wide spectrum is hard to achieve.

Nevertheless, at least some viable multi-driver headsets have appeared on the market, some of which convey a quite convincing illusion. Even in these cases, though, this sensation is interpreted subjectively and can never be transferred from one person to another.

5.1- or 7.1-channel sound delivered through headphones is always imaginary, and can only be achieved by the brain relying on previous experience. Even then, the result will only be two-dimensional at best. Not even proper loudspeakers are able to convey real 3D; they're unable to reproduce more than the X- and Z-axis.

So What Gives Us The Best Performance?

Personally, I prefer a good set of stereo headphones with detail-rich sound reproduction. Not only the reproducible frequency range and its linear character play an important role, but also the system's ability to accurately reproduce several acoustic events overlapping or occurring simultaneously.

The separation of individual sources and their precise spatial location within a big acoustic picture is often referred to, in typical hi-fi jargon, as the so-called sound stage. Its width describes the good spatial reproduction capability of headphones and their respective audio resolution. Without it, the result sounds murky and undifferentiated. As an example, if sounds and noises mix down into an indistinguishable acoustic mess, spatial representation collapses like a house of cards.

Is The Difference Apparent To Everyone?

The answer can be yes and no. We repeated all of our tests several times using recorded surround material with a total of six test subjects (three male and three female) aged between 16 and 50 years. They auditioned different headsets in a randomized order. Only two of our test subjects were able to spatially locate sound sources correctly using the "real" surround headphones. Three people got it right in at least some of the test cases, while one person could only guess. Using the virtual surround headsets with one driver per earpiece, only two candidates reported to have noticed anything at all. But nobody managed to get a 100% score.

More complex and louder ambient noise increased our margin of error. Furthermore, during our blind tests, none of the test candidates were able to tell whether or not the headset used was equipped with just one or up to three drivers (+bass) per earpiece. Interestingly, two people even claimed to have perceived surround sound while using just the stereo reference headphones. This, in turn, goes to show how all of this is actually the human brain at work. So, we can't definitively answer the original question with a yes or no.

Beyond their supposed claims of surround sound or the marketing-friendly Dolby certificates they get pinned with, headphones should first of all have some basic qualities like good sound resolution for individual acoustic events across the widest possible frequency spectrum, linear sound reproduction across said spectrum, inconspicuous transient response, and high sound level stability. With these, everything else falls into place.

Very good 5.1-channel systems with multiple drivers per earpiece may be able to provide the illusion of spatial representation, with some additional help provided by the human brain, but in turn they usually suffer from a lack of proper sound resolution and clean reproduction of complex scenarios with many overlapping sound sources.


MORE: Best Deals

  • the nerd 389
    In the article, you say that headphones always create the perception that the source is right at the head. This is incorrect. Newer techniques use either binaural recordings or HRTF approximations to reproduce entire auditory spaces.

    When employed correctly, these techniques allow for a more accurate reproduction of an environment than is possible with speakers.

    If surround headphones employ those techniques, the result is true 3D sound (two angles, as well as distance). That said, the technology needed to pull this off in real time is still a few years away. For now, binaural recordings are the only way to achieve this effect in practice.

    Finally, you should probably disclose the output impedance of your sources. Many headphones have drastically different frequency and phase response between low impedance and high impedance sources. The effect is most pronounced with low impedance headphones, like most gaming headsets (<64 ohms). Motherboards and sound cards usually have a high output impedance (10-100 ohms), while decent headphones amps have low impedance (<2 ohms).
    Reply
  • Spanky Deluxe
    I just don’t get Tom’s Hardware sometimes. They go out of their way to make top quality articles with great testing methodologies but at the same time make the website absolute garbage to use with half the stuff covered in abnoxious adverts, a terrible UI experience and then even more adverts for even worse stuff that is blatantly obviously scam sites like “How local <ISP LOCATION AREA> housewife earns $15962 per month from home” and the like. Sometimes adverts even stop you reading an article on mobile, i.e. actually block you pressing the next page button. I hardly ever read articles anymore here despite being a visitor for about 20 years as it’s just too much hard work to wade through the website design and adverts everywhere.
    Reply
  • icycool_q1
    Completely agree Spanky, I am coming here less and less because of the huge video playing throughout almost every article. In addition to that, I have very poor internet and it takes considerable time for this crap to even load. Hell, I'd be happy to pay them $5/year just to have NONE of this extraneous rubbish.
    Reply
  • Co BIY
    Tom's reviews headsets?

    No category under peripherals ? and I don't remember a recent review.
    Reply
  • FormatC
    This is the intro review to understand, what we test and how. The headset reviews will follow soon. I'm testing headsets for the German site since a few years (my latest). I think, it is a good compromise for the masses to get more info and to understand, what we write. Not too flat like an unboxing video on YT and and not too in-dept like for the golden ears. ;)

    Finally, you should probably disclose the output impedance of your sources.
    If measure with an good external amp, not on an onboard output. But with exception of a few pilot headsets I wasn't able to find a gaming headset above 32 ohms.

    Newer techniques use either binaural recordings or HRTF approximations to reproduce entire auditory spaces.
    As I wrote in the review, all this techniques are using your brain (and experience) to produce this immersion. But it doesn't help to make a bad headset better or to hear "more". ;)
    Reply
  • hannibal
    I agree the article. I have always have good Hi-Fi speakers and headphones. Many years back I did Also buy Gaming headset (expensive one) and used it a while and was deeply dissapointed by the sound quality. Did go back to old music Hi-Fi headphones and was much more pleased and the next time I did upgrade my music system I did by the best Hi-Fi music headphones that I could afford at that moment and have been really pleased the sound quality.
    Good bass even in low volume, good balance at low volume, same when you increase the volume upp. So balanced and straight linear sound in all situations!
    The multidriver system is interesting, but very difficult to make right. Have to check out some uber models From Senheiser and othe big audio manufacturers to see how well They can do it with their best systems.
    Reply
  • the nerd 389
    @FormatC

    I didn't mean to imply that you should measure with a low impedance source. Quite the contrary, actually. Most people don't have headphone amps, so it doesn't make sense to measure with one if that's your audience. All I meant was that the output impedance has a substantial impact on how headphones measure, so it's worth disclosing. It only takes a few minutes to measure if you have a decent multimeter and a couple of resistors lying around.

    Your Creative unit should have an impedance between 10 and 50 ohms. That's a big enough range that it's a pretty notable amount of uncertainty in any measurements you do with it, though. It's worth mentioning that of the units I've measured, I've found almost no difference in output impedance between onboard audio and gaming-oriented sound cards. Even the external ones usually have at least 10 ohms of impedance.

    Regarding the binaural and HRTF techniques, it's not really fair to say that they use the brain/experience to produce the effect. They simulate a physical phenomenon that occurs when sound hits the outer ear and head. You can even measure the effect (assuming you can sit still long enough).

    The wording you chose is akin to saying "the lights in your room allow you to see using your brain/experience to allow you to perceive objects." Again, it's not technically incorrect, but it's not really going to help anyone understand it any better. Also, I wouldn't mention it except for the fact that the HRTF is likely going to become a major factor in mainstream audio in the next few years. It's already used in some high-end audio solutions, and it's expected to be a major ingredient in the next phase of VR.
    Reply
  • FormatC
    I'm building audio systems since 1979 and I followed in the past each trend to understand (or very often not), what the industry and their PR tried to sell us. I'm visiting a lot of meetings and workshops and it is every time very interesting to hear, what scientists, engineers and medicals said. Alone the inner structure of a human ear is so complex, that we can found thousands of variances of a different spatial immersion. For one person this fakes are working well, for others not. Or simply different. I also saw a lot of blind tests and the results were every time within the statistic. A good 50:50 for stereo vs. artificial sound manipulations.

    What I would say with this small story is:
    Never believe, what the industry said. This is (mostly) pure PR and the utilization of the term Gaming, only to sell their low-end more expensive, is pure nonsense. Stickers and audio labels are nonsense too. Money makers. Good audio hardware can cost a lot of money but our job must it be in the first row, to find the the better pieces between all this crap. What we will do is a thing in the middle of this mostly senseless unboxing YT videos and the Hi-Fi magazines. Call it science for the masses. But we have in each case to take care, that we will stay understandable for all readers. Not so easy... ;)

    BTW: The Creative amp is running between 10 ohms and 600 ohms not bad. But I'm trying each headset also onboard (if it is not USB) to see, how it performs. I'm worked together with MSI for example to improve their mainboard audio solutions and a lot of other companies have now a bigger focus at the audio part of their mainboard design (components, positioning). This was also a follow of my investigations of all this VGA-related influences ("you can hear what you see") and the mainboard layout. The fact, that you can see now on a lot of VGA cards low pass filters for all rails is a direct follow of this work. Together with improved PSUs... ;)
    Reply
  • JonDol
    20524062 said:
    ... it’s just too much hard work to wade through the website design and adverts everywhere.

    There are already long years since AdBlockers became mandatory on THW in order to have a pleasant reading experience. Don't wait anymore before installing a few (not only one since some of them have agreements with ad networks and sneakily whitelist some ads to let them reach you). I personally have at least 3 AdBlockers on every web browser I use.
    Reply
  • Virtual_Singularity
    20524205 said:
    Completely agree Spanky, I am coming here less and less because of the huge video playing throughout almost every article. In addition to that, I have very poor internet and it takes considerable time for this crap to even load. Hell, I'd be happy to pay them $5/year just to have NONE of this extraneous rubbish.

    Hmmm... Totally, agree. Save for the $5 a year part. =P

    Edit: Good article all the same, as it's an area Tom's hasn't breached much, at least yet, as far as I've seen (or heard). Like the contributor with the outstanding DL/UL speed in the sig hints at: "not too flat like an unboxing video, and not too in depth like for the golden ears". I like that description. Though I personally wouldn't mind a "golden ears" type of review. But then, that'd likely remove or disqualify most gaming headsets from the topic altogether, along with readers? Though just so Tom's knows, some of us would indeed take the time to read the in-depth, "golden ears" stuff. I grew up in the analog world and usually still prefer it, maybe. Ha.
    Reply