Heard any good 4,200 year old sounds recently?

This is an educational read with some examples off site that you can build.

Binaural is a special type of stereo recording that simulates a human head, often even having fake ears and even a fake head. 

You listen to binaural with headphones because it's not "mixed down" to stereo - which typically uses three mics to *represent* a stereo soundstage on a pair of speakers. More usually today, we have (at least) four speakers as 3.1 with a centre speaker reproducing the centre stage.

Some Background

This section is for information you don't need to read or even understand it but it will make you really boring at parties if you can churn this stuff out. 🤣 

Cinema sound often extends this into 5.1, 7.1, 9.1 and more... if you have the dough and the room to put it in. But it's important to know that cinema sound is mixed from prepared sources and digitised into those 10 or more channels. Beyond that there are now projection systems like Dolby Atomos which includes "height" information so you hear things flying around you.

Dolby Atmos for headphones mixes the available digital channels down into a stereo effect that using something called HRTF - head related transfer functions into (wait for this): binaural stereo.

HRTF, which is also used in speaker based setups where results are usually poor, is based on the idea that your head is a particular shape and that your brain is sensitive to the fractional time delay that it takes for a sound coming from any given direction arrives at each eardrum (tympanic membrane) at a slightly different time and with a slightly different frequency composition.

Clever things brains (but they've been training themselves since before you were even born) and now with a lifetime of seeing (or not) the source of a sound and learning how to react to them our brain will tell us where something is in space relative to us with remarkable accuracy. Even as we age and the systems start to become less flexible (age related hearing loss) we are constantly adjusting how we perceive sound so the ability remains unless something goes badly wrong.

The other effect, which I won't cover here, is called psychoacoustics because hearing doesn't happen in a vacuum (that's a science joke and a bad one for which I should be quite ashamed) visual clues help us determine how a noise-making object is travelling in space. Ever "heard" a plane fly off the screen of your TV or a train come toward you and seem to keep coming even though you don't have rearward speakers? That's your brain using visual and audible clues to produce a soundscape that doesn't actually exist. It's an optical illusion that action directors use to great effect. (See the first Top Gun for an excellent example.)

Enter Binaural

Binaural only works with headphones because it's recorded by a pair of simulated ears. Those things on the side of your noggin are called "pinnae" and have a particular shape which is unique to you. This does have an effect on HRTF binaural sound which the top manufacturers are now trying to address but that is really outside the realm of what a home builder can do.

If we take an artist's view of a human head, we can see (and you can try this for yourself) that your ear canal is at the same level (middle of your head) as your eyes are. Evolution "designed" it that way along with the shape of the pinnae - which extend further upwards than they do down.

So we can plot an imaginary point right in the middle of our head in three dimensions. Front-to-back in this view is just a point. Many animals and birds use a system like this - a bird's ears are little holes covered by feathers because pinnae would cause drag and birds need to be able to fly. But much the same applies and a bird's tiny brain, along with it's side-of-head mounted eyes give it stereo sound and vision not entirely unlike ours. 

Returning to our own head take a closer look at the pinnae. You'll note that they are angled slightly forward. In some people (cough) King Charles (cough) this angle is quite severe often causing embarrassment because they... look like the handles of a jug. The actual angle doesn't matter but you can probably understand that a sound coming from behind hits the the fleshy part of the ear and gets muffled.

Sound from the front is caught by the pinna and directed into the ear canal largely unencumbered.

Now consider how sound reaches from directly above our below - and see how it has to take a different route. Unlike light which tends to move in straight lines, sound is a pressure wave and will move around objects.

Hear It For Yourself

A really creepy way to hear this effect without finding a binaural recording is to find a mono recording and play it through headphones: the sound appears to be coming from inside your head! Early stereo recordings would often mix a singer in perfect mono - separately recorded from the musical track - and listening through headphones the experience was similar and quite uncomfortable. Modern studios seem to have mastered the minute difference in the mix so this is far less apparent and even on headphones.

Putting It All Together aka TL;DR

What you can probably gather from all of this is that music isn't real - and this goes to explain why many popular artists sound dreadful live (even compared to the live *recording*).

But what if we could do this on an affordable home-made rig? Create recordings - made especially for headphones that reply sounds with amazing positional accuracy. Sounds coming from above, below, flying past... even the approximate distance?

I'm not sure if Matthew (@diyperks) would do this as a project but it's actually surprisingly simple. I've designed some modules to help (more of that later) that developed out of the USB-C microphone project. But making a very good approximation is very easy and you don't even need a head to do it.

Some people attach microphones to the side of their eyeglasses but recordings done this way tend to suffer from movement noise and, it probably goes without saying, when you replay these recordings you hear the person move around in physical space which can be disorienting.

The two realistic and realisable options are left. The dummy head (literally a life-sized human head)

Image: EJ Posselius

Or the Jecklin Disk as seen here: Jecklin disk - Wikipedia

Jecklin's disk might look a bit silly but the principal is sound (sorry, no pun intended) and it's not just a better than the dummy head for binaural, Jecklin disk recordings are said to produce a more accurate sound stage for stereo speakers.

Image: RevWillis.com

Cheap to build, these devices make a surprisingly accurate sound stage too and are excellent for outdoor work such as bird song and other wildlife.

There are plenty of plans online since it really IS just a circular baffle. The key to making a great Jecklin device lies in the buffer material.

This is another dip in the physics pond so if that isn't your thing I won't bury the lede and tell you up front, it needs to be soft, lightweight material - speaker baffle material works quite well.

Why?

This is where the physics comes in but I'll skip the fancy math and get down in the weeds. You might already know that it's high frequency sound that our brains use to determine directionality; that's why a sub-bass speaker is usually mono and "just works". Deep sounds like an explosion or an earthquake seem to come from everywhere but as soon as we get into the speech frequencies (above 3-400Hz) sound is increasingly directional.

The crack of a gunshot is easily located because it's mostly a high frequency bang. This is why movie snipers always have a sound AND flare suppressor/silencer. (Real snipers employ guns so powerful that the bullet finds a target long before they heard the sound.)

We can do this because high-frequency sound is more easily absorbed by your our skull and when off to one side, the other ear senses not just a delay but the dampened high-frequency parts. Cloth and layered materials trap high frequency sounds and because the fibres move (you can't see it with the naked eye but they do) that absorbs energy and reduces the volume by the time it reaches the other ear. 

So we now have even more clues - our brains can compare the frequency, the volume, the phase shift and even the time delay between that the signal arrives at each ear to create an amazing 3D audio "image". And we take all of this for granted.

It's worth noting that the "Dead Cats" which cover the "balloon" on high-end hyper-cardioid (highly directional) mics and the tiny you might see attached to small microphones for outdoor are actually largely transparent to high frequency but damped the wind by disturbing its flow in a similar way. 

Jecklin's disk needs this buffering because otherwise those high-frequency sounds will bounce off and hit the mic reducing directionality and causing all sorts of potential nasty phase effects which aren't immediately obvious but your recordings will sound "off." 

If you fancy making one there really are lots of plans online but if there's enough interest I'll hopefully be able to demonstrate a design that (using the same digitiser as Matt employed in the USB C mic, because it's excellent) produces great results and won't cost the earth to build.

There are a few other stereo designs that I've used over the years and some others, including a "boundary effect" microphone that's not just subtle but similarly produces excellent results even in a room with little furniture. Ideal for audio blogging, gaming and and places you don't want to take that beautiful piece of brass Matt's created for us. (I have a rather nosy cat for example and the last thing I need is to have my microphone crushed as he pounces on it!).

If you've managed to make it this far you should now be able to understand why we can't tell where a sound is coming from if it is being made in another room. Sure, we can tell if someone is in the shower because we hear the water running, but without that clue something like a voice might appear to be coming from (say) the living room when the speaker is in a completely different place in your home. 

Remember that if someone shouts to you upstairs from a room downstairs, that pressure wave has to go around multiple corners to get to you - it gets reflected off different surfaces, some soft like carpet, others hard like plaster and brickwork. Much the same applies in dense forest when someone gets lost. Sound bounces around the trees, gets absorbed by undergrowth just like the buffer material on a Jecklin Disc and by the time someone hears it, any chance of direction is distorted.

Keen nature recordists will experience in forests after a rainfall. The timbre of any sound we hear (or that is collected by the microphone) is directly affected by the water on the leaves, in the ground: indeed, even by the humidity. Again, you can experience this for yourself by familiarising your ears to a local park in spring or summer and then going back on a crisp winter morning. The hard ice on everything reflects high frequency sound and therefore, the sounds we hear are different, more defined.

This is why professionally designed auditoriums are invariably very large indeed. It's not just about getting a lot of bums on seats (that's a useful side effect) it's about creating a space where sounds can travel largely unencumbered from the performers to the audience.

The ultimate auditorium in this regard is the outdoors - something that was recently discovered at the famous Stonehenge stone circle as it appears that even those stone-age builders were aware that sound reflected off stones and may have been used to create specific effects. Scientists Map Stonehenge's Soundscape