7.1 vs brain

[Oktyabr]

Is there a good example of a game that does this well?

Unmodded:

Modded (JSRS3 Dragonfyre):

Modded (Speed of Sound):

[Oktyabr]

Each weapon round fired in ARMA 3 has three, distinct (and important to learn!) sound FX that are based on real speed of sound calculations and locations of all parties involved:  the sound of the weapon being fired, the sound of the round passing by you, and the sound of the round hitting something.  You can hear any of those in different sequences (and with different timing), with different locations, depending on your location relative to the weapon and the direction it is firing in.  And, of course, if you get shot in the head you might not hear anything at all

I can tell, with STEREO headphones, where someone is, how far away they are, which direction they might be moving, and even what type of weapon is being used simply by the sound.  Same with vehicles.  You can roughly tell how far away an explosion is by comparing the time difference between when you see the explosion (moving at the speed of light) and finally when you hear it (moving much slower, at the speed of sound).

Last but not least ARMA3 features "free look", meaning your head can move independently of the rest of your body (or your weapon), and it also has positional audio (HRTF) so it's possible to whisper to someone right in front of you or yell at someone across the street and your voice will sound like where it came from, with roughly the appropriate volume adjustment as well (not perfect yet due to variances in mic levels, but quite usable).

[Grrizz]

That might be a problem, using binaural recording for game sounds. They must have each and every recording for when you turn your head to the left 5 degrees, 10 degrees, 15 degrees, to the right, tilt up 5 degrees, etc. A simple FPS might have to take tenths of gigabytes for sound data only.

It wont be a case of recording the sound binaurally it will be a case of calculating the wave phase and deformation. Simulated surround does a good job of it (using pitch and phase primarily) but technically binaural would be slightly better, so slightly it hasn't been bothered with though due to the drastic amount of extra effort and processing. In fact now we have a good processing overhead I believe there are a couple of middleware companies doing it it just hasn't been "officially" adopted into any games yet AFAIK.