Two Ears Means Stereo, Right?
I teach an advanced audio production class at 8:30 am every Monday and Wednesday at CSU Dominguez Hills. The students are currently working on mixing a couple of tunes (one that I recorded and another of there own) in 5.1 surround. Exactly how they’re getting the work done is going to be a challenge for some because the one studio we have that is equipped for surround mixing is overbooked…by a large margin. Some students got in the studio early and got their projects done while others have waited until the last minute and are now struggling to find enough studio hours to finish. In light of this problem, I offered an option. What if they could mix the tracks in 5.1 surround using a set of headphones rather than a complete 5.1 surround mix studio? They liked the idea. Then the questions started…
The first was, “How can a standard set of headphones reproduce the sound of a sound coming from behind you or over your shoulder?” I responded by telling the student, “we all have only two ears, right? And if I were to move over to the corner of the classroom behind your right shoulder and start making some sounds, you’d hear me over your shoulder, right?” He agreed. That’s the answer I told him. Using our two ears we can spatially locate sounds coming from all directions. Reproduced music might be limited to stereo…a left and right speaker with a smooth soundstage between them…but that’s a technical limitation of our recording and playback systems not a constraint caused by having only two ears.
To reproduce a sound over your right shoulder using a standard (but high quality) set of stereo headphones requires some fancy signal analysis and digital processing. Think about it. We know it has to work because we experience immersive surround sounds all day as we go about our daily routines. The trick so to make is work with standard headphones.
The measurement process that I talked about yesterday succeeds because the Smyth Room Realizer (and other systems based on the same operations method) knows what music filtering, delay, and equalization to apply to individual signals to make them mirror the real world location of a speaker found in a 5.1 surround monitoring environment. A sound source coming from the right surround speaker will reach the right ear slightly ahead of left speaker because your left ear is further away from the source. The right ear will also receive a cleaner or unfiltered version of the source sound because it’s not masked by your head as the left ear is. And the timbral balance…the equalization…reaching each ear is unique as well due to the convolutions of our outer ear…the pinnae. So it’s possible to use a set of headphones and some virtual 3D panners to move sound around a 5.1 surround array AND even change the height of each signal.
If you’re curious, you can experience demos at New Audio Technologies and Robert Margouleff’s website.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Paul Horner’s Kickstarter Campaign will conclude in only 38 hours. His amazing record will be available as a standard CD, high-resolution stereo files, and if he reaches his stretch goal of $2000, in full 5.1 headphone surround. Since I wrote about his new initiative yesterday, he’s raised the funding level to over $1700…just $300 short of his stretch goal. If 10 more people pledge just $30, he’ll made the new goal.
I wouldn’t be helping Paul with his project unless I honestly felt a kinship with Paul and his amazing talent. The tunes are witty, engaging, fun, and thoughtful…just like the man himself. I mentioned Robert Margouleff in the post above…he produced the project and I assisted in a number of ways. If you’ve got a few stockings to stuff, this is the perfect item to acquire. It even features a very clever Christmas song titled, “Snowdrop (The Three Legged Reindeer)”…how could you go wrong with a song about a three-legged reindeer? As always, Paul and I appreciate your support. Don’t delay…there’s not much time left!
Hi Mark,
I still don’t understand why simulating the experience of a 5.1 surround system can be more interesting than the experience of being there, at the performance, which can be achieved with 2 microphones substituting our ears with binaural recordings? I can understand how it can be an interesting mixing task for students, but not see a better or more accurate result than binaural recordings, that don’t require mixing or simulating a sounfield, but only capturing it in a way that simulates our ears and hearing, with the head and torso simulators manufactured for that purpose.
Dr. Edgar Choueiri has even developed Pure Stereo, which consists in a filter that can overcome crosstalk with speakers, provided you have appropriate speakers, and that a technician installs the digital pure stereo BACHH filter and calibrates it to your listening room. Although not as practical or affordable as a headphone system, it can also rely on just two microphones and two speakers to produce a three dimensional audio experience.
I believe, from what I have read about binaural technology and implementations, from Jens Blauert, that the main complication of binaural recordings, is that both our pinnae and especially our ear canals, are unique, and factor in to the compensations that our brains perform to tackle those individual differences when interpreting and reconstructing a three dimensional soundfield, reason why not all binaural recordings work equally well for all listeners.
Nevertheless, I have always found binaural to sound more natural and convincing than surround mixes or DSP, applications, although your headphone surround mixes are by far the best I’ve encountered so far.
I would go with Dr. Choueiri’s premise of trying to put the listener in the venue with the performers, or among them, etc., rather than with putting the listener in a room with a surround system, and agreeing with Siegfried Linkwitz, that the latter attempt is rather dissapointing compared to the former, and also less rewarding and affordable.
Cheers!
Camilo, the goal of some engineers is to recreate a live event. That’s not my goal. I find binaural recordings and the work of Dr. Choueiri interesting but unsatisfying to my music sensibility. I prefer a much more intimate sound.
The goal to recreate the live event is doomed from the start because, even if the auditory recreation were perfect, psychological recreation is impossible. It is only *knowing that you are* at a live event, that makes the psychological experience of a live event possible.
The intricate co-synthesis between mind and sensory data means that we cannot separate what we think of the sound from what we know of the circumstances. Hence, even given a perfect sonic reproduction, we will insist that it is not so (not a perfect sonic reproduction), because we *know* we are not having the same experience as when we were at the live event. Communications research has shown that we actually perceive (process) the two events through different ‘channels’, as different realities, with different processing. Doomed.
That is why musicians and engineers need to understand, from communications theory, that a music recording for playback is a unique artwork and a unique creation. They have the opportunity *musically and artistically* to create, specific to the medium through which the participant experiences. Not many have grasped this and worked it as artists, although some no doubt do. Once that happens, then the goal (and it is a possible goal, this time) is for the consumer/listener to recreate in his or her home the sound that was heard in the mastering suite by the engineer (and, hopefully, artist/s).
cheers
True…thanks. I’m not sure why people want things to sound “live”…that’s an unnecessary compromise IMHO.
I completely disagree with this thesis, both the theory behind it and the results. If it were true and the technology worked, binaural recordings played through headphones would be ideal. But they’re not, they are far from it. They don’t work. The sound seems to be coming from inside your head. The reason has been known why it fails since at least the early 1960s and maybe a lot earlier. It is that when your head turns the sound turns with it. This makes the sound from headphones the equivalent of two scalars, not a vector field as an external source of sound would produce at your ears. The theory that the three factors invariably cited for directional detection, time of arrival between your ears, loudness of arrival, and HRTF are at best incomplete or at worst just plain wrong. Ironically Quadraphonic sound and its successors 5.1, 7.1, 9.1, X.Y all fail for the same reason. They do not accurately create the vector fields we hear from real sounds in real environments. Their vectors are all wrong. This is what started me on my own exploration over 40 years ago.
I listened to the clips in your link through headphones and got the same results I always get. I recommend listening with your eyes closed so as not to be distracted by what you see. Seeing could reinforce what you think you want to hear.
I’ve done lots of binaural recordings and don’t find the sound comes from inside my head. The turning of your head is a problem…on addressed by BACCH and Symth Research.