In last month’s column (“Sex & Audio”), I mentioned how there are distinct differences between male and female sound preferences when listening over both loudspeakers and headphones. I also mentioned that most people prefer to listen to a headphone/earphone response that isn’t flat, but instead is boosted at bass frequencies and attenuated at high frequencies. Whereas these days it might be argued that the measured differences between good loudspeakers is decreasing (they are all mostly starting to look pretty flat when measured anechoically), the same cannot be said for headphones.
When you consider that some 300 million pairs of headphones/earphones were sold in 2015, it is clear that this is big business, and getting it right is worth a lot of money. But what is “right?” What should the response of a pair of headphones be, particularly if it isn’t going to be “flat?” This interests me for several reasons.
First, when conducting the odd psychoacoustic experiment that I do from time to time, I need to know what the headphone response is so I can consider if it might be influencing the result of the test.
Second, we all have listened from time to time to a signal or some program material on cans, so we can listen in to the detail to see if we need to either reduce the ambient noise or reduce the acoustic influences of the surroundings.
Third, I have found that, sometimes, I prefer the sound of a different set of headphones to my reference ones when listening for fun, such as when on a plane or traveling on a train, etc. These days, this will be via a pair of good-quality, noise-canceling headphones; however, they sound very different from my reference ones and, occasionally, I have to try to mentally compensate for the additional bass and treble that they produce. When commuting or leisure listening (e.g., lying by the pool in some far-flung exotic location; well, if I close my eyes and switch on the noise cancelling, I can imagine I am), I usually rather enjoy the more dynamic sound and the additional bass. The boosted treble can also be useful to improve film or program dialog intelligibility, though on some music tracks I do find it getting a little “screechy” and annoying. (But what the heck; I am listening for fun, so don’t have to take it too seriously.) My only regret is that I wish noise-canceling headphones had been better developed 15 to 20 years ago when my children were small!
The fourth reason for my interest in headphones relates to binaural listening and the current developments in immersive and binaural sound.
I guess my final interest is the need to understand something about the measurement principles and technology so I can understand and “see” what I am listening to. Whereas many of us have spent almost a lifetime measuring sound systems with either a Real Time Analyzer or some rather more sophisticated Time Selective audio analyzer and have a pretty good idea of what the response of what we are hearing will look like (i.e., we have been calibrated over the years), few, if almost any of us, know what the corresponding measurement made on a set of headphones would look like. Luckily, few in the sound installation or design business probably care or have even considered this.
Sadly, a few years ago, I did, and it has been a thorn in my side ever since. Although it is true that I need to know if a headphone might affect a test I am doing, and so this serves as an excuse to legitimize my interest, it now goes far beyond that. For if I am going to listen on headphones, the sound has to be right. Well, it has to be good. As I have already said, there is considerable and varied opinion on what “right” is!
The other niggling thing I find with headphones is the “in your head” image that they produce. Why can’t we hear the sound from out in front or around us as it should be? Well, of course, under appropriate circumstances, we can. It’s all about the shape of our ears and associated Head Related Transfer Functions (HRTFs). Sticking a transducer halfway down the ear canal or covering up your ears under headphones completely screws up these patterns and your brain is not fooled into thinking that there is another world out there.
Binaural recording should and, at times, can fool the ear and a world of spaciousness can open up. However, to date, I have been very disappointed with the binaural recordings that I have heard or indeed made. The problem is, the dummy head’s ears are not my ears, and my brain knows this, as it knows precisely what my ears sound like! Yes, certain binaural cues are well preserved and are effective (I still instantly turn around every time I hear one particular track), but to me, the overall sound is disappointing. I was musing the other day, with 3D printers and laser scanners becoming readily available, about if I could take a scan or impression of my own ears and then 3D print a replica set for the dummy head. That way, any future binaural recordings I made could effectively be via my own ears. I will let you know if this idea comes to fruition, but it’s not the long-term solution. No, that would require a recording to be made in a standardized and agnostic way and then mathematically my HRTFs could be added to the recoded signal via digital filtering. If this technique could be developed commercially, it could have some interesting spinoffs, not least of which would be to enable test subjects to listen to headphone renderings via their own ears or, indeed, via anyone’s ears (I wonder if I would prefer the late Sir George Martin’s or George Massenberg’s ears to my own?).
In the meantime, I will keep measuring what my headphones sound like and see if others’ preferences agree. To put this in context, I have plotted the measurements I made on a couple of units I often listen to. The scales and format probably won’t line up with anything else you are likely to see, but the measurement rig enables me to at least generally compare headphones and correlate this with subjective impression. As you can see, the responses are wildly different. Which one is right? Well, you could say that they both are; it just depends what you like. However, I know which one I do most of my listening on. Now pass me my tequila and a new battery for the noise-canceling units…we will be landing soon.
Editor’s Note: The AES Convention in Paris (cleverly timed to coincide with InfoComm) will be holding several sessions about binaural hearing. Also, the AES is holding a three-day conference dedicated to headphone technology and listing, August 24-26 in Aalborg, Denmark.