First, thank you for your comments on the "Future of Headphones" article. I found them a very stimulating read. It turns out Jude had a particular direction he wanted to go with the panel, and I didn't get to bring up the points I had hoped to. However, your comments haven't gone in vain, Innerfidelity is a pretty good soap box all on its own, and I think they helped me get my mind open and thinking about the issue. Enough so that by the end of the panel discussion I, and a lot of people in the room, had a cool new perspective on the future.
The panel discussion today was moderated by Jude Mansilla of Head-Fi, and panelists included Alex Rossen of Audeze; Fang Bian of HiFiMAN; David Chesky of HD Tracks; Mark Waldrep of AIX Records; Paul Barton of PSB Speakers; and myself. You would have had to be there to experience the dialectic process and how it wove its way to the conclusion, so I'll just lay out some observations and then pose a possible future for high-end headphone listening.
Better Headphone Measurements
Paul Barton worked closely with Canada's National Research Council (NRC) in developing improved speaker performance standards. Barton rightly observes, in my opinion, that variability in the sound of speakers 20 years ago was far greater than it is today. He credits some of today's more consistently pleasing and similar sounding speakers to research done at the NRC using blinded listening tests. Subjects were asked to characterize the sound of speakers using lists of adjectives, and a robust relationship was found between various subjective impressions and the speaker's objectively measurable correlates. They also found people's sonic desires varied much less than one might expect. The NRC formulated an objectively definable set of speaker performance goals, and it worked. People preferred and purchased more speakers designed on these new standards, so as manufacturers trended toward the new standards, speakers began to sound more similar...and better in general.
The problem is, headphone measurements are notoriously difficult. Studies like the one above done with speakers aren't very practical without improved headphone measurements to correlate with subjective responses. PSB has thrown its hat into the headphone ring, and Barton is now collaborating with the NRC in an effort develop a pleasing headphone standard. The great news is that NRC researchers have a long history developing artificial ear and ear canal simulators, and have recently developed an improved acoustic coupler. Barton claims existing headphone measurement methods are only meaningful to 4kHz, while the new coupler will deliver reliable information to about 14kHz. This is a hearty step toward an objective standard and better sounding headphones.
Digital Ear Impressions for Custom In-Ear Monitors
At the JH Audio booth on Friday, ear impressions were being taken with a Lantos Technologies 3D Digital Ear Scanner. This system performs a scan of the ear and canal with lasers in about 60 seconds. Data delivered is better than traditional ear impressions because the subject can perform jaw movements, and the dynamic characteristics of the customers ears can be captured. This will allow custom in-ear monitor makers to improve the reliability of fit with custom IEMs. The digital file from the Lantos system can be instantly transmitted to the custom headphone makers.
Headphone Acoustic Virtualization is Getting Better
Recent high quality binaural recordings from Chesky are compatible with techniques developed by Professor Edgar Choueiri of the 3-D Audio and Applied Acoustics (3D3A) Laboratory at Princeton University that permit very convincing imaging to be achieved on speakers. This technology demonstrates that binaural to speaker transfer functions are within reach.
Conversely, the Smyth Research Realiser comes with small microphones to measure one's ear canal response. From the measured response, a custom head related transfer function (HRTF) is created to synthesize virtual speakers when listening to headphones. It, too, is very convincing.
Fang Bian of HiFiMAN said he has been talking with Smyth Research to find a way to include acoustic virtualization in future high-end player products.
The Parrot Zik headphone has acoustic room synthesis DSP built into the headphones, and has enough processing power to deliver surprisingly satisfying acoustic spacial simulation.
Future High-End Headphone Purchases May Start at Your Audiologist
Shake all of the above in a tin can for a bit, and you might catch a glimpse of a future where the purchase of a high-end headphone starts at your audiologist. For custom IEMs a physical impression would be made of the ear, and the scanned data sent to the earphone maker of your choice. An acoustic impression also might be made of the ear using small microphones and a calibrated sound source. The acoustic model might be a file you load into an iPhone app, headphone amp, or even the headphones themselves.
In five or ten years, headphone enthusiasts might be walking around with seriously good custom virtual listening rooms on our head...and I really didn't think that probable until today.
The future looks bright for headphone enthusiasts!
) With the custom IEM market reaching maturity and the enthusiasm Waldrep and Fang expressed towards the Smyth products there's a strong incentive to buy the realizer, not only for HT/videogames, but the new generation of recordings. Chesky's idea of bringing DF eq to speaker reproduction of binaural material also sounded interesting. Although I think his idea behind a pinna-measuring app using smartphone cameras was a little too ambitious.
