Click to view our Accessibility Statement or contact us with accessibility-related questions
ralphgonz
50
Nov 6, 2018
I apologize in advance for my characteristic skepticism, but it is based in a math, engineering, and acoustics background... Conventional over-ear headphones require only a fraction of the power handling of loudspeakers, and also eliminate dispersion and room interaction concerns as well as the challenge of controlling enclosure resonances. This allows a single high quality driver to cover the entire audible frequency range, and allows a set of $500 Sennheiser HD-650's to compare to $8,000+ full-range loudspeakers. (Check out the prices of Stereophile's Class A recommended components in these two categories.) In-ear monitors take this even further, reducing power requirements due to the extreme proximity of the driver to the ear canal and further reducing bass cancellation and ambient noise effects. This should translate to further reductions in price, yet manufacturers have managed to convince buyers they need to spend a lot more to get meaningful improvements in sound quality and comfort. The most ludicrous "feature" is multiple drivers: A tiny IEM transducer should be able to operate as an near-ideal piston over the entire audio range, so why shoehorn the complexity, cost, and sonic degradation of a crossover network into the design? It is a much greater engineering feat to make a great sounding pair of loudspeakers for $1000 than it is to make great sounding IEMs for -- say -- $100. IMHO diminishing returns set in quickly above this price point.
phoenixsong
1055
Nov 7, 2018
ralphgonzInteresting perspective- personally I have a preference for single dynamic driver iems as well. However, I wouldn't say it is an entirely fair comparison that you made. Following the line of your argument, shouldn't loudspeakers only require a single woofer and do away with their tweeters?
ralphgonz
50
Nov 7, 2018
phoenixsongGood question -- it would certainly be desirable to use a single driver in loudspeaker design as well, to avoid dealing with the expense and complexity of combining multiple drivers using passive or active crossover networks. Unfortunately this is not possible except in narrow applications... Firstly, producing bass frequencies in a typical listening room requires a driver that can move a lot of air. A high-excursion driver with diameter of 5" or 6" is the bare minimum, and producing deep, loud bass comparable to inexpensive headphones requires a subwoofer class 10" to 15" driver. Secondly, producing wide dispersion treble frequencies requires a low-mass diaphragm with well-controlled breakup modes and diameter of 1" or less. That said, speakers using a single 3" to 4" driver have had a cult following for decades and are suitable in some applications. For example, desktop speakers intended for low-volume nearfield listening in a fixed listening position can forego power handling and dispersion, particularly with bass-light music recordings. And hobbyists often experiment with speakers consisting of arrays of many identical small full-range drivers arranged as a 6-foot or 8-foot line source, with no need for crossover components. This configuration provides good horizontal dispersion and is tolerant of vertical listening position, while the combined surface area provides much greater power handling. This is an expensive design however, with limited commercial appeal.
phoenixsong
1055
Nov 8, 2018
ralphgonzI see, so the extra demands are mainly due to the distance involved
ralphgonz
50
Nov 8, 2018
phoenixsongIn a nutshell. We perceive loudness according to sound intensity level (not sound pressure level). This falls according to the inverse square law with distance, so doubling the distance to the source reduces perceived loudness by a factor of 4. If a IEM diaphragm is 1/2" from your ear drum compared to a speaker that is 128" away, then the speaker requires 65,536 times as much power or 48 dB on the decibel scale.
EdinNJ
271
Dec 15, 2018
ralphgonzNope. Not true. I've got five pairs of CIEMs that I love, they sound amazing. I've got four pairs of IEMs where I had custom silicone tips made. Only one of the single driver IEMs sounds anywhere near the quality of the multi-driver CIEMs, and that's the top end Etymotic. And they're not $100. I've got $150 Etys and they're not even close - and, I'm using the same custom fit silicone tips that Ety recommends on both, so the tips aren't the difference. That said, I've got one set of CIEMs with 10 drivers a side, one of five per side, one of three per side. The top of the line Etys, single driver, are very, very close to the three per side. They're also only about 40% less in cost (including custom tips... and the perfection of fit is a HUGE determinant of in ear sound quality) than the three driver units. Really, it's very much like my speakers. I have five listening spaces (photography studio, editing and printing work area, family room, bedroom, living room) with speakers. I've got a set of very exotic single driver speakers. I've got a couple two driver, one three driver, and one 8 driver speakers. The single driver units are not the best, even though they cost more than the two driver, and three driver speakers. Engineering skill (versus engineering theory) isn't insignificant. Besides what I talked about above, I also have three sets of CIEMs that I never use, because they suck. Live in my desk drawer. All of them were made by people who were legends in the CIEM world. They weren't cheap at all... although one of them I got at a huge kickstarter discount, the other two were sent to me for review. One one driver, one two driver, one three. Really, they're awful. Two drivers and incoherent sound. One driver and incoherent sound. Driver count doesn't define anything other than driver count. Not even coherence. My point is - I have real world experience that says your well argued theory doesn't match the real world. As a lifelong quant - meaning, I've worked in a crazy number of disciplines creating or fixing the analytics - I have a rule of thumb. Any position that posits a linear relationship, particularly based on some concept of physics, is flawed. We can talk quantum effects in molecular biology if you want... that was my last quant gig. but simpler yet... Multiple drivers, with well engineered crossovers and well engineered driver placement, will outperform the simplicity of single drivers. Speakers, IEMs, CIEMs.
ralphgonz
50
Dec 16, 2018
EdinNJThanks for the detailed reply and for the description of your experience with IEMs. I just want to point out that I did not say that single-driver speakers are better than multi-driver speakers. My post was about headphones and IEMs. You can not make blanket statements that apply both to headphone design and to loudspeaker design. In headphones and IEMs a single driver is able to provide suitable listening levels while covering the entire audible frequency band. Real world multi driver systems are much more complex than single driver systems and as you correctly pointed out cannot be modeled as linear systems. Why would manufacturers add this complexity? I believe that in many cases this is used as marketing one-upmanship such as digital camera manufacturers used to do with megapixels. Nonetheless I'm sure there are high-end manufacturers who produce legitimate multiway designs and I'm sure these can sound wonderful. Incidentally without getting into my own academic and professional qualifications you are incorrect in implying that my post was based on theory rather than decades of real world engineering experience.
(Edited)
PRODUCTS YOU MAY LIKE
Trending Posts in Audiophile