Obviously you are basing your argument mainly on coherence. Then tell me, if a single full range dynamic transducer is capable of making good sound, why is there still plenty of multi-ways system out there? If you take ER4S for an example, it has a single transducer designed to be as accurate as possible (and Ety did study and published paper on it), then why we still need any other IEM for the last 19 years? I think the obvious conclusion is, there are those who do not look (at least as a major concern) for coherence nor accuracy in the music, just as there are those who complain ER4S being too cold or lean on bass.
Perhaps in minimizing whatever flaw there are in a multi-ways systems, something else can be achieved. I am not suggesting that multi-ways is the only way to go. What closer to the truth is perhaps when you are looking for the most coherence sound, the designer might be trying to solve other set of problems based on whatever resource (s)he has or is most familiar with. Most importantly, (s)he is trying to develop something that the market can accept, not what theoretically best.
Originally Posted by amnsiac /img/forum/go_quote.gif As for Anaxilus' comment on manufacturers, I'll have to disagree with you a bit there. The case in point is the Audio Technica CK10, for me anyway. How did they get it so right? Multiple BA was hardly a new concept when they came out. The answer is design R&D. Plenty of manufacturers had lobbed a few BAs into a housing before, but AT (a company with huge experience) outdid them without doing anything different, just doing it right.
I never said no manufacturers do R&D to make a proper product. I said it is false to assume ALL manufacturers do. Big nuance there. The CK10's are gorgeous from an engineering standpoint. I could point to others that are far from it.
"but it is assumed the manufacturer has thought about and worked through this."
My comments still hold true about the affect of marketing on product development IMO. There was a post somewhere about how the Multi driver fad took off in Loudspeakers awhile back. The whole more is better thing.
@Marcan
I think there are just different sound signatures for different people. I think you like myself probably prefer the emotionality we get out of our music and others are more analytic about their music. This is why I'm a Dynamic driver kind of guy, those moving armatures are pretty fantastic too.
Originally Posted by ClieOS /img/forum/go_quote.gif Obviously you are basing your argument mainly on coherence. Then tell me, if a single full range dynamic transducer is capable of making good sound, why is there still plenty of multi-ways system out there?
When you need a high level of acoustic pressure (loudspeakers in general or headphones in live situation) or for marketing purpose: people believe that several drivers will be better than one and are ready to pay several times the price.
Originally Posted by Anaxilus /img/forum/go_quote.gif @Marcan
I think there are just different sound signatures for different people. I think you like myself probably prefer the emotionality we get out of our music and others are more analytic about their music. This is why I'm a Dynamic driver kind of guy, those moving armatures are pretty fantastic too.
Being audio engineer (mainly mixing and production), I'm in both camps: Analytic and emotional. When you are mixing/producing, you have to work on the intention of the artist and in music, emotion is a crucial part of the intention. In the other hand, you need to analyze the music and particularly the sound. Actually, while I liked to be in front of big full range monitors in studios, I discover that small speakers with one drive (I use Avantones for small monitors) gave me a coherence that I couldn’t get through my big monitors.
Originally Posted by marcan /img/forum/go_quote.gif Being audio engineer (mainly mixing and production), I'm in both camps: Analytic and emotional. When you are mixing/producing, you have to work on the intention of the artist and in music, emotion is a crucial part of the intention. In the other hand, you need to analyze the music and particularly the sound. Actually, while I liked to be in front of big full range monitors in studios, I discover that small speakers with one drive (I use Avantones for small monitors) gave me a coherence that I couldn’t get through my big monitors.
Feel free to continue buying the single driver IEMs, clearly there is a market for them just as there is for the multi-driver models. But it is absurd to expect all the manufacturers to bow to your whims when people are very clearly pleased with the multi-driver designs. Perhaps one day we'll have a single driver IEM that can complete with the JH13, or perhaps it's already out there but somehow not discovered. In any case, just because there are technical issues that potentially compromise sound does not mean they can not be greatly mitigated. We would still be the stone age if we just gave up every time we hit a hurdle.
Do you also feel we should throw out all class B and AB amplifier designs due to their inherent non-linearity? We should just use battery zapping class A designs in all devices?
In Marcan's defense. I don't think he is attacking technological progress or innovation. I think he is making a simple point about what delivers better SQ in his opinion. I think its a fair point worthy of discussion rather than dismissal.
Has anyone else heard a noticeable weakness in the crossover between drivers? I can't recall listening to a track of music thinking, a single driver would have rendered this better.
This is a fascinating thread. It got me wondering just how much geometrical phase difference there might be in an IEM based on relative differences between tube lengths, not on the cross over phase shift which is likely to happen. Finding the speed of sound is 343000 mm/sec at 20°C, this would indicate a full 360° wave at 20Hz is 17150 mm, at 200 Hz it's 1715 mm, at 2000 Hz it's 171.5, and at 20k Hz, it's 17.15 mm. Assuming tube length variations are ~4mm (just picked a number out of the air), the phase difference between two tubes at 20kHz is a significant 84°. 200 Hz, it's hardly more than 0.8°. A shorter tube imbalance will result in a correspondingly lower degree of phase distortion. It would seem that doubling bass with two drivers has little risk of amplitude distortion when they are mixed together. The higher the frequency, the more critical the alignment. The drivers and cross overs have to be carefully selected to ensure little mixing from the different drivers. A 24db/octave crossover filter possibly?
There is a lot of controversy regarding whether or not individuals can hear phase differences at frequencies above 300 Hz. This controversy stretches back to the 1800s. It seems some can hear phase differences or distortion above these frequencies and others can't--maybe genetics are involved.
Early on in the life of the Ipod Classic (6G), a smooth phase difference of almost 180° was measured for this device across a frequency sweep from 20 to 20kHz, meaning there is quite a bit of source dependence on phase distortion, as well as the inevitable phase alteration caused by crossovers.
I have never read anywhere that phase relationships are maintained during while compressing music into MP3s or AACs. Certainly, phase relations of the harmonics have to be maintained from frame to frame to avoid discontinuities at the frames which would show up as high frequency noise, but I'd sure like to know how phase relationships are maintained and stored during FFT compression of the frames.
If I recall correctly, in one of Wendy Carlos' commentary tracks which are included in her boxed set (or maybe "Secrets of Synthesis"), there is a sample of a waveform she made on her additive synthesizer by adjusting the phase of the sign waves to form two very different waveforms. The amplitudes of the sign wave harmonics were identical, just variations of the phase. To my ear and to hers, there was no difference in the tone.
Anyway, when a driver can't handle the entire frequency range without some drop out at one end or the other, two drivers can make up for the deficiency at the cost of phase coherency. I find it interesting that my three driver UM3X delivers some of the best high frequency ambiance of my IEMs, something which is said to depend heavily on the phase of the drivers. I didn't notice this at first with the UM3Xs--it was something which seemed to burn in.
Anyway, more questions than answers in a debate which stretches back to some of the earliest audio investigations.
Thank you IpodHappy for raising the tone of the debate.
I didn’t know Wendy’s test.
I did a phase test few years ago. I couldn’t distinguish a linear phase shift (same phase shift on all the frequencies), but I could distinguish a frequency dependant phase shift, particularly on voice, piano, bass and drums.
On a constructive note, I thought about it and I came to the conclusion that a solution for the problems induced by crossover might be to conceive drivers that has complementary frequency response. Basically, you wouldn’t need a crossover because the physical limitation in frequency response of each driver would make it useless. For example, a low band driver starting at 20 hz and with a 6 db roll off at 300 hz and a second driver with a roll off of 6 db below 300 hz.
I don’t know if it’s possible, but etymotic is able to change the frequency characteristic of their driver with an electric resistance, so I guess it might be doable.
This might be a solution for loudspeakers, but my bet is that for the same price than a top multi driver headphone, it should be possible to make an even better heaphone with one very good driver. Everything you add between the source and your ear has a sonic and financial price...
Originally Posted by mesasone /img/forum/go_quote.gif Do you also feel we should throw out all class B and AB amplifier designs due to their inherent non-linearity? We should just use battery zapping class A designs in all devices?
My point of view is, crossover is unfortunately a necessary evil (we talked about phase issue but another consequences are distortions, ripples, time modifications) for loudspeakers, so why use it when it’s not necessary for headphones…
Originally Posted by marcan /img/forum/go_quote.gif so why use it when it’s not necessary for headphones…
Because there are people who are not satisfied by any of the single driver IEM in the market?
There are a few single driver design that can get pretty close to multi-drivers setup, yet they are neither priced much cheaper nor recognized as much superior to multi-drivers, which means the market will not adapt them over multi-drivers design any time soon.
Originally Posted by ClieOS /img/forum/go_quote.gif Because there are people who are not satisfied by any of the single driver IEM in the market?
There are a few single driver design that can get pretty close to multi-drivers setup, yet they are neither priced much cheaper nor recognized as much superior to multi-drivers, which means the market will not adapt them over multi-drivers design any time soon.
Indeed, who can tell the future?
But as I said, I think you can make a better IEM with one driver for the same price than a multi-driver. If we talk about the market, that's another story. In my book, market doesn't mean quality, but a good information and an open debate do.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
