[kiteki]

 
Quote:

I've never tried anything higher than dual driver, but I do know this. The crossovers in IEMs are R/C networks - the most simple form of crossover circuit possible.
 
And with a simple crossover like that, you get all kinds of phase shift, and harmonics and so on and on.
So IMHO, I think for simple IEM's with the most consise sound, I would prefer a two way, (single crossover) that way, there's less to go wrong.
 
The future for complex multidriver IEM's is with external digital crossovers.

There's different types of crossovers in custom IEM's I believe, for example the TS842 uses "a hard crossover of the 2nd order with 12dB step increments for a more natural sound", whatever that means.
 
As far as disharmonics and phase shift goes, I haven't been listening for that in particular with multi-BA stuff, but definitely haven't picked on any of it either.
 

 
Quote:

I thought the most detailed IEMs were ER4, 2X-s and UERM?

Not hardly!
 
The weakness of the ER-4 is it's a single driver unit so the imaging is far too concentrated to be able to layer and dissect the details, and the 2-XS (well, K2 SP, exact same unit...) is dull, bright, lacking high-end extension and just plain jane boring.
 
Not sure about the UERM.
 
 

 

[Pianist]

Hey guys. Thank you very much for all your input. There are some really interesting thoughts. I don't have time to read it all right now. Will do that tomorrow. In the meantime, keep it comin'!

 

[alphaman]

Quote:

Not hardly!
 
The weakness of the ER-4 is it's a single driver unit so the imaging is far too concentrated to be able to layer and dissect the details...

...or it's mostly due to that old Knowles BA driver ER still uses ... but newer ER's have the same charc. sound (dunno how diff. their BA drivers are from the 4S orig.??). Disagree about the single-driver vs. imaging comment -- imaging should theoretically be superior from a linear point-source -- single or multi. True, v. carefully aligning/tuning multi BA (or dynamic) drivers can achieve this. But, given the totality of effects from myriad factors -- x-over, physical driver alignment, etc. -- makes this a complex issue!
Measurement science for ear-/headphones is still in the dark ages comparatively speaking. When we start seeing measurements like this, we can better argue the multi- vs. single issue:

 

 

[kiteki]

 
Alphaman I'm not sure if I understood your post, I just mean that the imaging of the CK10 is more expansive and accurate than the ER-4S, so my guess was that it's due to 2 drivers versus 1.

 

[nikp]

More drivers = more accurate audio reproduction. 
 
1 Driver = multitasking (Highs, mids, lows all mashed up into one)
 
3 Drivers (3 Way Crossover) = 1 for the high freq, 1 for the mids and 1 for the bass = Much more cleaner sound is being reproduced.
 
Think of violinists. 
 
A single violinist playing a tritone. 
 
3 violinists playing an individual tone to make up a tritone.
 
All the violinists in an orchestra playing a tritone.
 
Did you notice a difference between them? Need I say more?
 
 

 

[kiteki]

 
Thanks for the analogy nikp, very well put.
 
 

 

[nikp]

Quote:

 
Thanks for the analogy nikp, very well put.
 
 

 

 

[Xymordos]

Not sure, but from what I've read, the UERM is a detail beast. Many people consider it to be UE's flagship instead of the UE18s

 

[alphaman]

 
Quote:

More drivers = more accurate audio reproduction. 
 
1 Driver = multitasking (Highs, mids, lows all mashed up into one)
 
3 Drivers (3 Way Crossover) = 1 for the high freq, 1 for the mids and 1 for the bass = Much more cleaner sound is being reproduced.
 
Think of violinists. 
 
A single violinist playing a tritone. 
 
3 violinists playing an individual tone to make up a tritone.
 
All the violinists in an orchestra playing a tritone.
 
Did you notice a difference between them? Need I say more?
 

How about ... three poorly-chosen (and/or mediocre-performance) drivers vs. one well-chosen (and/or high-performance) driver? In some "purist" audio circles, minimizing (or ditching) x-over for loudspeakers is chosen deliberately. Also ... most full-sized cans (moving coil or ES) are full-range.
 
Spectrum-specific BA drivers have been around for much longer than high-performance IEMs (they were orig. designed for hearing aids; not purpose-designed for high-fidelity audio/music [methinks BA IEMs generally have a "cupped" or "midsy" or "nasaly" sound  'cause this is where all-important human-voice-speech occurs ... what you want from a hearing aid]). And they have been adapted or tweaked for IEM (= hi-fi audio/music) apps. Hence, they might not be ultimately be ideal for IEM design -- they just happen to "pre-exist" and are relatively easy to incorporate and cost-effective.
 
Not saying multi-driver approach is necess. wrong ... just presenting counterpoint.
 
The violin(ists) analogy is confusing ... for one, the earphone driver is an electro-mechanical transducer with v. important physical diffs. And it's not a musical instrument.
 
 

 

[kiteki]

 
It doesn't matter if they were once designed for hearing aids when the current high-end custom IEM's can present an overall better musical experience than the Ultrasone Edition 8.

 

[alphaman]

Quote:

Alphaman I'm not sure if I understood your post, I just mean that the imaging of the CK10 is more expansive and accurate than the ER-4S, so my guess was that it's due to 2 drivers versus 1.

You didn't mention the CK10, so I'm confused ???
 
IAC ... I think you're connecting too many dots.
 
To get to the bottom of the issue, you'll need to know/identify the actual BA drivers used in the IEM you want to evaluate. The specs sheet of the BA driver will give you their curve (FR or range). The graph I posted above is of a three-way loudspeaker and clearly shows three curves. You can follow the link and see the combined FR for that speaker.
 
The few cans that are tested via the dummy head/ear (by sonove, Head Room, Inner Fidelity, etc. An all-too-rare practice for cans that definitely needs to change!! ) only show combined FR graphs. TTBOMK, no ear-/headphone tester presents multi-curve graph like the one for loudspeakers. Maybe the DIY IEM discussion thread (elsewhere at H-F) will come to the rescue. Otherwise, were debating in a vacuum.
 

 

[kiteki]

 
How am I connecting too many dots?
 
The UM Miracle (or any hexa-driver custom IEM) has imaging 10 stories higher than the ER-4S, I think a fish could hear the difference.
 
 
I like the testing that sonove does a lot, and innerfidelity too, and I agree that it should be done more, but like you're saying, the testing is limited, at least for now.

 

[nikp]

Quote:

 
How about ... three poorly-chosen (and/or mediocre-performance) drivers vs. one well-chosen (and/or high-performance) driver? In some "purist" audio circles, minimizing (or ditching) x-over for loudspeakers is chosen deliberately. Also ... most full-sized cans (moving coil or ES) are full-range.
 
Spectrum-specific BA drivers have been around for much longer than high-performance IEMs (they were orig. designed for hearing aids; not purpose-designed for high-fidelity audio/music [methinks BA IEMs generally have a "cupped" or "midsy" or "nasaly" sound  'cause this is where all-important human-voice-speech occurs ... what you want from a hearing aid]). And they have been adapted or tweaked for IEM (= hi-fi audio/music) apps. Hence, they might not be ultimately be ideal for IEM design -- they just happen to "pre-exist" and are relatively easy to incorporate and cost-effective.
 
Not saying multi-driver approach is necess. wrong ... just presenting counterpoint.
 
The violin(ists) analogy is confusing ... for one, the earphone driver is an electro-mechanical transducer with v. important physical diffs. And it's not a musical instrument.
 
 

I would say a well chosen/high performance driver beats three poor/mediocre drivers anytime. But generally, triple or sextuple (high performance) drivers will provide better instrument separation than a single (high performance) driver iem. The violinists analogy is confusing for individuals who knows very little about violins. An earphone driver produces sound just like a musical instrument. And yes I get your point that more doesn't mean better. But the statement you made,"A quick answer: marketing; selling to those with a "keeping-up-with-the-Jonses" psychopathology,  etc." is too exaggerated.

 

[wilzc]

 
Quote:

More drivers = more accurate audio reproduction. 
 
1 Driver = multitasking (Highs, mids, lows all mashed up into one)
 
3 Drivers (3 Way Crossover) = 1 for the high freq, 1 for the mids and 1 for the bass = Much more cleaner sound is being reproduced.
 
Think of violinists. 
 
A single violinist playing a tritone. 
 
3 violinists playing an individual tone to make up a tritone.
 
All the violinists in an orchestra playing a tritone.
 
Did you notice a difference between them? Need I say more?
 
 

 
What happens when the conductor (crossover/tuning) is bad? it'll be more obvious with more violinists
 
Also, what happens in the case of CK10? where both violinists play the exact same thing
 
Klipsch image x10 (i didnt like them) and the Sleek sa6 as well as the ER4p have been very revered in some parts of the head-fi world and they do hold their ground. I do agree however like kiteki explained that they're all sharply imaged but never as expansive as a good multi driver. A coherently implemented multi way always hits me at the sweet spot where i feel butterflies in my ears. A fantastic single driver never does that.
 
 
 
 

 

[kiteki]

 
Quote:

Originally Posted by wilzc /img/forum/go_quote.gif

[...]
A coherently implemented multi way always hits me at the sweet spot where i feel butterflies in my ears. A fantastic single driver never does that.