[tomscy2000]

People need to keep in mind that Etymotic is not about pushing the limits of audio performance. The company is about listening accuracy and hearing protection. They really couldn't give two thoughts about so-called 'audiophiles'.
 
Don Wilson is an employee of Etymotic. He doesn't control what products get rolled out. If there's anyone that has a vested interest in Etymotic and the direction of the company, it's Mead Killion. I highly doubt he cares about maximizing the performance of the ER4 over something that may potentially be of societal benefit, such as the EtyKids and its accompanying iPhone app.
 
We can safely assume that Etymotic will never create an IEM with a 'colored' response. Every single one of their IEM efforts since the ER4 has been to recreate the ER4's response with lower cost materials and manufacturing.
 
But yes, there can easily be improvements to the ER4 design, and it doesn't even need to be with a dual-BA setup. Mead Killion used to work for Knowles. He likely knows all the people there (that matter, at least). He can easily commission a single-BA with the correct diaphragm stiffness and sufficient back volume to yield a frequency response with even higher accuracy capability. The ER4's biggest sonic problem (in my opinion) is ultra-low frequency roll-off. That can be corrected with the right venting of back volume. High frequency deficiencies can be rectified by introducing well-placed perforations within the diaphragm. Better vibration reduction can be done by designing a dual driver with the same armature to back volume ratio as the ED. Perhaps the slightly larger housing would even spur them to design a more ergonomic housing (not exactly their strong suit).
 
However, if the ER4B is able to achieve 92% accuracy (this is based off Killion's own older study, which may not be the final word in the best possible averaging of diffuse field eardrum responses amongst the general populace, as not enough peoples' individual HRTFs have been measured), then that's good enough for Etymotic. Why spend many more R&D dollars into proprietary-spec drivers for an extra 3-5% in accuracy? Business-wise it doesn't make sense, as it doesn't feed into the company ethos of protecting peoples' hearing.

 

[supersleuth]

Quote:

Is there room for improvement? I believe so, specifically at the extreme frequency ranges. A good way for them to keep their curve while getting a more visceral bass presentation, will be to introduce a dedicated woofer, a real "woofer", one that will increase regions only below 100hz.

 
 
I don't think the Ety engineers would regard that as an improvement. For better or worse according to one's taste, they don't seem to subscribe to the idea that a bass boost (relative to the flat low-frequency response that they in fact achieve- let's be clear, none of their IEMs rolls off significantly until you get below the 20 Hz threshold of audibility) is psychoacoustically desirable to compensate for the absence of "felt" bass. Those of us who buy Etys know what we're getting, and it's what we want. I detest artificially boosted bass.

 

[tinyman392]

Quote:

 
 
I don't think the Ety engineers would regard that as an improvement. For better or worse according to one's taste, they don't seem to subscribe to the idea that a bass boost (relative to the flat low-frequency response that they in fact achieve- let's be clear, none of their IEMs rolls off significantly until you get below the 20 Hz threshold of audibility) is psychoacoustically desirable to compensate for the absence of "felt" bass. Those of us who buy Etys know what we're getting, and it's what we want. I detest artificially boosted bass.

 
I actually feel that Etymotic can get flatter bass.  To me, it sounds like they have a slight mid-bass focus.  The slower decay helps it, but isn't a solution in my eyes.  
 
Actually, scratch that, I need to hear the ER4 before I make statements like that.  That statement above was in response to the HF series.  

 

[Inks]

However, if the ER4B is able to achieve 92% accuracy (this is based off Killion's own older study, which may not be the final word in the best possible averaging of diffuse field eardrum responses amongst the general populace, as not enough peoples' individual HRTFs have been measured),

  This was reaffirmed by a newer 2008 study by Moller and Hammershoi. Their average HRTF closely matches the one used by Killion, this time based on 47 subjects total, the most done yet I believe. 

 

[tomscy2000]

Quote:

  This was reaffirmed by a newer 2008 study by Moller and Hammershoi. Their average HRTF closely matches the one used by Killion, this time based on 47 subjects total, the most done yet I believe. 

 
I know. However, 47 subject is not enough.

 

[Inks]

I don't think the Ety engineers would regard that as an improvement. For better or worse according to one's taste, they don't seem to subscribe to the idea that a bass boost (relative to the flat low-frequency response that they in fact achieve- let's be clear, none of their IEMs rolls off significantly until you get below the 20 Hz threshold of audibility) is psychoacoustically desirable to compensate for the absence of "felt" bass. Those of us who buy Etys know what we're getting, and it's what we want. I detest artificially boosted bass.

That is is partly what they make as a case. Though the boost below 100hz does keep their curve in mind and I'm sure if they were to implement, it wouldn't be too extreme. 

 

[Inks]

 How would evaluate that its "not enough". How many subjects will it take and how was the number derived?

 

[Anaxilus]

Quote:

 
I know you did, which is why I'm scratching my head trying to figure out what you're on about.

 
Read the thread title.  I'm scratching my head trying to figure out how only ER4 users are relevant to this topic.
____
 
Obviously the UERM is a custom and not produced to represent an average of 47 different ears, only the actual user.  Anyway, based on my listening it's more accurate and detailed than the ER4 so the placebo from the extra drivers and crossover seem to be working for me.

 

[supersleuth]

Quote:

Read the thread title

 
I did. The question has  been (speculatively, but plausibly) answered many times over already. And I'm not an ER4 owner so why would I think only ER4 owners should comment in this thread?
 
UERM (tuning tailored to the purchaser) represents a completely different design philosophy from Ety's (which is to design everything to approximate an empirically derived "ideal"  EQ curve as closely as possible at a given price point.) Again, if you don't like Ety's approach, you're welcome to continue to not buy Etys. But your insistence that they should do things your preferred way and not theirs is pretty pointless.

 

[tomscy2000]

Quote:

  How would evaluate that its "not enough"

 
Well, for starters, I could say, I'm 95% confident that 99% of elements in the Kuiper Belt lie between Neptune and Pluto. I can also say that I'm 99.9% confident that 100% of elements in the Kuiper Belt lie between Neptune and Proxima Centauri. Which is the better statement?
 
I haven't read the paper itself, as I'm not subscribed to that journal. However, I can say what questions I would keep in mind before reading such a paper. Firstly, the methodology of selecting subjects is highly important. Are they volunteers? Which socioeconomic stratum do they arise from? What is the demographic spread? I would have to imagine that if these researchers are sufficiently competent, the selection criteria is based off of other studies as well, to choose subjects that fit within the a criteria of say, within two standard deviations of the average height/width/depth of the external acoustic meatus, concha, and auricle. You can even throw in the head shadow if you want.
 
However, even that isn't in complete accordance with experimental methodology, as subjects should be chosen at random and then outliers that don't fit within the statistically significant standards are thrown out. Such a method would require a study of thousands of people, if not tens of thousands or millions. It should also be chosen from people of every continent. Such a study is never realistic, but that'd be the experimental ideal. Yet, every study should strive for that ideal, and I can guarantee that 47 subjects would not come close to this ideal.
 
Keep in mind, also, that the point of the study you referenced is not to plot out the most accurate average diffuse eardrum response of the general population. It is to verify whether or not mannikins' measurements correlate closely to those of real humans using in-ear microphones, and that is a direct comparison between two metrics. Extrapolating the plot they graphed out to say, 'this is the most accurate graph available' is not founded on sound logic. There can be dozens of confounding factors that obfuscate the discrepancy between the responses of current mannikin models and real ears. None of these factors seem to have been discussed in the paper, at least from what I see in the abstract.
 
Anyways, this is all OT, so respond to me via PM if you wish.

 

[tranhieu]

Quote:

 How would evaluate that its "not enough". How many subjects will it take and how was the number derived?

I have not read that paper but in statistics to be able to use Normal/Gaussion distribution we need at least few hundreds of sampling units. 47 is a bit lacking in terms of population size. I have seen some analysis used as few as 100-150 but 47 might not tell you much.
Or you can toss Central Limit Theorem out of the window and it doesn't matter if you have 47 or 1000 people for testing.
 
Before talking about whether the samples are unbiased, the maths doesn't stand its ground to begin with.

 

[planx]

One day of not being on Head-Fi sure makes you feel left out of the community

 

[jhelsas]

I wanted just to say that I'm impressed that my (which I thought to be simple) question developed in such interesting thread, unfortunatly I'm still reading about the things you people posted and trying to understand them, and that's the main realy why I haven't posted any replys for a long time here.
 
But, again, thank you all for the great discussion!

 

[Spyro]

Maybe we are all over-thinking this.  It seems insanely illogical that Ety couldn't improve upon ER4 considering what UERM, JH10, JH13 sound like.
 
ER4 has been a real niche and great product run for them.  They just raised the price to $300 and must be making huge margin.  I presume sales are still decent.   No one would dare pay more for a single driver but what would they charge for a dual driver?  Is someone gonna pay $450-$500 for a dual driver from Ety and have it pull away all those ER4 sales?  Heck, the UM2 already uses the same Knowles ER4 driver yet adds a bass driver and it still sells for less than the ER4.  Notice every product they have made since has been step downs.

 

[ianmedium]

I will repeat what I said on the FAD thread, give me a well designed and thought out single BA driver IEM over a multi-driver any day. There is a simplicity to the delivery of music that I find so appealing. With multi-drivers in my experience it is like when you go into a store selling TV's and they have jacked up the contrast and colour to make things pop so as to look more spectacular but when you get the set home you can't last long with it like that before it gets tiring.

I find the finals I have deliver a sharper sound and the ETY's a more neutral sound that adds nothing that is not already there and I have been listening to them quite a bit more these past few days since seeing this thread and find myself being drawn into how they portray music more and more again.

Listening to Mozart's clarinet concerto K622 was sublime with the Ety's.

I think I must be one of the odd balls as well that find the ER4's very comfortable with long listening sessions!