[robm321]

That's it.

Sent from my SM-N910T using Tapatalk

 

[tomscy2000]

Hi @EtyDave,
 
I appreciate your being in the forum and interacting with everyone. While I've known about the ER4SR and ER4XR for a number of months, I haven't had time to really look into them too much because of time issues.
 
However, I do have a few questions to pose, in order to gain a better understanding:
 

• What are Etymotic's performance standards with respect to distortion response? Balanced armature drivers are known to exhibit more odd-order distortion products than their dynamic coil counterparts. The quantity of odd-order distortion when standard-size BA drivers such as the Knowles ED or Sonion 2300 series are driven to 100 dB SPL@1 kHz is much higher at frequencies under 1 kHz. It looks like the ER4SR exhibit <0.5% THD for most of the audible spectrum when driven to 100 dB SPL, but the ER4XR instead exhibits just under 1% --- would that necessarily be below the hearing threshold? While I do not have links to justify the claim, many audio engineers and audiophiles claim that odd-order harmonic distortion products are easier to detect and less mellifluous to the ear than are even-order distortion components. There is also some anecdotal evidence that, when driven to higher drive voltages, balanced armature drivers begin increasingly producing higher-order harmonic distortion, which is much easier to detect than lower-order distortion. Factoring in trained ears (whether they be audio engineers or audiophiles) on top of all these concerns, would something like the ER4XR have inaudible bass distortion at typical music listening levels?
• From what I see, the QC measurements for the ER4SR/ER4XR frequency response and channel matching are performed at 200 mVrms. It seems the test levels are too loud for real-world listening (105 dB @ 1 kHz would not be a comfortable listening experience). How does the FR change, if at all, when the drivers inside the ER4SR/ER4XR are driven to, say, 100 mVrms? I know that companies like Sonion test their driver responses at 100 mV; what led Etymotic to test at 200 mV?
• While we're on the issue of output levels and audible distortion, what is Etymotic using to test these target responses for both R&D and QC? Although the "711 coupler", such as the GRAS RA0045 or B&K 4157, has been the de facto standard for occluded ear simulator measurements for many years, it's also widely accepted that it has some shortcomings, such as the high-Q resonance centered around 13.5 kHz that might misrepresent a real ear's canal resonance. GRAS recently took the initiative to try to modify this response with the 43BB system and equalizing away the resonance, and also introducing a very low noise acquisition system in order to better represent human hearing threshold levels. Has Etymotic been using such a system, and if not, what would the reasons be for not adopting it? If Etymotic decides to adopt such a system, how would the target response curve accuracy score be altered?
• Do the ER4SR/ER4XR utilize equalization with Helmholtz resonators (i.e. acoustic side branch) like the MC5/EK5/MK5? I believe the ER4B/ER4S/ER4P's target responses were achieved via high-pass RC and resistive underdamping, respectively. How do the ER4SR and ER4XR differ?
• Please correct me if I'm mistaken: If the ER4XR and ER4SR utilize a similar-sized driver (assuming it's still an ED-sized driver), then I can only assume that the ER4XR utilizes an open-vent variant driver (which also helps explain the added distortion). Looking at the controlled rise in the base response in the context of such an acoustic design, I can also only assume the space allowed for venting is restricted in volume. If so, would that mean that some head-fiers who desire to "reshell" the ER4XR in a custom-molded shell risk having extra bass (because of extra back volume) in the CIEM shell?
• With the rapid miniaturization of microphone components, how does Etymotic feel about utilizing evoked otoacoustic emissions to approximate a listener's hearing profile and adjusting EQ accordingly? Doing so can potentially compensate for otosclerotic changes and impaired hair cell function in the case of presbycusis or NIHL, in a limited fashion.

 

[luisdent]

TheDreamthinker,

While the er4sr isn't perfect, it is substantially superior to everything i have heard so as to make most earphones unnecessary for me. The xr is odd beacause the bass is great but the mid bass is a tad too much for me. The treble is noticeably better on the sr/xr compared to the s to me as well. Not drastically, but noticeabley. So either is an improvement in treble in some regard. Upper mids and mids and bass? Maybe not so much. Although the impression of an improvement in bass is there due to the overall response difference.

Much of this is preference too. I listen to bass oriented music on my sr and prefer it over the xr. To me it can be energetic or relaxed. It reflects the recording appropriately. I have said over and over the er4 lacks low sub bass volume. Not extension and not quality. It's just a touch low in sub amplitude.

Not everyone hears it this way. But i do. I rarely feel the need for eq on the sr. And I'm a huge proponent of eq to improve any earphone/headphone/speaker.

Anyway, there's nothing wrong with having multiple iems to suit whatever preferences one may have. i just don't feel that need myself. Until something better comes along, the sr is indeed the holy grail for me. I like the xr, and it may have it's place, but I always want to go back to my sr. It's phenominal.

I say that people should worry less about what is perfect and more about what they enjoy. The er4sr gives me the most enjoyment of any earphone. If it doesn't for you, try something else. Or find multiple earphones if each makes you enjoy one type of music more. For me, I'm happy with my er4sr.

 

[luisdent]

Make sure you replace you eartips now and then. I didn't notice anything wrong until the eartip spun too easily in my ear. You can guess where the other half is...

 

[vapman]

I've been using many of the same silicone tips for years, but i wash them with dish soap every few weeks too.

 

[Marc Lian]

Make sure you replace you eartips now and then. I didn't notice anything wrong until the eartip spun too easily in my ear. You can guess where the other half is...

Sheesh that's frightening. Hope all is well.

 

[mdiogofs]

One thing I noticed recently is that it's impressive how the SR can be resolving and relatively clear, while not being thin sounding. The notes actually have considerably weight and thickness. I can't explain why this really happens right now but the SR sound is quite addictive. Maybe because I like a bit of note thickness and surprises me how the sound stays clean and with nice separation. Addictive like when you put them in it's not easy to take them out. I listened to Joe Satriani playlist two times in a row. The guitar sound is really addictive in these.

 

[castleofargh]

I can't find much info about the drivers in the ER4 variants. I know that they are balanced armature types but are they just one driver configurations?

I read that Etymotic believes the lack of crossovers and driver networks is a plus point. While other IEM manufacturers think the opposite and try to tack on as many drivers as possible, curious to know which school of thought is better.

 crossovers make it easier to fine tune a signature. it's like using an EQ, the more points you have, the easier it is to get the frequency response you were aiming for. the problem obviously is that you need to find an effective way to band limit each driver and that's not as clean as we wish it would be.  
for speakers, both for signature and for frequency range, I believe nobody keeps trying to do great systems with 1 driver per side.
but for headphones it's the other way around, multidrivers have sucked one after the other.
for IEMs, we have both. with dynamic drivers that can extend pretty well on their own, usually adding more drivers isn't that beneficial. now BA drivers are more limited(because of less air flow?) so the degradation from the crossovers may be out weighted by the extra frequency response. I guess it's all a game of finding what works without too much counter effect.
 
  now etymotic can get pretty close to the signature they want with only one driver, so when you can do without, you do it I guess ^_^.

 

[MarkF786]

Can anyone recommend a dealer who offers a discount on the new ER4SR's?
 
Thanks!
 
Mark

 

[pkshan]

ER4B/SR bass is the best I've heard, in terms of quality & accuracy
 
3 reasons why there are not enough bass on ER4:
1. Tips are not deep enough
2.Bad quality source ,.ER4 needs a good vinyl or digital source to get proper bass.(In digital sources,  only a well made TDA1540/1541 CD player/DAC will produce the real,full & deep bass)
3.You‘re a basshead

 

[pkshan]

 
Quote:

   
I'm really glad you like the ER4B.  While it's not as popular as the other varieties of ER4, the ER4B fans tend to be very passionate about it. 
 
The anatomy of the human ear creates a resonant peak at 2.7-2.8 kHz (on average).  This can be confirmed (and has been by many people) with probe measurements at the ear drum.  If you have a deep insertion earphone, you lose that resonance and the earphone needs to put it back in.  Now, that accounts for for why it's there, but does not account for individual preferences.  If one prefers to reduce the peak via EQ (or higher value damper) there's certainly nothing wrong with that.
 
The peak at 16kHz seems unnatural high on that particular graph.  Did they say which artificial ear simulator (or coupler) they used to make that measurement?

 

 

Thanks very much for the reply, sorry for my grammar mistakes, English is not my mother language.
the peaks were shown in the official FR graphs too.
 
I play guitars, when I listening to ER4S, because the 6khz-10khz roll-off, the dynamic is compressed, timbre is losing power, music isn't involving,
from an instrument player's perspective, this sound is unrealistic. from a music lover's perspecive, the music is not fun, not moving me.
 
I think the ultimate goal of a transducer, is it  "sounds like a real thing", "reflected the recording" ,
not: "sounds like a hi-fi speaker" or "a compensation to the commercal recording".
the difference here is huge.
a flat response transducer without the peaks & roll offs will be well appreciated

 

[AnyDave]

  Hi @EtyDave,
 
I appreciate your being in the forum and interacting with everyone. While I've known about the ER4SR and ER4XR for a number of months, I haven't had time to really look into them too much because of time issues.
 
However, I do have a few questions to pose, in order to gain a better understanding:
 

• What are Etymotic's performance standards with respect to distortion response? Balanced armature drivers are known to exhibit more odd-order distortion products than their dynamic coil counterparts. The quantity of odd-order distortion when standard-size BA drivers such as the Knowles ED or Sonion 2300 series are driven to 100 dB SPL@1 kHz is much higher at frequencies under 1 kHz. It looks like the ER4SR exhibit <0.5% THD for most of the audible spectrum when driven to 100 dB SPL, but the ER4XR instead exhibits just under 1% --- would that necessarily be below the hearing threshold? While I do not have links to justify the claim, many audio engineers and audiophiles claim that odd-order harmonic distortion products are easier to detect and less mellifluous to the ear than are even-order distortion components. There is also some anecdotal evidence that, when driven to higher drive voltages, balanced armature drivers begin increasingly producing higher-order harmonic distortion, which is much easier to detect than lower-order distortion. Factoring in trained ears (whether they be audio engineers or audiophiles) on top of all these concerns, would something like the ER4XR have inaudible bass distortion at typical music listening levels?
• From what I see, the QC measurements for the ER4SR/ER4XR frequency response and channel matching are performed at 200 mVrms. It seems the test levels are too loud for real-world listening (105 dB @ 1 kHz would not be a comfortable listening experience). How does the FR change, if at all, when the drivers inside the ER4SR/ER4XR are driven to, say, 100 mVrms? I know that companies like Sonion test their driver responses at 100 mV; what led Etymotic to test at 200 mV?
• While we're on the issue of output levels and audible distortion, what is Etymotic using to test these target responses for both R&D and QC? Although the "711 coupler", such as the GRAS RA0045 or B&K 4157, has been the de facto standard for occluded ear simulator measurements for many years, it's also widely accepted that it has some shortcomings, such as the high-Q resonance centered around 13.5 kHz that might misrepresent a real ear's canal resonance. GRAS recently took the initiative to try to modify this response with the 43BB system and equalizing away the resonance, and also introducing a very low noise acquisition system in order to better represent human hearing threshold levels. Has Etymotic been using such a system, and if not, what would the reasons be for not adopting it? If Etymotic decides to adopt such a system, how would the target response curve accuracy score be altered?
• Do the ER4SR/ER4XR utilize equalization with Helmholtz resonators (i.e. acoustic side branch) like the MC5/EK5/MK5? I believe the ER4B/ER4S/ER4P's target responses were achieved via high-pass RC and resistive underdamping, respectively. How do the ER4SR and ER4XR differ?
• Please correct me if I'm mistaken: If the ER4XR and ER4SR utilize a similar-sized driver (assuming it's still an ED-sized driver), then I can only assume that the ER4XR utilizes an open-vent variant driver (which also helps explain the added distortion). Looking at the controlled rise in the base response in the context of such an acoustic design, I can also only assume the space allowed for venting is restricted in volume. If so, would that mean that some head-fiers who desire to "reshell" the ER4XR in a custom-molded shell risk having extra bass (because of extra back volume) in the CIEM shell?
• With the rapid miniaturization of microphone components, how does Etymotic feel about utilizing evoked otoacoustic emissions to approximate a listener's hearing profile and adjusting EQ accordingly? Doing so can potentially compensate for otosclerotic changes and impaired hair cell function in the case of presbycusis or NIHL, in a limited fashion.

 
Hi Tom, 
 
You ask a lot of very good (and detailed questions).  I'm going to answer some of them, but some require a level of detail sharing beyond my comfort level of posting on an Internet forum.  I hope that you understand that I might be slightly vague in some areas.
 
1.)  The limit for THD on the ER4 line is actually 2% with a 200mV drive level.  They usually come in well below that.  As you suggested, THD does vary with drive level and it would measure a bit lower at 100mV drive.  There really isn't an industry standard.  I've seen those who test at 100mV drive, I've seen those that set it according to output level of the earphone (usually about 100dB SPL).  I understand that this can be frustrating when trying to get comparative data.  The ER4 test at 200mV is a bit harsher of a test than 100mV, so you may want to keep that in mind when making comparisons.  As far as audibility, part of it depends on source material.  Users can sometimes pick up sub 1% distortion if listening to sine waves but I've seen no data that shows that people can hear anything close to that on music.  In fact, studies I've seen point in the opposite direction.  Distortion levels certainly do go up with voltage, but so does the distortion of the human ear.  It's been a while since I've read the study, but I seem to recall that the distortion of the ear approaches 10% once you are over 100dB SPL (but certainly don't quote me on that).  You are right that dynamic drivers do have, on average, lower distortion levels.  For some applications, that can prove beneficial.  As far as the audible distortion of the bass on the ER4XR, I've not observed or measured distortion at what I would consider typical listening levels (remember our THD measurements push 100dB or more at 1kHz).
 
2.)  The shape of the frequency response between 100mV and 200mV will be very similar.  It does shift a little bit with level, but you'll need more of a level shift than that for a pronounced difference.  That is inherent in using balanced armature receivers.  The 200mV for the ER4 is actually a carryover for the last 25 years.  We actually use 100mV for our other consumer earphones (like the HF, MC, and EK).  An argument can be made for making them all the same, but we stuck with 200mV to stay consistent with the way we tested the previous generation ER4 earphones.
 
 
3.)  The IEC-711 coupler is actually a recent change for us in our production department for ER4 production.  Previously, we  used the DB100 (aka the Zwislocki) coupler, which is a four branch coupler instead of the two branch 711.  We switched for a couple of reasons:
           A.) The 711 is the most commonly used coupler and it's easier for others to replicate the curves we provide with couplers they have access to.
           B.) The Zwislocki coupler (what we have historically used) is long out of production and increasingly difficult to support for production test systems.
In R&D we have a number of different couplers to choose from (including ones that use 1/4" microphones).  As you identified, there can be issues with the higher frequencies (although the DB100 is better than the 711 in that respect).  This is more important on an engineering level than on an OQC level, however.
 
4.)  The new ER4 line uses a similar approach to the acoustic design as the previous generation with improvements to the response by a combination of tweaking the driver and changing the impedance.  I think I've been pretty careful not to overstate the driver changes on the ER4SR.  It is different,but it's not a massive redesign.  The difference between the ER4SR and ER4XR is that they use two different drivers.
 
5.)  I will admit that we didn't consider reshelling the earphones when we designed the ER4SR and ER4XR earphones.  There is a percentage of our users that get custom sleeves made for them and we did consider this, but the reshell market is very small (at least as far as I can tell).  I am curious how many people do this, however so I'll look into it a bit more.
 
6.)  As you most likely already know, Etymotic is heavily involved in Otoacoustic emission testing and design.  The idea of incorporating it into earphone design has been discussed numerous times and it's a very interesting idea, but I can't really comment more than that on it.
 
I hope that helps.

 

[AnyDave]

   

 

Thanks very much for the reply, sorry for my grammar mistakes, English is not my mother language.
the peaks were shown in the official FR graphs too.
 
I play guitars, when I listening to ER4S, because the 6khz-10khz roll-off, the dynamic is compressed, timbre is losing power, music isn't involving,
from an instrument player's perspective, this sound is unrealistic. from a music lover's perspecive, the music is not fun, not moving me.
 
I think the ultimate goal of a transducer, is it  "sounds like a real thing", "reflected the recording" ,
not: "sounds like a hi-fi speaker" or "a compensation to the commercal recording".
the difference here is huge.
a flat response transducer without the peaks & roll offs will be well appreciated

 
Your reply was quite clear.  Please do not worry about your grammar.  You write very well, especially considering English isn't your mother language.
 
Just curious, are you live monitoring your guitar or listening to recordings of guitars?  If you are live monitoring your guitar, are you using a direct out or are you using some sort of cabinet simulator? (I also play guitar).  If you are primarily listening to non-mastered recordings of instruments, I can see why you might prefer an ER-4B response.  Especially, if the recordings are binaural, uncompressed recordings.  The first ER4 we ever released was the ER4B and is most accurate to the raw frequency response of the human ear.  The ER4S (and the adjusted target curve) came later in response to significant market pressure from our users who liked the ER4B but simply felt it was too bright.  The ER4P then came to pass because the sensitivity of the ER4S was too low for most portable users (back then portable media players and CD players had lower output levels).  At the time, we had to compromise on the high frequencies with the ER4P because we ultimately needed enough sensitivity that people could use them.
 
As far as tweaking the peaks and roll-offs, it's can be done acoustically if you have unlimited space in the earphone, but it's far more difficult to do in the small space of an ER4.  There are often trade-offs.  If we can find a way to get rid of the ultra-high frequency peaks without causing problems in the lower (and much more audible) frequencies, we will do that.  The ER4B in that graph does have more of a high frequency peak than I remember, so I'll have to revisit that.
 
In the end, the majority of our customers prefer the high frequency performance of the ER4S/SR/XR.  There is a customer base that does prefer the ER4B (sometimes very strongly prefer) so we make that earphone for them.  We appreciate all of our customers, no matter which version they prefer; we're just happy they chose something we made.  My hope is that there is enough demand for a new ER4B version that we can come out with that, too.

 

[Francisk]

Hmmm....I'm one of those rare ones who loves the ER4B 

 

[JackRabbitSlims]

any comments from users paring the FiiO X1 with ER4SR??
 
I'm currently using the X1 with UERM's but want something smaller and more pliable for a cycle touring trip!