[money4me247]

just a note it is actually a commonly documented phenomenon that our brains always prefer the loudest sound when all other things being similar. that is why it is important to volume match when demoing or else you just think the loudest one sounds best regardless of how true that is.

 

[thyristor44]

Brilliant article mate, realy enjoyed it.
thyristor44

 

[SilverEars]

  Just as a side note, I think it's really strange that any IEM would be designed to be so low impedance and so sensitive. Most gear won't have power problems with IEM's, there's channel imbalance with pots at lower levels where you'd be more likely to be in, noise becomes an issue, you could accidentally blow them out much easier if you bump the volume knob, you're more limited in gear to stuff that has a really low output impedance as described (especially true with IEM's due to crossovers and varying impedances across the frequency spectrum when trying to get a proper damping factor and frequency response), and many amps become quite unstable/distorted/current-limited with loads even under 20 ohms.
 
Designing anything below 32 ohms is just a bad idea to me. Higher impedance headphones means you can use many more sources to greater success, as it is usually flaws in sources that cause most of the issues I've explained. If you're having that many noise issues, you could maybe try an impedance adapter - http://www.head-fi.org/t/601669/impedance-adapters-cables-explained-listed/45 - or something like Heir Audio's Tzar series IEM's with impedances of 90 or 350 ohms.

If you are talking IEMs.  I've only seen low impedance dynamics drivers which are flat impedance.  BA varies in impedance characteristics which means FR will be affected depending on it's reactance which you will see much more variations with BA iems with crossovers.  UERM may spec 30ohms, but there could be impedance drop to 10ohms at the treble.  ER4's impedance adapters boosts the treble since it has a typical single BA impedance response with flat impedance with rising impedance at high frequencies.  
 
What I don't understand is how the impedance adapter drops the hiss without dropping the the signal level.

 

[stv014]

What I don't understand is how the impedance adapter drops the hiss without dropping the the signal level.

 
It drops both, but to compensate, you need to turn up the volume, and that typically increases the hiss less than the signal (because noise that is added after the volume control is not affected by it), so you get better SNR at the same loudness.

 

[castleofargh]

  Just as a side note, I think it's really strange that any IEM would be designed to be so low impedance and so sensitive. Most gear won't have power problems with IEM's, there's channel imbalance with pots at lower levels where you'd be more likely to be in, noise becomes an issue, you could accidentally blow them out much easier if you bump the volume knob, you're more limited in gear to stuff that has a really low output impedance as described (especially true with IEM's due to crossovers and varying impedances across the frequency spectrum when trying to get a proper damping factor and frequency response), and many amps become quite unstable/distorted/current-limited with loads even under 20 ohms.
 
Designing anything below 32 ohms is just a bad idea to me. Higher impedance headphones means you can use many more sources to greater success, as it is usually flaws in sources that cause most of the issues I've explained. If you're having that many noise issues, you could maybe try an impedance adapter - http://www.head-fi.org/t/601669/impedance-adapters-cables-explained-listed/45 - or something like Heir Audio's Tzar series IEM's with impedances of 90 or 350 ohms.

I +10000000 this opinion that a headphone shouldn't be designed with too low an impedance or too high a sensi!!!! as a consumer I do not need that shhhhhiiiiiiiiii....
but it's done, and we see it a lot more than some years back. I really don't know enough about electricity to tell if they are retards messing with us, or if most likely they have something to win with those designs in term of audio on multi BA crossovers?
 
 
for the link (it is interesting indeed) there is the "too good to be true" suspicion in my head, with all the useless meaningless junk sold all over headfi, if something like this worked ideally why isn't it sold everywhere by some guys selling the "1000$ 3resistors special" to audiophiles for IEMs?
 
stv014 I cast a magic zobel spell and call upon your spirit to appear on this topic!!!
plz explain moarrr, and if it's all good for IEMs and 5ohm daps(but the interest really being the voltage gain loss, not so much the damping gain), would you take me as a partner to make the 2015 elite headfier's toy that will lend us billions of undeserved $$$$?
 
edit::: damn I should have F5 my browser, or maybe my casting worked so well it made you come back in time before I asked??? ^_^(now now I'm a science guy, can't you tell?)

 

[SilverEars]

   
It drops both, but to compensate, you need to turn up the volume, and that typically increases the hiss less than the signal (because noise that is added after the volume control is not affected by it), so you get better SNR at the same loudness.

Yes, it makes sense that the noise added after the volume control attenuated with the resistor, but you can increase the signal level still.  I have experienced hissing increasing with increased volume control with amps, why is this?

 

[GrindingThud]

Most likely because the source component has a noise floor that is getting amplified. Short the input to an amp to ground and listen to it with the volume down, that is the lowest noise configuration. Moving the volume knob up from there will add thermal noise from the pot itself into the input to be amplified. Highest noise from the pot is usually heard at around 1 o'clock then decreases again beyond that point.

I have experienced hissing increasing with increased volume control with amps, why is this?

 

[bigshot]

  I think you're confusing an attribute of IEM's physics (low impedance) with an inherent flaw with their design, or some kind of failure to deliver.

 
IEMs are primarily for mobile use. So I would think they would be designed to work with most mobile DAPs.

 

[SilverEars]

   
IEMs are primarily for mobile use. So I would think they would be designed to work with most mobile DAPs.

Mobile devices have insignificant impedance when compared to something like a dynamic driver portable phone of around 32ohms.  But iems have impedance characteristic that is not leveled and some dip low at certain frequencies than others.  Some DAPs have low impedance some relatively higher that may not work well with iems with varying impedance response.  One extreme example is the SE846.  You see where it drops to 4ohms?  Iphone 5 has like 4.5ohms output impedance.  There is no doubt 846 was designed for mobile use.
 

 

[Strangelove424]

   
IEMs are primarily for mobile use. So I would think they would be designed to work with most mobile DAPs.

 
I’m a big headphone guy, but from my limited experience in IEMs, I know there are two driver types: typical dynamic, which can produce good low end but get lost on detail, or BA drivers which function like planar magnetics with a sheet between two tiny magnetic structures. The BAs do detail great, but they don’t do low end, so they often combine multiple BA drivers. It’s the BA drivers that present the harder to drive loads that stress the mobile DAPs I believe, and they come in about 16ohm. The typical dynamic drivers are the ones we’re all used to that work fine on the mobile phones and stuff, but I think it is the slightly more exotic/precise BA drivers that present the problems for DAPs. If I were in that boat, I'd check out DAPs with less than 1 ohm output, and with the iPod Nano's .3 ohm output (atleast the generation I looked up) I would give that a shot. They're like $50 on Ebay.  
 
Edit: Nevermind, the Nano is about 1 ohm. That's still vastly lower than the iPhone though. 

 

[money4me247]

  Mobile devices have insignificant impedance when compared to something like a dynamic driver portable phone of around 32ohms.  But iems have impedance characteristic that is not leveled and some dip low at certain frequencies than others.  Some DAPs have low impedance some relatively higher that may not work well with iems with varying impedance response.  One extreme example is the SE846.  If you see where it drops to 4ohms?  Iphone 5 has like 4.5ohms output impedance.  There is no doubt 846 was designed for mobile use.

 
I have read that all headphone impedance changes with frequency. The greater your output impedance is above zero, the greater voltage changes will to delivered to your headphones as the frequency changes, so for DAP/mobile phones/amps with higher output impedance, you will get greater frequency response deviations. There is a 1/8 rule of thumb stating if the output impedance is greater than 1/8th the headphone's impedance, you will get variations in frequency response. The variations are more extreme in balanced armature or multi-driver designs of IEMs. Balanced armature IEMs (such as models from shure, etymotic, and ultimate ears) can sound significantly worse with higher output impedance devices as their actual impedance varies with frequency very drastically from their rated impedance on the spec sheet.
 
If you are passionate about multi-driver or balanced armature IEMs, you will need to invest in sources with really low (near zero) output impedance to eliminate the audible distortions.

 

[Head Injury]

   
I have read that all headphone impedance changes with frequency. The greater your output impedance is above zero, the greater voltage changes will to delivered to your headphones as the frequency changes, so for DAP/mobile phones/amps with higher output impedance, you will get greater frequency response deviations. There is a 1/8 rule of thumb stating if the output impedance is greater than 1/8th the headphone's impedance, you will get variations in frequency response. The variations are more extreme in balanced armature or multi-driver designs of IEMs. Balanced armature IEMs (such as models from shure, etymotic, and ultimate ears) can sound significantly worse with higher output impedance devices as their actual impedance varies with frequency very drastically from their rated impedance on the spec sheet.
 
If you are passionate about multi-driver or balanced armature IEMs, you will need to invest in sources with really low (near zero) output impedance to eliminate the audible distortions.

The 1/8 rule isn't a rule, it's a recommendation. The frequency response will change with any output impedance, even 0.1 ohm. 1/8 of the headphone impedance just happens to be a good number to aim for that achieves inaudibility almost all of the time. Most full-sized headphones will cope with much more. Some IEMs might need even lower.
 
Also AFAIK planar magnetic drivers have flat impedance curves. There was some debate in another thread about whether or not damping factor mattered at all for them.

 

[SilverEars]

  The 1/8 rule isn't a rule, it's a recommendation. The frequency response will change with any output impedance, even 0.1 ohm. 1/8 of the headphone impedance just happens to be a good number to aim for that achieves inaudibility almost all of the time. Most full-sized headphones will cope with much more. Some IEMs might need even lower.
 
Also AFAIK planar magnetic drivers have flat impedance curves. There was some debate in another thread about whether or not damping factor mattered at all for them.

Yup, and if anybody has looked at impedance curve of dynamic drivers of lower impedance for portable over ears or iems, you will see flat impedance.  Even look at the Sennheiser Momentum, it's flat.  When you get to high impedance designs like the Beyers or Sennheisers with 300ohm nominal the impedance curve shows significant resonance hump.  That hump is resonance where the peak point is where the inducatance and capacitance meet and cancel each other out to be resistive.  That means the left side where it rises is inductive and the right is capacitive.  The Q factor of the resonance changes with the added resistance which shape the resonance.  Within the resonance area is where the added resistance affects the FR.  And a measurement I've seen of the HD800 shows that it happens only when the resistance is pretty significant like a value that is equal to it.  180ohms shows only slight boost.  And, there was minimal FR changes in the region that is close to flat in the mids to high frequencies even with 600ohm output impedance.  Also I read that if you short out the speaker terminal, the cone resists movement, which means that with a resistance in between, it moves more freely.  That means the output impedance would cause less damping and cause less control of the driver.  Close to 0 output impedance would be like a short at the terminals creating maximum damping.  I believe stv posted some distortion measurments of the Beyers showing much higher distortions at the resonance frequencies with added output impedance.
 
I think damping factor is for full sized headphones or speakers that has the characteristics like the Sennheisers or Beyers with the resonance hump and only matters because of the resonance.  I think low impedance iems with non-flat response is different.  Not sure if "damping" really applies to BA iems with skewy impedance response.  Since they are uneven some parts are going to affect the power output relative to others that would change the FR compared to output impedance close to 0.  I would think the reactive areas would be affected differently from resistive areas, but don't know how.

 

[castleofargh]

 

   
I have read that all headphone impedance changes with frequency. The greater your output impedance is above zero, the greater voltage changes will to delivered to your headphones as the frequency changes, so for DAP/mobile phones/amps with higher output impedance, you will get greater frequency response deviations. There is a 1/8 rule of thumb stating if the output impedance is greater than 1/8th the headphone's impedance, you will get variations in frequency response. The variations are more extreme in balanced armature or multi-driver designs of IEMs. Balanced armature IEMs (such as models from shure, etymotic, and ultimate ears) can sound significantly worse with higher output impedance devices as their actual impedance varies with frequency very drastically from their rated impedance on the spec sheet.
 
If you are passionate about multi-driver or balanced armature IEMs, you will need to invest in sources with really low (near zero) output impedance to eliminate the audible distortions.

The 1/8 rule isn't a rule, it's a recommendation. The frequency response will change with any output impedance, even 0.1 ohm. 1/8 of the headphone impedance just happens to be a good number to aim for that achieves inaudibility almost all of the time. Most full-sized headphones will cope with much more. Some IEMs might need even lower.
 
Also AFAIK planar magnetic drivers have flat impedance curves. There was some debate in another thread about whether or not damping factor mattered at all for them.

yup, in fact the 1/8 or 1/10 is a middle ground between power efficiency and damping factor impact. in truth having a damping ratio of 1/100 is just as good for sound so it should be called  the "at least 1/8 rule". 
still following it is a no brainer. the fact that it doesn't always blow up when we don't do what's expected isn't reason enough to do whatever we want. it's still electricity and will always work best when we do as advised. if they make headphones and amps for impedance matching in mind, I'll be glad to follow a 1/1 rule instead.
at least for electricity I do as I'm told ^_^.
 
@money4you
  etymotic is a bad example as the er4 is the most famous in ear for getting "better"(er4s) with added impedance. I think I've come across 3 or 4 IEMs thinking that a little more impedance could sound good(like the W4). for the rest, it either sounds like crap, or doesn't matter (I have the ety mc5 in my ears right now and they don't seem to care about impedance much, +100ohm seems to sound pretty much the same to me(minus the huge loudness difference).

 

[Audio-Omega]

http://www.popsci.com.au/science/medicine/this-woman-sees-100-times-more-colours-than-the-average-person,396736