1. 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.

    Dismiss Notice

Is impedance or sensitivity more influential in hiss for an IEM?

Discussion in 'Sound Science' started by baskingshark, Sep 24, 2019.
  1. baskingshark
    Hi friends, for some of my low impedance and high sensitivity IEMs, I get a hiss with smartphone and desktops. But it disappears with an inline volume controller or impedance mismatch device or an amp. This hiss is more marked in some IEMs than others (using the same source and same 8 core copper cable from NICEHCK):

    These are my IEMs with hiss:
    - TFZ No. 3 (impedance 20 ohms, sensitivity 108 dB)
    - KZ ZS10 Pro (impedance 30 ohms, sensitivity 111 dB)

    No hiss:
    - Audiosense T800, (impedance 9 ohms, sensitivity 90 dB)
    - Toneking nine tails (impedance 16 ohms, sensitivity 100 dB)
    - Hisenior B5+ (impedance 20 ohms, sensitivity 102 dB)
    - Westone W30 (impedance 30 ohms, sensitivity 107 dB)
    - Semkarch SKC CNT1 (impedance 32 ohms, sensitivity 108 dB)

    Is hiss factored by both low impedance and high sensitivity? Which is more influential in hiss among the 2 factors?
    I'm no audio expert, but I'm wondering how come the IEM with very low impedance (9 ohms in Audiosense T800) can have no hiss compared to the 30 ohm KZ ZS10 Pro with hiss?

    Would appreciate any advise, thanks in advance!!!
    Last edited: Sep 24, 2019
  2. Scgorg
    Sensitivity to hiss is due to both impedance and sensitivity (mostly sensitivity). What you have to realize is that when you turn a volume knob to minimum there is still a slight signal passing through (for analog controls at least, probably a lot of digital controls too) and on some amps you can even hear a little bit of sound with volume knob turned all the way down such as on the Auralic Taurus. As you go lower in power the amplifier typically gets a worse SNR (signal to noise ratio), if you want some kind of graphic on this then checking out the 50mV output measurements on audiosciencereview could be enlightening.

    Basically, when the amplification is turned all the way down it still sends out a weak signal that is riddled with noise. As you increase the signal, noise often stays the same/decreases because the amplifier gets more linear at slightly higher volume (better SNR). This is why you may often feel that the hiss is the same or at least similar on most volume settings when there is a quiet section in a song (but it can of course increase, especially when you get close to max output).

    This is merely my understanding of the subject, some of it may not be 100% correct or properly explained but this should still give you a general idea.

    Edit: Also using an attenuator in form of something like the IEMatch simply attenuates the signal (noise and music signal) but when you turn it up the noise stays mostly the same while signal is better, this is why the hiss disappears in this case.
    Last edited: Sep 24, 2019
    baskingshark likes this.
  3. baskingshark
    Thank you for your input, I think what u say is correct about the SNR, when the amplification/power is higher the hiss is less.
  4. castleofargh Contributor
    -short answer, it's complicated.
    -short answer that's correct most of the time: sensitivity is what matters. just try when possible to avoid ludicrously low impedance IEMs(you should check at all frequencies, not just the specs at 1kHz), but otherwise focus on sensi.
    -now the long ineffective attempt at explaining the situation without really providing a concrete answer:
    hiss can have various origins so you can't be sure to take care of all cases of audible hiss with only one approach. the very basic and typical issue is that if an IEM is very sensitive, it will require low voltage to sound loud. and that applies to low voltage noises as well, making them louder into a sensitive IEM. another important variable is how high is the noise in the first place obviously. those 2 variables are pretty much why I wrote the second line of this post. quite often those 2 variables will let you avoid audible hiss. now for the fun extras. the sensitivity of the IEM is at 1kHz, but it will have its very own frequency response so if the noise happens to spike let's say in the 5kHz region, from one IEM to the next, you may end up with that hiss being subjectively twice as loud despite the sensitivity specs showing the same number. to make things even more annoying, some DAPs will have an increase in noise when driving a low impedance load, so now the impedance response of the IEM becomes relevant, but doesn't explicitly tells us anything unless we can measure the DAP with the IEM plugged in. so we can at best act a little cautious/paranoid and try to avoid IEMs with extremely low impedance. hopefully the amp struggling with low impedance load is only going to be a rare scenario, but in a world where TOTL IEMs are more and more reaching below 8ohm somewhere, it wouldn't be reasonable to simply dismiss the possibility.
    as to noise sources, some will increase in a linear manner along with the gain applied to the music(everything is controlled by the volume knob), other noises won't. if using an in-line attenuator solved your issue, then it's likely that the noise is fairly constant and doesn't increase along with the music output. meaning a low sensitivity IEM is very likely to be your answer for that device. but depending on the amp design and type of volume control, we can have various behaviors. I have a portable amp where the hiss becomes maximum when the volume setting gets high, but the hiss is inaudible even on my most sensitive IEM while set at a lower level(making this amp a nice IEM companion but a pretty poor amp for low sensitivity headphones. it's rare but that happens too on occasion).
    and of course there is us. we're more sensitive to noise at some frequencies(midrange) than others, we have our own listening habits, our own FR preferences, the type of music and listening level will all be more or less likely to mask background noise. for big loud noises we're all the same, but more often than not, we're discussing rather small noises that would become barely audible or go unnoticed if it was just 6 or 10dB lower. so all those parameters may influence our final impressions and how mental we're going to get over some hiss.

    ps: yet another unnecessary mess is caused by how sensitivity isn't always expressed the same way by manufacturers. we kind of expect dB for 1mW so quite often we won't pay much attention to the unit or worry about the lack of clear nomenclature. many websites fall for that and display an output for 1volt or 0.1V or whatever, as if it was for 1mW, so I always suggest to look for a website that measured it themselves or check on the manufacturer's webpage. other places might not be correct.
    Last edited: Sep 28, 2019
    TronII and baskingshark like this.
  5. dazzerfong
    Forgive the ramble, but here's a general overview of hissing. To understand hissing, the following equation must be understood:

    Power = Voltage x Current

    Where's sensitivity in that? Sensitivity governs how much power* is required to achieve a certain volume. High sensitivity means lower power is required to achieve the specific volume, and vice-versa.

    How about hissing? Hissing is a function of the noise floor for a device, and is generally fixed in voltage. Ergo, if your amp is running at the lowest volumes, the noise floor might be very close to the waveform of your music, making it more audible. This concept is called signal-to-noise ratio (SNR): the higher the SNR, the less prominent noise is in your system.

    With that in mind, here's the general consideration:

    High sensitivity + low impedance = lower voltage, higher current draw
    Low sensitivity + high impedance = higher voltage, lower current draw

    Noise is a function of voltage; the current draw is a consequence of what comes after and whether or not your amp can supply that current. As noise floors are very low in voltage, your amp can supply the current necessary to 'drive' that noise. Therefore, sensitivity is the one that governs whether or not something hisses or not.

    However, as castleofargh pointed out, this isn't made easy by the fact that people use different units for sensitivity. There's generally two ways manufacturers express sensitivity, some use dB/mw, others like Sennheiser use dB@1 Vrms. You can convert these two units over, but it's not quite an honest conversion as it relies on impedance to convert over, and impedance measurement is generally done on only 1 frequency point. That being said, both units are perfectly valid units of measurements - I can't remember which one's the newer 'standard' way of doing things.

    There are attenuators like the iEMatch and iEMBuddy which reduce all voltage: therefore, the hissing also decreases. SNR is still in theory the same, but because your music signal is much higher now, it 'claws' back the 'missing' bits. I can personally testify that they do indeed work: while they do almost completely eliminate hissing, a consequence of these attenuators is that you must jack your volume all the way up to compensate.

    * You might've caught on that there's two ways to express sensitivity. This assumes it's in dB/mw.

    TL;DR: sensitivity governs if an IEM/headphone is prone to hiss or not. Some maths to explain why that is.
    Last edited: Sep 25, 2019
    TronII and baskingshark like this.
  6. danadam
    Those sensitivity values are probably in dB/mW (so I'd rather call them efficiency, not sensitivity) and if you convert them to dB/V (actual sensitivity), e.g. using Headphone Sensitivity & Efficiency Calculator you'll get:
    • TFZ No. 3: 125 dB/V
    • KZ ZS10 Pro: 126 db/V
    • Audiosense T800: 110 db/V
    • Toneking nine tails: 118 db/V
    • Hisenior B5+: 119 db/V
    • Westone W30: 122 db/V
    • Semkarch SKC CNT1: 123 db/V
    So those with hiss are in fact the most sensitive.
    baskingshark likes this.
  7. baskingshark
    Thanks @danadam , @Scgorg @castleofargh , @dazzerfong . Didn't realize that there are different units for sensitivity and all your explanations make sense!
    Last edited: Sep 28, 2019
    TronII likes this.
  8. jsmiller58
    This thread needs to be a sticky on all IEM threads... Thanks to the OP for asking the question, and those who took the time to answer.
    TronII and baskingshark like this.
  9. dazzerfong
    That's an interesting tool you got there, certainly means I don't need to crack the calculator out from time to time!

    However, I'm a bit wary of converting it over, especially for IEMs. IEMs, especially balanced armatures, have wildly swinging impedances which make conversion much less apples-to-apples. For things like planars and FIBAE's flat impedance IEMs, however, it's much more accurate.
    TronII likes this.

Share This Page