Headphone & Amp Impedance Questions? Find the answers here!

[LarsHP]

It could be the amp circuitry, not the DAC. You would need to use an amp to balance levels and guarantee identical impedance. You’re also not doing a blind test if you have to see to plug it in. There are many reasons they might sound different that have nothing to do with the DAC.

The things you need for a fair test, aside from an apples to apples comparison, is level matching, direct A/B switching and blind comparisons with multiple trials. I’m guessing you don’t have any of these.

The FiiO and Hifiman are DAPs, in other words, integrated units with both DAC and amplifier. It is in this use, the difference appears clearly.

I should add that when using a headphone I also have, the tonal difference between the FiiO and Hifiman is minor. This clearly suggests that there is something going on which has to do with amplifier output impedance impacting frequency response. When feeding an external amp, using the (proper) line out from the Hifiman and the FiiO from it's combined HO/LO, the difference is not much to talk about.

As I stated above, even if the loudness level is slightly higher from the FiiO, the Hifiman still sounds darker. I don't need further suggestions regarding blind testing etc. etc., thanks. The difference between the FiiO and Hifiman headphone outputs to the UM Miracle is real.

 

[LarsHP]

As suggested there is the matter of listening to the same thing at the same volume level in a somewhat unbiased way.
But let's pretend that's not as much of a problem as it is and focus on gear. There could be sound differences for more reasons than just Impedance.
But let's pretend that's not the reason for your impression. Impedance could indeed affect the FR. How is tricky to know without actual measurements. I couldn't find an impedance graph online for that IEM, and that's a big issue to make any sort of guess as the IEM's impedance curve determines how much and where the FR would be affected by a different amp's impedance. The 15.9ohm spec is most likely the value measured at 1kHz, but with a multidriver IEM we know for sure that the value changes a lot between 20Hz and however high that IEM goes. If 15.9ohm was the lowest impedance point for the IEM in the audible range, then as you say, there shouldn't be more than 1dB difference between 2ohm and lower value amps. But if the IEM actually reaches, say 5ohm somewhere, now we could expect a bigger FR change in that frequency area when changing the same amps.

In case this is not annoying enough for you yet, consider that the impedance of the amp is in fact rarely flat over the entire audible range and can even change based on what you test it with. Hopefully this rarely matters outside of very low or very high frequencies and we can try to wishfully ignore it and stick to the given impedance number because it shouldn't do too much at an audible level. But it's there.

Long story short, find someone with that IEM who knows how to measure the impedance curve(maybe some of the reviewers who make measurements already have it?). Or get yourself like a 100Ohm resistor some wires, plugs and a soldering iron, and you can measure it yourself with Room Eq Wizard(free software).

For the short story long, you can check https://www.head-fi.org/threads/feedback-about-gears-stop-doing-it-wrong-impedance.866714/ where I show 2 IEMs and 4 "amps"(pretty extreme cases picked on purpose but not the worst that can happen) and how vastly different the FR impact can be depending on the IEM's own impedance curve.

From the link you shared, I can see that even though two of the amps have markedly less than 1 Ohm output impedance, i.e. 0.4 and 0.6 Ohms, the difference in frequency response from your JH13 is 1dB below 1kHz and up to 0.5dB above. This principle mirrors what I am hearing, except I don't have the impedance curve of my UM Miracle, so I can't tell which amp has more or less output impedance, but the impact I am hearing seems to be markedly more than 1 dB difference.

Can we assume that, generally, multi BA IEMs will have an impedance curve that is low in the bass and higher in the treble?

 

[castleofargh]

Can we assume that, generally, multi BA IEMs will have an impedance curve that is low in the bass and higher in the treble?

Sadly there is no clear rule for that. A single BA will be like you describe with the impedance flying high up in the sky at high frequency. But with several drivers it's not really possible to know what 'base' value each of the selected ones will have(unless you manage to know exactly what drivers are inside?).
We can't even make educated guesses about how stacking drivers in parallel reduces impedance because in your case it's a 2+2+2 3way IEM.

You saw the JH13 with low impedance in the bass making the bass "quieter"(relatively to the rest) with higher impedance amp. An opposite case as I probably mentioned somewhere, is the Shure SE846 where it's now the bass that increases in amplitude when you plug in a higher impedance amp.

IEMs usually don't go as low as this one, but a handful do. Many have impedance curves that wiggle like crazy but because the lowest value is still in the 15-20ohm, the actual impact stays moderate on most amps.
Unless you get the graph or try measuring impedance or FR yourself, I wouldn't dare bet on anything. If you don't feel like that's the kind of stuff you can do but you're still curious, you could play some noise file loaded on the DAPs and try to record from a cellphone stuck on the nozzle with some RTA spectrum app of sort. And hope the difference is big enough to show up clearly. Rather easy to try, not that easy to make sure it's correct(just the low frequency noise from the environment is likely to drown whatever the mic tries to pick from the IEM's bass signal).
Just don't go kill your IEM just to make them loud enough for the measurements to show clearly! Leave killing IEMs with measurements to the experts(stepped sine measurements anyone?)

 

[bigshot]

Impedance is a wild card, especially with IEMs. You need to match the transducers properly or you won’t get the right results. Use the right ones and it’s the response that was intended. The way to even it all out is to use the DAP’s line out to feed an amp that can properly drive the cans if that’s what they need.

The problem in a case like that is the cans or IEMs, not the DAP itself, because with the right cans it performs as expected.

 

[Lumithium]

Most amps perform significantly worse at really low output levels. UM miracle is a very sensitive IEM. It is possible for that to skew the results.

 

[LarsHP]

Sadly there is no clear rule for that. A single BA will be like you describe with the impedance flying high up in the sky at high frequency. But with several drivers it's not really possible to know what 'base' value each of the selected ones will have(unless you manage to know exactly what drivers are inside?).
We can't even make educated guesses about how stacking drivers in parallel reduces impedance because in your case it's a 2+2+2 3way IEM.

You saw the JH13 with low impedance in the bass making the bass "quieter"(relatively to the rest) with higher impedance amp. An opposite case as I probably mentioned somewhere, is the Shure SE846 where it's now the bass that increases in amplitude when you plug in a higher impedance amp.

IEMs usually don't go as low as this one, but a handful do. Many have impedance curves that wiggle like crazy but because the lowest value is still in the 15-20ohm, the actual impact stays moderate on most amps.
Unless you get the graph or try measuring impedance or FR yourself, I wouldn't dare bet on anything. If you don't feel like that's the kind of stuff you can do but you're still curious, you could play some noise file loaded on the DAPs and try to record from a cellphone stuck on the nozzle with some RTA spectrum app of sort. And hope the difference is big enough to show up clearly. Rather easy to try, not that easy to make sure it's correct(just the low frequency noise from the environment is likely to drown whatever the mic tries to pick from the IEM's bass signal).
Just don't go kill your IEM just to make them loud enough for the measurements to show clearly! Leave killing IEMs with measurements to the experts(stepped sine measurements anyone?)

Thanks. This seems to confirm my suspicion that the UM Miracle has a quite varied impedance curve, and that the general trend is like the JH13. (The SE846 reacts counterintuitively.)

I just tested a few tracks with my smartphone (Samsung Galaxy S9+) which makes the sound even lighter than the FiiO M11 Plus Ltd. While I haven't found any official numbers, there are users that claim the output impedance of the S9+ to be 2.6 Ohm and around 5 Ohm, respectively. It is to be expected that a mobile phone will have a higher output impedance than a modern high quality DAP.

So, my tentative conclusion is that the D7 Sidewinder has the lowest output impedance, followed by the HM901S, then the M11 Plus Ltd and the S9+ has the highest (as expected).

 

[LarsHP]

Impedance is a wild card, especially with IEMs. You need to match the transducers properly or you won’t get the right results. Use the right ones and it’s the response that was intended. The way to even it all out is to use the DAP’s line out to feed an amp that can properly drive the cans if that’s what they need.

The problem in a case like that is the cans or IEMs, not the DAP itself, because with the right cans it performs as expected.

In this case of the HM901S, I can actually change the internal amp, since it is designed for it with various amp cards available. The IEM amp card has 2 Ohm output impedance (unlike the Balanced amp card, which has 1 Ohm), but I don't think I want to invest further in the HM901S. I will probably sell it, now that I have the M11 Plus Ltd.

Right now, I am wondering why it isn't standard for headphone amps to have various output impedance settings. I remember seeing that in a desktop (headphone) amp ten years ago.

 

[redrol]

Get a E1DA 9038D which has around 0.5ohm OI. Also the Radsone ES100 has 0.5ohm OI.

 

[LarsHP]

Get a E1DA 9038D which has around 0.5ohm OI. Also the Radsone ES100 has 0.5ohm OI.

Very low (much lower than 1 Ohm) OI appear to be the problem, not the solution, regarding the UM Miracle.

 

[redrol]

What?! No. Impedance has nothing to do with your noise.
[edit] wrong thread, my bad

 

[LarsHP]

What?! No. Impedance has nothing to do with your noise.

You obviously haven't read what I wrote and asked about. I have said nothing about noise. Where did you get that from? My issue is change in frequency response with various 1 Ohm (or less) OI amps.
(All my mentioned sources are dead-silent.)

 

[redrol]

You obviously haven't read what I wrote and asked about. I have said nothing about noise. Where did you get that from? My issue is change in frequency response with various 1 Ohm (or less) OI amps.
(All my mentioned sources are dead-silent.)

Oops, sorry, i conflated 2 posts. The device should be 8x - 10x lower than the impedance of the iem itself so as to not change the FR tuning from the manufacture. IF you like it better with a higher impedance, fine but you are acutally changing the FR compared to what it is supposed to be.

 

[LarsHP]

Oops, sorry, i conflated 2 posts. The device should be 8x - 10x lower than the impedance of the iem itself so as to not change the FR tuning from the manufacture. IF you like it better with a higher impedance, fine but you are acutally changing the FR compared to what it is supposed to be.

You really need to read the conversation above. It's more complicated than what you are assuming. The listed impedance is 15.9 Ohm, and my amps are specified as having less or around 1 Ohm OI.

The point is that UM Miracle appears to be extremely picky regarding what OI it "gets". Apparently, when it gets near zero Ohm OI, it sounds way, way too dark. When it gets around 1 or 2 Ohms, it sounds about right. If it gets several Ohm OI, it sounds too bright.

 

[bigshot]

Man! This thread is a cornucopia of gear designed to not work well with others! With all the headphones, amps and DACs that work fine out of the box, I don't know why people put up with stuff like this. It's like people go looking for equipment with esoteric incompatibilities.

 

[redrol]

You really need to read the conversation above. It's more complicated than what you are assuming. The listed impedance is 15.9 Ohm, and my amps are specified as having less or around 1 Ohm OI.

The point is that UM Miracle appears to be extremely picky regarding what OI it "gets". Apparently, when it gets near zero Ohm OI, it sounds way, way too dark. When it gets around 1 or 2 Ohms, it sounds about right. If it gets several Ohm OI, it sounds too bright.

So you are describing what happens when the bass is too much basically. Think of it like this: 0.5 Ohm OI will almost never influence the tonality where as higher can/will. TBH I like some IEMs around 1-2ohm, and some 0.5. Not all iems change their bass profile with different impedances, but a lot do.