[reginalb]

Thanks Chris, I shall do some side by side listening then, as I have seen the Plenue S being described as having a warmer signature as compared to the QP1R.

You quoted the second part of my question, did you mean to answer it?

 
You will find that a lot of descriptions of sound signatures don't sit with the facts when you measure them. You'll also hear tomes written about the difference between DAC's. Good luck even being able to tell them apart during a proper ABX. And all the hand wringing about the problems with the stress created by blind tests are fine, except that you if you can't even tell them apart, then the difference isn't enough for all the silly flowery descriptors bandied about. 

 

[CraftyClown]

   
You will find that a lot of descriptions of sound signatures don't sit with the facts when you measure them. You'll also hear tomes written about the difference between DAC's. Good luck even being able to tell them apart during a proper ABX. And all the hand wringing about the problems with the stress created by blind tests are fine, except that you if you can't even tell them apart, then the difference isn't enough for all the silly flowery descriptors bandied about. 

 
Thank you Reggie. I've started to get quite fed up reading most of the equipment threads on Head-fi for this very reason. I was starting to think I was either crazy or partially deaf as I couldn't hear these apparent night and day differences people talk about, but the more time I spend in the sound science forums, the more I realise I've swallowed the red pill and I can see the Emperor standing buck naked in front of me 

 
It just surprises how many people seem quite convinced they can hear these huge differences and are willing to pay thousands of pounds to keep playing the game.

 

[CraftyClown]

I took some reading from my Samsung Galaxy S7 as well
 

 

[hakuzen]

C-Media 6206 (external USB soundcard, 7.1)
 
In Chinese videos about IEMs measurement, saw some of these cheap USB external soundcards being used; they also include it together with measurement gear in a pack. So got one to try for 7€ in aliexpress.
(no opamp, all capacitors are Chongx 100μF 10V)
 

 
 
Unloaded and loaded with Pai Audio MR3: [used Realtek ALC888 (PC soundcard) line-in, and 16-bit 48KHz signal]
 

 
Can't figure how can they use this to measure IEMs and BAs.. Even with a calibration to try to fix the coloration, the bass roll-off when loaded starts too early.
 
If we use its SPDIF out signal, we could use another source, of course.
Let's see if its microphone or line-in are usable: [used iBasso DX90 as source, and 16-bit 48KHz signal]
 

 
same bass roll-off
 
new edit: forgot to add loopback test (unloaded and loaded)
 

 
 
Measures (@ 1KHz, 0dBFS), using DMM:
Output impedance: 1Ω.
Max power unloaded: phones out, 1.538Vrms.

 

[HiFiChris]

   

 
Ahhh, the good old problem of having a pretty low output impedance but nasty capacitors in the signal output path that mess with the bass of low-impedance headphones and in-ears.

Thank you, I've added it to the overview.
 
 
The Line In and Mic In performance doesn't surprise me too much, as many of those cheap interfaces have got a more or less strong roll-off on one or both ends of the frequency spectrum.
 
 
As for the high-frequency wiggle: Can you figure out whether it is just the interface's bad FR or a sampling rate mismatch of the output and input signal?

 

[hakuzen]

   
Ahhh, the good old problem of having a pretty low output impedance but nasty capacitors in the signal output path that mess with the bass of low-impedance headphones and in-ears.

Thank you, I've added it to the overview.
 
 
The Line In and Mic In performance doesn't surprise me too much, as many of those cheap interfaces have got a more or less strong roll-off on one or both ends of the frequency spectrum.
 
 
As for the high-frequency wiggle: Can you figure out whether it is just the interface's bad FR or a sampling rate mismatch of the output and input signal?

A new thing I learn from you, once again: the nasty capacitors action. Thanks!
 
About the highs wiggle, I don't think it's a sampling rate mismatch; I've checked the config of both output and input (16-bit, 48KHz), and also the test signal used (same rate).
Since @castleofargh told me about the convenience of using 48KHz to get extended highs, I've changed the rate in all my gear thoroughly.
The only possible mismatch would be I used ASIO4ALL (for output test), and set "Always Resample 44.1KHz <-> 48KHz" on. I don't know if it has resampled the 48KHz signal to 44KHz (my ignorance, again).

 

[castleofargh]

it's really a trial and error thing to find how to get the best of each device. the worst they are, the more you have to tweak.  on the cheapest inputs I have, I set the input as high as I can without regard to what I set in RMAA or other test. obviously doing so can go against some measurements, so this is really specific to trying to get a reliable frequency response.
 
the low freqs dropping like this is indeed often a matter of caps. if you have different known loads, you can try them and check if the lower the impedance, the higher the roll off. that would be an even better hint at caps playing high pass filter with the headphone/IEM.

 

[Samueru Sama]

  C-Media 6206 (external USB soundcard, 7.1)
 
In Chinese videos about IEMs measurement, saw some of these cheap USB external soundcards being used; they also include it together with measurement gear in a pack. So got one to try for 7€ in aliexpress.
  
 
 
Unloaded and loaded with Pai Audio MR3: [used Realtek ALC888 (PC soundcard) line-in, and 16-bit 48KHz signal]
 

 
Can't figure how can they use this to measure IEMs and BAs.. Even with a calibration to try to fix the coloration, the bass roll-off when loaded starts too early.
 
If we use its SPDIF out signal, we could use another source, of course.
Let's see if its microphone or line-in are usable: [used iBasso DX90 as source, and 16-bit 48KHz signal]
 

 
same bass roll-off

 
 It would be great to know the capacitance of the output capacitors, open the thing up and check the capacitors. And also the output opamp (To know the model). It can easily be fixed by replacing the output caps for bigger ones. (>1000 uF).
 
It's likely that the output stage directly works with the 5 Volts of USB without inverting charge pump or other DC-DC converter (the output capacitor says a lot). A better PSU decoupuling may fixed the ripple at high frecuencies without load (and reduce hiss with sensitive headphones, if you have it in first place).
 
It's also likely that absolute phase is inverted 180 degrees, there's no problem with that. It's just very common in that type of designs (the inverted opamp draws almost no current from the 2.5V bias point in that configuration).
 
Edit: Well seems there's no dedicated output opamp, the datasheet of the DAC states that it has a integrated headphone buffer. xd

 

[Samueru Sama]

http://image.prntscr.com/image/36a5bed4120c44fb83104eccfb73e343.png
 
What?
 
Anyway xd, C1, 2, 3, 4, 5 and 6, two of those capacitors have to be in series with the headphones jack, Also it seems they added bleeding resistors (R14, R5, etc) (they reduce the "pop" when you connect the headphones).
 
Well hopefully you will try to replace them, to do so you need a multimeter, with its continuity mode you can figure out which are the capacitors and need to be changed, and also after changing the caps make sure there's no DC voltage at the output, what that capacitor does is block the Bias voltage (this case 2.5V) of the chip. Without it your headphones would die in a matter of seconds.
 
measuring the pin of the cap that "goes to the chip" (without anything playing) should give 2.5V DC. The leg that goes to the output jack should measure 0V (+-20 mV). And if you want to know if those are bleeding resistors, check if they´re connected between the output jack traces and ground.
 
If they're not bleeding resistors, you will need to add a small load (600 Ohm) before measuring the DC offset at the output. Without a load you may read a high DC offset. Actually you should do it even if there are bleeding resistors.

 

[hakuzen]

  http://image.prntscr.com/image/36a5bed4120c44fb83104eccfb73e343.png
 
What?
 
Anyway xd, C1, 2, 3, 4, 5 and 6, two of those capacitors have to be in series with the headphones jack, Also it seems they added bleeding resistors (R14, R5, etc) (they reduce the "pop" when you connect the headphones).
 
Well hopefully you will try to replace them, to do so you need a multimeter, with its continuity mode you can figure out which are the capacitors and need to be changed, and also after changing the caps make sure there's no DC voltage at the output, what that capacitor does is block the Bias voltage (this case 2.5V) of the chip. Without it your headphones would die in a matter of seconds.
 
measuring the pin of the cap that "goes to the chip" (without anything playing) should give 2.5V DC. The leg that goes to the output jack should measure 0V (+-20 mV). And if you want to know if those are bleeding resistors, check if they´re connected between the output jack traces and ground.
 
If they're not bleeding resistors, you will need to add a small load (600 Ohm) before measuring the DC offset at the output. Without a load you may read a high DC offset. Actually you should do it even if there are bleeding resistors.

i'm not going to try to fix/improve this 7€ soundcard by my own; i'll let my brother -electronic/software/audio engineer- to do it, if he likes, so shall keep all your hints for him. thank you so much!
 
i'll consider to get a good external usb soundcard, with neutral flat response both output and input, and very low output impedance, for measurement. Meanwhile, i'll keep using iBasso D14 DAC/amp as output source, and Realtek ALC888 (soundcard embedded in my PC's mainboard) as decent input (flat and extended input, some wiggle in highs though).

 

[hakuzen]

liked to show my PC soundcard measures. i use its line-in/mic for input in all measures, flat enough, although slight wiggle in the high end. output is unusable for iems (i use iBasso D14 DAC/amp for that).
 
Realtek ALC888 soundcard
 
Input test:  used iBasso DX90 as source, but also loopback test (after stating its unloaded output is valid)
Output test: Unloaded and loaded (Pai Audio MR3)
 
Input                                                                                                 Output (different scale!):
 
 
Measures (@ 1KHz, 0dBFS), using DMM:
Input impedance: 16.5KΩ.
Output impedance: phones out, 82Ω; line out, 151Ω.
Max power unloaded: phones out, 1.459Vrms; line out, 1.387Vrms.
 
 
 
After seen this, please comment if any of these upgrades can improve future measurements enough to justify the high spent:
 
1- StarTech ICUSBAUDIO2D 38€, to improve mic input; maybe a bit flatter, plus adjustable ghost power (?)
2- Creative SoundBlaster E5 external soundcard 120€ now. ZOut: 2.2Ω (not so low). Or a more suitable one (I don't know how flat is its input/output, measures?).
 
Already owning some DAC/amps, (1) looks enough, but can't find the StarTech cheaper. The soundcard is 3x more expensive (but quite more versatile/better).
 
Edit: added measures at 1KHz

 

[bartzky]

After seen this, please comment if any of these upgrades can improve future measurements enough to justify the high spent:

1- StarTech ICUSBAUDIO2D 38€, to improve mic input; maybe a bit flatter, plus adjustable ghost power (?)
2- Creative SoundBlaster E5 external soundcard 120€ now. ZOut: 2.2Ω (not so low). Or a more suitable one (I don't know how flat is its input/output, measures?).

Already owning some DAC/amps, (1) looks enough, but can't find the StarTech cheaper. The soundcard is 3x more expensive (but quite more versatile/better).

It depends on what you want to measure.
If you want to measure output impedances a multimeter and a few resistors would be more suitable than RMAA measurements. I use a UNI-T 61D for that purpose. As a side benefit you can use a DMM to volume match sources and to do RMAA measurements at known levels.

In my opinion your soundcard's input is descent enough for frequency measurements.
If you want to measure the others things like THD and so on you could buy an used EMU 0404 - at least I did so. But mind that you will need a DMM first to know at wich levels you are measuring, otherwise the results will be quite useless

 

[HiFiChris]

Sabaj Audio D1:
 
 
Headphone output, unloaded:
 


Permanent moderate roll-off in the lows that is unaffected by the load impedance.
 
 
Headphone output, loaded (Triple.Fi 10):
 

 
Very low output impedance. Apparently slightly higher in the right channel than in the left.
 

---

 
Other noteworthy things: really loud hiss, very powerful output.

 

[hakuzen]

It depends on what you want to measure.
If you want to measure output impedances a multimeter and a few resistors would be more suitable than RMAA measurements. I use a UNI-T 61D for that purpose. As a side benefit you can use a DMM to volume match sources and to do RMAA measurements at known levels.

In my opinion your soundcard's input is descent enough for frequency measurements.
If you want to measure the others things like THD and so on you could buy an used EMU 0404 - at least I did so. But mind that you will need a DMM first to know at wich levels you are measuring, otherwise the results will be quite useless

thanks!
i recently got a multimeter and also this headphone impedance test box (i'll have to figure how it works, i guess it has just some resistors and 3.5mm i/o):
https://world.taobao.com/item/41340690348.htm
i want to measure frequency response mainly, but also impedance, impulse response, and THD.
it's a hobby, so i don't want to spend too much in it now. that's why i haven't bought a DMM yet (i adjust levels by ear, lol). but i 'm taking good notes about it (DMM and E-MU 0404), for a possible future. than you so much for all the info and for confirming that the input of my soundcard is decent enough for FR measures.

 

[Samueru Sama]

 
thanks!
i recently got a multimeter and also this headphone impedance test box (i'll have to figure how it works, i guess it has just some resistors and 3.5mm i/o):
https://world.taobao.com/item/41340690348.htm
i want to measure frequency response mainly, but also impedance, impulse response, and THD.
it's a hobby, so i don't want to spend too much in it now. that's why i haven't bought a DMM yet (i adjust levels by ear, lol). but i 'm taking good notes about it (DMM and E-MU 0404), for a possible future. than you so much for all the info and for confirming that the input of my soundcard is decent enough for FR measures.

 
You can't adjust levels by ear. How do you know that whatever device that's being tested is outputting say 1 Vrms? You could in theory adjust levels with the ADC of your PC if you have a device with a known max output voltage at 0 dBFS. But my god, it is going to be a complete hassle. With a multimeter you can do it in a couple of seconds.

It seems the impedance test box is actually more than just resistors. By looking at the graphs looks like it emulates the impedance of balance armature IEMs. You will probably find an RLC circuit inside.
 
By the way here's a bit of useful links:
 
http://reference-audio-analyzer.pro/en/
https://www.seeko.co.kr/zboard4/zboard.php?id=m_device&page=1&select_arrange=headnum&desc=asc&category=&sn=off&ss=on&sc=off&keyword=&sn1=&divpage=1
http://nwavguy.blogspot.com/
http://www.innerfidelity.com/headphone-amp-measurements
http://archimago.blogspot.com/ This guys tests ADC's
http://www.innerfidelity.com/images/AllAmpsJuly2013.pdf
http://ko.goldenears.net/board/GR_Amp_DAC
http://goldenears.net/board/GR_Mobile
http://monoadc.blog64.fc2.com/blog-entry-99.html
https://www.reddit.com/r/headphones/wiki/tech/output_impedance_database
 
Edit, just wanted to share this, it is a A58M-E, it has a high output impedance (82 Ohm @ 1 KHz), and measured even higher at low frequencies (output coupling capacitors), so I gave it a try. With 32.5 Ohm load (completly resistive):
 
 

 
 
 
In case anyone wonders, the max output voltage at said load was 0.354 Vrms at 1 KHz, the limit was 0 dBFS, not clipping.