General Headphone Audio

Raymond Lin · 2026-01-30T10:38:36+0000

I thought about getting an amp for the office, but I think I will just bring my Xduoo in. It's not like I will ever need more features than what it offers.

Angst · 2026-01-30T12:50:44+0000

Timber said:
Indeed. For example even a few weeks ago I confidently picked out individual differences between the K13 R2R and K15 I had arrive and going into details on what I was hearing. But I'm genuinely quite intrigued as to whether this is my imagination. I suppose the only way to know is to try and have a go at a blind test.

That is a slightly different story from blind testing the differences between different versions of delta sigma dacs though, the Fiio r2r dac is shown to be measurably different.

mrk · 2026-01-30T12:58:39+0000

And not all measurements are audible either, whilst some measurements that are the same between 2 devices do have audible differences such as soundstage, so the rabbit hole continues!

Which is why I said above that people shouldn't get too engrossed into the whole thing as variances and context are so broad between devices and ears that nobody will ever be in whole agreement on one particular area.

Angst · 2026-01-30T13:12:12+0000

The discussion in the linked reddit thread was about the delta sigma dacs, talking about r2r dacs in comparison is a different discussion. The point I was trying to make was that the differences between that R2R dac and delta sigma dacs are measurable and those measurements are audible. Whereas the differences people describe between different delta sigma dacs are not really measurable or are far beyond human hearing and as shown in pretty much every blind test study done arent distinguishable.
Its very easy to convince yourself you can hear differences when A/B testing or have your brain just trick you and tell you you can hear them when there arent any but the fact is they are usually so small or as said just a trick of the brain or volume mismatch that when blind testing you cant reliably pick them out.
Another thing that really needs doing is taking the amplification stage out of the mix and using the same amplifier when comparing multiple DACS.

mrk · 2026-01-30T14:13:55+0000

I have considered buying the Toping EXT90 which adds 3x additional XLR pairs to the A90 Discrete, means instant DAC switching is possible for no-delay A/B testing. It is £250 though for such a device that will only get used so little unless I fancy using multiple DACs (I do not), it's a lot of £ just for the sake of testing something I kind of do in a roundabout way already anyway (short delay but not zero delay).

At least it needs no extra power I guess...

Angst · 2026-01-30T16:40:28+0000

Heres another good blind test just been dragged up for an article.

"Audiophiles Can't Differentiate Audio Signals Sent Through Copper, Banana, and Mud in Blind Test"

Audiophiles Can't Differentiate Audio Signals Sent Through Copper, Banana, and Mud in Blind Test

The test was simple, the files were clean, and confident listeners still ended up guessing.

www.headphonesty.com

mrk · 2026-01-30T16:56:57+0000

I just came here to post that when it popped on to my google now feed

Timber · 2026-01-30T17:05:19+0000

OK so I managed to find some time this afternoon and do a blind test with the support of my other half. I don't think it was perfect but it did produce some interesting results. Apologies for the wall of text.

I used my K13 R2R, K15 and Fosi K7. I choose these because they are already out on my desk and behave the same in Windows and don't adjust Windows volume (unlike the Qudelix 5K, or Creative SB X1) . Originally I was just going to test the K7 and K15 so as to test a £180 vs a £500 box, but I thought it might be interesting to see if I could spot the K13.
I used the HD 660S2 headphones (balanced). Using 4.4mm phone out on all three DAC/AMPS.
I didn't include the TA-66, or iFi stack given we already know their distortion impacts.
Volume matching was hard. I placed my mobile phone in-between the 660S2 earpads, didn't move it and adjusted volume to get a 20.4db reading on the left/right audio test at the The Ultimate Headphones (and Earphones) Test website with each DAC/AMP.
Volumes set to: K13 R2R = 40, K15 = 30, K7 = 21. All low gain, UAC 2.0 mode, no EQ. Fosi K7 'Bypass' set to on. NOS mode on the K13. No Peace APO, or EQ software is installed on the PC. Windows Spatial Audio off. Fiio & Fosi drivers installed.
3 songs played in this order 1. Firework - Jennie Lena, 2. Get Lucky - Pharrell Williams & Daft Punk & 3. The Sidewinder Sleeps Tonite - R.E.M. All via Spotify Lossless. I do have the later two tracks on FLAC, but didn't use these.
I wore a blindfold and left the office each time the headphones were unplugged and connected.

Summary

My testing wasn't perfect, I think you need a better way of matching volume and ideally a quicker way of switching headphone out. I also don't have a more expensive DAC/AMP to test which would be interesting.
Even though I was listening for quality I also thought I would try and identify each DAC/AMP. I could only correctly identify the DAC/AMP 3 times out of 9.
For the first two songs I could only correctly identify the K7. For the final song I could only correctly guess when the K13 was being used.
What really surprised me is that I couldn't identify the K13 and it's so called distortion, or warmness. And in general struggled to tell the DAC/AMPs apart.

Eyes (or ears) opened...

mrk · 2026-01-30T17:14:00+0000

What kind of music was it out of curiosity? Different genres can favour certain types of amps depending on how they are mastered whilst others are just fine with everything. Be interesting if the same is observed with a pair of non veiled headphones as I could not tell much difference at all with the HD600 in my earlier testing but it was noticeable enough with the HE1000se due to how expansive and detailed the whole FR is on them.

This was what I meant by a whole lot of variations of things need to be factored into the overall context, it's not as black and white as it would seem at first glance.

Timber · 2026-01-30T17:15:33+0000

mrk said:
What kind of music was it out of curiosity? Different genres can favour certain types of amps depending on how they are mastered whilst others are just fine with everything. Be interesting if they same is observed with a pair of non veiled headphones as I could not tell much difference at all with the HD600 in my earlier testing but it was noticeable enough with the HE1000se due to how expansive and detailed the whole FR is on them.

This was what I meant by a whole lot of variations of things need to be factored into the overall context.

Question for me, or that Banana article from Angst ;p

(If it was me, I did list the songs).

mrk · 2026-01-30T17:17:54+0000

Huh completely missed the songs! Alright I will try something similar tonight with the same songs with the hifimans!

Timber · 2026-01-30T17:24:45+0000

mrk said:
Huh completely missed the songs! Alright I will try something similar tonight with the same songs with the hifimans!

Np.

I think what I will say is that blind testing versus A-B testing everything seems to be much less pronounced. I've said this before, but I've listened to the K13 and K15 enough to be able to describe the differences in sound. But that goes out of the window in blind testing. It's probably an exercise worth doing for anyone interested.

I did have to move the K7 to be in the same spot as the K15/13, but I suspect it would have been better to have moved everything into another room so as to have better test conditions.

Angst · 2026-01-30T20:01:08+0000

Timber said:
Np.

I think what I will say is that blind testing versus A-B testing everything seems to be much less pronounced. I've said this before, but I've listened to the K13 and K15 enough to be able to describe the differences in sound. But that goes out of the window in blind testing. It's probably an exercise worth doing for anyone interested.

I did have to move the K7 to be in the same spot as the K15/13, but I suspect it would have been better to have moved everything into another room so as to have better test conditions.

Memory for sound is notoriously terrible in humans only lasting a matter of seconds so what you are experiencing makes sense in regards to telling apart sublte differences.
Sounds like it was an experiement worth carrying out.

Timber · 2026-01-30T20:40:47+0000

Angst said:
Memory for sound is notoriously terrible in humans only lasting a matter of seconds so what you are experiencing makes sense in regards to telling apart sublte differences.
Sounds like it was an experiement worth carrying out.

I think so, very revealing. Certainly makes me want to try it again, or expand the selection, but it isn't the easiest of things to setup and find time to do.

Jenz · 2026-01-30T21:04:28+0000

I am very happy with my X9. I bought it for the single unit and features.

I’ve found a new power amplifier for @mrk to upgrade to.

BAT REX 3 Amplifier

“The world’s finest products offer a level of intimacy and involvment that sets them apart from the rest. That’s what the REX 3 delivers - a level of sonic intimacy that you won’t want to let it go if it grabs you. And you won’t find it anywhere else." -Jeff Dorgay, Tone Audio A truly different...

www.kronosav.com

mrk · 2026-01-30T22:49:50+0000

Timber said:
OK so I managed to find some time this afternoon and do a blind test with the support of my other half. I don't think it was perfect but it did produce some interesting results. Apologies for the wall of text.

I used my K13 R2R, K15 and Fosi K7. I choose these because they are already out on my desk and behave the same in Windows and don't adjust Windows volume (unlike the Qudelix 5K, or Creative SB X1) . Originally I was just going to test the K7 and K15 so as to test a £180 vs a £500 box, but I thought it might be interesting to see if I could spot the K13.

I used the HD 660S2 headphones (balanced). Using 4.4mm phone out on all three DAC/AMPS.

I didn't include the TA-66, or iFi stack given we already know their distortion impacts.

Volume matching was hard. I placed my mobile phone in-between the 660S2 earpads, didn't move it and adjusted volume to get a 20.4db reading on the left/right audio test at the The Ultimate Headphones (and Earphones) Test website with each DAC/AMP.

Volumes set to: K13 R2R = 40, K15 = 30, K7 = 21. All low gain, UAC 2.0 mode, no EQ. Fosi K7 'Bypass' set to on. NOS mode on the K13. No Peace APO, or EQ software is installed on the PC. Windows Spatial Audio off. Fiio & Fosi drivers installed.

3 songs played in this order 1. Firework - Jennie Lena, 2. Get Lucky - Pharrell Williams & Daft Punk & 3. The Sidewinder Sleeps Tonite - R.E.M. All via Spotify Lossless. I do have the later two tracks on FLAC, but didn't use these.

I wore a blindfold and left the office each time the headphones were unplugged and connected.

Summary

My testing wasn't perfect, I think you need a better way of matching volume and ideally a quicker way of switching headphone out. I also don't have a more expensive DAC/AMP to test which would be interesting.

Even though I was listening for quality I also thought I would try and identify each DAC/AMP. I could only correctly identify the DAC/AMP 3 times out of 9.

For the first two songs I could only correctly identify the K7. For the final song I could only correctly guess when the K13 was being used.

What really surprised me is that I couldn't identify the K13 and it's so called distortion, or warmness. And in general struggled to tell the DAC/AMPs apart.

Eyes (or ears) opened...

Alright so I did a similar test between HE1000se and HD600 and comparing between A90D and X8 headphone amps, volume matches as previously noted in X8 feedback, X8 in its most pure mode (bypass+NOS)

1. Firework - Jennie Lena
2. Get Lucky - Pharrell Williams & Daft Punk
3. The Sidewinder Sleeps Tonite - R.E.M

All three have mastering artefacts and on the HD600 it's impossible to tell either amp apart tbh. With the HE1000se only Get Lucky had an obvious tell in favour of the A90D to my ears. The cymbals that clap have a little bit more lift and pop out on the Discrete. Everything else between all three songs sounded about the same between both amps on HE1000se, including soundstage. I did not plug in the Warmer here because no point, it's got clear and present colouration so expected results would be obvious so left the DAC the same for this test and just swapped HPA outputs.

The one song where low end bass slam was obvious between amps is in this:

Through the entire song and again only on the HE1000se, it's a slam that I feel inside my head rather than just hear, the type of bass only a planar seems to be capable of. With the X8 the slam isn't as clear, although it's still excellent, I think even on my Arya Stealth I'd not notice this difference as the slam is something that only came after getting the HE, so I guess yeah the headphones do take centre stage after all.

I tried to find other songs that had extended low end rumble that could be felt on the HE and came across this:

It's also a really great song with a lot of instrumentation and superb mastering, well worth adding!

I will say though, like @Timber said, being blind to the source swapping does add an element of guessing error and that makes it harder because now you have no visual bias which I agree plays a roll. Though having said that, playing some of my select songs off the hifi playlist there is still some noticeable differences between both amps, the A90D leans into the upper mids more and sparkle with no means of adjusting anything, whereas the X8's HP amp leans on smooth and warm and slightly recesses the upper mids, but this is only noticeable with the HE1000se, with the HD600 both amps sound identical, and I think this is to do with the veil on the Senns and the lower layering capabilities of them vs the planars.

Also this little exercise made me appreciate the X8's HPA a bit more now too, not that I didn't like it anyway lol. And that display and the features, excellent value for what you're getting.

Jenz said:
I am very happy with my X9. I bought it for the single unit and features.

I’ve found a new power amplifier for @mrk to upgrade to.

BAT REX 3 Amplifier

“The world’s finest products offer a level of intimacy and involvment that sets them apart from the rest. That’s what the REX 3 delivers - a level of sonic intimacy that you won’t want to let it go if it grabs you. And you won’t find it anywhere else." -Jeff Dorgay, Tone Audio A truly different...

www.kronosav.com

Lol, I am not getting rid of my speakers until they start to fault, otherwise they are here for life, so a power amp with more power than what I have now would be completely wasted. I got that CA MXW70 recently remember, supposed to be more refined on paper than the PA7, yet the sound on my speakers didn't gel at all and I had to return it. A prime example of lack of system synergy with that combo.

Timber · 2026-01-31T00:11:58+0000

mrk said:
Alright so I did a similar test between HE1000se and HD600 and comparing between A90D and X8 headphone amps, volume matches as previously noted in X8 feedback, X8 in its most pure mode (bypass+NOS)

1. Firework - Jennie Lena
2. Get Lucky - Pharrell Williams & Daft Punk
3. The Sidewinder Sleeps Tonite - R.E.M

All three have mastering artefacts and on the HD600 it's impossible to tell either amp apart tbh. With the HE1000se only Get Lucky had an obvious tell in favour of the A90D to my ears. The cymbals that clap have a little bit more lift and pop out on the Discrete. Everything else between all three songs sounded about the same between both amps on HE1000se, including soundstage. I did not plug in the Warmer here because no point, it's got clear and present colouration so expected results would be obvious so left the DAC the same for this test and just swapped HPA outputs.

...

I will say though, like @Timber said, being blind to the source swapping does add an element of guessing error and that makes it harder because now you have no visual bias which I agree plays a roll. Though having said that, playing some of my select songs off the hifi playlist there is still some noticeable differences between both amps, the A90D leans into the upper mids more and sparkle with no means of adjusting anything, whereas the X8's HP amp leans on smooth and warm and slightly recesses the upper mids, but this is only noticeable with the HE1000se, with the HD600 both amps sound identical, and I think this is to do with the veil on the Senns and the lower layering capabilities of them vs the planars.

Sorry! Didn't mean to send you down the testing rabbit hole as well. Interesting to see your results though.

One thought from your findings vs. mine, or the original OP on reddit is that you are using much more expensive headphones. That OP was mentioning Edition XS, DT 1990 and DT 880 600's on their Atom and X9 which are all in a similar price range to the HD 660S2 I used. Obviously as mentioned headphones do make a bigger differences and means you can notice the differences more. With the HD 6x0's the biggest way I've improved them was putting them on a tube amp, or the EQ on the iFi Zen Signature set (which you can also do with digital EQ).