• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Can we trust our ears?

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,383
Likes
24,749
Location
Alfred, NY
What is the saying about fighter aircraft? Designed by guys with PhD's, built by guys with Masters degrees, flown by guys with Bachelors degrees, and maintained by people with GED's.

Since two of my closest friends are a former fighter/bomber pilot and an Air Force maintenance engineer, and I came out of Lockheed's advanced aeronautics group, this rings painfully true.
 

cshake

Member
Forum Donor
Joined
Jul 20, 2019
Messages
37
Likes
41
Location
NY State
I was thinking about how to blind test headphone cables correctly and think there may be a way, though it requires a third cable in the system that is believed by all participants to be transparent.

Make a box with a tiny pigtail that goes to the amp, and a jack for the 'control' cable going to the actual headphones. The box has 6 test jacks: A out (1/4 TRS), B out (1/4 TRS), 2x A in (R/L with connector matching the headphone connectors), and same for 2x B in. Add two ganged 3P3T switches that force going through a non-connected state for each change. Wire it the logical way internally to put each test cable in the signal path. Plug each test cables into the jacks as loopbacks. Ensure the second person who is flipping the switch goes to neutral even if they go back to the same cable to cause switch noise and delay so as to not cue the listener. Maybe add an internal connection that bypasses the test cables in the neutral switch position for a true control test.

This is still susceptible to the level matching problem if one of the test cables is high resistance, but fixes the head positioning problem. If you really wanted to control level perfectly, you could measure DC resistance of each cable and put tiny inline resistors with the better (lower Z) cable, but anyone who goes that far probably already knows that measurements can show the difference (e.g. @solderdude's article).
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
A true audiophool and cable believer will object to the usage of extra wire and switches in the signal path.
He will instantly dismiss the test results because connectors and switches have a sound.
When they can't hear any differences in a well performed blind test they will simply deny the test results or claim this or that for their failure.

You can't convince (subjective only) audiophools. Even those that use expensive test equipment and know something about electronics. These trust their ears and findings more than their test gear and fall in classical traps as well.
Most audiophools don't believe measurements nor level matched blind tests anyway and claim that those pesky cloth eared engineer/science/design guys only perform standard tests that say nothing about how it performs with music and rig tests to trick them.
They (the engineer/science types) can't even measure differences between 2 different resistor types and cables yet they all know how different their 'house signature' is.
Even their wives can hear it.
There is no convincing audiophools

- end of rant -
 

StevenEleven

Addicted to Fun and Learning
Forum Donor
Joined
Dec 1, 2018
Messages
581
Likes
1,188
Well, as Edgar Allen Poe wrote in 1845, presciently paraphrasing the lyrics to I Heard It Through the Grapevine:

Believe nothing you hear, and only one half that you see.
 

josiah

Member
Joined
Apr 16, 2019
Messages
47
Likes
24
I just took this test with my D50s, Atom, and Sundara combination. My questions is wouldn't this test be influenced by windows mixer?

I have troubles telling the difference between Uncompressed and the hire compression. I never guessed 128kbps. am very new to this...Spotify user. Maybe, I just don't have the gift?

This did help me. At the very least, I know I can cancel Amazon Prime HD if they do not add Wasapi or Asio support. Since I have troubles telling the difference without those.
 
Last edited:

Phrangko

Member
Joined
Oct 15, 2019
Messages
57
Likes
22
In our listening room when we retire to face our audio setup our only reliable companion is our ears. Trust them.
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,004
Likes
3,998
Location
Amsterdam, The Netherlands

Wombat

Master Contributor
Joined
Nov 5, 2017
Messages
6,722
Likes
6,459
Location
Australia
"Here we go, again" takes up a large part of many threads.

A small number of well thought out and presented subject 'stickies', for quick diversion to, could promptly lessen lots of the bumph/constipated threads. o_O

These could be purposely written or refer to credible articles elsewhere. The onus on credible disagreement is then placed on the contrarian, i.e. put a case up-front or stop continually quibbling.

Of course this does not aid the desire for more forum 'hits'.
 
Last edited:

BDWoody

Chief Cat Herder
Moderator
Forum Donor
Joined
Jan 9, 2019
Messages
6,948
Likes
22,625
Location
Mid-Atlantic, USA. (Maryland)
"Here we go, again" takes up a large part of many threads.

A small number of well thought out and presented subject 'stickies', for quick diversion to, could promptly lessen lots of the bumph/constipated threads. o_O

Maybe a well done FAQ with the standard list of 'but I hear a difference' questions with solid answers including references? There are some threads with mountains of reference links, but a lot of this could be summarized to a big 'start here if you think you hear differences' button?
 

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,967
Likes
2,523
Location
Iasi, RO
I'm adding here @pma link for an interesting A/B test anyone can take:
https://www.diyaudio.com/forums/eve...be-distortion-listening-test-post5589641.html

Also, REW's ability to add harmonics up to 9th order will ease our in-house hearing tests: https://www.roomeqwizard.com/help/help_en-GB/html/siggen.html. Simply use REW's signal generator and add the distortions you might think your ears can hear; later you can increase the distortion % until you'll hear the difference.

harmonicdistortioncontrols.jpg

Testing can be done for what frequency you want and checking/unchecking the check-boxes on the fly is the easiest A/B test possible.
 

JustPoo

Member
Joined
Oct 30, 2019
Messages
67
Likes
77
Location
UK
I stumbled across that test a few months ago and consistently got the Coldplay track completely backwards, 128kbps always sounded the highest quality. My niece had the same problem. I couldn't get my head around it and while I got most of the others right, I still wasn't 100%. That was using IEM's and my phone.

Skip forward a few days and I tried it again using my full headphone setup. Only then did I 100% it every time. Thing is, it wasn't that I was picking out subtle differences between 320kbps and lossless - it was that lossless just sounded louder. I must have done the test ten times in a row and after a few tries I was clicking through them in seconds; whichever sounded loudest was the lossless file every time. I explained it away at the time as a flawed test, that they just hadn't volume matched them properly, but it seems like such a fundamental thing to get right I'm doubtful that's actually true. Unless of course they wanted to skew the results but that doesn't seem to be the case.

Forgive me if this has been raised earlier in the thread, I haven't read it all. I'm curious - if the lossless files aren't actually louder then what's going on to make me think they are?
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
I'm curious - if the lossless files aren't actually louder then what's going on to make me think they are?

If you can download the files, one could compare them easily -- although, even if they were not matched, using some kind of 'replay gain' will allow you to play them at the same volume. For the link posted by trl the files are readily provided so I made a quick comparison:

1575403303508.png


Looks close enough to me.
 

JustPoo

Member
Joined
Oct 30, 2019
Messages
67
Likes
77
Location
UK
That's very good of you, cheers for checking that. I don't understand why I can still cheat the test if the files are near as damn it the same volume. Is there any way that a higher res file can subjectively sound louder even if it's objectively not? I have only a surface understanding of how lossy codecs work, so is it ridiculous to think that with less information there's less sheer "mass" of sound?! Akin to heat vs temperature?. That sounds daft to even me now I've written it down. I'm baffled by it. Add it to the list :)
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
That's very good of you, cheers for checking that. I don't understand why I can still cheat the test if the files are near as damn it the same volume. Is there any way that a higher res file can subjectively sound louder even if it's objectively not? I have only a surface understanding of how lossy codecs work, so is it ridiculous to think that with less information there's less sheer "mass" of sound?! Akin to heat vs temperature?. That sounds daft to even me now I've written it down. I'm baffled by it. Add it to the list :)

I haven't listened very closely to those specific tracks. Maybe I'll try it later tonight. As far as I recall with my own tests 256 vs 320 vs flac original (VBR and CBR) in the past, couldn't really notice any apparent volume difference. *There's a noticeable difference sometimes, but not in the overall volume.
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
The Coldplay et al. track comparison from the original post was different from the second link where files were provided e.g. tube vs solid state. I expect there to be much more obvious perceptual differences there than the ones originally provided in the OP article and listened by JustPoo. But one can’t really download those samples, and make sure that they’re level-matched in the first place — it would be difficult to determine that his perceived change in “loudness” was not caused by a level mismatch.

*Though I think it would be unlikely that those samples are really mismatched by a significant amount to be perceivable to what amounts as an actual change in loudness.
 
Last edited:

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
That's very good of you, cheers for checking that. I don't understand why I can still cheat the test if the files are near as damn it the same volume. Is there any way that a higher res file can subjectively sound louder even if it's objectively not? I have only a surface understanding of how lossy codecs work, so is it ridiculous to think that with less information there's less sheer "mass" of sound?! Akin to heat vs temperature?. That sounds daft to even me now I've written it down. I'm baffled by it. Add it to the list :)

Okay, so I finally took a listen. Half of the time I chose 320kbps, the other half lossless. I found the classical piano piece very difficult... and I chose the 128kbps compressed sample in that one. To me, loudness was the same in all samples. Used an OL-DAC + O2 AMP + HD650 (equalized to flat).

*I have to add a complaint: the freaking player was laggy as hell. Difficult to jump and skip between samples and tracks instantaneously.
 

JustPoo

Member
Joined
Oct 30, 2019
Messages
67
Likes
77
Location
UK
I can't remember exactly what gear or headphones I was using when I first did it. The reason I mention it is because I tried to replicate the results this morning with my phone headphone out into a THX and had a more difficult time. Particularly with the three tracks from Mozart, Vega and Coldplay. With IEMs direct into the phone my results dropped again. I plugged my phone into the RME by USB and my results went up, An odd thing is that if I don't choose instinctively, by listening to the first second or so only, then my success rate plummets; ie the more I compare the tracks the more unsure I become. Three tracks are easiest to spot - Neil Young, Jay-Z and Katy Perry. I'm trying to remember exactly what equipment I was using back then, but it really shouldn't make much difference. Except the headphones, so I'm going to go through them.

If the tracks are indeed properly level matched - and I've no reason to think they're not - I can't explain what happened. I don't have special hearing, or even good hearing, I'm 40 with tinnitus and I max out about 14khz. There was clearly a reason for it, and for why I'm now failing to get it every time, but I'm drawing a blank. It could be that I saw an imaginary correlation and happened to luck out ten times in a row, but that seems statistically improbable and I'm still getting higher than you'd expect, using the same method.

Unless I can replicate my results I'm going to have to chalk it up to a blip, I wish I'd have screenshot it back then because now I'm half beginning to think it was a cheese dream :). I'm going to keep going because as the Internet wisely says "pics or it didn't happen". Until I can produce ten results and screenshots in a row of 6/6 I can't be sure I wasn't just on the end of some phenomenal luck, despite what my senses seemed to tell me.
 
Top Bottom