Head-Fi.org › Forums › Equipment Forums › Sound Science › AES 2012 paper: "Relationship between Perception and Measurement of Headphone Sound Quality"
New Posts  All Forums:Forum Nav:

AES 2012 paper: "Relationship between Perception and Measurement of Headphone Sound Quality" - Page 6

post #76 of 130
Quote:
Originally Posted by xnor View Post

Read the blog post linked on the previous page:

 

The correct order appears to be:

HP1: LCD2r2
HP2: K701
HP3: Bose QC15
HP4: K550
HP5: Beats Studio
HP6: Crossfade

 

Taking the perceived spectral balance plots into account some listeners had problems getting proper seal with the AKG cans. This would also explain how some reviewers say their K701 has enough bass ("bass-heavy" version*) and others say it is lacking bass.

*) The old "explanation" was that there are huge manufacturing variations leading to bass-heavy and light versions..

 

Yes, I think the paper shows that bass leakage was more of a problem on certain headphones than others. See slides 34 to 39.  HP2, HP4 and HP6 tended to have more variation in measured in-ear response across subjects due to headphone  fit / air leaks.  You see the same variance in listener preference for different IE monitors if they don't have a good seal. 

 

Once people are presented the same signals at their ear drums, they will agree more on what headphones they prefer. Easier said than done.


Edited by Tonmeister2008 - 4/25/13 at 2:26pm
post #77 of 130

Yeah, in case of IEMs the lack of seal can completely destroy the sound quality.

 

Btw, was the Crossfade the "LP" model?

post #78 of 130

That's a good bet since the rest are cirumaural and the M100 didn't come out until after the study. Certainly not the best showing from V-moda, even on the bassy front.

 

I was somewhat surprised by the lack of Senn's 5x8 or 6xx models as they seem to be natural competitors to the K701 and offer admirable measured performance.


Edited by anetode - 4/26/13 at 2:01pm
post #79 of 130

+1 on that.  I'd love to see how the HD 600 fares.

post #80 of 130

first thx for the very interesting paper. I stumbled upon it thx to rin choi's blog and only yesterday did I find this topic. the pdf is very clear and interesting even for an audionoob like myself.

but I got a little problem with the lcd2 to be honest. I suspect there is a lot more than just frequency response pushing people to prefer it.

I believe your way to experiment is better than just EQ 1 headphone. but I would have loved to see at least another top class headphone. or simply if you avoided planar in the middle of the rest, so we could rule out the simple fact that people preferred the best headphone they had, and really concentrate on frequencies.

 

Quote:

Originally Posted by Tonmeister2008 View Post

For both headphones the new target based on an in-room loudspeaker response was the most preferred target response curve. 

does this means a perfectly neutral signature modified only to simulate how high freqs get attenuated with distance in the air? is my understanding correct or is there more to it like reflections on the walls?

 

again if i'm not already lost in space, do you think it comes from the simple fact that we're used to hearing loudspeakers since childhood or is it really some kind of natural balance that suits a human body? (off we go to test people in some amazonian forest who never heard a loudspeaker)

post #81 of 130
Someone please kindly advise me:

In this AES "relationship......." paper, wouldn't the comfort/weight/clamping of headphones allow listeners to distinguish headphones without listening? LCD-2, K701 and K550 all feel distinct when worn, so much so that a listener would be able to tell, say LCD-2 vs K701, simply by weight instead of listening. If listener knew he was wearing LCD-2 before listening test, then he might be biased? 

Sorry I have not read the above paper, so I don't know enough about those listeners/test subjects. I would be wrong in saying " biased", if all test subjects had not worn or known of LCD-2 before that test -- that way, they would only feel "this pair is heavier" but would not know "this is the expensive LCD-2!". Otherwise it is not really a blinded test.

The virtual headphone experiment obviously removed the comfort variable and is a better blinded test. Regarding the two headphones with poor correlation coefficients (r=0.05 and 0.69), did the virtual LCD-2 vs real get r=0.69? And was the preference trend in this virtual test the same as the real headphone test? I apologize if I miss some points. Thank you in advance.
post #82 of 130

I'm sure its possible outside of head-fi to find people who have never seen in person, up close, any $1k headphone - much less worn them

 

but yes comparing different headphones by wearing, listening is always going to difficult to properly "blind"

 

I suppose you could add weights to the others...

post #83 of 130
^ Thanks for your comments. Yes, I guess one can add/normalize clamping too. However feel of pads (e.g., firm leather of LCD-2 vs cloth of K701 vs soft memory foam pleather of Bose and Beats) may be difficult to blind.
post #84 of 130

But changing the clamping force or the pads would change the frequency response. Guess you'd have to measure each stock and modified headphone on a dummy head to document the differences.

post #85 of 130
Question for Dr. Olive or any knowledgeable Headfier:

Was the rank (among listener rating of all headphones) of virtual LCD-2 remain the same as real LCD-2?

Also is distortion of different headphones considered in the virtual test? LCD-2 has significantly less bass distortion than the other 5 dynamic headphones. Did the test songs contain enough low bass frequency? And what is the audible threshold of bass distortion? 5%? 10%?

Thanks!
post #86 of 130
Quote:
Originally Posted by zzffnn View Post

Question for Dr. Olive or any knowledgeable Headfier:

Was the rank (among listener rating of all headphones) of virtual LCD-2 remain the same as real LCD-2?

Also is distortion of different headphones considered in the virtual test? LCD-2 has significantly less bass distortion than the other 5 dynamic headphones. Did the test songs contain enough low bass frequency? And what is the audible threshold of bass distortion? 5%? 10%?

Thanks!

No, in the virtual test the LCD-2 was ranked one place lower than in the standard test. We speculated this may have been due to biases in the real test (its weight and feel) but we can't be certain without repeating the standard test with the weight/feel removed as a nuisance variable.

 

We didn't include headphone nonlinear distortion in the virtual test but given the conservative playback levels and these headphones,  I doubt that this was a significant influence in the results. In cases where the nonlinear distortion is a factor the virtual method via equalization wouldn't be a method of choice.

 

The music selections did have generous low frequency content based on our program analysis.  

 

The  masked threshold of distortion in a transducer/system can be quite high depending on the spectral/temporal content of the program and the playback level. Also, the percentage of distortion is not a reliable indicator of its audibility or perceived degradation.

post #87 of 130
Quote:
Originally Posted by Tonmeister2008 View Post

The  masked threshold of distortion in a transducer/system can be quite high depending on the spectral/temporal content of the program and the playback level. Also, the percentage of distortion is not a reliable indicator of its audibility or perceived degradation.

 

Is this attributed to frequency masking, more sensitivity to distortion at particular frequency ranges...? What distortion parameters so far have been shown to best correlate to it's audibility?

 

In advance, sorry if these are too basic questions.

post #88 of 130
Quote:
Originally Posted by ultrabike View Post

Is this attributed to frequency masking, more sensitivity to distortion at particular frequency ranges...? What distortion parameters so far have been shown to best correlate to it's audibility?

In advance, sorry if these are too basic questions.

I am wondering about the same questions.
Dr. Olive, please kindly enlighten us. Thank you in advance.
Edited by zzffnn - 7/9/13 at 9:48pm
post #89 of 130
My question for Dr. Olive would be: How well would this curve agree with his own if measured in a similar manor? http://dtmblabber.blogspot.com/2013/06/finally-ultimate-frequency-curve-for.html
The first link is sort of interesting wink.gif
post #90 of 130
Anyone who has read Dr. olive's paper care to comment? Seems a cat's got his tongue. :/
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › AES 2012 paper: "Relationship between Perception and Measurement of Headphone Sound Quality"