[purrin]

Actually you can one interesting pattern pattern emerge:
 
Higher scoring headphones tend to have a cleaner waterfall plot floor and faster decay (< 1ms) with smoother or smaller ridges.

 

[Focker]

Quote:

So are you saying these measurements are useless?

 
The measurements of the headphones aren't useless, no. But I don't think you can glean anymore from the polling of 30 people on their opinions of 7 headphones than you can from reading these forums. But like I said, it's always interesting to discuss this topic. 

 

[purrin]

Agreed that some samples were too small, but it's important to understand the point of this study. It's not a popularity contest with these headphones. If a headphone was rated 2.0, that doesn't mean that everyone voted 2.0 for it. There was substantial variation.
 
The point of this study is as follows:
 

• Is there correlation of certain measurement patterns to headphones which on average are subjectively rated good or bad?
• If there is, could it be possible that measurements are of use?

 
This is just the tip of the iceberg. There are a lot of measurement patterns and behaviors which I feel people would find very interesting once they understood how they correlated with the subjective experience. One of the more interesting things I've come across: the varied and polar reactions toward the T1 and DT1350. Each one measures differently! LOL I just got in a third DT1350 today. TOTALLY DIFFERENT from the other two. 
 
Measurements also help to solve huge subjective discrepancies like this!

 

[SanjiWatsuki]

Quote:

 
The measurements of the headphones aren't useless, no. But I don't think you can glean anymore from the polling of 30 people on their opinions of 7 headphones than you can from reading these forums. But like I said, it's always interesting to discuss this topic. 

 
I think it's much more easier to digest than what Head-Fi has, if only because most people don't give numeric assignments to their impressions unless they did a full review. 

 

[TMRaven]

LCD2 vs HD800 isn't a popularity contest?  

 

[Focker]

Quote:

Agreed that some samples were too small, but it's important to understand the point of this study. It's not a popularity contest with these headphones. If a headphone rated 2.0, that didn't mean that everyone voted 2.0 for it. There was substantial variation.

 
Yes, but all you really did was ask thirty people from head-fi about which headphones they like. I'm not sure how it goes beyond that. Some may like one over another because they are a fanboy whereas another saw that one headphone measured better and they want to have objective evidence to back up their opinion, etc, etc. There are just way too many variables at play here for it to really be able to say anything beyond which headphones these 30 people prefer. I do think it's an interesting topic, though, because I find both the objective and subjective points of view interesting. 

 

[ocswing]

Quote:

One of the more interesting things I've come across: the varied and polar reactions toward the T1 and DT1350. Each one measures differently! LOL I just got in a third DT1350 today. TOTALLY DIFFERENT from the other two. 

 
So is the 3rd one a good one?

 

[Focker]

Quote:

 
The point of this study is as follows:
 

• Is there correlation of certain measurement patterns to headphones which on average are subjectively rated good or bad?
• If there is, could it be possible that measurements are of use?

 
 

 
You can't really investigate that by using people who come into the study with pre-conceived notions about their headphone preferences. You could still use the head-fi population, but you'd need to randomize the trials and have some sort of isolation of variables. Your study also can't account for people who are straight up lying  

 

[SanjiWatsuki]

Quote:

 
You can't really investigate that by using people who come into the study with pre-conceived notions about their headphone preferences. You could still use the head-fi population, but you'd need to randomize the trials and have some sort of isolation of variables. Your study also can't account for people who are straight up lying  

 
I doubt there is anyone who has heard several $1000 headphones without having preconceived notions about their headphone preferences. 

 

[purrin]

Quote:

 
So is the 3rd one a good one?

 
Depends on how you like your graph to look.

 

[rhythmdevils]

Quote:

 
Yes, but all you really did was ask thirty people from head-fi about which headphones they like. I'm not sure how it goes beyond that. Some may like one over another because they are a fanboy whereas another saw that one headphone measured better and they want to have objective evidence to back up their opinion, etc, etc. There are just way too many variables at play here for it to really be able to say anything beyond which headphones these 30 people prefer. I do think it's an interesting topic, though, because I find both the objective and subjective points of view interesting. 

 
Is there any possibility that your evaluation of this study is biased because the results of the study just so happen to show your favorite headphones being less popular and measuring in a way that you do not want to see? 
 
Would you be reacting differently to the study if it showed that on average this group of experienced HF'ers preferred headphones with peaky treble and resonance throughout the spectrum?  Or if your T1 measured to be more clean and neutral? 
 

 

[Focker]

Quote:

 
I doubt there is anyone who has heard several $1000 headphones without having preconceived notions about their headphone preferences. 

 
Exactly. That's why you account for it when you design your study. 

 

[Focker]

Quote:

 
Is there any possibility that your evaluation of this study is biased because the results of the study just so happen to show your favorite headphones being less popular and measuring in a way that you do not want to see? 
 
Would you be reacting differently to the study if it showed that on average people preferred headphones with peaky treble and resonance throughout the spectrum?  Or if your T1 measured to be more clean and neutral? 
 

 
No, because I'm basing my evaluation of his study on my background in statistics, not my headphone preferences 

 

[SanjiWatsuki]

Quote:

 
Exactly. That's why you account for it when you design your study. 

Account for it how? By having people that have never heard high-end headphones listen to them? Having them go through some type of ear training program? (I think the AES actually had a study about teaching people to listen and discern differences between different audio equipment, but I'm not 100% sure that I'm remembering that correctly. Slightly unrelated note.)
 
The only people on hand that have heard these headphones at meets are otherwise are the only currently available data points. This is an informal study, not one trying to find a mathematical model to represent the subjective rating of headphones correlated with the cleanness and overall tonality of their waterfall CSD plots with 95% confidence. I'm just happy that this information is here, assuming you understand that it isn't an end-all study. I think you're reading too much into this.

 

[rhythmdevils]

Quote:

 
No, because I'm basing my evaluation of his study on my background in statistics, not my headphone preferences 

 
 
And what does your background in statistics tell you about your ability to accurately judge this study considering your personal bias towards not wanting it to be accurate?
 
What are the chances, using your understanding of statistics, that a HF'er who loves the ED8 or T1 would wind up finding fault in this study?  I'd say the chances of that happening were about 99.99%  Give or take .1%