How and why do members fall in love with second tier headphones?
May 6, 2015 at 5:23 PM Post #361 of 483
That's reasonable.   You can compare different products at the same table (assuming you can avoid the temptation to bring expectations about products you can see).  Walk to a different table and the likelihood that you can reliably compare the products you listened to at the previous table becomes suspect. Forget about comparing the products you listened to at the last meet.


Qualitative assessment is powerful enough but doesn't scale past the products you have in front of you.



And reading from resources can help you narrow the testing field at a Head-Fi meet.


Still I have always tried to keep an open mind and have learned to respect other people's conclusions even though they may not be my own.

At a meet is where I am always amazed at the personal side of this hobby and science.
 
May 6, 2015 at 7:03 PM Post #362 of 483
And reading from resources can help you narrow the testing field at a Head-Fi meet.


Still I have always tried to keep an open mind and have learned to respect other people's conclusions even though they may not be my own.

At a meet is where I am always amazed at the personal side of this hobby and science.

 
All true. Opinions and reviews help you prioritize the products you audition. But in the end, if you don't audition everything, you can easily miss the best product for you.  

Quantitative assessment moves us toward categorizations that are far more reliable. Assessments from large numbers of independent reviewers are correlated with every measurement and data point we can get. Innovative new measurements and standard tests are  equally valid so long as they match assessments from a sufficient number of reviewers. Audiophiles who can really hear differences have nothing to fear. Their observations are easily validated. Tube rollers and wire twisters who rely on magic fairy dust may not fare as well if sufficient numbers of reviewers do not independently support their assertions.  

Predictive analytics fuel recommender systems that gather metrics on every measured product and offer suggestions that beat the best recommendations from a single reviewer who can't help but follow their physiological and mental biases (again proven by double blind tests virtually 100% of the time). 
 
 
May 6, 2015 at 7:04 PM Post #363 of 483
 
   
Regression typically uses samples as well (how often do we have all the data?), and uses stochastic assumptions when things like confidence intervals are required. Plus there are plenty of models not labeled as "regression" that can be used to describe dependence relations among variables.

While we typically have access to more data that we used to, regression models (and related models - for those who insist on more  pedantic dissertations) need to work with data that has not been encountered before. Otherwise, why would we bother? 

Traditional sampling is different and is used more often in predictive analytics where current data points offer insight into future behaviors.
 
Regression analytics (and related analytics) will let us reliably associate human perception with quantitative measurements. The more data, the better (and less sampling - which is aways preferred) 
 
Predicative analytics is less common with audio but still has applications. We could, for example, predict that headphone X will play a particular piece of music in a pleasant way (i.e., where pleasant way is a human sentiment correlated with a complex set pof measurements) based on assessments of other genres.  

don't we already have that in quantity?
from my point of view the difficulty isn't to go from a measurement to an associated human perception, we're actually becoming pretty good at that. and I myself find it easier with each year passing and more measurements coming to me. 
what I find mighty difficult is as soon as one positive measure meets one negative measure. to be able to predict if the end result will be nice or not is so hard because the thing measured are not directly related so I can't mix pleasure estimations as if they were values with the same unit.
I don't know if I'm clear enough about that. when I look at measurements, I can tell that a specific FR will please me or feel neutral to me(not the same in my case). then I can look at some massive distortion measurements and depending on the order, amplitude, and at what frequency they show up, I usually get an idea that is pretty close to what I will hear. at least in term of "I like it".
but as soon as I'm presented with something that has the right FR I love so much, but some bad distortions I know I will notice and if so will dislike, it stops being measurements translating into human perception. instead it turns into a struggle between subjective and subjective, and I have just no idea how to deal with that only with measurements.
that's my limit.
 
but telling if something globally superior will have better fidelity, or even if I will like it better. that I could do for years.
I knew long before trying it that I would not like the hd800's signature at all. just like I knew thanks to all those measurements, that it had some pretty amazing specs for a dynamic headphone. and I have a great many examples of stuff I saw in measurements before listening to them(of course then I have to wonder if I'm not hearing what I expect, but that's another problem for another time ^_^).
 
at the moment I can make a relatively good prediction for one kind of measurement, and tell if it will most likely be below audible level, and when audible, if I will like it or not. and if it's good or not(again, not always the same thing for me).  so my problem is really to be able to put several measurements giving opposite "pleasure" signals, together into one unique subjective idea.
 
from experience I start to make some rules for myself, but that's the limit if it. me me me.  like I know that I prefer a FR I like and an early roll off in trebles(say 12khz), to a slightly less ideal(to me) FR and good trebles extension. it's obviously something very personal and I can't publicly judge a headphone based on those values. because they're mine. 
I do hope that "big data" will be able to turn all that mess into 1 subjective scaling some day. because I have no doubt that there actually are a lot less "preferred" sounds than what the audio hobby make it to be. most people just think they like something because it's the best they've had(best to their own taste), not because it's actually their favorite possible sound.
so I suspect that it is possible to rank headphones at some point(not now), and that when something like that comes up, it will do the same thing it did for speakers. all the alternative sounds will disappear and we'll at long last have our "neutral" high quality sound.
 
 
 
 
 
Quote:
The best way to learn is to bring your own equipment to a Head-Fi meet. You listen to your own music on other rigs and bring equipment into play in your own system, if you can.

it's certainly a fun thing to do, and it does clear the "I like" / "I don't like" once and for all for all for many headphones.
but looking at a meet in a practical way, it's really not a good listening process. you always have a lot of noise, you can't really hope to AB anything efficiently unless a guy organized the test for you and matched levels beforehand.
to me meetings are a great discovery moment(and fun!!!), but a really poor judge of audio quality IMO.
 
May 6, 2015 at 7:25 PM Post #364 of 483
 
How do you know your system is closer to your goal?  Or that now is closer than it used to me?  Or that it is closer after the next upgrade?

 
I identify specific limitations and bottlenecks in my system through critical listening, research the science behind the cause of the problems, then research the most effective way of addressing them. Once I know what the problem is and how to address it, I apply what I know to correcting the problem. Then I revisit my critical listening to see if the specific limitations and bottlenecks are dealt with.
 
Too many people never do critical controlled listening. They rely on subjective impressions and just assume that a more expensive piece of equipment or one with better specs will automatically improve sound quality. They randomly swap equipment in without ever defining the problem that needs a solution. That kind of random flailing about never gets anybody any closer to their goals.
 
May 6, 2015 at 7:28 PM Post #365 of 483
 
At the same time, well developed, repeatable, objective testing should be something that Sound Science focuses on, not just tuning systems by ear.


I was talking about the purpose of the testing, not testing itself. You need to define what it is you are testing for, and there has to be a goal in mind for how to apply the test results to real world improvements in sound quality. It isn't science for its own sake... it's applied science.
 
May 6, 2015 at 7:40 PM Post #366 of 483
Well, all this boils down to IS the community IS missing a grouping graph for headphones.

They could just have it by price a graph charts. The closest thing on Head Fi is the thread and chart for Headphones for Metal Music

There are statistical measurements and graphs for almost everything but it is somehow missing here. Even a rudimentary graph showing bass headphone, mid-centric headphones and high treble detail headphones along with price factors.

A new person is jumping into an ocean with 400 unidentifiable fish swimming that takes years to understand.
 
May 6, 2015 at 7:52 PM Post #367 of 483



http://www.head-fi.org/t/715478/headphones-for-metal-music-ultimate-solution




You go to your cheaper fast food restaurant and they give you a combo that all goes together. You go to a fancy place and they suggest a group of food and maybe even wine that goes together well.

The graph is the little request but still new members would gain from complete system choices with price options in order? Why not? They do it in most industries. Here it it almost confusing by construction. Most stores would love for new members to flounder buying wrong headphone, wrong amp forever and ever.:dizzy_face:
 
May 6, 2015 at 8:12 PM Post #368 of 483
http://www.headphone.com/pages/build-a-graph



Here you build a graph.
 
May 6, 2015 at 8:28 PM Post #369 of 483
http://www.headphone.com/pages/build-a-graph



Here you build a graph.



So then the argument is the complete system then changes the headphone sound. I am in total belief of the fact. Warm amp warmer headphone sound, cold analyst sound, colder headphones!


So groups of source and amps along with headphones can be tested together and charted together as a whole, just placed in groups by sound signature and price.
 
May 6, 2015 at 9:40 PM Post #370 of 483
.......and as a result discount the performance of superior models?


1: the onus is on you to prove that people fall in love with second tier headphones.

2: that they do so even after experiencing something they acknowledge to be better.

3: that there is indeed something they "ought" to like more.

4: assuming you achieve the above, now you have to prove two further things: that the same people actually do discount "superior models" (superior according to who?) and also that this is as a result of said romantic entanglement.

The whole question is poorly considered, is itself loaded with assumptions and biases, is badly formed and overall smells like a newly deposited jobbie.

Even if giving the questioner the benefit of many doubts in his assumptions, presumptions and biases there still remains the biggest ommission: maybe it's possible that people who buy "second tier" or "unapproved by redcarmoose" headphones just buy the stuff that they can afford, is available locally, and suits them, not the stuff that is available to try at a store hundreds or thousands of miles away but which suits some ghastly snob they never met?

Maybe fabulously expensive headphones only sound better if you keep your nose pointing resolutely skywards? Discuss.
 
May 6, 2015 at 10:22 PM Post #371 of 483
 
http://www.headphone.com/pages/build-a-graph



Here you build a graph.



So then the argument is the complete system then changes the headphone sound. I am in total belief of the fact. Warm amp warmer headphone sound, cold analyst sound, colder headphones!


So groups of source and amps along with headphones can be tested together and charted together as a whole, just placed in groups by sound signature and price.


as I find FR to be the main factor in me loving a sound or not(not the factor but clearly above all else), I would be up for something making a classification by signature. but I'm afraid it would still be very complicated to implement right now. let's take 2 headphones with pretty much the same signature but one with jerky graph and the other with a smooth one. should they be equal?
then if one is electrically flat, and another is V shaped with as much bass as trebles, are they both placed at the same level? or do we need to create a special category from warm to bright inside the V shaped category itself?
it's really a complicated thing to organize clearly. and that would only sort 1 parameter. it's scary the load of work that classifying headphones would demand to be done right.
 
about sources, to me it's easy to see anything colored as a deviant. if we start with the opinion that the highest quality is high fidelity, anything that isn't transparent fails to pass the requirement and should be kept in a category of it's own.
as for transparent sources, they will sound the same so measurements only are enough to try and sort them by rank, but it really doesn't matter much.
and if we don't go with the idea that high fidelity is the main criteria, that good for me, we can go by warmth or how non aggressive and pleasant they are or whatever other choice. but then a hd800 has nothing to do in the top tier of the headphone list. I also don't mind, but we should a least keep the same set of values for sources and headphones.
 
 
maybe you should just make a list of what you see as top tier, so that we can discuss why one headphone should or shouldn't have its place in the A team. instead of talking about abstract things that don't have the same meaning for everybody.
HD800? T1? DT880? LCD2?3?X? abyss? HE-4?5?6? I guess at least a few stax? k701? hd600?650? just some "famous" stuff on the top of my head. what is in the "top tier"?
 
May 7, 2015 at 12:15 AM Post #372 of 483
The graph is the little request but still new members would gain from complete system choices with price options in order?

 
That graph might actually be useful if the ratings were based on real measurements.
 
May 7, 2015 at 12:19 AM Post #373 of 483
1: the onus is on you to prove that people fall in love with second tier headphones.

 
You missed my earlier post where I mentioned that a friend of mine got a set of Sennheiser HD595s. I own Oppo PM-1s, but with five minutes of playing with an equalizer, I realized that I could get to pretty much the same place with headphones that cost a fraction of the price. I'm not sure whether that's because the Oppos are overpriced or the Senns are underpriced, but the bang for the buck ratio there is WAY out of whack. With a little more effort to EQ, people could save about $750.
 
May 7, 2015 at 12:20 AM Post #374 of 483
  as I find FR to be the main factor in me loving a sound or not(not the factor but clearly above all else), I would be up for something making a classification by signature. but I'm afraid it would still be very complicated to implement right now. let's take 2 headphones with pretty much the same signature but one with jerky graph and the other with a smooth one. should they be equal?

 
It also depends on where the jerky stuff occurs... at 3kHz it could be a huge issue, at 15kHz and above, not so much.
 
May 7, 2015 at 3:23 AM Post #375 of 483



It is really a big project as in this graph you can see he rated headphones in terms of price and we have 4 groups. He then rates how they perform in each sub-genre of metal, due to the different genres of metal requiring different response from each model of headphone. Still with in that there is a rating system which shows just how good each headphone is in the sub-genre.


So to try and start to choose first tier and second tier you have maybe to know the genre, as some headphones are all around good and some only have a use in excelling at a genre. Take AKG k701s as they are known to measure flat, have a following still in classical circles but have fallen from a top flagship to 2nd tier. And again it IS a matter of opinion?:p


I am also not sure if headphones sound the same at all volume levels. Each person has a different volume setting and I still believe there is resonance that changes what a graph would point out at a single volume level.

Again rev a car and put your hand on the hood. That vibration changes frequency response, I would think? This factor would be interesting if there was a choice to slot the headphones into cataloging which is flat and accurate in relation to headphones that simply measure bad?
 

Users who are viewing this thread

Back
Top