• 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!

"Bias" of some members towards headphone measurements?

I think we need the moderators, @AdamG , @BDWoody , @Jimbob54 , @RickS and even (perhaps) @amirm to provide us with guidance in this matter. I wholeheartedly agree that we not make their job more difficult. Still, doesn't some bit of accountability seem in order?

Thrashing about in the dark isn't the way to resolve the issue.
We are trying to keep up. We also try to stay out of the way and let the members resolve their issues and differences.
 
Using only rock music in a study protocol limits the generalizability of the study findings to rock music. It doesn't "invalidate the research."
So you've asserted, but presented no evidence.
 
The refusal to synthesize with the rest of the audio community at large is a bit sad.

If "to synthesize with" means to tolerate and accept the widely shared falsehoods and anti-scientific sentiment of the "audio community at large", then thanks, but no thanks. Don't feel sad about it at all. Not a bit.
 
The refusal to synthesize with the rest of the audio community at large is a bit sad. I dont mean to not stand ones ground, but hey! Painting oneself into a corner with this.
Large swaths of the audio community are openly hostile to science, so who's refusing to synthesize with who?

Head-fi literally bans science outside of a specific sub-forum.

Audiogon is full of people who are convinced audio over ethernet is different than other types of data over ethernet, who think SINAD is a ridiculous metric but think uncontrolled listening is better, "bits aren't bits", etc.

Youtube is largely populated by the same ideas.

The problems in the audio industry and audiophile community are widespread, obvious, and not new. It all goes back to the audio "press" realizing they couldn't make a living (ad revenue) on rigorous measurements and blind testing, and instead hitching their wagons to subjective uncontrolled listening as the primary means of evaluating gear.

If you realize this and understand the science even superficially, you're not going to want to "synthesize" with points of view that fail to. This is meaner than it needs to be, but I don't want to synthesize my understanding of gravity with flat earthers' either.

It's hard to have productive discussions with people who can't or won't acknowledge the existence of cognitive effects on hearing, or how electronics actually work, for example.
 
Last edited:
...
I think one should start from what the headphone is supposed to do. German academic Günther Theile argued, that the in-ear sound, delivered by a headphone, should closely resemble the in-ear sound that is generated by a diffuse field.

A head and torso simulator was designed to measure the in-ear sound generated by a diffuse field. To test a headphone it is seated on that HATS, and the same in-ear sound is expexcted, as is present with the diffuse field.
...
Multiple studies since then have shown diffuse field (DF) target is not the most preferred. We don't listen to music nor most other things in reverb chambers. See:
target_curves_ratings.png
 
The right measurements, when analyzed by a computer, can account for 74% of the variation in listener preferences, according to Harman research. 74% is not 100%. Furthermore, the predictive power would be predicted to be less than 74%, when the study conditions are not replicated (i.e. source material is not rock music, speaker to be tested is not one of the 70 tested, etc.).
This is good to point out, I guess my response is just that I'd like to also talk about the remaining 26%, whether it's all down to personal preference, attributes of speakers not accounted for in the model, shortcomings in rock music, or something else. I don't think that the Harman research is the final word, in science nothing is really supposed to be final, but where would we take the next step in improving it?
 
So you've asserted, but presented no evidence.
Correct, I have asserted that a research study performed with a specific protocol (in this case, rock music) is not necessarily generalizable to situations different from that protocol. I follow basic principles of science and research interpretation.
 
I've had to increasingly bite my tongue over the last 2 or 3 years.
What did you expect, a site composed of 100% logical Vulcans?
Live Long and Prosper. ;)
 
This is good to point out, I guess my response is just that I'd like to also talk about the remaining 26%, whether it's all down to personal preference, attributes of speakers not accounted for in the model, shortcomings in rock music, or something else. I don't think that the Harman research is the final word, in science nothing is really supposed to be final, but where would we take the next step in improving it?
My personal takeaway is that if Harman researchers, using computerized regression analysis, precision measurement capabilities, and a suite of spinorama data, can only explain 74% of the variation in blind listener preferences, that tells us that maybe, just maybe, the average forum participant looking at a 2D FR chart for 5 secs is probably going to do a lot worse than 74% in predicting how good ("preferred") a loudspeaker will sound.
 
Correct, I have asserted that a research study performed with a specific protocol (in this case, rock music) is not necessarily generalizable to situations different from that protocol. I follow basic principles of science and research interpretation.
If what you said is the case, where did Harman get their data for this?
Program Influence on Listener Performance.png
 
But while the ASR community is partly a bastion against snake oil and subjectivism, I think some ASR users tend to overrate the meaningfulness of measurable data when it comes to headphones.
A genius move on the side of Harman in particular was to acknowledge, that there are no standards, and to go for 'preference' instead. That made a new standard for speaker measurements, but still it is preference based, hence allows for some wiggle room.
With headphones things are messed up a bit. The engineering (not science by the way) lacks self-reflection, stating a standard to soon, me thinks. Some additional work is on the way, as far as I've heard.

But people take stereo etc way to serious, spending to much money. One person's vinyl pickup cartridge is the other person's house ... nearly. That adds onto the difficulty of 'preference--what is it? In case I *love* the music, the stereo isn't that important, in case not, why that much effort put into the stereo?

That said, when it comes to me personally, if cost is kept at bay, I really love good engineering as such, as a piece of Fine Arts. Stereo delivers on that at a moderate price, see above. The powerful yet smooth engine in my car is way more expensive.

ps: essentially, 'preference' is not understood ;-)
 
maybe, the average forum participant looking at a 2D FR chart for 5 secs is probably going to do a lot worse than 74% in predicting how good ("preferred") a loudspeaker will sound.
Well, the Harman numbers in this context are predicting how preferred a given speaker is going to be among a group of people, not what the speaker sounds like.

If you show me a batman frequency response curve, or one missing bass, or one with elevated treble, I can tell you something about how the speaker will sound. That's a different exercise than predicting whether people would like the speaker. I'm not actually aware of any research that directly shows how well people can predict sound quality from reading graphs, but it would be interesting.
 
Multiple studies since then have shown diffuse field (DF) target is not the most preferred.
Thanks, I should have known that. Otherwise, my comment on headphone evaluation was a bit misplaced anyway.
 
If "to synthesize with" means to tolerate and accept the widely shared falsehoods and anti-scientific sentiment of the "audio community at large", then thanks, but no thanks. Don't feel sad about it at all. Not a bit.
One just has to slide into one ear and out the other.
 
Large swaths of the audio community are openly hostile to science, so who's refusing to synthesize with who?

Head-fi literally bans science outside of a specific sub-forum.

Audiogon is full of people who are convinced audio over ethernet is different than other types of data over ethernet, who think SINAD is a ridiculous metric but think uncontrolled listening is better, "bits aren't bits", etc.

Youtube is largely populated by the same ideas.

The problems in the audio industry and audiophile community are widespread, obvious, and not new. It all goes back to the audio "press" realizing they couldn't make a living (ad revenue) on rigorous measurements and blind testing, and instead hitching their wagons to subjective uncontrolled listening as the primary means of evaluating gear.

If you realize this and understand the science even superficially, you're not going to want to "synthesize" with points of view that fail to. This is meaner than it needs to be, but I don't want to synthesize my understanding of gravity with flat earthers' either.

It's hard to have productive discussions with people who can't or won't acknowledge the existence of cognitive effects on hearing, or how electronics actually work, for example.
There are Ok ppl at all those sites you listed. Just as some ppl here doesnt stare themselves blind!
 
There are Ok ppl at all those sites you listed. Just as some ppl here doesnt stare themselves blind!
I agree, it's not all terrible on the other sites and the people are generally nice unless you start asking them about blind tests. I don't think they're bad people, they've just often bought into unsupportable theories of sound quality, or treat it as mostly unquantifiable, which is not interesting to me.

For me it's not really that deep. I enjoy discussing audio on the internet for some odd reason, but on other forums if I call out nonsense I would get attacked, or ignored. And that's not as fun as being here and calling out nonsense with other people who see it for what it is.

ASR is not the only forum that's science-heavy and hostile to nonsense, (hydrogenaudio, *arguably* gearspace or stereonet are too, DIYaudio is at least not entirely hostile to skeptics) it's just the most famous audio forum of that type in the present day.
 
OK folks. Let’s do less Crass and snark here. Be respectful and treat others with dignity and kindness. Maybe we should consider methods of communication that will get the nonbelievers attention? We are outnumbered and alienating them is not going to improve our situation. More and more people are joining ASR. We could look at this as an opportunity to spread the awareness and knowledge of Audio Science and grow our community. But we have to present our thoughts and opinions in a manner that is not off putting and talking down to the “Flat Earthers” to borrow a term just used to differentiate between the enlightened and the cave dwellers.

Let’s look for ways to win them over. Not chase them off and concrete their minds that Science Bad. Sound/Listen Good! :p

Suit up…Did I mention that today is Taco Tuesday?
 
It makes sense as it is ASR, Audio SCIENCE review. To meet evidence with evidence. Well everybody who´s a scientist, put your arms in the air. To work in a laboratory is an actual education. So where does that leave us?
I think parts of ASR has been taken over by paranoia and hostilty. I emphazise parts of since ASR is pretty diverse with many threads and sections. The refusal to synthesize with the rest of the audio community at large is a bit sad. I dont mean to not stand ones ground, but hey! Painting oneself into a corner with this.


as in thesis-antithesis-synthesis?
 
Large swaths of the audio community are openly hostile to science, so who's refusing to synthesize with who?

Head-fi literally bans science outside of a specific sub-forum.

Audiogon is full of people who are convinced audio over ethernet is different than other types of data over ethernet, who think SINAD is a ridiculous metric but think uncontrolled listening is better, "bits aren't bits", etc.

The Steve Hoffman forum bans discussion of blind testing.

On the other side, Hydrogenaudio has done just fine for decades while 'banning' otherwise unsupported subjectivist blather as violation of its TOS.


'meeting somewhere in the middle' on matters of fact isn't generally how science works
 
Back
Top Bottom