The "we all hear different" thread.

Here's the image from the other thread, but flipped (like how it should be):
If the headphone measured like this on my coupler, it would measure flat on my head which is my target for now. Hence these would be my coupler targets for both headphones.

Note that 2db is a big difference. The peaks above 8kHz are just the dips from the IEM measurements and should be ignored. For a real coupler target, I would leave the area above 8kHz flat. What concerns me most is the big difference between 240Hz and 1.9kHz. This is more than 1db for the most part. I left the cursor at 500Hz where the difference is 0.8db. A 1db difference over such a big region is certainly noticeable. Also, as I said above, the 4kHz dip with the HD800 on couplers is partly an artifact, but it's interesting that the HD600 "target" also dips around 4-5kHz.

 

Those hearing tests only test for the hearing threshold, which I'm pretty sure is different from the perceived FR balance. I also think that (depending on the actual tests and methods) these tests are only a rough indicator. For example my threshold when I was a little kid was measured at about 10-20db (which is pretty bad) and when I had it measured again half a year ago the results were perfectly normal (0db or less for every frequency in both ears).

 

^ great comment.

Peoples hearing can have huge swings in difference. Some may perceive too little detail at certain frequencies while others may hear enormous amounts. Of course transducers reproduce the same waveform whatever our own "curve" is. That was my only point. It might seem like "just tonal balance preferences" but when your hearing at a particular frequency is -6dB or more below someone else's, that can certainly effect how much detail you perceive at that frequency, and the surrounding frequencies, for example (oversimplifying here) training helps a lot of course : )


Source: audiology, physical acoustics, physics, psychoacoustics, audiometry study at a college/conservatory.

 

Targets, and all the gear absolutely does make a difference! We're in total agreement here, I think. It just means that some of us can experience the same gear in very different ways. And there's nothing wrong with that.

If you're curious, There are plenty of great Audiometry resources out there.

Cheers : )

 

i think we don't all hear that much different personally. when you tune a guitar to be tuned. the results is all the same and consistent between person a to person for tuning that guitar by ear. same occurs with music and just everyday life. i think the only difference between individuals is most like due to some extent of hearing loss or something else. i even had well trained older gentlemen come to same conclusion as i have when it comes to speakers and instruments but, one thing they always mention is when i complain or mention a distress in my hearing due to a really bad peak somewhere in the upper end of top-end and air fundamentals. or when i complain bout monitors,security systems, and some cable modems and even tv's due to my sensitivity still surprisingly in the upper range above 16khz. that is the only difference i ever noticed with anyone and any ever limitations i saw in a persons overall hearing ability but, in most cases everyone i think will hear the overall same way.

 

Funny I had a similar discussion about "we all hear the same / differently" and "true accuracy" with friends just a few days ago.
Here's from an email I sent to them:

"Anyway, here are some of the graph examples I promised yesterday:

Eardrum measurements (mic at eardrum level):


Open entrance measurements (mic at ear canal entrance, canal open):


Blocked entrance measurements (mic at ear canal entrance, canal blocked):


Grey areas are standard deviation, the thick grey line is mean value. There's a good correlation of mean values among the majority of studies, but overall variance / deviation is anything but negligible imo. Blocked ear canal measurements show the least variance, but are probably farthest from what we really hear. Eardrum level measurements are likely closest to the truth (since they take the ear canal into account), but show the highest variance.

From what I've read, most studies just settle for blocked canal measurements, because they're easiest to perform and yield similar mean results to the other methods. (It probably doesn't hurt either that the variance is lowest that way ;-) I think I already showed you measurements from the Olive-Welti study some time ago and these tell a similar story in my book. Even though O-W performed only blocked canal measurements, deviations have been quite substantial too, for some headphones.

Bottom line, my personal conclusion is we all seem to gravitate towards a universal mean HRTF, but conclusively measured deviations across subjects make me believe that "true accuracy" can only be achieved on an individual basis. In other words, while I may hear roughly the same as you, our individual HRTFs will probably still differ up to several dBs here and there, and as a result our perception of "accuracy" is bound to be slightly different."

 

Essentially what I wanted to say, but you replied earlier than me.

But when you are at the point of over-thinking, then it might be okay to think about how different angles of the sound source and direct vs diffuse energy shape the FR at the eardrum. I'm sure that the FR differences between 60° speaker angle and 50° speaker angle will not be the same for every person and this could potentially also explain some preferences (in terms of FR, but not spatial representation), but I largely think that this is overthinking the issue and that it doesn't really matter.

What does matter, I think, is the differences in the FR at the eardrum for a good set of speakers versus one universally agreed upon good set of headphones for different people. Essentially the differences between speaker neutral and headphone neutral for multiple people (with headphone neutral shifting based on different interactions with the ears).
What I'm saying is that while it may be possible to have one reference room with a pair of speakers that may sound very close to neutral for everyone, the same may or may not be true for headphones. Again, like Hands said, the differences won't be huge (nearly everyone agrees that the HD6X0 are very close to neutral), but they may explain slight differences in perception. With some more exotic headphone designs the differences may end up being bigger.

 

Yes, but if we change the angle things may become a little more complicated. Still, those differences won't be huge. I've measured the differences a different angle makes with a pair of speakers at my ear canal opening (ear canal opening blocked). I'm not sure if I still have them. It does make a not-so-negligible difference and if you listen for it you can easily hear the tonal changes that even moving your head only slightly makes. The differences won't be the same for everyone. Still, as I said, this is overthinking to me. Speakers should be fairly neutral and you should position them to project an image that you like and not based on tonality, atleast that's what I think.

 

Yes, I should have been more clear. I meant changing the angle of the speakers compared to the original source. You will never place your speakers at the same exact angle that the original source was. There is no truth.

 

I just meant that a sound infront of your head reproduced perfectly by two stereo speakers will not sound the same as the original sound infront of your head produced by one sound source. If there was a perfect speaker, it would only work for reproducing sound from the exact same position where the speaker is standing.

I'm not in favor of measuring speakers with dummy heads. However there are some things that such measurements could capture more accurately according to how we hear them, like room effects. The head is not an omni mic. Problem is of course that there is no perfect compensation.

I feel like a lot of this stuff is stuff that can also be read on Changstar, just using different words for saying the same things.

 

I agree, for "ideal speakers" and relative to the original sound source, individual HRTF shouldn't matter.

(Edit: your ideal speakers would still need the same radiation characteristics as the original source, no?)

So, what's your idea of a "perfectly recorded sound" for speaker reproduction? The best approximation of true-to-source perfection I personally have come across are binaural recordings, but those are meant for headphones, not speakers.

 

This is also part of what I was getting at. Stereo reproduction/soundstage is never natural in reproduced audio. The only exception to this may be making your own binaural recordings and listening to them with IEMs. I tried that and the result was very 3D and life-like spatially.

But really, the main thing I was getting at was that most likely there couldn't be a perfect headphone for everyone, even if there could be a perfect speaker that would sound perfect to everyone. This is very similar to the different measurement setups produce different results thing. Different heads will produce different results and may interact just ever so slightly differently with different headphones. That is to say that the difference between HD650 and HD800 may not be exactly the same on different heads. These differences will most likely not exceed 5db, at least from the data that I have, and will most likely be even smaller than that.

(http://www.superbestaudiofriends.org/index.php?members/sphinxvc.201/#profile-post-2540)