Ideal headphone frequency reponse graph?

Quote:



A few reasons. The first is that the mix would normally be done on and for speakers, not headphones. Therefore the equal loudness curves would differ, particularly in the bass, because a headphone doesn't have the advantage of bone/body conduction and therefore needs a boost in the bass response to maintain the same apparent power.

Secondly, imagine a perfect mix made on perfectly flat monitors. Then imagine listening to that mix with your ear right up against the speaker. Clearly it would sound different. A similar thing would happen if a headphone were to "flatly" reproduce the original, and is the main reason most headphones roll-off in the high frequencies (generally above 12khz).

The third reason has to do with HRTF;

Quote:



No single headphone can restore the neutrality of the piece to everyone. A particular headphone to a particular listener can, if its FR when perpendicular to the ear and 'filtered' by the ear, happens to be close to the FR of "real" sounds far from the ear. See Smyth for a customised HRTF headphone system that is almost indistinguishable from speakers from what I've read.

Quote:



FR is one of the ways we determine direction and distance of sounds. The soundstage ofcourse is closely linked to the apparent direction and distance of instruments. Changing the FR will therefore have an effect on the soundstage. More info here.

Correct me if i am wrong please but do the frequencies not attenuate at different rates during air travel? Such you should be able to hear some frequency up close but then further away you cannot, but you can still hear some other frequency.

If so is that a reason why headphones have such specific frequency response?

Correct me if I am wrong, but are not the FR graphs you refer to just the drivers ability to accurately reproduce just one given frequency at a time, in a sweep from low to high?... not ten or twelve unique frequencies at a time, much like the music we listen to. I think that the graph should be supplimented by another graph with some kind of standard musical passage as played by a standand amplifier. Perhaps a test passage for each genre, and a few levels of amps.

Quote:



If the speakers were perfectly flat and the room was perfectly flat the response heard from the speakers and that heard from the perfectly flat headphones should be the same I would think. Sound attenuates as r^2 and for each frequency this would be true 3 cm and 3 m away. There is a frequency dependence on the attenuation due to the viscosity of the air etc but I used this calculator:

Atmospheric Sound Absorption Calculator

And got an attenuation of about 1.5 db for 20kHz @3m from the speakers (much less than most un-equalized phones vary from flat).

Anyway as far as the HRTF point goes though I had not thought of that one and it does look that having the speakers in front of you vs 90 degrees to your ears (in phones) does seem to have a fairly large bump in the frequency response. They usually equalize the room with measurement mic which doesn't get this effect but the mastering is done by a human so we'd still have that effect in there.

I don't see any real way around that though. In the end I think that most people prefer rolled off highs (that sound harsh etc) and that's why it's done not for the reason you give.

^What he said.

Etymotic claims otherwise on the reason behind rolling off highs, and makes a passing reference of studies to back them up.

Quote:



First, that's incorrect, because the effects of HRTF that occur during speaker listening would not occur with headphone listening.

Also, given the assumption of a flat room response, yes, there would be relatively little roll-off. But the room is not flat. Typical walls absorb high frequency energy (among doing other things to the sound). This is known as

Quote:



(Technical Crew Dictionary - diffuse field response)


With re. to Etymotic, I would need to see the studies to determine exactly what those graphs represent, but the text generally agrees with what I said:-
Quote:

Old thread, but important.
 
There is no ideal, as Keith Howard well states in his Stereophile 2008 article:
 
Between the Ears: the art and science of measuring headphones
http://www.stereophile.com/features/808head/index.html
 
However, there are some approximations. The following is a combined free field and diffuse field averaged frequency response measured at ear canal entrance level:
 

It's from a classic paper by Møller from 1995 and I have it lying around here somewhere. Oh, please discard the Figure title, that is not in the original paper as far as I remember.
 
Caveat emptor: if you take that and start comparing it to headphone measurements done by others, you can easily lead yourself astray. The proof is in the pudding.