Head-Fi.org › Forums › Equipment Forums › Sound Science › Why does headphones FR graph not look neutral like a straight line like output of devices? Why is the high end have lots of peaks?
New Posts  All Forums:Forum Nav:

Why does headphones FR graph not look neutral like a straight line like output of devices? Why... - Page 2

post #16 of 24

I feel sorry for people to whom a flat response sounds unnatural. That probably means that the whole world around them sounds unbalanced, since real sound is by definition flat response. It would be particularly bothersome when listening to live acoustic music.

post #17 of 24

Home Cinema sound tracks are generally recorded with the cinema X-Curve compensation.  Most excellent surround amplifiers supplied with a microphone will calibrate the system similarly to this curve, or a variation of it.

 

A typical response of a flat speaker system when accounting for typical room reflections will typically have a response like the B&K house curve.  Commercial recordings have been recorded with this in mind for a long while now.

 

This article by Tyll at Inner Fidelity will help to clear the confusion that is being propagated:

 

http://www.innerfidelity.com/content/headphone-target-response-curve-research-update

 

I managed to dig a lot of information out from the comments section of that article also.

post #18 of 24
Quote:
Originally Posted by bigshot View Post
 

I feel sorry for people to whom a flat response sounds unnatural. That probably means that the whole world around them sounds unbalanced, since real sound is by definition flat response. It would be particularly bothersome when listening to live acoustic music.

 

Bigshot,

what baseline curve do you normally use for obtaining a flat response?

 

I'm assuming that the ideal response curve should be something close to the equal-loudness contour?

Because then the question is, a tone might be 'equally loud' as 1kHz but not 'pleasing', especially in the high frequencies.

 

So then, what constitutes a 'target response'? A combination that's 'pleasing' or 'represents reality' the best?

post #19 of 24

I don't use numbers or established curves. I calibrate using tones and my ears and recordings of acoustic instruments that I know what should sound like. The goal is representing reality.

post #20 of 24

If a pair of headphones has a perfectly flat measurement they will sound brighter than a pair of speakers with a perfectly flat response. This is because high frequencies are absorbed by air more than low frequencies. Since headphones are closer to the ear than speakers a gentle roll-off in the highs is preferable to simulate the fact that speakers are not on your ears.

post #21 of 24

I'm always referring to audibly flat, not mathematically flat.

post #22 of 24
Thread Starter 

Have you guys have done a frequency ressponse sweep of your ears?  Can be done at the audiologist if you are curious how flat yours are. lol.   I wonder if there are graphs of many people on record somewhere, I would like to see how they all differ.

post #23 of 24

What's the difference? I'm listening to my own stereo with my own ears? You can listen to your own with yours.

post #24 of 24
I think that these developments are extremely interesting - from InnerFidelity.

Across every headphone I can now see a very clear, and very accurate (for the most part) visual representation of their sound. With the previous curve you had to “know about” headphone curves to know what you were seeing. We have got used to seeing a dip between 1-2k as “neutral”, which is you think about it, makes absolutely zero sense. We have also all convinced ourselves that headphones measuring roughly flat between bass and mids means the right bass amount, when it has always been obvious from simple listening tests that this is in fact shelved down. 

Of the ones I have listened to -

Sennheiser HD600 – with a few wobbles, a largely linear set of headphones, with rolled off bass response. Exactly what I hear.

Sennheiser HD800 – slightly low-bass light, largely linear midrange with a bit of swell in the lower mids for a smooth, natural sound. That smoothness ruined by a clearly shelved up treble response, with three obnoxious peaks and drips. Exactly what I hear.

Stax SR009 – slightly bass-light, but incredibly linear through the mid-range and most of the treble. A significant “shelf up” above 10k explains the increased sense of “detail” over some headphones, and also why they can sound slightly strident on some tracks which feature plenty of information at the 10-20k region. Exactly what I hear.

Shure SE535 – slightly light on deep bass, but with a swell to the lower midrange providing a smooth, natural sound, quite seriously attenuated treble. Exactly what I hear.

Audeze LCD-2 – While I would describe from listening as having prominent mids than shelved down bass (two sides of the same coin), they measure with noticeable attenuated upper treble (providing its trade mark “rich and creamy”, or slightly dark, sound signature). The emphasis of the 10-20k region was a “penny dropping” moment for me. I have always noticed that despite attenuated treble and dark sound, detail retrieval was high, breath noticeable and cymbals crisp on the LCD2; it’s actually one of the things I like about them – how they can sound laid back but very clear at the same time. The 10-20k emphasis perfectly explains this. Again, a win for the new curve.

Koss PortaPro – Big swell to the upper bass/lower mids giving a “bassy” sound, but lacking in sub bass. A lively, prominent (but not harsh due to being lower level than the mid/bass swell) treble region with a few dips and peaks. Another home run.

AKG K712 – Lacking in sub bass, coloured mid-range and shelved up treble, getting very peaky in the 10-20k region. Sums up these headphones really.

Focal Spirit Pro – Very linear through the bass and lower mids, relaxed in the upper mids, with some added energy to the treble, and the 10-20k region. Yep.

Beyer DT880 – lacking sub bass, slight swell to the upper-bass, shelved and peaky treble with a huge spike in the 10-20k region. Again, sounds right to me. 

Grado SR60 - next to no sub bass, quite linear lower mid-range, horrifically peaky and shelved up upper-mids and treble. Bingo.

I think this is a pretty promising development in the search for neutrality in headphones.
Edited by EddieE - 3/24/14 at 1:22pm
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Why does headphones FR graph not look neutral like a straight line like output of devices? Why is the high end have lots of peaks?