[spinali]

As an illustration: I have a pair of K701s that claim a frequency of range of 10-39,000. I've had DT880s that claim 5-35,000.

If human hearing has a considerably narrower range, what is the value of such extravagant figures? How much of this figure will be "overtones" with relevance only for ad copy? Or does this really have an effect on the subtleties of sound?

None of the IEMs I've tried seem to exhibit overtones - even though I'm not supposed to notice the difference. Is this more than a "golden ear" phenomenon?

 

[kramer5150]

frequency reponse specs are meaningless without a bd-down point. So yes its entirely possible a K701 can reporoduce 5hz. It may be 30db down from the next adjacent octave however... and thus practically non-existent. Yes it may also consist purely of harmonic resonances and overtones, frequency response specs do not distinguish between fundamental tones and resonant overtones.

There really is no value/use to the end user. Companies publish such specs to generate revenue and keep up with the rest of the market. There is no governing body to regulate the publication so manufacturers are free to publish whatever specs they desire.

 

[Duggeh]

There are also several headphones that claim 10-100,000. Personally I think its more to do with being able to claim that under certain conditions that the driver will be capable of such a measurement for the purposes of marketing / looking impressive on the box than it is to do with ultimate sound *performance*. However a driver with extreme frequency response shoudl be abel to better resolve when given SACD or DVD-A to reproduce, whie the human ear only hears roughly 20-20k, the subsonic and supersonic ranges also have an effect upon our percetpion of the sound.

Of course, you wont get these frequencies out of a CD, and whatever the analogue purists might claim, youll never get them properly out of an LP either. So IMO, the best thing to do is to note them, then pay them no heed at all. Certainly people shouldnt buy headphones based upon their frequency range claims, otherwise everyone would own HE90s, Qualias and H2s.

 

[b0dhi]

Quote:

Originally Posted by spinali /img/forum/go_quote.gif None of the IEMs I've tried seem to exhibit overtones - even though I'm not supposed to notice the difference. Is this more than a "golden ear" phenomenon?

If by overtones you mean harmonics, then you would've heard the harmonics produced by the Etymotics. They're well into the audible range (almost -40dB), and it appears, at least by HeadRoom's measurements, that the whole ER series has the same response.

Compare this to the Shure E5-C and the HD-650, both of which have relatively low THD:

 

[ericj]

On top of the usual issues with frequency response measurements, the measured response curves are different if you've got the headphone hanging in mid air vs. up against a flat surface vs. up against a human ear vs. up against something like my avatar that just pretends to maybe be similar to a human head.

It's all lies, one way or another.

 

[b0dhi]

Quote:

Originally Posted by ericj /img/forum/go_quote.gif On top of the usual issues with frequency response measurements, the measured response curves are different if you've got the headphone hanging in mid air vs. up against a flat surface vs. up against a human ear vs. up against something like my avatar that just pretends to maybe be similar to a human head. It's all lies, one way or another.

Those measurements aren't exactly FR measurements. The harmonic distortion of a driver isn't significantly affected by different ears/heads. More importantly though, those measurements were made with the same dummy head and equipment, so the comparison is valid. The differences, as you can see, are huge.

 

[tbonner1]

The range is more useful if it includes + or - db specifications.

 

[Canman]

Most manufacturers quote a wider bandwidth as the model in the range gets higher. cough Grado cough. A layman in a store looking at two different headphones in a box will say "what's the difference?" and "why is this headphone $100 more?" Of course, the expensive one goes to 35 kHz and the cheaper one only goes to 30 kHz. 5 more kHz! It's like saying "but this one goes to 11!"

 

[qazwsx]

Frequency responses tell little to nothing about how a headphone will sound. And most of the time, they are inaccurately done.

For a short example, a headphone can be rated at 5-30,000Hz (like the Sony V700) and have numerous peaks and valleys throughout the response, while a headphone at 20-20,000Hz can be 100% audiophile grade, yet I've seen some dollar store earbuds boasting that range. It's how a headphone performs in that range that gives it its sound quality.

 

[enjoiflobees]

If a headphone can respond at such low and high freq that will do something to the sound even if you can not hear at those freq. An easy example is bass. If a headphone can respond to low bass you can "feel it" which would help in the SQ of the headphone right?

 

[qazwsx]

Quote:

Originally Posted by enjoiflobees /img/forum/go_quote.gif If a headphone can respond at such low and high freq that will do something to the sound even if you can not hear at those freq. An easy example is bass. If a headphone can respond to low bass you can "feel it" which would help in the SQ of the headphone right?

It can show the amount of extension, and that is only if it is measured at +/- 3db.

And even if a headphone boasts a good range, it can still sound very crappy. A range only shows were the range begins and where it ends. It doesn't show what's in between, which could be extremely crappy. (Again why extremely cheap 'phones even boast as well as 20-20,000Hz.)

 

[enjoiflobees]

I agree. It should be something you look at but defiantly not something that make you buy a headphone.

 

[musicmind]

hi guys

Some valid points raised here already, esp. about corresponding -3db or -6db annotations being required for any such frequency specifications.
Obviously its common knowledge that our hearing doesnt extend above 20Khz, and many are still unconvinced about whether the >20khz frequencies have any tangible effect on our hearing. I'll leave that to the experts.

I tend to view those frenquency specs from a band-pass filter point of view, in that the wider the bandwidth of the device, the less it is going to be affected by the ripples and non-linearities that occur as you approach the cut-off frequency. So a speaker/amplifier "over-designed" to 35khz for example, may be more linear in its audidble range frequency response, than one with a 20khz cut-off, where we might start getting some ripple effects as we approach the higher frequencies.
I'd be interested to hear your views on this.

Thanks, cheers.

 

[Zenja]

Let me clarify a bit. Normally when you see frequency response specs, they're completely meaningless because it says nothing about what level of output you're getting at those lower and upper frequencies. For example, if a headphone is 20Hz-20kHz, it could be at like -30dB at 20kHz and you wouldn't even be able to hear it, but yet they'll publish it as that anyway.

What IS meaningful is when they say something like 20-20000Hz @ +/- 3dB. This truly means that they'll have a frequency response to that level. It means that even at those extremes at 20Hz and 20000Hz, you'll get output that's only 3dB quieter. I don't know what that's with respect to though. It could be -3dB compared to the average levels in between, or peak, or something else.

 

[vvs_75]

Quote:

Originally Posted by qazwsx /img/forum/go_quote.gif Frequency responses tell little to nothing about how a headphone will sound. And most of the time, they are inaccurately done. For a short example, a headphone can be rated at 5-30,000Hz (like the Sony V700) and have numerous peaks and valleys throughout the response, while a headphone at 20-20,000Hz can be 100% audiophile grade, yet I've seen some dollar store earbuds boasting that range. It's how a headphone performs in that range that gives it its sound quality.

Yep! And incontestable proof of it - Sony R-10! (20-20KHz)