[fburdelliv]

Hey, I can't find this anywhere on the internet despite much Wiki lurking. 
What am I exactly looking at when I look at frequency response charts and what are they suppose to look like? And what does a good headphone look like? Or is it variable for what you need per headphone?

Thanks All

 

[Shike]

Depends.  I go by flat measured with DF equalization (ala Headroom).
 
Some will say treat headphones like instruments and others will say seek accurate reproduction of the signal at all costs.  I'm in the latter group.

 

[nick_charles]

Quote:

Hey, I can't find this anywhere on the internet despite much Wiki lurking. 
What am I exactly looking at when I look at frequency response charts and what are they suppose to look like? And what does a good headphone look like? Or is it variable for what you need per headphone?

Thanks All

You are looking at the way in which the item responds to differing input frequencies. A flat frequency response means that the item responds the same to all frequencies i.e it does not boost or attenuate any range of frequencies more than any other.
 
This is normally deemed desirable in the signal chain up to the transducer stage (headphones/speakers)  at this point opinions vary markedly and the topology of the ear means that a flat FR is not necessarily a good thing. But an amplifier or DAC or CD player that had the FR of even the flattest headphones or speakers available would be pretty poor.

 

[WrxSTI]

Hmm. I've actually read a lot of people here claim that neutral response is not "natural". But why? Let's take for example this scenario;
 
You have a live instrument playing, let's say a guitar, or a piano. You are in the same room, the sound is as natural as it is ever going to get.
You use a microphone with a theoretically perfectly flat frequency response, it will capture all of the frequencies of the instrument just as they are without boosting or reducing any of the frequencies > sound still stays natural.
The sound is then converted with a theoretically perfectly flat ADC into digital format and onto a CD > sound stays natural.
You put this CD into your CD player whose perfectly flat DAC sends the now analog signal into your perfectly flat headphone amplifier and the sound at this point is still exactly unchanged, i.e. as natural as the real thing.
You hear the sound through your perfectly neutral headphones and in essence you are hearing the exact same natural sound that the instrument has played.
 
Needless to say this is all ideal-case scenario. But what exactly is the argument against neutral headphones here? You mention the construction of the ear (which, yes, is responsible for everyone's different hearing). But, like I said, provided you were in the same room,
the "sound chain" would basically be: Instrument > your ears. Whatever frequency imperfections your ear design has would be the only thing changing the sound. When listening to recorded music, if all your other gear is neutral, then again only your ears are changing the sound. However, when listening through substantially colored headphones you are in essence adding another layer of frequency imperfections so how can this be any more natural?
 
I'd really like to know. Hope I'm getting this right

 

[chinesekiwi]

Well, you need to know a little bit of psychoaucoustics.
 
1) As canalphones and in-ear monitors bypass the pinna, you cannot fairly compare the FR responses of them vs. headphones.
 
headphones are a special case as the pinna affects the frequency response. a balanced frequency response is not flat on headphones.
As stated because of the pinna, there has to be a significant bump at the 10k mark or else you get recessed treble, particularly noticeably with drum cymbals.
Also you are looking (for balanced sounding headphones) at it's relative line.
i.e. How much does the frequencies deviate from one another.
 
e.g.
 

 
From here, you can tell the AD700 is not only bass light, but also quite unbalanced sounding as the gap between the lwoer frequencies and other frequencies is quite high while the DT880, known to sound balanced, has a relatively low deviation. With headphones, due to the pinna, there has to be about a +2-3dB in the bass region or else it will sound abss recessed. In fact even the DT880/600 shows this as it is a tad sub-bass light and the FR line proves this.

 

[SP Wild]

Quote:

Hmm. I've actually read a lot of people here claim that neutral response is not "natural". But why? Let's take for example this scenario;
 
You have a live instrument playing, let's say a guitar, or a piano. You are in the same room, the sound is as natural as it is ever going to get.
You use a microphone with a theoretically perfectly flat frequency response, it will capture all of the frequencies of the instrument just as they are without boosting or reducing any of the frequencies > sound still stays natural.
The sound is then converted with a theoretically perfectly flat ADC into digital format and onto a CD > sound stays natural.
You put this CD into your CD player whose perfectly flat DAC sends the now analog signal into your perfectly flat headphone amplifier and the sound at this point is still exactly unchanged, i.e. as natural as the real thing.
You hear the sound through your perfectly neutral headphones and in essence you are hearing the exact same natural sound that the instrument has played.
 
Needless to say this is all ideal-case scenario. But what exactly is the argument against neutral headphones here? You mention the construction of the ear (which, yes, is responsible for everyone's different hearing). But, like I said, provided you were in the same room,
the "sound chain" would basically be: Instrument > your ears. Whatever frequency imperfections your ear design has would be the only thing changing the sound. When listening to recorded music, if all your other gear is neutral, then again only your ears are changing the sound. However, when listening through substantially colored headphones you are in essence adding another layer of frequency imperfections so how can this be any more natural?
 
I'd really like to know. Hope I'm getting this right

 
Putting your head inches away from a guitar string, mouth etc would sound very unnatural, if you want my opinion.

 

[ClieOS]

Quote:

...
I'd really like to know. Hope I'm getting this right

It is not just the ears to consider, but the whole head and upper torso. It is called the HRTF, or Head Related Transfer Functions. Here are something to read: wiki and Headwize.

 

[Shike]

Quote:

Well, you need to know a little bit of psychoaucoustics.
 
1) As canalphones and in-ear monitors bypass the pinna, you cannot fairly compare the FR responses of them vs. headphones.
 
headphones are a special case as the pinna affects the frequency response. a balanced frequency response is not flat on headphones.
As stated because of the pinna, there has to be a significant bump at the 10k mark or else you get recessed treble, particularly noticeably with drum cymbals.
Also you are looking (for balanced sounding headphones) at it's relative line.
i.e. How much does the frequencies deviate from one another.
 
e.g.
 

 
From here, you can tell the AD700 is not only bass light, but also quite unbalanced sounding as the gap between the lwoer frequencies and other frequencies is quite high while the DT880, known to sound balanced, has a relatively low deviation. With headphones, due to the pinna, there has to be about a +2-3dB in the bass region or else it will sound abss recessed. In fact even the DT880/600 shows this as it is a tad sub-bass light and the FR line proves this.

Isn't that what DF equalization is for?  The FR graphs you access on HeadRoom be default are DF equalized, only if you look at the uncompensated graphs are you getting the true raw measurements . . .

 

[xnor]

But DF equalization is only a rough estimate. And then there is the problem with bass impact of headphones. => Different strokes for different folks.
 
Anyway, if you ignore the balance between bass, mids and treble I'd look out for an overall smooth frequency response (no excessive "ringing", no deep dips or uncontrolled peaks ..) and decent extension on both frequency extremes.

 

[Shike]

Quote:

But DF equalization is only a rough estimate. And then there is the problem with bass impact of headphones. => Different strokes for different folks.

 
DF maybe rough, but it's also the best available considering.

 
Trying to emulate bass impact with headphones is one of the most amusing and infuriating concepts to me.  There is none, it's inherent to headphones.  Why, why are you trying to fix something that can't possibly be fixed!
 
*pulls out hair*

 

[Antony6555]

http://en.wikipedia.org/wiki/Equal-loudness_contour
 
To be accurate, headphones have to take into account the countours of the human ear. The equal loudness countour was developed to determine what "accurate" means for headphones. Essentially, a flat fr is not accurate for headphones. Rather there must slight treble and bass emphasis (look at the graph on the page for exactly what this means).
 
The equal loudness contour describes what the international scientific community considers an accurate response for headphones.

 

[Shike]

Quote:

http://en.wikipedia.org/wiki/Equal-loudness_contour
 
To be accurate, headphones have to take into account the countours of the human ear. The equal loudness countour was developed to determine what "accurate" means for headphones. Essentially, a flat fr is not accurate for headphones. Rather there must slight treble and bass emphasis (look at the graph on the page for exactly what this means).
 
The equal loudness contour describes what the international scientific community considers an accurate response for headphones.

Not exactly . . . loudness curves are just loudness interpretation based on the frequency and not just used for headphones (though they provide the easiest way to measure it).  DF equalization is probably the most accepted for correction based on the ear, FF being the next, raw being the least.

 

[xnor]

Quote:

http://en.wikipedia.org/wiki/Equal-loudness_contour
 
To be accurate, headphones have to take into account the countours of the human ear. The equal loudness countour was developed to determine what "accurate" means for headphones. Essentially, a flat fr is not accurate for headphones. Rather there must slight treble and bass emphasis (look at the graph on the page for exactly what this means).
 
The equal loudness contour describes what the international scientific community considers an accurate response for headphones.

Far from it. "slight treble and bass emphasis"??? If you'd actually taken a look at the contours you would have noticed the +30 to +40 dB at 20 Hz...

 
Quote:

Trying to emulate bass impact with headphones is one of the most amusing and infuriating concepts to me.  There is none, it's inherent to headphones.  Why, why are you trying to fix something that can't possibly be fixed!
 
*pulls out hair*

 
Dunno how you define or what you connect with the term impact but it's not true that headphones cannot have bass impact.
I think that some headphones actually have skull-crushing impact, but of course there's trade-offs. Heck, headphones are a matter of trade-offs.

 

[Shike]

 
Quote:

Quote:

Trying to emulate bass impact with headphones is one of the most amusing and infuriating concepts to me.  There is none, it's inherent to headphones.  Why, why are you trying to fix something that can't possibly be fixed!
 
*pulls out hair*

 
Dunno how you define or what you connect with the term impact but it's not true that headphones cannot have bass impact.
I think that some headphones actually have skull-crushing impact, but of course there's trade-offs. Heck, headphones are a matter of trade-offs.

Most bass response is felt and not heard when listening to a speaker system.  In that sense when most people think impact it's something they feel in a very literal sense.  Headphones largely aren't capable of this, all you doing by increasing the bass response is trying to emulate something that really isn't there.
 
This is why I got a tactile transducer for use with headphones.  Best of both worlds.

 

[chinesekiwi]

Quote:

Quote:

But DF equalization is only a rough estimate. And then there is the problem with bass impact of headphones. => Different strokes for different folks.

 
DF maybe rough, but it's also the best available considering.

 
Trying to emulate bass impact with headphones is one of the most amusing and infuriating concepts to me.  There is none, it's inherent to headphones.  Why, why are you trying to fix something that can't possibly be fixed!
 
*pulls out hair*

The JVC FX500 and FX700 in-ears laughs at you However they have birchwood shells and a birchwood based driver.