[Sonic Atrocity]

This is what I wrote in the HD600 Appreciation Thread (and I'm currently interested in the HD600 since they are often referred to as being neutral):
 
When I think of "neutral" gear I don't think of gear that corrects for the dips and spikes caused by the very shape of my ear. After all, how my ears are shaped has determined how I hear throughout my entire life. Indeed, how I 'hear' a song will be different from how everyone on this planet 'hears' a song because of the differences in the shape of our ears. As such, it isn't that I want to hear exactly what the studio producer was hearing when they produced the album - because I can't. I don't have their ears, and as such can only listen with mine. But what I want to hear is no different from the objective material that was produced in the studio...
 
So, let's say that my ears were conducive to different spikes and dips than what is demonstrated in the link so kindly provided by Palmfish, I wouldn't want my gear to correct for my ears' abnormalities. Even if it means I am hearing things vastly different from how others are that is fine by me if it is just a result of my weird ears. Though, all this being said, I think my ears are fairly "normal". 
 
So, when I say that I want neutral gear what I mean is this: I want gear that plays the material (the music) back to me for what it is. No more, no less. I don't want my gear to place an emphasis on any frequency for my enjoyment or for a presentation that is different from the source material. It is the objective product that came straight from the studio that I want to hear, and the only changes to that product that are okay are the changes caused by my very own ears. 
 
So, there ought to be some objective standard for what is the most neutral playback possible. This is why we almost unanimously agree the Beats are further from neutral than the HD800. If the conception of neutral was contingent upon subjective factors then there wouldn't even be a wide standard for what constitutes neutrality. It'd be entirely arbitrary.
 
Edit: What are your thoughts?

 

[bigshot]

How'd that go over in the Headphones forum? Not well, I'm guessing since you're here...

I agree with you a hundred percent. All you need is perceived flat for human ears. It doesn't even have to be exact. Close enough for government work is fine.

Or save a lot of money and just buy a set of headphones that are reasonably good, but not totally flat and EQ them to flat. It will sound the same.

People try to overcomplicate frequency response. But it's really simple. It's a graphic representation of sound, so if you compare the graphic representation with a baseline based on the thresholds of human perception, you should be able to get a very good idea of how the cans sound.

The trick is how you measure. Using a mic is more accurate, but it involves complications of having to add compensation curves to make what the mike hears more like what human ears hear. If you measure by ear, all of that is built in.

In any case, you're thinking along the right lines, and you'll end up with a great sounding set of headphones I'm sure.

 

[White Lotus]

 

Spoiler: snip

How'd that go over in the Headphones forum? Not well, I'm guessing since you're here...

I agree with you a hundred percent. All you need is perceived flat for human ears. It doesn't even have to be exact. Close enough for government work is fine.

Or save a lot of money and just buy a set of headphones that are reasonably good, but not totally flat and EQ them to flat. It will sound the same.

People try to overcomplicate frequency response. But it's really simple. It's a graphic representation of sound, so if you compare the graphic representation with a baseline based on the thresholds of human perception, you should be able to get a very good idea of how the cans sound.

The trick is how you measure. Using a mic is more accurate, but it involves complications of having to add compensation curves to make what the mike hears more like what human ears hear. If you measure by ear, all of that is built in.

In any case, you're thinking along the right lines, and you'll end up with a great sounding set of headphones I'm sure.

 

 
Great response! I was hoping someone would come along and hit the nail on the head.
 
I think the common misconception is: 
 
"If a headphone is measured, and shows a graph representing a ruler-flat frequency response, it therefore should sound flat to my ears"

 

[superjawes]

One way to think about it is filters. In electrical engineering, a filter is an electronic device specifically meant to pass power on certain frequencies while blocking the rest. Now we could get into filter specifics, but all you really need to know is that the curves you're looking at represent how much power is passed at certain frequencies.

Now in terms of getting music to sound "correct" at your ears, you might think that you want something completely flat, but you have to keep in mind that every step in the process acts as its own filter. Your amplifier will power certain frequencies more easily than others, your speakers or headphones will respond to certain frequencies more than others, and your ears and brain will respond differently to different frequencies as well. (And that's not even getting into the frequency content of the original music).

Unless you are at a concert, you are only listening to a representation of the original music, and your speakers (or headphones) are trying to replicate that music. Where there were once instruments at a certain distance, there are now speakers (probably at least two) at a distance in a completely different room. Yes, the room will have its own filtering effect, too.

So what does this mean for headphones? Well in some ways, who knows? Harman is working on a target headphone response curve with some interesting results (I've been mostly reading the summaries at InnerFidelity). This is very interesting because we are starting to get data on how humans experience sound through headphones and what we (collectively) prefer. Basically, they are figuring out the frequency response of your ears with headphones.

I hope I didn't derail the conversation, because I really do want to echo bigshot's comments more than anything. I think the best bet to finding a response curve that will sound transparent (to you) is to try a few headphones, pick which one sounds "right" to you, and then look at the data. That is probably the best way to identify the measuremnets that will sound "right" (to you) as long as the person making the mesurments is consistent.

 

[bigshot]

In general, it really isn't difficult to find a response curve that is correct for human ears. You just use your ears to balance the response instead of measuring equipment. It isn't as precise, but it's precise enough for ears. When you use equipment, you need to do all sorts of adjustments to make up for the difference between the way a microphone hears and they way that human ears hear. That's fine for putting charts on a page, but if your goal is to EQ your system, it's more direct to just use ears and not have to bother with applying complicated compensation curves. If your hearing is normal, your hearing won't be all that much different from other people with normal hearing, so one curve does fit all.