[tbger99]

Cheers everyone,
 
I'm a little confused about this issue.  The more confident I was, the more confused I then became.  I thought someone here might be kind enough to clarify and organize things for me.
 
As far as I understand at this time, sound perception procedure goes like this:
 

There are two other elements I didn't know where to place:
 
Head related frequency transfer, to my understanding is the impact our ears and ear-canals make to the sound passing through them.  It is specific and can be measured (kind of) only per individual.
On the other side, there're the equal loudness curves, which, I suppose, tell us how loud each frequency has to be in order for the listener to perceive a flat sound signal.
Do they actually play the same role for us ?
 
Also where or how do DF / FF come into play ?
 
 
I guess it's pretty basic but I just can't seem to understand the whole picture which makes it hard to step any further trying to understand other things.
 
 
 
Thanks in advance

 

[Head Injury]

You don't really want a system which accounts for the equal loudness curve. Nature doesn't account for it, and nor would a live music show. It might be a good idea to account for it slightly in the sub-bass, because headphones don't shake the body like real life.
 
This is a good post on field equalization. Basically, free field only tries to account for the effects of the outer ear and attenuation from distance, and diffuse field also tries to account for certain room reflections. Which one sounds more accurate is really up to the individual, because we all have different ears. If the LCD-2 is any indication, free field equalization works best for me. Most headphones are diffuse field equalized.
 
Even more simplified, free field is like listening to pure music from a distance, and diffuse field is like listening to the musicians in a really good room.
 
If you look on Inner Fidelity and compare the compensated graphs with the raw graphs, you can see how the outer ear affects sound. The largest difference is in the upper mid-range, which is greatly amplified by the ear. All the raw graphs should have big hills there, even when the headphones are perceived to be recessed.

 

[rroseperry]

Cheers everyone,
 
I'm a little confused about this issue.  The more confident I was, the more confused I then became.  I thought someone here might be kind enough to clarify and organize things for me.
 
As far as I understand at this time, sound perception procedure goes like this:
 

There are two other elements I didn't know where to place:
 
Head related frequency transfer, to my understanding is the impact our ears and ear-canals make to the sound passing through them.  It is specific and can be measured (kind of) only per individual.
On the other side, there're the equal loudness curves, which, I suppose, tell us how loud each frequency has to be in order for the listener to perceive a flat sound signal.
Do they actually play the same role for us ?
 
Also where or how do DF / FF come into play ?
 
 
I guess it's pretty basic but I just can't seem to understand the whole picture which makes it hard to step any further trying to understand other things.
 
 
 
Thanks in advance

One possible reason for confusion is that those different graphs don't represent the same information. Put in the coordinates and units and you'll see that the translation isn't as direct as the graphic suggests.

 

[tbger99]

It does look like headphones are equalized more like the Fletcher Munson curves rather than the equal-loudness curves by ISO.  I'm still confused on where DF / FF comes into play here, but I'll read further on that on the link there and this might clarify the point.  Thanks for that
 
As far as I understand Tyll's methods, the frequency plots are compensated by subtracting the HRTF coordinates from the raw graph and the result is the frequency response of the given headphone.  Please, correct me if I'm wrong.
 
Telling my 'story' in more depth might explain my point.  I'm trying to find a reference can for equalization of my field recordings.  For this I'll have to make sure a headphone is reliable in terms of tonal balance and accuracy.  If it tells me the recording is warm, then I'll decrease the corresponding frequencies, so that it sounds great on both warm-sounding speakers and also bright-sounding ones, and virtually on any other kind of system.  Users seeking for a good sounding headphones can go audition them and see if they like it or not.  I don't seek for good sounding cans, but accurate ones that will serve my purpose.  I can't really bring my projects there and see how the headphone works for me in 'real' circumstances.  I don't have the cash for full size system so I'll have to stick with headphones for the next few years being a student.

 

[gregorio]

tbger99 - If I've read your last post correctly, what you are looking to do is effectively a mastering engineer's job with cans. It cannot be done, not even by a mastering engineer! the experience of wearing cans just does not translate well to the experience of listening to speakers and frequency response is only part of the equation. The only advice I can offer, is to get as good a set of cans as you can afford and get access to as good a speaker system as you can to use as a reference. Then, with time, work out the differences between the reference speakers and your cans so that when you are wearing your cans you consciously make allowances for those differences. This is how audio professionals often have to deal with weaknesses in their systems when dealing with translation issues. I think you are over thinking the problem in the hope of discovering a solution which does not really exist.

G