EQ almost impossible? (Need to filter the sine wave)
Sep 11, 2009 at 2:55 PM Thread Starter Post #1 of 21

skynetman

New Head-Fier
Joined
Sep 8, 2009
Posts
32
Likes
0
Guys i just read the sticky about headphone EQ.
I tried with my cheap AKG K55 and the sine wave generator.
Holy crap! If i keep the volume at 1-1.5KHz as reference, during the sweep i listen to at least 4 spikes and 5 holes!
Do i have to kill only 5.6 and 7.2 KHz spikes and forget about the rest?
I have big holes for example at 3077, 4562 Hz!!!
Ok ok i really need new cans!
confused_face.gif


Has anyone tried with a K701 a 3.5D or better cans?

Question: If i EQ to have constant volume up to 15 KHz don't i compensate also for my ears FR? That would be no good since music is supposed to be listened "as it is", right?
If i use my ear to EQ the headphones then i sum cans FR with ear FR!
Is there any way to filter the sine wave with the inverse of human ear Frequency Response curve (Fletcher & Munson) so that i have to hear constant volume from 20 to 20K?
This way (-FR_ear_filter + FR_cans + FR_ear) = FR_cans >>> to be constant
Strange no one made the software for this....
 
Sep 11, 2009 at 5:42 PM Post #2 of 21
Just because there are 'holes' doesn't mean that the headphones are bad, actually some expensive headphones do.

Check out how headphone.com do their measurements..
 
Sep 11, 2009 at 7:47 PM Post #3 of 21
A dummy head with a mic will never be like our head with real ears.
It can only be useful to compare different headphones but not to describe the real behaviour or to do an EQ....

About holes in FR i disagree.
Every hole i listen with real live sine wave sweep is a missing frequency.
Every missing frequency colours the sound and changes of you feel music.
I am not talking about holes in a chart measured with a dummy head and a mic, but a hole i can feel with my head and my ears
L3000.gif


Anyway you are wrong also according to headphone.com papers:
Code:

Code:
[left]"Frequency response is the measure of a headphones ability to reproduce all frequencies equally. Theoretically, this graph should be a flat line at 0dB."[/left]

 
Sep 11, 2009 at 8:33 PM Post #4 of 21
I like my holes and peaks, as a perfectly flat response would be boring and unengaging.
 
Sep 11, 2009 at 9:01 PM Post #6 of 21
@iriverdude: ears frequency response has no holes at all only fading at extremes and max at 1-4 Khz....
If a certain frequency goes -6/-8 db (you listen to an hole) and then 500 Hz higher or lower comes back at the right volume then it's a can problem!
That's why you are not supposed to mix with headphones but with reference speakers, that do not have holes in the frequency chart (if placed in a controlled room where resonance are killed).
 
Sep 11, 2009 at 9:15 PM Post #7 of 21
Quote:

Originally Posted by skynetman /img/forum/go_quote.gif
That's why you are not supposed to mix with headphones but with reference speakers, that do not have holes in the frequency chart (if placed in a controlled room where resonance are killed).


FWIR one shouldn't use headphone to mix because they don't give the spatial feeling and generally different balance of bass/mids/treble compared to speakers.
 
Sep 11, 2009 at 9:20 PM Post #8 of 21
Quote:

Originally Posted by skynetman /img/forum/go_quote.gif
About holes in FR i disagree.
Every hole i listen with real live sine wave sweep is a missing frequency.
Every missing frequency colours the sound and changes of you feel music.
I am not talking about holes in a chart measured with a dummy head and a mic, but a hole i can feel with my head and my ears
L3000.gif


Anyway you are wrong also according to headphone.com papers:
Code:

Code:
[left]"Frequency response is the measure of a headphones ability to reproduce all frequencies equally. Theoretically, this graph should be a flat line at 0dB."[/left]




And I'm not wrong. Theoretically it should be flat, yes, but in reality I guess that everyone would hate such headphones...

Let me think out loud about this.. If you use speakers that have a flat FR it doesn't mean that your eardrums will receive the same flat sound, because there are a lot of things that have certain effects like distance from/between the speakers, angle, sound insulation of the room, overlapping/crossfeed etc.
With headphones it's a similar situation but depends on other things like shape of your head (clamping pressure), sealing, age of the earpads etc.

So I think that there are a lot of problems doing such measurements..


edit: Maybe the best way to eq headphones 'right' would be to compare the sine waves levels between headphones and reference speakers? Though, results would be very very inaccurate.
 
Sep 11, 2009 at 10:00 PM Post #9 of 21
Quote:

Originally Posted by xnor /img/forum/go_quote.gif
edit: Maybe the best way to eq headphones 'right' would be to compare the sine waves levels between headphones and reference speakers? Though, results would be very very inaccurate.


Yeah you are right. I read that K701 is chosen often for mastering cause they feel it's similar to reference speakers....
Or probably is easier to learn how much you have to set the mix off with headphones so that it can be good on speakers
smily_headphones1.gif
 
Sep 11, 2009 at 10:06 PM Post #10 of 21
Yup, however everyone in the business would laugh at you if you did mixing/mastering with headphones only.
From what I've heard the results will be awful if you play it on speakers eventually.
 
Sep 11, 2009 at 11:45 PM Post #11 of 21
There are no headphones at any price range that I've heard that don't have lots of peaks/valleys. Even a theoretically flat headphone can have them because it may not match your own HRTF. Buying a new pair of headphones won't make the problem go away.

Quote:

Question: If i EQ to have constant volume up to 15 KHz don't i compensate also for my ears FR?


If you have an EQ with enough bands and lots of time on your hands, sure. You'd end up with excellent sound. Don't forget you'd need to do both left and right channels.

Quote:

That would be no good since music is supposed to be listened "as it is", right?


Don't know where you heard this but it's nonsense.
 
Sep 12, 2009 at 12:08 AM Post #12 of 21
Quote:

Originally Posted by b0dhi /img/forum/go_quote.gif
Don't know where you heard this but it's nonsense.


I'll make an example.
If you are at a concert live, instruments create some music.
Then music is received by your ears and becomes something else that is music + ears frequency response, you can't do anything about it.

When you record something you want to recreate ORIGINAL sound.
If we suppose that recording equipment is calibrated and does not affect sound, we can play it with loudspeakers or headphones.
We do not want EQ to compensate for our ears, because our ears would have been at the concert live as they are in our room now.
We only want to compensate for non flat frequency response of our speakers or headphones (plus distortion caused by headphones on our ears that does not exist when live at the concert).
So if you use a sine wave without first applying our ears frequency response you will compensate TOO MUCH, as if our ears were PERFECT at every frequency, something that does not exist and that is wrong.

I'll share my sine wave file modded with ear frequency response given by izotope ozone 4 loudness preset ad 100%:
Sine wave with loudness

The problem is that i do not have a sine wave generator that has this curve hardcoded inside, so i can only find frequency spikes with it and then find the right amount of dB trying.

sinewaveloudnessskynetm.png
 
Sep 12, 2009 at 1:25 AM Post #13 of 21
It's not correct to say that headphones should have flat FR to get neutral sound. They should have flat FR of the waves that reach your eardrums. That means you'll need equalization of some sort (DF for example) to have roughly flat perceived FR - even then everyone's ears are different so headphones equalized to your ears won't be perfect to someone else (thus the benefit of custom iems which don't have to bounce off of your ear lobes).

Since ear lobes have roughly predictable FR you're best off doing the method described in the tutorial of equalizing by ear. You do care about the sound that actually goes to your brain, not the sound a (non-binaural) mic would measure, right?
 
Sep 12, 2009 at 6:41 AM Post #15 of 21
Quote:

Originally Posted by skynetman /img/forum/go_quote.gif
I'll make an example.
If you are at a concert live, instruments create some music.
Then music is received by your ears and becomes something else that is music + ears frequency response, you can't do anything about it....



Ok. I thought you were talking only about headphones. If you're listening with speakers it's a different story. Your HRTF is bypassed (mostly) with headphones; not so with speakers.

Equalising so that the headphones sound flat to you is a good thing depending on what you want to do. If you want the music to sound like it's coming from around you, you need to measure your HRTF and adjust the EQ based on that. Using a perceptually flat EQ still has a lot of advantages though - instruments will become much more focused, things will sound more like real instruments. But the big problem is that unless you measure your HRTF and adjust based on that, you might end up with a slightly worse soundstage, which is the result of "over-compensating" as you say.

Then you could forget all that and just buy a Smyth
tongue.gif
 

Users who are viewing this thread

Back
Top