How to equalize your headphones: A Tutorial

[PiccoloNamek]

Quote:

actually using sine.wav you can easily know where your correction points are, just set them to the max gain and see where they are on your media player transport bar

I am not sure what you mean here. Certainly, if you already know where your correction points are, you can see them on the media player viewer, but it will not help much because it will not tell you how much to cut. If you don't already know where your problem spots are, there is no way to find them without either listening or using a measuring microphone placed at the eardrum. (Which still would not give the same results, as the microphone would have a different acoustic impedance than your eardrum.)

Quote:

Originally Posted by Bengt77 This is all very interesting for sure. But please, guys, do yourselves a favour. Just listen to and enjoy your headphones.

I think headphones are more easily enjoy when some inherent problems are fixed.

I also feel that I did myself a huge favor by undertaking my equalization experiments. Now I enjoy my headphones more than ever!

 

[leeperry]

Quote:

Originally Posted by PiccoloNamek /img/forum/go_quote.gif I am not sure what you mean here. Certainly, if you already know where your correction points are, you can see them on the media player viewer, but it will not help much because it will not tell you how much to cut. If you don't already know where your problem spots are, there is no way to find them without either listening or using a measuring microphone placed at the eardrum. (Which still would not give the same results, as the microphone would have a different acoustic impedance than your eardrum.)

I mean you can know at what timing your correction points are.

anyway, I can lower the trebles resonance quite a bit, they they're still very much audible w/ sine.wav...did you manage to make them vanish completely ?!

BTW, I think your 2 correction point on the first page work much better than mine, it takes some drastic settings to remove all that crazy sibilance

so you don't like digital mode, eh?

 

[PiccoloNamek]

I like what I read about the S-plane type filter. The manual even recommends that you use it. I don't think Linear mode is quite the same as plain analog mode though, either.

 

[leeperry]

ok I think I'm all set....for today at least

I kept your second point as is, and put the first one on 8250 instead of 7400...seems to do wonders

so you didn't answer me, does it fix ALL the resonance for you when you play sine.wav?
if that's a trickly question, feel free to ignore it

 

[PiccoloNamek]

Oh, sorry, I forgot.

When I play the sine wave file in Foobar2000, the response is essentially uniform, yes.

 

[leeperry]

Quote:

Originally Posted by PiccoloNamek /img/forum/go_quote.gif When I play the sine wave file in Foobar2000, the response is essentially uniform, yes.

ah, well mine resonates A LOT in the upper medium >12K and has some silences from 8 to 11 due to my 2 correction points......too bad this can't be measured w/ a commercial sensor

oh well, it's way better than before anyway....I think I'll wait for my phones to break in before spending weeks on this

 

[PiccoloNamek]

It helps if your phones are a lot fairly flat to begin with. Without EQ, mine are almost completely flat from 20hz to 2kHz. When I remove the 7400Hz peak and the 13kHz peak, only a few minor and almost unnoticeable non-linearities remain in the high frequency response, that really aren't worth the trouble of removing.

 

[FSonicSmith]

My view of equalization is that the cure may be worse than the malady. In the old days of analogue hardware based equalizers, one had to wonder if introducing a fairly long signal path didn't cause significant degradation of the signal. It is always important to at least consider the theory that not all aspects of recorded sound and playback can be measured or quantified or objectively confirmed. I don't see why it would be any different when using software and computer based equalization. It's a tool to be tried, but don't be a "tool" focusing on one aspect of sound playback to the possible detriment of others. Spacial cues, soundstage width and depth, imaging, PRAT, micro and macrodynamics and detail, liquidity, cohesion, jump, and the intangibles that cause-at one extreme-the listener to get bored and at the other extreme-to want to stand up and boogie/dance/wag the butt can all be adversely affected with or without scientific explanation. Just my 2 sense.

 

[hybris]

Quote:

Originally Posted by FSonicSmith /img/forum/go_quote.gif My view of equalization is that the cure may be worse than the malady. In the old days of analogue hardware based equalizers, one had to wonder if introducing a fairly long signal path didn't cause significant degradation of the signal. It is always important to at least consider the theory that not all aspects of recorded sound and playback can be measured or quantified or objectively confirmed. I don't see why it would be any different when using software and computer based equalization. It's a tool to be tried, but don't be a "tool" focusing on one aspect of sound playback to the possible detriment of others. Spacial cues, soundstage width and depth, imaging, PRAT, micro and macrodynamics and detail, liquidity, cohesion, jump, and the intangibles that cause-at one extreme-the listener to get bored and at the other extreme-to want to stand up and boogie/dance/wag the butt can all be adversely affected with or without scientific explanation. Just my 2 sense.

One huge difference to consider between analog and digital EQing, is that the latter is essentially lossless - at least if you only reduce and not increase (increasing will reduce available dynamic range). Thus the sound quality will not be reduced.

This is of course only true if you perform the equalization on the original digital signal before your DAC, not do another AD/DA-conversion on the analog signal.

 

[PiccoloNamek]

All aspects of recorded sound and playback can be quantified and confirmed. I believe that one day we will be able to quantify and confirm the perception of sound in the brain as well.

But this kind of discussion is better suited for the Sound Science forum. I would like to keep the discussion here strictly on topic.

 

[leeperry]

Quote:

Originally Posted by FSonicSmith /img/forum/go_quote.gif My view of equalization is that the cure may be worse than the malady. In the old days of analogue hardware based equalizers, one had to wonder if introducing a fairly long signal path didn't cause significant degradation of the signal.

this is the 21st century my friend, we got 32 float workflows, and Ozone4 works in 64 float internally....the cure is definitely worth being implemented

I think I prefer Ozone4 over Electri-Q, now that I know what I'm looking for.....which took a while to find out actually

you wanna find the limit between ugly resonances and muddy sound....which changes depending on the song...anyway, this is as good as it gets I think :

 

[FSonicSmith]

Quote:

this is the 21st century my friend, we got 32 float workflows, and Ozone4 works in 64 float internally....the cure is definitely worth being implemented

Quote:

All aspects of recorded sound and playback can be quantified and confirmed.

Please guys, I want to keep this friendly and constructive. That said, and with all due respect, the second assertion quoted above is just flat out incorrect. It is unfortunate (to my mind, not yours obviously) that you earnestly believe this.

The first assertion quoted above is much more interesting (again, to me). Digital sound is not yet perfect. Sinewaves do not get reproduced with perfect smooth arcs, as they do with analogue gear, but in stepped notch like patterns that only resemble sine waves if you're drunk or legally blind. While digital sound has come a long way, it still causes listener fatigue to many. And switching gears here to something else worth considering, it may be the 21st century, but last time I looked the transducers in your headphones are still based on magnets and pistonic motors, pushing air in much the same fashion as in the days of Alexander Graham Bell.
Now please don't get me wrong and assume I am some kind of hairshirt. I love listening to music over my laptop with wifi, with the liberty to stream audio from virtually any radio station in the Country or World, happy that I can bypass the D/A in my MacBook Pro since I happen to think that the D/A in my RSA Predator is far better, and I'm listening to one of my favorite radio stations thousands of miles away as I type this. But please don't tell me that with some software equalization, all shortcomings in a set of cans can be cured without any corresponding trade-offs.

 

[Max F]

Why don't you try it yourself. I don't think anyone is going to convince you otherwise.

I like how this thread has been more about "Practice" than "Theory". I hope it stays that way as it's rather refreshing.

 

[leeperry]

Quote:

Originally Posted by FSonicSmith /img/forum/go_quote.gif please don't tell me that with some software equalization, all shortcomings in a set of cans can be cured without any corresponding trade-offs.

why don't you try it?
it's not like we force you or sumthing

quite clearly I won't turn Ozone4 off, as it KILLS a lot of resonances in the trebles and makes the sound a lot more "natural"(as playing on speakers).

16integer EQ wouldn't make sense, but 32/64float is cool IMHO...it won't add useless distortion to the source.

it's like gamut conversion for movies, you don't know that you need it.....until you try it

look here, and try it if you dare(yes it works in 64float too) :
https://forum.doom9.org/showthread.php?t=139389
Color Correction with a HTPC - Simpler solution and now it really works! - AVS Forum

but it makes me wonder about all these ppl who mix their music in Nuendo/Cubase w/ headphones, because they don't quite hear the same as w/ speakers

 

[b0dhi]

Quote:

Originally Posted by leeperry /img/forum/go_quote.gif why don't you try it? it's not like we force you or sumthing quite clearly I won't turn Ozone4 off, as it KILLS a lot of resonances in the trebles and makes the sound a lot more "natural"(as playing on speakers). 16integer EQ wouldn't make sense, but 32/64float is cool IMHO...it won't add useless distortion to the source. it's like gamut conversion for movies, you don't know that you need it.....until you try it look here, and try it if you dare(yes it works in 64float too) : https://forum.doom9.org/showthread.php?t=139389 Color Correction with a HTPC - Simpler solution and now it really works! - AVS Forum but it makes me wonder about all these ppl who mix their music in Nuendo/Cubase w/ headphones, because they don't quite hear the same as w/ speakers

Using 32 or 64 bit data length wont do much to improve the quality of an EQ in this scenario. It's only useful in a studio situation where you're stacking lots of effects. You still get ringing, phase distortion (if you're in analog mode), and if your Q is sharper than 0.6 you're likely to get audible distortion outside of the intended EQ band.