Equalization - capabilities and limitations

[desik]

One thing I'm surprised about on head-fi is that there's ton of discussion about sound signatures but relatively little about equalization. Why it is not possible to pick a technically capable headphone and equalize it to any signature you want?
 
I'm a borderline basshead I guess. Every time I tried to increase the bass in headphones I would usually loose details to the point that unequalized they would always sound better. In more extreme cases on cheap IEMs some rattling noise would add on top of the music, like if the IEMs were defective.
 
Now for example I have a LCD-2F, which is amazing sounding in theory, but I want more bass. If I try to equalize it, could I get a sound closer to non-fazor LCD-2?

 

[RRod]

  One thing I'm surprised about on head-fi is that there's ton of discussion about sound signatures but relatively little about equalization. Why it is not possible to pick a technically capable headphone and equalize it to any signature you want?
 
I'm a borderline basshead I guess. Every time I tried to increase the bass in headphones I would usually loose details to the point that unequalized they would always sound better. In more extreme cases on cheap IEMs some rattling noise would add on top of the music, like if the IEMs were defective.
 
Now for example I have a LCD-2F, which is amazing sounding in theory, but I want more bass. If I try to equalize it, could I get a sound closer to non-fazor LCD-2?

 
There are linear problems (frequency response, phase) and non-linear problems (distortion), and your typical EQ can only address ½ of the linear problems. That is, you can control the frequency response, but too much adjustment can mean issues with phase and distortion. Then consider the sometimes massive differences in FR between headphones, and you can see how sometimes you just can't turn all problems into nails for the EQ hammer.

 

[jcx]

I've mentioned some "EQ" options a few times: http://www.head-fi.org/t/677809/the-stax-thread-iii/3225#post_10869915

 

[desik]

   
There are linear problems (frequency response, phase) and non-linear problems (distortion), and your typical EQ can only address ½ of the linear problems. That is, you can control the frequency response, but too much adjustment can mean issues with phase and distortion. Then consider the sometimes massive differences in FR between headphones, and you can see how sometimes you just can't turn all problems into nails for the EQ hammer.

Thanks! I understand what FR is. What is phase and how does it lead to problems?

 

[RRod]

  Thanks! I understand what FR is. What is phase and how does it lead to problems?

 
You've got a bunch of sine waves floating around: FR tells you how loud they are, phase tells you how shifted/delayed they are. Here's a little bit on audibility. It doesn't lead to horrendous problems; I was just being overly cautious in what might still differ between headphone responses after correcting FR using a typical (minimum-phase) EQ.

 

[castleofargh]

I may be stating the obvious, but the felt resolution of a headphone may have little to do with the actual resolution. many headphones pass for detailed/resolving headphone with one or 2 ugly spikes in the mids/trebles. while very competent headphone with a warm signature will end up branded as slow/unresolving/... 
the signature affects our interpretation of sound at a level that goes beyond just making a frequency louder, so it's easy to ruin something while thinking we're just adding more of what we like. bass IMO can very much be like that. I do like my rumble, but it's easy to add too much and ruin the rest. EQ like everything else, is best used in moderation when we don't know exactly what we're doing.
 
if you're a basshead you should always go for a headphone that has a lot of bass to start with. it's always better to lower too much bass or to add only a little, than to try compensating to get +20db of subs that the headphone may never be able to output without huge distortions. you need to pay attention to the global digital gain(peak meter) to see if you're not clipping the signal with your EQ. and you also need to make sure your amp can deal with what you're asking(any digital attenuation will have to be compensated by the amp to give an equivalent perceived loudness)

 

[desik]

I understand. What about harmonic distortion? How can it increase if you, say, boost bass through EQ?

 

[RRod]

  I understand. What about harmonic distortion? How can it increase if you, say, boost bass through EQ?

 
Pick a frequency and amplitude, say 100Hz at 75dBSPL, and it's blaring through your headphones. Now you put on an EQ that boosts 100Hz by 10dB, so now you're hearing 100Hz at 85dBSPL. Your distortion products will thus be those associated with 100Hz@85dBSPL instead of 75dBSPL. Those may be relatively higher or lower to the fundamental, depending upon the specifics of the speaker/headphone

 

[Roseval]

In more extreme cases on cheap IEMs some rattling noise would add on top of the music

That is called clipping, you boosted too much.
Instead of boosting the bass you better lower the rest.
Gives you more bass too but without the clipping

 

[desik]

  That is called clipping, you boosted too much.
Instead of boosting the bass you better lower the rest.
Gives you more bass too but without the clipping

That means clipping happens in the digital stream, right? I suppose for IEMs whether you boos bass of reduce the rest + increase the volume, it should not make any difference.

 

[RRod]

 
That means clipping happens in the digital stream, right? I suppose for IEMs whether you boos bass of reduce the rest + increase the volume, it should not make any difference.

 
You can always apply a pre-cut before the gain to (mostly) prevent clipping. Clipping can happen to the digital and analog signals.

 

[lawlbear]

Same here man. I love bass. I usually equalize my headsets to give me a perfect balance and then boost up the bass. Even though this hurts the rest of your frequencies it gives u a more tuned sounds. 

 

[desik]

Tried to play with EQ a bit last week. Couldn't make Fazor sound like Pre-fazor. Might try again just for fun in 3 weeks, when I'll be over with a terrible deadline. I should probably sell those Fazors instead, but just have an obsession to try everything else first.

 

[pila405]

Maybe this is due to the variation in the Auduze lines. You have on "innerfidelity" many FR graphs for the LCD2, so it should be easy enough using a proper equalizer (and also because the LCD2 FR is very smooth), like FabFilter Pr-Q 2 etc'.

And here is some information regarding what RRod stated:
http://www.audioholics.com/room-acoustics/human-hearing-phase-distortion-audibility-part-2
http://www.audioholics.com/room-acoustics/human-hearing-amplitude-sensitivity-part-1
http://www.audioholics.com/loudspeaker-design/audibility-of-distortion-at-bass

 

[castleofargh]

not sure an EQ can do that. AFAIK the fazor thing was to improve on air flow disturbances(did I get that right?), so if it does as advertised that should help as much for distortions than it may change the signature.
also using online measured graphs are of limited interest as you never know that the signature of your headphone is the same as the pair measured. for a hd800 it shouldn't be too far off as they really try to hit the target specs. but for audeze, between the eternal tweaking going on under the same model name, and usual manufacturing variations, it's hard to say that one pair sounds like another pair.