[ultrabike]

A CSD does not depend on multiple impulse responses taken a different times. It is really dependent on a single impulse response to which a sliding window is being applied. The sliding window steps corresponds to the CSD decay plot. The decay will depend on both impulse response and type of window applied.
 
In this sense a CSD is valid as long as the impulse response on which it depends is valid. Which means linear behavior needs to be dominant and the system needs to be roughly time-invariant... therefore equalization should be able to offer some of it's corrective magic to CSDs as well... within the limits of the system being linear time invariant in the first place (i.e. a well behaved and trained headphone

)

 

[musical-kage]

I think EQing is a weird one.
 
I can for example turn all EQ off straight away and be happy with the audio UNTIL I EQ the bass and treble up to create a U curve. That is when I notice the difference, and if I then try going back to flat, I don't like the sound.
 
I think if you are used to something, you will find that sound naturally better, if it is closer to what you feel is right, or not.
 
If you turn EQ completely off one day, you probably won't notice it until you turn it on.

 

[ultrabike]

Before I sold it, EQing made my DT990 tolerable for long listening sessions.

 

[bigshot]

I can for example turn all EQ off straight away and be happy with the audio UNTIL I EQ the bass and treble up to create a U curve. That is when I notice the difference, and if I then try going back to flat, I don't like the sound.

That isn't equalizing. That's coloring the sound. It's very unlikely that a U shaped curve would correct for the imbalances in your headphones, because a lot of cans have bass bumps and treble boosts built in. A U shaped curve just makes it more out of whack. The idea of equalization is to cancel out imbalances and achieve as flat a response as possible. Flat response is what good engineers mix and master to, so if your response matches theirs, it will sound the same, even if you're using different speakers or headphones. EQing is a bit of work, but it's worth it.

A nice flat response makes a world of difference. I can bypass my EQ settings on my speaker system and the sound sounds full and bright, but the detail is all muddy because frequencies are competing with each other instead of working together.

 

[Eisenhower]

EQ works very well at tailoring a headphone's sound signature, yet is highly underrated amongst audiophiles. I think this is in part because it is so easy, and feels like cheating to some people.

It has real limitations though. For example, open dynamic headphones have increasing THD below 100Hz because of the freely moving driver. This is not a big deal because these headphones generally have a bit of bass roll off:





However if you go and attempt to boost the bass with EQ (or by decreasing everything else and increasing the volume), you may start to have audible distortion in the bass frequencies. I imagine this happens quite often, since bass is no doubt the most desirable frequency region amongst most listeners, and open dynamic headphones are extremely common.


I disagree about EQ not being able to increase detail. The ability of a headphone to reproduce details is highly dependant on how much treble it has. There are other factors that matter too, like distortion and damping, but most decent open dynamic headphones are quite sufficient in these areas. The so-called Sennheiser veil, for example, is entirely due to a modest treble response, and can be easily "fixed" with EQ.
It is certainly more effective than buying a $500 tube amp.

 

[musical-kage]

Quote:

That isn't equalizing. That's coloring the sound. It's very unlikely that a U shaped curve would correct for the imbalances in your headphones, because a lot of cans have bass bumps and treble boosts built in. A U shaped curve just makes it more out of whack. The idea of equalization is to cancel out imbalances and achieve as flat a response as possible. Flat response is what good engineers mix and master to, so if your response matches theirs, it will sound the same, even if you're using different speakers or headphones. EQing is a bit of work, but it's worth it.
A nice flat response makes a world of difference. I can bypass my EQ settings on my speaker system and the sound sounds full and bright, but the detail is all muddy because frequencies are competing with each other instead of working together.

If that is the case, I guess I wouldn't like flat response, as I typically do have the EQ always in a U curve when on portable players. On my PC, I have it flat with a bit of emphasis on treble, but only because my sub has external controls.
 
In my opinion, EQing this way adds some sparkle to what otherwise could be a boring set of cans. I don't mess about with EQ settings for every genre. I simply keep it as is throughout. My brain just gets used to the difference.

 

[bigshot]

You haven't heard what a balanced response does for music. It makes a big difference for the better. MUCH better than just boosting frequencies at the edges of your headphones' ability to reproduce..

 

[AstralStorm]

There's U and there's U. Miine typically end with <50 Hz and >15 kHz, depending on IEM/headphone and its rolloffs. Usually the bass roll is <= 6 dB. Air roll tends to be much sharper in most cases.

 

[budget minded]

a lot of people don't like the sound of truly flat audio... it sounds too thin and bright to them. they buy tube amps to warm up their sound with more midbass or get speakers that have a frequency response that they like. EQing only undoes the conscious decisions they made. i don't get it myself. nowadays with digital room correction hardware and even freeware, you can tweak your system to within half a decibel or better of perfectly flat with much less distortion than parametric and graphic analogue EQs that audioophiles might still associate with distortion. just like with old school EQs though, if you don't like the sound of flat, you can boost the bass, roll the treble off, give it the BBC midrange bump, or make an "ideal" downward slope from bass to treble that many prefer.
 
i really want to get into digital room correction as i'm about to get a western digital media tank to eventually use with a high end DAC on my new speakers. they don't just improve the jagged response of every speaker's frequency response, but they can also tame even worse room artifacts. i recently read one review where a person that used a DRC liked the flat sound he made a pair of cheap bookshelf speakers he bought more than his expensive quads.
 
you don't need expensive mics either. behringer has a calibrated mic, probably designed to go with it's $350 "bargain" DEQ2496 that many use and dayton audio has a similar model that i think even comes with an individual calibration file for about the same price. really, you can get into the game for just $50 as i'm seeing there's DRC freeware even out there now! i came here to find out about that and was going to create a thread about it as i want to find DRC freeware that i can use to apply directly to sound files stored on a media tank so i don't need the $350 + $50 for mic behrigner system i wanted after reading a couple glowing reviews for that unit. i don't want to stream off my computer. really, i might be able to get away with the sure PZM mics i already have as they're omnidirectional and flat through most of the frequency range, but $50 for a proper calibrated mic is a cheap upgrade to turn entry level speakers into high end room corrected flat gear.
 
as i'm using 5 1/2" energy 2 ways with their ports plugged, i know i could use at least some bass boost in my system and the bass boost on my receiver makes them sound boomy in the midbass even at 5dB.
 
digital room correction is the most bang for your buck you can get i think. i can't wait to hear the difference it makes. it can also compensate for amps and preamps effects on frequency response and DACs too if you have one that you can use with your PC's USB outs. for just $50 i could make my already nice speakers sound $1000+ better? sounds like a plan to me.
 
not just that, but maybe, i can even get flat EQ out of my system, at least down to my 40Hz limits, and then re-record from my system binaurally with a dummy head i was given by a beauty college (that REALLY freaked a passerby out one night i was waiting to record a train with) as well as compensate for my PZMs AND headphones for EQ flat mp3s with no more hole in the head imaging. if you've never heard a binaural recording, they can be scary realistic, especially at capturing ambient cues. that'd be esoteric and complicated to pull off maybe, but it's worth a try. i've already tried it with cheesy aluminum realistic minis and got a really thin sound EQing the entire signal chain would help a lot. there's a big difference between radioshack minis and energy minis too.

 

[bigshot]

A flat response doesn't have a "sound". It's a *calibration* so your home system precisely matches what the engineers created in the studio. People who say they don't like the "sound" of a flat response don't know what they're talking about.

I suspect that the reason people think that flat response sounds thin is because microphone based automatic settings are often extremely inaccurate when it comes to bass. When I ran the auto EQ on my system, it turned my subwoofer off completely.

 

[Eisenhower]

Quote:

A flat response doesn't have a "sound". It's a *calibration* so your home system precisely matches what the engineers created in the studio. People who say they don't like the "sound" of a flat response don't know what they're talking about.
I suspect that the reason people think that flat response sounds thin is because microphone based automatic settings are often extremely inaccurate when it comes to bass. When I ran the auto EQ on my system, it turned my subwoofer off completely.

 
I think you might be underestimating the variability in subjective preferences

 

[bigshot]

It doesn't have anything to do with preferences. It's calibration to a standard. Once you're calibrated, you can use the bass and treble to set it the way you prefer if you like a different sound from the one the artists who made the recording intended. But you should start from a zero point so your calibration is the anchor to your settings. That way you can always get back to calibrated.

But calibrated doesn't sound any particular way. It just sounds the same as the engineers heard.

 

[AstralStorm]

Quote:

But calibrated doesn't sound any particular way. It just sounds the same as the engineers heard.

We hope so - it's possible the engineer used uncalibrated and/or lousy hardware.
 
DRCs aren't super accurate on their own - most attempt to do their best to control resonances though. If you want to try it, first eq by ear, then run DRC (will correct resonances), then eq again.
I found that to work better than straight DRC. (sample size of 1)

 

[RPGWiZaRD]

Quote:

It doesn't have anything to do with preferences. It's calibration to a standard. Once you're calibrated, you can use the bass and treble to set it the way you prefer if you like a different sound from the one the artists who made the recording intended. But you should start from a zero point so your calibration is the anchor to your settings. That way you can always get back to calibrated.
But calibrated doesn't sound any particular way. It just sounds the same as the engineers heard.

 
But why should one calibrate to flat response if that's not what the person enjoys the most? I want more bass than flat, I like aprox 7~9dB or so boost depending on the resonance and all that stuff, the Q40 have about 8dB boost in bass which is quite perfect to me.
 
I love EQing and have used it for years to EQ accordingly to my personal preferences, when I first got into headphone listening my preferences were that I only wanted a bassheavy sound, then later on I wanted V-shape sound, very heavily so. Today I despise V-shape sound as I can't stand it when highs are pushed in front of the mids. Today I want a little closer to flat but I still want a good 7~9dB or so boost in bass and especially lower-midrange, 300Hz - 1000Hz I'd say I optimally prefer about 3dB louder in average compared to the highs but around the 8~10kHz area it's ok with a small spike.
 
I can usually tell just listening to a song, whereabouts I need to EQ as by now I've learnt quite a lot how the difference frequencies and the different instruments etc sound and lies at so I know before I adjust it roughly what kind of result I will get.

 

[AstralStorm]

Quote:

 
But why should one calibrate to flat response if that's not what the person enjoys the most? I want more bass than flat, I like aprox 7~9dB or so boost depending on the resonance and all that stuff, the Q40 have about 8dB boost in bass which is quite perfect to me.
 
I love EQing and have used it for years to EQ accordingly to my personal preferences, when I first got into headphone listening my preferences were that I only wanted a bassheavy sound, then later on I wanted V-shape sound, very heavily so. Today I despise V-shape sound as I can't stand it when highs are pushed in front of the mids. Today I want a little closer to flat but I still want a good 7~9dB or so boost in bass and especially lower-midrange, 300Hz - 1000Hz I'd say I optimally prefer about 3dB louder in average compared to the highs but around the 8~10kHz area it's ok with a small spike.

 
More bass where? Are you actually sure your bass is flat?
Most of headphones and IEMs actually roll off in subbass... I hate bass boosts and upper bass boosts though, they dull the sound of everything bassy.
 
What you're describing is likely reasonably close to a perceptually flat eq for Q40. Check it. No really, please do.
 
Edit: Yep, it is quite close, if the boost is subbass, then you're just turning a hump into a much less obnoxious slope. You might be listening far too loud if you're actually boosting lower midrange with these, unless by lower midrange you mean 300-1k - boosting there would counteract the bass boost.
Source: http://en.goldenears.net/9049