How far can EQ really go towards truly equalizing headphones? - Page 4

+1. I've never liked the sound of a headphone that I can't stand to wear; and it has nothing to do with FR or anything else - if they're a painful experience, they sound bad. Illogical as it may seem.

To the broader question - I think the short'and'sweet is that radiation patterns, distortion, and aesthetics/fit aside you can probably get most quality headphones to aim at "perceptionally flat" albeit in some cases it may require substantial lifting power on the front-end, and it may require boosts from EQ that exceed the abilities of whatever set. So in the sense of making say, an MDR-SA5000 sound more like an HD 650 - sure you can venture down that road with reasonable results. But getting it to sound like an HE60? Probably not.

 

Yeah the MDR-SA5000 doesn't seem like a good choice for flat (HE-6) and deep bass. Not only would boosting bass cause a lot of distortion, it seems like it's too open to "contain" the pressure. The other way around would be much easier and yield better results.

I was thinking in terms of treble/midrange response - the SA5000 has a very clean mid/top-end and you can boost or cut there with relative impunity (it doesn't have tons of resonance like the PRO2900 or many Grado designs). The driver can handle a reasonable amount of power and is relatively capable, but indeed, it won't be able to replicate the HE-6 in terms of bass response. This is partly along the lines of my "EQ can only go so far" argument. You can get *some* bits to change (usually the treble is least worrisome), but other bits are restricted by whatever the driver element can do. Basically what came to mind was that you could take their generally "bright" nature, and tone it back more in-line with something like the HD 650 (and I've actually applied such a top-end roll-off to those cans, and they do take on a fairly smoothed out demeanor as a result; still not quite an HD 650).

If you take something that's fairly clean and flat (and that can handle a lot of power) like an ESP or ortho, you can play with it a bit more, but you'll still run into a wall at some point. It'd be interesting to take something like the LCD-2 which is supposed to handle hilarious amounts of power (what is it? 15W or something like that?) and put an amplifier behind it that can actually dish out an equally huge amount of power, and see just how far you can wrestle it into shape - I think you could go further than a lot of purists would expect, but probably not quite as far as being able to re-create any/every high-end headphone simply by changing DSP profiles. I'm also curious if it'd be possible to have consistent results as you changed output intensity, or if it'd have to constantly re-calculate itself (in other words, say you get it to sound like an HD 650 at 100 dB, can it do it at 50 dB?).
Edited by obobskivich - 6/9/12 at 3:42pm

Headroom claims that the square wave response is a substantial factor determining the sound of a headphone. (See http://www.headphone.com/learning-center/about-headphone-measurements.php, towards the bottom of the page.) Is this claim false? If it is false, please elaborate on why the claim is false. Bigshot said on the first page of this thread that "time based attributes will probably be pretty insignificant in most cases" but did not go into much detail as to why this would be the case.

 

If headroom's claim is not false, then square wave response would be another factor that cannot be equalized, to the best of my understanding.

It's a fine test for what it is, but it doesn't relate entirely to real world situations because music doesn't contain square waves. Good speakers sound much better than headphones, but try to get speakers to deliver a clean square wave. All a speaker can make of a square wave is a mess of distortion. Other factors are much more important to sound quality than how well a transducer reproduces entirely artificial and theoretical sound.
Edited by bigshot - 6/10/12 at 1:58am

The "music has no aquare waves" argument is not an original one - read the links I proveded (some of which are published by Tyll) - the sqaure wave test allows phase and frequency to be observed indirectly. You cannot ignore everything but magnitude response and assune you're getting the whole picture.

A good example would be comparing the SA5000 and T70 - both are bright and have bumps around 10k but one rings and sounds harsh. Their CSDs and SRs are quite different; but the FR looks similar.
Edited by obobskivich - 6/10/12 at 7:00am

 

Why look at something that tells something else indirectly if you can look at it directly?

 

SA5000 vs T70 .. their FR imho looks very different which also causes the square wave response to look quite different.

My understanding is that measuring phase is a real PITA, especially for headphones. Again, check the IF links. If this is mistaken, show me someone who is measuring, and publishing, phase response and other information that SWR gets at (even for speakers would be fine - I know such measurements are done for amplifiers, and DtoA converters, and other devices like that). CSD is not an indirect measure - it looks right at resonance and decay; and that's where the SA5000 and T70 argument makes sense; they're both very bright and have a ~10 dB bump around 9k. The difference is the SA5000 doesn't have a few treble ridges up there to compete with the T70. One therefore is just really bright, one is harsh and aggressive and really bright. And I think it would be remiss to declare one of them "better" than the other - there are people who enjoy both equally; is one group "unwashed?" Time domain does matter. Another great example is looking at the "Grado house sound" which relies on resonance. And just because you set this one up for me:

Why look at something that tells something else indirectly (FR) if you can look at it directly (CSD)?

And that's in response to:
"Resonances usually go hand in hand with peaks in the frequency response."

Really, do we have to keep going around and around? If I just tell you that you're the smartest person in the history of the world and that everything you say is the way, truth, and light, will that be enough? [/sarcasm]

Tyll stated in one of the articles I linked (and I won't keep re-linking it because I'm guessing nobody who's arguing with me is clicking anything I link, and is instead just blindly dismissing anything I say) that his measurements are attempting to look at three phenomenon:

And this brings me back to what I've been saying thus far (which I don't think anyone disagrees with, until it means that we can't keep posting about it) - you can modify some, but not all, of this because acting on the time domain is incredibly expensive and in some cases impossible without (potentially extensive/impractical) physical modification to the cans in question (for example, I'm sure it's *possible* to modify the Pro2900 to ring less, but I doubt it's *simple*). If you had a can that measured both absolute ruler-flat from 10-100k and produced an absolutely perfect loopback on CSD, with absolutely perfect phase and zero distortion, then yes with appropriate DSP processing you could probably make it sound like a lot of different cans or maybe even a lot of different speakers. But no such device exists, anywhere, for any amount of money. So it seems like a moot point to debate the merits of such an approach.

In terms of real-world cans, they will *always* have some flaws in those domains; the very best products try to minimize them (and often do a good job of it), but they are still not spot-on perfect. So the best you can do with equalization is attempt to address those flaws, not take a crack at transmutation. What's the problem with this reality?

To use an example brought up earlier, the HD 590 was mentioned - I found some measurements (in Japanese) for that can: http://www.geocities.jp/ryumatsuba/hd590.html

It actually looks pretty clean (I should qualify that Ryumatsuba measurements are suspect in terms of accuracy and what kinds of smoothing are applied - if someone can fluently read Japanese and could find out if he/she talks about their measurement methods, that'd be awesome; if you want to see an example of this discrepancy, there's an SA-5000 measurement on that site - compare it to Tyll's measurements of the SA5000. The assumption is that those measurements are relatively, but not absolutely, accurate.). And I'm guessing that with a bit of tweaking you can get it even cleaner sounding, wrt your individual hearing acuity, tastes, etc. But it still has flaws that prevent it from being transmuted into something else. From user reviews, it's supposed to be similar-ish to the HD 580.
Edited by obobskivich - 6/10/12 at 9:00am

It can be calculated from the magnitude response for a minimum phase system, which most headphones are.

 

Thought we were talking about square waves and measured response, oh well. You cannot just look at some common peak and say they are similar. They both show huge differences in the FR, but in the spectral decay not so much.

 

From purrin's thread (note the different ranges):

 

 

And this only shows stuff above 500 Hz.

 

"Grado relies on resonances", I agree, but those show clearly in the FR.

 

I don't go around and around, I was talking about square waves and not CSD.

 

Which confirms what I've written before.

 

Sure, CSD is nice to have, but you don't need it for EQing.

 

 

Actually, with a bit of DSP knowledge it's quite simple. All you need is the impulse response and a convolver plugin (a couple of free ones available as VST or even as fb2k plugin). With this you can apply arbitrary magnitude and phase response in realtime with quite low CPU usage.

Sure, everything has its limits, but we've talked about that before. An no, it doesn't need to be perfect decay and zero distortion because no other headphone has that and as I've said before, it's questionable whether small differences in spectral decay or distortion are audible.

 

All you need to address is the FR really, assuming you're using a minimum phase EQ and it's not a multiple driver headphone with crossovers or noise cancellation or other weird stuff.

 

What flaws?

Edited by xnor - 6/10/12 at 9:55am

Ah - didn't catch that you were looking at (and only at) SWR.

On to everything else: I'm not on-board with hand-waving when it suits my position, so unless there's something that quantifiably demonstrates the imperceptibility of differences in the time domain it seems like a cop-out answer. Basically "well this part doesn't agree, so it doesn't matter." When you're treating a room, you're absolutely trying to attack problems in the time domain. And it is audible. With headphones it's the same argument - you want less resonance (ideally - if we're going to assume that flat and technically accurate is our goal; if you want whatever "flavor of the month" (or any of the other reasons bigshot gave) then all bets are off). Sure, there's probably some JND limits (just like there is for everything else); I don't know if they've been documented and if so, where they sit, I also don't know if they change relative to frequency, radiation pattern, HRTF, HRPFR (and that's probably not the proper acronym), etc etc.

On the DSP thing - did I say anything that directly disagrees with what you said? I mean really. You also haven't really addressed any of the other points - how is DSP going to magically change anything else? So you can shift frequency and phase, but how are you going to influence resonance and radiation? Is it somehow going to decrease ringing by simply wishing it away? If you weren't so hell-bent on being contrary to anything that's said, I think that either there'd be nothing more to discuss, or we'd be more or less seeing eye to eye.

To the phase Q: neat. But that doesn't really answer the rest of the question.

To the flaws question - I think I'm done with this thread. Literally, I could've said that it's perfectly flat and the ideal headphone (and it may very well be; BigShot says it sounds flat, and he's the only user review I've found for them apart from the GoodCans review (which is just there to hawk Grado cans) - so I have no reason to say it's not flat sounding), and you'd want to attack that assertion as well. Like I said, we just have to going back and forth over minutiae.

Cheers.
Edited by obobskivich - 6/10/12 at 10:12am

Why do really good speakers sound so great when their square wave performance is many times worse than the worst headphones? Square waves turn into a mess of distortion on even the best speakers.

The goal isn't just to get accurate reproduction of theoretical sound. It's to get accurate reproduction of music.

No it isn't. The time problem isn't audible unless you live in an echoey train station and the delays reach very large bits of time. Time isn't audible. The effect the time problems has on the *response* is audible. Frequencies are what we actually hear.

You know how I went about choosing those cans? I didn't look at numbers. I put them on and listened to music and compared them to other models. I was looking for headphones that sounded nice and flat like my speakers. I've put enough work into balancing my speaker setup, I know what balanced sounds like. I don't think many people do.
Edited by bigshot - 6/10/12 at 10:36am

What if resonance doesn't matter. What if magnitude response is the single most important factor to sound quality in a headphone?

 

From the "Headphone equalization using DSP approaches" paper: "Generally, users preferred equalized headphones and it is even more so for cheap headphones."

 

Btw, I haven't claimed that perfect "transmutation" is possible.

 

 

I have to disagree, and others (^) do as well.

 

You can take a look at beyerdynamics website where they write that mechanical/electrical options for changing the FR of a headphone are limited, so they cannot equalize them perfectly (diffuse field EQ). FR is much bigger of a problem than some resonances.

 

Yes, you said "acting on the time domain is incredibly expensive" which is not the case at all. Also that you need a perfect headphone to make it sound like a different headphone and since such perfect headphone doesn't exist it's not possible, or something like that..

 

You don't change resonances or the way the driver radiates, you EQ the frequency response. Say there's a resonance that is distinct to your ear canal, simply cut the affected frequency range and be happy. What's so complicated about that?

 

What are the rest of your questions? Please end them with a question mark.

 

 

Seriously, what are you on about? It has a very clean impulse response, a relatively flat FR, low distortion.. so what are the flaws?

Edited by xnor - 6/10/12 at 11:52am

 

I would like to provide my point of view on square waves and then dig a little into how well they can predict music reproduction.

 

An analog square wave has infinite bandwidth and cannot be reproduced 100% by a headphone. It is possible to generate a reasonable band-limited version of the square wave that the headphone, given its bandwidth, should be able to reproduce to some accuracy. The lower the frequency of the square wave fundamental, the more of its harmonics the headphone will be able to reproduce given the headphone bandwidth limitations (note amplifiers, storage and other things in the reproduction chain are band-limited,... but then so are our ears and likely more so).

 

With that said, a square wave is a relatively easy sound test for speakers and headphones IMHO. For example, if a headphone can only reproduce 5 Hz to 20 kHz, a 300 Hz band-limited (to 20 kHz) square wave will only have 33 tones, and 95% of the power will be within the fundamental and the first 4 harmonics. That is, a 300 Hz square wave, band-limited to 20 kHz, resembles a 5 tone test (a 300 Hz fundamental tone with it's first 4 odd harmonics at 1/3, 1/5, 1/7, and 1/9 of the fundamental  amplitude). Compare that to the complex frequency response a typical sound file will present to the headphone. If a headphone has issues with a simple 5 tone test there, consider the coloration it will strike to more complex music passages.

 

It is therefore my opinion, that one cannot expect accurate reproduction of music from a headphone if a simple 300 Hz sinusoidal along with it's first 4 odd harmonics cannot be accurately reproduced by it... where accurate at some point needs to be quantified BTW (say mean square error, or just listen and compare to a reference).

 

As far as equalization... Well, I don't think it can solve all of our problems... but it can definitively improve the sound coming out of our gear

Edited by ultrabike - 6/12/12 at 3:32pm