[shamu144]

So do the response below 1000Hz if you take the Beyer measurements for the DT48.
 
Quote:

I'd say, the highs and upper mids look similar, but that's about it.

 

 

[aluweer]

 
Quote:

How do you not get distortion from 17db of boost? I had to set my computer volume to 2% to not get distortion using this curve. This EQ made the sound extremely muddy and unlistenable with any genre.  

To avoid effect of sound degrade you need use 32 bit or ASIO output in your soft player. 1bit=6db, in my eq graph I set gain -18db is equal -3bits, then rest 29bit it is overkill for 16 or 24bit signal. Also soft players not have correct volume control, for these purposes need use professional VST equalizers.

 

[aluweer]

 
Quote:

Regarding EQ, I understand it when done to fine tune and match the headphone frequency response to the user HRTF, meaning mostly above 5khz... Other than that, If you don't like the frequency response of the DT48, then just buy another headphone.

I agree.Default dt48 have no competitors for vocal music, nice for chamber classical and many live instruments, but they terrible for contemporary music. For this I always using AT A900, but I was interesting to get this ability on dt48 and i got very nice result. dt48 may have bass like A900 and with more quality (!)

 

[shamu144]

A short read available here on headphone calibration... What retains my attention is the following advices given at the end of the article:
 
Dr. Dalenback offers the following advice:
 
“Be aware that no filter can perfectly compensate for a headphone since the transfer function may vary considerably :

1. * between headphones of the same type (due to manufacturing/wear)
2. * between different persons wearing the same headphone (different coupling to the head/ear)
3. * between replacements of the same headphone on the same person (different coupling to the head/ear)

 
The variation typically starts from 5 kHz and may be as large as +/-15 dB for the highest frequencies (some models are worse than others). Just like HRTFs, headphone eqs are ideally personal.
 
Headphone_calibration_for_auralization
 
I wouldn't be surprised if this was one of the cues to explain why the DT48 can generate completely different perception between individuals (many reports find it rather harsh sounding and treble heavy). Its tortured frequency response above 1000Hz is similar to that of a free field calibration which is known to provoke a much higher deviation between individual than a diffuse field calibration. So, more than ever, YMMV.

 

[Deep Funk]

Oh my, more variations...

 

[svyr]

A short read available here on headphone calibration... What retains my attention is the following advices given at the end of the article:
 
Dr. Dalenback offers the following advice:
 
“Be aware that no filter can perfectly compensate for a headphone since the transfer function may vary considerably :

1. * between headphones of the same type (due to manufacturing/wear)
2. * between different persons wearing the same headphone (different coupling to the head/ear)
3. * between replacements of the same headphone on the same person (different coupling to the head/ear)

 
The variation typically starts from 5 kHz and may be as large as +/-15 dB for the highest frequencies (some models are worse than others). Just like HRTFs, headphone eqs are ideally personal.
 
Headphone_calibration_for_auralization
 
I wouldn't be surprised if this was one of the cues to explain why the DT48 can generate completely different perception between individuals (many reports find it rather harsh sounding and treble heavy). Its tortured frequency response above 1000Hz is similar to that of a free field calibration which is known to provoke a much higher deviation between individual than a diffuse field calibration. So, more than ever, YMMV.

you'd think 3 is not that relevant the longer you own a pair of hp - since you'd explore their sound in different positions on your head and pick a level of fit/sound/isolation you want

man, you really gotta start distinguishing between DT48 and DT48e and the newer versions of DT48e. I presume the post above was about DT48e (new)

 

[shamu144]

I agree with you that 3 can be reduced to less significant influence. This statement was for headphones in general, not specific to the DT48. I think it is worth to put a real value to those HRTF variations between individuals, because +/-15dB is indeed huge, probably more than many imagined. Not saying it happens with all headphones, nor at all frequencies though. 
 
Quote:

you'd think 3 is not that relevant the longer you own a pair of hp - since you'd explore their sound in different positions on your head and pick a level of fit/sound/isolation you want

man, you really gotta start distinguishing between DT48 and DT48e and the newer versions of DT48e. I presume the post above was about DT48e (new)

 

 

[shamu144]

Take a look at those measurments for the DT48 (I assume the E model but not sure) published in a technical reports from the equivalent of the Japanese Audio Society:
 

 
When using a standard ear coupler, the frequency response below 100Hz follows exactly the same pattern as measured by Tyll on InnerFidelity. Fascinating. So it really seems that when the seal is perfect, a strange resonance can occur around 80Hz. Curiously, Beyer own coupler measurement doesn't show this. Of course, the actual-ear response measures nothing alike, but the sharp roll off in the lower frequencies seems a little excesive though I would guess due to a bad seal.
 
However, it is striking to look at the standard deviaton measured for the actual-ear response in the sub 100Hz frequencies for those 4 subjects: up to 10dB differences. Here we have in my opinion a sheer evidence of the importance of an excellent seal needed to bring the low frequency to decent levels.
 
And look at the standard deviation for the actual-ear response above 5khz: up to 10dB as well meaning consistant diffferences in individual perceptions. And this was only involving 4 human subjects... I guess we can assume that the standard deviation with a more representative population could easily reach the 15dB mentionned above.
 
The full document source is available for free in pdf here
 
Where I get lost is that looking at the actual-ear response measured, it doesn't look very similar to the free field design goal shown in the previous page. And in fact, according to this other study available here for free using the loudness comparison method, the DT48 do not really appear as a free field calibrated headphone, nor diffuse field either. Deviations in the frequency response are perceived in both cases...
 

It is just amazing to see how complex those things can become. But it is also amazing to realize Beyer designed such a natural sounding (those who agree at least) headphone 60 years ago, before the age of computers, when most of those studies and concept were not even available.

 

[Tyll Hertsens]

Nice post, thanks for that! 
 
I think you need to get off the "free-field" design goal thing.  What they want to do is make headphones that sound like real sound does when you put your head in it. The free-field thing is a particular curve you get when sound is coming from directly in front of you in an anechoic environment.  That's a very particular HRTF that rarely occurs in real life.  That was a starting point quite some time ago.  Since, they've figured out that designing cans with that simple goal doesn't really work and nets headphones that don't sound right.
 
What does sound "right?" Heck if I know ... and I don't think "they" (whoever that is) knows either.
 
 

 

[shamu144]

Thank you Tyll. Hey, I didn't have time yet to post that on your site but will do soon.
 
You are certainly right when you say that "they" probably don't know exactly what frequency response model sound most natural to the ears, especially considering large deviation in individual HRTF which complicate things even more. At least, G.Theile from the AES did an attempt to standardize the diffuse field response for studio headphones (diffuse field is less subjetc to interindividual HRTF differences). A link to that paper can be find here. I would guess each manufacturer have their own understanding on what the frequency response should look like and must balance nowadays the demand of their acoustic engineers with the requirements of their marketing department. Something the DT48 was certainly not spoiled with.

 

[shamu144]

More reflexions on sound perceptions with headphones, and more particularly, the lowest frequencies... If you look at the latest equal loudness curves published (ISO:2005) and threshold of hearing in the low frequency, you come up with the following figures:
 

 
 
Well, I don't know for you, but 78dB for 20Hz to be audible is indeed quite loud... How many headphone do reproduce 20Hz sounds at much lower average listening levels than that... Plenty. Speaking of low frequency reproduction accuracy. So even in real life condition, it sounds like the lowest frequency are not easily heard, or at least, not without reaching pretty loud levels. And still, many complain of the bass roll off for the DT48...

 

 
I say it is not very surprising to have a natural roll off perception of low frequency if we look at the equal loudness contours curves: And yet, many manufacturers do attempt to trick the headphone low frequency response for plenty of slam... I guess it all depends on what you value in the music, accuracy and natural timbres, or fun.

 

[Soaa-]

I don't think the dummy head measurements take into account equal loudness contours. The actual perceived response would be a compound effect of the dummy head response and the equal loudness contour.
 
With that said, I'm no longer convinced of the DT 48's neutrality in terms of frequency response. Granted, its presentation of mids is absolutely phenomenal, and I have no complaints regarding the highs, but I've always had trouble with its reproduction of the low-end. The way it presents bass lines, either synth or acoustic, just isn't satisfying. It doesn't sound real, like the mids and highs do. The measurements now prove that.
 
With that said, the DT 48 still occupies a space in my collection of headphones.

 

[Deep Funk]

Have you tried various set ups with amplification and DACs? Amplification especially can sometimes make a bigger difference then you expect. 

 

[Soaa-]

I've already discussed source differences in this thread. Came to the conclusion that my Samsung P2 produced bass roll-off, but not my Clip+, Yamaha power amp nor my Yamaha DAC. I'm sure the electrical signal going to the cans is already as flat as can be; the problem lies in the fact that the DT 48 is too dependent on a perfect seal to achieve perfect bass extension.
 
I'm looking at the Ultimate Ears Reference Monitors now. Funny thing is, they're described in the same way the DT 48 is described here: sonic chameleon, perfectly neutral and shining mids. Steep investment though.

 

[judy1992]

Everything in audio is subjective, even with measurements.. It's like a dog running in circles trying to catch it's tail.. It's wasted energy. I'm sure the UE will have better lows & more accurate bass lines. Maybe even better, period.. With apologies to DT48 fans. Not sure how accurate or neutral the 48's are.. Too many variables.. But NAGRA thinks highly of them.. & they are used to test peoples hearings, used by labs, DR's,, testing, & acoustic investigation. So maybe they are the real deal?