[Catharsis]

Quote:

Really? I thought it would add a virtual 100 ohm load to the headphones.

 
Nope, I don't think so (though I could be wrong).  There apparently is a way to increase the load of the headphone (artificially make a 250ohm DT880 become a 600ohm DT880) according to Jan Meier on his website, but my understanding that a resistor / adaptor would increase the theoretical output impedance of the amp, making the DT880 midbass get muddier and muddier as you increase resistance.
 

 

[Shike]

Quote:

The difference in treble ranges from 2.5 - 5dB which is significant and in my opinion is not due to manufacturing tolerances. I tested the 2 versions before using the same 250 ohm set and a different 600 ohm set and the treble difference was still there. If they were closer, like 1dB, then one could argue it's due to manufacturing tolerances.

 

What's the process you use for measuring?  Is it a multiple run with averaging or something else?  What is the accuracy from run to run?  Is there anyway headphone positioning could have caused this?
 
I know I'm asking a lot of questions, but it's surprising to see such different results in relation to Headroom.

 

[Donald North]

 
Quote:

Can you make some graphs of each headphone separately with and without the resistor?
 
I would also be interested in seeing data using an amp with a higher output impedance, as that would (I think?) have a noticeable affect on damping factor, although I'm not sure if that would show up on an SPL graph.

Here are the individual frequency response graphs without level matching offsets:
 
DT880 250 ohm version, 0 ohm source impedance

 
DT880 600 ohm, 0 ohm source impedance

 
DT880 250 ohm, 100 ohm source impedance

 
DT880 600 ohm, 100 ohm source impedance


These are all the measurements I have. Yes, the added series resistance interacts with the headphone's impedance, modifying the frequency response and reducing the low frequency damping. However with the 100 ohm resistor, I don't see any low frequency peaking which is an indication of the bass being underdamped.
Quote:

What's the process you use for measuring?  Is it a multiple run with averaging or something else?  What is the accuracy from run to run?  Is there anyway headphone positioning could have caused this?
 
I know I'm asking a lot of questions, but it's surprising to see such different results in relation to Headroom.

Headroom uses different equipment than I have. For my tests, 5 measurements were taken and averaged for each curve. I positioned the headphones as identically as I could for each measurement. There is some difference in low frequency SPL in the range 20-80Hz on the 600 ohm curves which may be due to the 18.5Hz data point increments and MLS signal. It's easier to get consistent low frequency measurements using swept sine.

 

[Catharsis]

Wow!  Thank you for those measurements. 
 
Right off the bat, I would have to summarize the DT880 250ohm and DT880 600ohm graphs as "nearly" identical insofar as FQ response is concerned.  Sure there are slight fluctuations in the order of 1db in the odd spot, and I'm not sure whether such differences could be accounted for by measurement deviations.  I would have to conclude that if there are REAL differences between the 600ohm / 250ohm versions in terms of sound quality that they are incredibly minimal, and certainly not different enough to account for the "wild" claims of how utterly superior the 600ohm headphone is compared to the 250ohm headphone.
 
If there are perceived sonic differences, I would likely attribute it to source / amp variations between users, or placebo.  You have provided remarkable proof that both impedance versions of the DT880 have nearly identical quantities of bass, midrange and treble.  Great JOB!

 

[Donald North]

When level matched, the two versions measure nearly identical in the bass and midrange but diverge in the top two octaves. At some frequencies there's a 5dB difference with the 250 ohm version generally having more treble energy (dotted yellow curve below). Depending on music spectrum and hearing acuity, such differences are easily noticeable. However I don't see one version being obviously superior to the other. To me it's personal preference.
 

 

[sid1712]

Thanks for the graphs. Does it apply to the DT770's too ? Since some people have reported huge sound differences in the sound quality of the DT770 250 and 600. 

 

[Donald North]

I've never had the DT770 600 ohm to listen or test, so I can't comment how it compares to the 250 ohm. Now I have heard the 250 ohm versus the Pro 80 ohm version and preferred the latter.

 

[maverickronin]

The 770s =2231&graphID[1]=713&graphID[2]=2131&graphID[3]=&graphType=0&buttonSelection=Compare+Headphones]vary greatly between the different versions.  =953&graphID[1]=963&graphID[2]=2751&graphID[3]=&graphType=0&buttonSelection=Compare+Headphones]Much more than the 880s.

 

[skeptic]

Just wanted to say thanks to Catharsis for starting and driving this thread and to Donald North for all of his insight, explanation, and testing.  This is really one of the better discussions I can recall having read on this forum, and as a long time fan of the DT880's, it answers some questions I've had for quite a while about what I might really expect to hear if I were to pick up a pair of the 600 ohm version.
 
The test results you posted do raise one further question in my mind.  Is there any possibility that the frequency differences between the 250 ohm DT880's and the 600 ohm DT880's would translate into a difference in the timbre of treble instruments?  One of my few issues with my DT880's, when jacked into the mid-fi rig in my signature, is that trumpets and violins in certain recordings lack texture and sound a bit electric.  My Sennheiser HD650's, by contrast, do not have this issue when playing the same recordings via the same source and amp.  Is it possible that the reduced treble energy in the 600 ohm Beyers would translate to a perceived more natural timbre (to my ears, in my rig), or would I be more likely to simply hear softer trebles of the same timbre as the 250 ohm version?
 
Thanks in advance for any thoughts you have on the matter. 

 

[xnor]

Quote:

The 770s =2231&graphID[1]=713&graphID[2]=2131&graphID[3]=&graphType=0&buttonSelection=Compare+Headphones]vary greatly between the different versions.  =953&graphID[1]=963&graphID[2]=2751&graphID[3]=&graphType=0&buttonSelection=Compare+Headphones]Much more than the 880s.

I also noticed this. Might have something to do with their closed construction.
 
Have you taken a look at the 50 Hz square wave measurements? The 600 ohm version overshoots a lot more than the lower impedance versions. 
What does this tell us about "precision"?

 

[maverickronin]

If you believe some people, =863&graphID[1]=2231&graphID[2]=2131&graphID[3]=&graphType=3&buttonSelection=Compare+Headphones]that overshoot is precision.  I remain unconvinced.

 

[upstateguy]

Would someone care to summarize?  A lot has been said.
 
@ Donald North:  Do you consider the 80 ohm the best of the 880 line regardless of which amp is used?
 
USG

 

[Donald North]

 
Quote:

@ Donald North:  Do you consider the 80 ohm the best of the 880 line regardless of which amp is used?
 
USG

Sorry for the confusion: I was referring to the DT770 Pro 80 ohm. I like them over the 250 ohm and they're on my Christmas wish list

  I haven't heard the 32 or 80 ohm version of the DT880.

 

[chinesekiwi]

frequency graphs are all well and good but does it tell anything about the quality of the driver?
 
Any electrical engineer care to chime in?
As people know, sound signature is not equal to sound quality.
I'm not doubting the graphs, I'm just doubting the conclusions people come up with using those graphs.

 

[xnor]

There is nothing else to really worry about I think.
Headphones are minimum phase and distortion usually is very low.
 
FR > the rest
 
Different headphones have successfully been emulated by adjusting the frequency response only. (equalization)