Solderdude, what is the output impedance of the amp you ran the Superluxes from when you measured them? InnerFidelity's HD681 measurements show a bump in the impedance curve around 40 Hz, right around where the bass on your measurements peaks. I'm wondering if the excessive bass response on your measurements is due to you using a high impedance output amplifier for your measurements.
I also wondering that. The measurement from golden ear and innerfidelity is much more linear in bass reagion, the peak in bass region is only about 5-8db higher than 1khz.
You are both quite right about resistance which can have an effect on certain headphones.
How MUCH effect depends on the impedance plot and how high the impedance gets opposite the value around 1kHz.
The HD681 (from Tyll's plots) increases from 35 Ohm nominal to around 38 Ohm around 40Hz (the resonance point of the HD681).
When we calculate the difference in amplitude between these points using a 120 Ohm output R we get an exact
0.5dB increase in lows and slightly less in the top end (increase in impedance because a voice coil is inductive)
To easy the mind 2 plots taken with (OPA-88 amplifier) in 0 Ohm and 120 Ohm setting.
To be complete I will also add the waterfall plots.
0 Ohm 500Hz - 20kHz:
below 120 Ohm, 500Hz to 20kHz:
below 20Hz to 20kHz 0 Ohm:
below 20Hz to 20kHz 120 Ohm:
Note the time scale differences on the last 2 plots.
I did overlays and they showed no real differences (they do differ slightly)
The reason for that is that the HD681 is acoustically damped mainly and thus electric damping doesn't change much at all.
Damping factor is not relevant for this headphone.
The reason why the mk-III filter (the one with bass reduction) does need a low resistance amp the drive it is because the lows filter section that is affected by a high output resistance.
All my measurements are always done on 0 Ohm impedance.
This doesn't explain the difference in FR though and the differences are rather big.
1: The EVO is not the same as the original one and has more lows. The careful observer will notice it slopes downwards from lows to 1kHz which doesn't make it sound bassy as only a high peak around 100Hz would do that. The slope gives it an overly warm sound, with big bass though not exactly 'tight'.
2: I have modded quite a few (at least a dozen) HD681's and listened to all of them before and after modding (didn't have a test rig then) so had to use my ears. I know not very reliable BUT I had a reference (or multiple ones) and found HD681's could be varying between BIG bass to bass shy (in very few) and in between.
No proof here in the from of plots but only anecdotal, from memory and compared to the HD681 I already owned. I also found that SOME HD681 changed the bass in the first few minutes and the worst one took about an hour to 'improve'. It never reached the same value.
Strange part is this was true for BOTH channels and not imagined (AFAICT) and given the matching on the EVO I would not be surprised if Superlux did some form of matching.
The peak in the highs was about the same in all of them.
Anyway... Tyll's (and thus also Headroom's) plots show a really flat and non bassy HD681 with a roll-off of -10dB for 20Hz.
Have a look at the HD668B, which is supposed to have a 'Balanced' sound with less bass.
It has more bass than the HD681 in that plot.
Those that ever heard them side by side (i have) or have read many impressions will know the HD681 is a LOT bassier than HD668B
Something is really fishy with Tyll's plot but assuming he does know how to measure (I do NOT agree with his 'HRTF compensations though) it stands to reason he had a 'bad' one. Consider the HD681-B and HD681=F which, aside from the trimming, seem otherwise similar makes me suspect that production processes aren't that well controlled OR are very well controlled.
I suspect (not KNOW) they 'match' what they produced and divide them into 'classes'.
The low bass ones for the F versions, the medium ones for the B and the bassy ones for normal version.
It could also be they tightly control the process (for E 15 retail ?) and 'tune' them.
Below the GE plot which, agreed, shows less bass than the one I have but looks MUCH closer to my plots compared to those of GE where the highs are higher in amplitude than the peak in the lows.
a Normal HD681 sounds this way... bass and piercing highs and 'sucked out' in the mids.
It doesn't sound as 'flat' as Tyll's plots seem to suggest.
Superlux's own measurements:
They are close to those of GE
Mine show more (2-3 dB) lows compared to Superlux measurements, 5dB more lows compared to GE and 15dB more compared to Tyll's.
As GE plots have smoothing I smoothed mine as well (1/6 oct instead of 1/3 for GE)
the same driver, except the red trace is the 'smoothed' one.
IMO Tyll's are way off and mine are slightly off to the other side.
The secret may lie in the measurement method.
Tyll, GE and (I suspect Superlux) use expensive HATS and I use a home made rig that does NOT have the shape of a head and doesn't have an ear canal nor pinnae.
These influence the lows measurements substantially (dB's)
Personal audio (Russian site) uses BOTH methods with some headphones.
This is one of them as an example.
You can see the lows are affected MOST and can easily reach 5dB.
If you browse their website (http://personalaudio.ru/raa/opisanie_testov/), using a translator, you will find that in some cases the lows measure quite the same and in other cases substantial differences exist.
All of the above is most likely the reason for the differences in measurements.
measurement method, production spread...
My disclaimer is I do not feel my measurements are absolutely correct for all headphones in an absolute sense but know they are quite close to reality.
In any case my measurements can be used to compare amongst my others.
So the difference between my HD681 and EVO are very valid.
Considering the Bass could be a few dB too much just 'imagine' the EVO with T50RP pads with yet just a tad less bass.
They sure SOUND excellent with those pads for SURE and the 'dark' sound is gone.
P.S. I still not done experimenting and will investigate a passive lows filter.
Also need to try 940 pads and K240 pads.
Sound-wise the easiest fix would be to fit T50RP pads and to remove the felt (it is only glued at the edges of the felt).
www.frontendaudio.com/Fostex-T40RPMKII-T50RP-Ear-Pad-p/9999-13915.htmwww.bhphotovideo.com/c/product/836475-REG/Fostex_1416902601_REPLACEMENT_EAR_PADS_FOR.htmlstore.majormusic.com.au/SpareParts/Fostex/T40RPmkII-T50RP-Replacement-Earpads-pairnetworkaudio.com.au/t40rpmkiit50rp-replacement-earpads-pair/www.fullcompass.com/product/380760.htmlthis guy (Dan):
www.mrspeakers.com/MrSpeakers-Mad-Dog-Fostex-T50RP-Headphonesshould have a few spares he is likely to bin anyway, perhaps ask him.. ?