The difference in sound between headphones equalized to one target

[outerspace]

This recordings are results of interaction between headphones and artificial copy of ear, roughly equalized to one target. In this case the target is flat response so recordings equalized to source audio. Recordings was made by youtuber "oluv's gadgets" on inexpensive measurement stand.

Some problems of this test is low S/N and high distortions caused by POST processing (it should be PRE equalized for best results). If I could make equalization directly on system output and in case of use better mic/ADC (e.g. from GRAS or Bruel & Kjaer) I'll get better results with better signal-to-noise ratio and even lower distortions (phase difference between left and right is the biggest problem).

When you listen original source track you hear only distortions of your audio reproducing chain. When you listen recordings you hear distortions of your audio reproducing chain with addition of distortions of tested headphones and measurement rig. So when you listen recordings and original file you actually compare signal with headphones distortions vs. signal without these distortions.

Test include recordings of Takstar Pro 82 ($80), Creative Aurvana SE ($60), HiFiMan Sundara ($350) and the original track, all hidden under letters (A, B, C, D). Unhidden reference also included.

File is "too large" so I can't attach it to post. Download link.

Any thoughts? Maybe someone who have good measurement rig can made similar test for public with better realization? I can help with it if needed.

 

[solderdude]

This works on the assumption that a measurement method is 'exact' and not position dependent. Alas this is not the case.
So even if one gets 'measured' exactly flat results this doesn't mean they sound exactly the same on anyone's head.
The headphones for sure will sound more similar to each other than without EQ.
It also depends on which transducer one listens to the results.

 

[outerspace]

This works on the assumption that a measurement method is 'exact' and not position dependent.

Yes, FR is position depended. Sometimes hugely depended. But this mean every time you put the headphones you get more or less different sound. From the same device. I don't think this is a problem in this case. I mean you can get same level of perceived similarity from other headphones as from different session with same headphones.

So even if one gets 'measured' exactly flat results this doesn't mean they sound exactly the same on anyone's head.

Yes.

It also depends on which transducer one listens to the results.

It's can be important factor if transducer is not full range and maybe masking effect can affect results. But eventually you compare [transducer distortions] vs. [transducer distortions + recorded headphones distortions] so the difference you hear is [recorded headphones distortions]. I have made some experiments on this topic and can say on pure tones I can detect 1% vs 0% THD as easy as 11% vs 10%, so addition of 10% of THD doesn't matter in this case. But don't know about real music.