1. So far as I understand, the recording you made for the ED9 would be one, if you were playing white or pink noise, with a flat frequency response once you apply this EQ curve, right?
2. Hence, for the specific recording of the ED9 you are not hearing how does the ED9 sound, but instead you hear the exact same sound as in the original audio from the song, right?
3. Why do you apply this ED9 specific EQ curve to the other IEM's recordings? Would you get flat frequency response for all IEM's using the ED9 specific EQ curve? Wouldn't you need a specific EQ curve for every earphone? (This is probably the main aspect I don't understand).
1. yes.
2. right, this effectively eliminates the ED9 from having any kind of character because it IS the source = flat.
3. The ED9 was a reference so it is EQed flat, one of them HAS to be the reference because there is no other way I am aware of making the comparison sound good like the source or what your ear would hear.
All other earphones have the ED9 EQ so you hear the difference in contrast to the ED9, this indeed eliminates the ED9 from the comparison because it is just flat as the source.
But the point is to have an idea what the other earphones sound in comparison to one another.
If I would have chosen the ZN1-Mini as a reference which has absolutely no treble response, all other headphones would have sounded razor sharp in comparison.
1. Ok, then. If you have the uncompensated graph for the frequency response of each IEM, recorded with a hose, that would be analogous to have the recording of each IEM's sound at the eardrum? Is that so? Even when your hose is only 1 cm long, but the human ear canal is 2.5 cm long?
2. The Harman curve stands for the frequency response you would get at the eardrum coming from loudspeakers in a room and being filtered by such room and then by the interaction of the sound with the head, torso and ears (pinna), and then by the ear canal. Hence, you are calling the ED9 to be the most approximate to the loudspeaker case since the ED9's frequency response using your ear canal simulator (hose) gives the most similar curve? Do I get this right?
1. Yes BUT, this is a rough approximation.
Since the original measurements I've posted I changed the hose length to something like 1.4cm.
I player a pink noise on my monitors and looked at a spectrum analyzer to see the response with a 2.5cm hose, it was around 2k, I had to shorten the hose bit by bit to have a 3k peak, this is based on the known fact the the human ear resonates at 3k.
Note that the capsule size of my mic tice the diameter of the canal and that also has an effect on the measurement.
The smallest hose (black) is the hose that simulates the ear canal all other are just adapters to fit the mic.
2. Yes.
But, keep in mind that these readings are far from accurate and can not be considered scientific in any way.
What I tried to see in those measurement is how smooth the frequency response is, because Harman's curve is smooth without peaks or variations.
The smoothest one from the bunch is the ED9 and I think this is why it sounds the most natural practically to us all.
I can also roughly guesstimate from this graph what the Bass to Treble balance would sound like and indeed it reflects what I hear quite well.
Again, in my opinion and from what I see in Harman's curve, it's the smoothness of the curve that matters the most and will translate to how good the earphones sound.
Is this Auto EQ plug in part of Waves plug-ins? Or is it from somewhere else?
http://www.meldaproduction.com/MAutoEqualizer
I agree on the part where the longer hose has a lower resonance frequency, and a shorter one gets a higher frequency. However, I would have predicted the resonance frequency to be lower when the hose gets wider, not narrower, since you have more volume. Do you have an equation, link, source, book, reference for this? (thanks in advance).
Nope, just experimenting.