The high frequency boost in IE response is simply due to the couplers. Assuming people eqed the measured iems flat in ears even after 10khz. It's not going to be flat in couplers.
The coupler is designed to be accurate measuring at reference plane which is equivalent to the second bend in the human ear canal. Etymotic's earphones are designed to be inserted to such depths and is expected to perform the same way as measured. If the insertion depth is changed the response will change in the ear canal.
In measurement rig it's even more difficult. Because the coupler is only designed to be measured at such depth if you don't insert that deep it's not representing the human ear canal. The main issue is the ear canal extension. It's arbitrarily designed and does no good. Remember the standard iec 711 was really really old. At that time no one except etymotic was into accurate iem design. Etymotic used Swizlocki coupler for the initial er4 design. The coupler later on matched the iec 711 standard which happened after the coupler was designed. The coupler was owned by Knowles (yes the famous BA manufacturer) at first then sold to GRAS that we know for this. Anyway, the response is only accurate up to 6khz for shallow inserting type as the coupler's resonance move down around that region from 13.5khz. That's why you see most IEMs have a peak around 7-8khz. And it's both artificial and truthful. In human ear you will hear a peak around 6-7khz but what happens afterwards are not going to be meaningful. Harman's IE target is this thing's fault.Then GRAS went ahead and tweaked the response to reduce the peak instead of redesign the coupler to match the impedance regardless of insertion depth making everything even worse. Jude's z1r measurement is caused by this.
The coupler is designed to be accurate measuring at reference plane which is equivalent to the second bend in the human ear canal. Etymotic's earphones are designed to be inserted to such depths and is expected to perform the same way as measured. If the insertion depth is changed the response will change in the ear canal.
In measurement rig it's even more difficult. Because the coupler is only designed to be measured at such depth if you don't insert that deep it's not representing the human ear canal. The main issue is the ear canal extension. It's arbitrarily designed and does no good. Remember the standard iec 711 was really really old. At that time no one except etymotic was into accurate iem design. Etymotic used Swizlocki coupler for the initial er4 design. The coupler later on matched the iec 711 standard which happened after the coupler was designed. The coupler was owned by Knowles (yes the famous BA manufacturer) at first then sold to GRAS that we know for this. Anyway, the response is only accurate up to 6khz for shallow inserting type as the coupler's resonance move down around that region from 13.5khz. That's why you see most IEMs have a peak around 7-8khz. And it's both artificial and truthful. In human ear you will hear a peak around 6-7khz but what happens afterwards are not going to be meaningful. Harman's IE target is this thing's fault.Then GRAS went ahead and tweaked the response to reduce the peak instead of redesign the coupler to match the impedance regardless of insertion depth making everything even worse. Jude's z1r measurement is caused by this.