Sean Olive on the lower-mids dip in the Harman Target curve:
I found and participated in an interesting twitter thread. Thought I'd repost and summarize here for discussion purpose:
@dealuxmusic:
The worst part of the Harman IE target is the bass by far. The way the low shelf is achieved makes no sense. The result is a headphone that can sound too bassy and bass light at the same time, sometimes even in the same song.
@seanolive:
It's based on science, not the arbitrary opinion of an audiophile. We did controlled listening experiments where 10 trained listeners could adjust the level and frequency of a 2nd order LF shelving filter. The IE Harman Target was defined from that study. We validated the Harman IE Target by testing it against 30 other IE headphones with 71 listeners. It was preferred over the other headphones. You don't have to like it. But we have some hard unbiased data that many people do. And one of those 30 headphones in the test included your favorite IE headphone and I'm sorry to say that listeners preferred the Harman Target over that too.
@dealuxmusic:
The mids and highs are perfect because in-ears can typically achieve a much smoother response (perhaps with multiple drivers) but there are many people in this hobby who feel that the sub bass is excessive relative to the 100-300 Hz region. The way it feels to me subjectively is like a hollowness in the 100-300 Hz region preceded by a very resonant sub bass region (that muddies up the texture/timbre of bass sounds). I was unable to achieve such an effect with EQ on over-ear headphones by following the OE target.
@seanolive:
Your taste in bass below 100 Hz is very personal and depends on quality of the reocording, age, training, hearing loss. That's why we have filters included with the N5005 and EQs in our apps. Personalization of sound will become a more common feature in the future. It may sound hollow to you because 90% of all IE headphones have too much energy between 200-500 hz. So you are accustomed to that excessive bloom or coloration. When listeners had that option they preferred the lower mids to be flat to 125-200 Hz with a rise below that.
@dealuxmusic:
I think the shelf would be fine as it is at +6 db. +10 to me creates too much contrast between sub and the region where kick drum punch comes from in a lot of music. So it can feel like listening to speakers in a room where the sub is out of control.
Me:
Absolutely agree with OP on this very criticism. I sold my N5005 after about a month after I realized this gimmick in tuning. It gives a clean powerful bass not bleeding into the midrange at first listen. However, when you (rightfully) leave the midrange flat and (rightfully) only boost the bass, the lower-mids *will* be relatively quiter. The "correct" design choices inevitably lead to a hollowed-out lower-mids if the overall SPL is constant. This is common sense. How could it be when you were scientific? Well, my guess is that in your listener's test, the listeners were given the option to boost the bass, without controlling for overall SPL, hence they didn't hear the hollowness in the lower-mids, only the excitement from the extra bass.
@seanolive:
We did a bass adjustment study where we compared adjustments made with and without loudness normalization. Negligible effect.
Original thread: https://twitter.com/dealuxmusic/status/1351162757245239298
I found and participated in an interesting twitter thread. Thought I'd repost and summarize here for discussion purpose:
@dealuxmusic:
The worst part of the Harman IE target is the bass by far. The way the low shelf is achieved makes no sense. The result is a headphone that can sound too bassy and bass light at the same time, sometimes even in the same song.
@seanolive:
It's based on science, not the arbitrary opinion of an audiophile. We did controlled listening experiments where 10 trained listeners could adjust the level and frequency of a 2nd order LF shelving filter. The IE Harman Target was defined from that study. We validated the Harman IE Target by testing it against 30 other IE headphones with 71 listeners. It was preferred over the other headphones. You don't have to like it. But we have some hard unbiased data that many people do. And one of those 30 headphones in the test included your favorite IE headphone and I'm sorry to say that listeners preferred the Harman Target over that too.
@dealuxmusic:
The mids and highs are perfect because in-ears can typically achieve a much smoother response (perhaps with multiple drivers) but there are many people in this hobby who feel that the sub bass is excessive relative to the 100-300 Hz region. The way it feels to me subjectively is like a hollowness in the 100-300 Hz region preceded by a very resonant sub bass region (that muddies up the texture/timbre of bass sounds). I was unable to achieve such an effect with EQ on over-ear headphones by following the OE target.
@seanolive:
Your taste in bass below 100 Hz is very personal and depends on quality of the reocording, age, training, hearing loss. That's why we have filters included with the N5005 and EQs in our apps. Personalization of sound will become a more common feature in the future. It may sound hollow to you because 90% of all IE headphones have too much energy between 200-500 hz. So you are accustomed to that excessive bloom or coloration. When listeners had that option they preferred the lower mids to be flat to 125-200 Hz with a rise below that.
@dealuxmusic:
I think the shelf would be fine as it is at +6 db. +10 to me creates too much contrast between sub and the region where kick drum punch comes from in a lot of music. So it can feel like listening to speakers in a room where the sub is out of control.
Me:
Absolutely agree with OP on this very criticism. I sold my N5005 after about a month after I realized this gimmick in tuning. It gives a clean powerful bass not bleeding into the midrange at first listen. However, when you (rightfully) leave the midrange flat and (rightfully) only boost the bass, the lower-mids *will* be relatively quiter. The "correct" design choices inevitably lead to a hollowed-out lower-mids if the overall SPL is constant. This is common sense. How could it be when you were scientific? Well, my guess is that in your listener's test, the listeners were given the option to boost the bass, without controlling for overall SPL, hence they didn't hear the hollowness in the lower-mids, only the excitement from the extra bass.
@seanolive:
We did a bass adjustment study where we compared adjustments made with and without loudness normalization. Negligible effect.
Original thread: https://twitter.com/dealuxmusic/status/1351162757245239298
Last edited: