so there is definitely something very deliberate about the midrange dip.
I believe so. I've been using the Sony MDR-Z1R as my main closed-back headphone for the past two weeks and I've found that its midrange dip (between 4KHz and 8 KHz) is part of what gives the Z1R its very interesting and compelling spatial qualities. I went into more detail about it on the Z1R thread in
this post. The gist is that the massive dip in the mid-treble forces vocals behind the main set of instruments, which matches the layout of a choir in a symphony orchestral performance. In such performances, the choir is usually positioned behind the orchestra, and the Z1R emphasizes this quality in a very immersive way. All of the spatial cues are exaggerated in a way that makes it seem like you're standing in the conductor's position and can hear the orchestra all around you, the soloists up front, and the choir in the back. It's not actually a matter of soundstage size, but of spatial positioning within that space. This works extremely well for orchestral music and I feel very immersed in the music. Everything sounds right about its tonal balance in combination with the spatial imaging in those orchestral recordings.
One of the things that the treble dip does is that strings and vocals sound somewhat muted or recessed. This can result in the Z1R sounding strange in studio-recorded music where the inherent spatial properties of the track don't align with the intended spatial tuning of the headphone. If I EQ out that treble dip, then strings and vocals cut through the mix more, they're positioned closer in, and they have significantly more texture, but those spatial qualities are diminished or lost. Then you're positioned like an audience member with the music in front rather than all around. And part-way EQing out that treble dip is also difficult, because half-measures often just result in a very strange sound where vocals and strings are more prominent, but their prominence doesn't make sense within the spatial presentation. It wouldn't make sense for the violins to be so clearly heard when the bass drums and brass are booming and blaring at the same time.
With the Yamaha, I'm guessing that they tried to go for a similar sort of spatialization, but with a key difference in how the midrange is presented. The YH-5K has a big FR spike at 1.5 KHz vs the Z1R's spike at 3 KHz. This is interesting because it's like the emphasis is on one harmonic series lower (fundamental vs 2nd harmonic, or 2nd harmonic vs 4th harmonic, etc.). 1.5 KHz is roughly where many of the throat-generated frequencies lie in vocals, so it should result in a richer, throatier tone to vocals rather than a "breathy" tone like the Z1R or many of the Hifiman models. I think, from my demo, that this is pretty accurate to the sound; the vocals are distant vs the instruments, like the treble dip would imply, but they are deeper/richer. However, this combination can sometimes sound like the vocalist is singing through a megaphone.
I've also drawn similarities with how sound is presented on the Stax SR-L700mk2. That one actually measures along the same lines as the Yamaha with a peak at 1.2 KHz and a general recession between 2-4 KHz. The Stax thus also has an interesting spatial presentation to vocals vs instruments, kind of a triangular soundstage with vocals far away at the point (imagine your head inside the triange and looking towards one of the points). But it's not as extreme, so it doesn't have as much of that megaphone quality as the Yamaha.
, and neither does the bass roll-off like the graph suggests. Although I'm finding that only my SS amp is able to bring out the sub-bass the yh5k can be capable of.
