I finally had the chance to hear the YH-5K for myself at The Source AV. I used two sources: the first is my personal Japanese SKU Sony Walkman ZX707 via the 4.4mm port, and the second is my Walkman acting as a DAC for the Woo Audio WA22 and its 6.35mm output (seen above). Low impedance output was selected on the WA22. The YH-5K is very light, quite compact, and very comfortable. Unfortunately, they didn't have a Stax SR-X9000 for demo for me to compare them, but I think they're in the same overall league. I suspect that the larger earcups on the Stax will prove more comfortable in the long run as they avoid putting any pressure on certain bones in my cheeks, but I wish I had the opportunity to test (I was told the store's X9K demo unit was stolen in a recent burglary). I think the tips of my ears only lightly touch the insides of the headphone due to the angled drivers and the contact wasn't hard or uncomfortable.
Sound-wise, it's probably the most interesting headphone I've heard in a while. With every test track, I went "huh, I haven't heard it like that before", and yet (or perhaps thus), I continue to be intrigued. It is quite difficult to pin down exactly what it sounds like, because it can provide different, sometimes contradictory characteristics with different tracks. To me, the first thing that stood out was its vocal spatialization. Vocals sound like they're coming from further away, deeper within the soundstage. This is something that I generally like. Yet the Yamaha doesn't do in the same way that other headphones do it. Many of the headphones that I've heard good vocal spatial depth on have had withdrawn midranges with a peak in the presence region. Think of the Hifiman open-back tuning; a dip in the 1-2KHz region and then elevation in the 3-5KHz presence region. That results in a thinner, "breathier" sound that has good spatial qualities while also being present and not muffled. The Yamaha does not do that. It tries to maintain the "body" of the vocals while also placing them further in front of the listener. So vocals don't have the thin and breathy sound of a Hifiman or Sony headphone. It has this more full-bodied or richer tone with more of those throat-generated frequencies present, but the vocals aren't shouty or forward (close to the listener). Some other listeners have called this a "retro" or "vintage" sound, and I can see that. The downside is that some vocals can sound "nasally", which is what happens when there's too much of a boost around 1.5KHz.
Bass is odd, because it can sometimes sound "pillowy", but other times sound thunderous. It doesn't have the immediacy of impact of something like the electrostatics, and its more rounded attack reminds me of a dynamic driver like the Sony MDR-Z1R. On a song like 'Obstacle' by Karra, the large electronic bass hits at 1:33 don't have the sudden and sharp impact that I hear on my Shangri-La Jr, so I was a bit underwhelmed (though I use a Ferrum Erco DAC which has a punchier sound). But on 'my strange addiction' by Billie Eilish, the bass there reminded me of when I heard that song on a pair of Wilson Audio Sasha V speakers driven by Dan D'Agostino power amps. Bass was thunderous on that speaker setup, and the Yamaha came closest to that same thundering presentation of any headphone I've heard it on (though I haven't listened to that track on an Abyss AB 1266).
The combination of bass and vocals is curious. Taking 'my strange addiction' again, vocals are distant yet full-bodied and the bass is powerful and localized slightly closer than the vocals and on either side. This reminds me of how I hear that combination on speakers. Vocals generally appear between the speakers, bass tends to image as coming from the speakers. There's no "between the speakers" effect for me with bass. So I think the Yamaha comes the closest to replicating the sound positioning that I hear with two-channel speakers. Most headphones image the vocals as being closer than the bass, or the bass being too far to the sides of my head rather than in front of me flanking the vocals.
"Did you replicate speaker imaging?"
"Yes"
"What did it cost?"
Well, maybe not everything, but it required quite a few trade-offs. First, tonality overall is pretty weird. I'd generally characterize it as a W-shaped sound signature. If a listener really values their tonal balance, this headphone WILL NOT be for them. The combination of distant yet full-bodied vocals seems to force a strange tuning in the midrange, with the slight nasal tone due to the emphasized body yet withdrawn presence region. Curiously, in one of my piano tracks 'La Gazza Ladra' performed by Chelsea Guo, at around 3:23 at the high point of that piano interlude, there seems to be a bit of distortion in those piano notes. I think the FR and distortion measurements also show some weird tonality and distortion issues. Treble also seemed to have some peakiness, though my usual upper treble peak detector track 'If It Isn't You' by Nurko between 1:29 and 1:57 didn't show any single sharp treble peaks. It's probably a series of smaller peaks distributed throughout the lower/mid treble rather than a single big peak. Jason at TSAV mentioned that there's an oddity at around 4KHz, and that seems plausible, though I didn't run measurements or a sine sweep. Treble definitely isn't as smooth as on a current-gen Dan Clark or my Shangri-La Jr, but never quite reaches irritation levels, at least for me with generally low listening levels. I'll need to listen to more tracks for sure. Bass is very present, and it shows in the FR as well. It's a bit like the MDR-Z1R in bass presentation, like quantity, attack, and decay. It doesn't have the usual "planar speed" or planar bass extension. It's honestly hard to put into words what exactly the headphone sounds like. It just sounds weird, a bit distant on some parts, a bit boomy in others, a bit lean in some parts, a bit strident in others. It's strangely compelling, as for any song in my library I don't know how it will sound on the Yamaha, but I'm not sure if it will hold the appeal after I've gone through it all.
For synergy and stuff, I'll concur that tubes can help. Comparing my Walkman to the WA22, the tube amp softens some of the upper midrange strangeness and stridency. It sounded a bit less weird overall, though I realize that description isn't particularly helpful. The stridency of the upper mids can be a distraction, and mellowing that out helps me enjoy its spatial presentation with less interruption.
For a comparison, I think the Yamaha is like if the Sony MDR-Z1R and the Stax SR-L700mk2 got kitbashed together. The YH-5K has the bass of the Z1R, but a midrange more similar to that of the Stax. The Stax has a similar vocal to bass presentation with bass clearly localized to two position surrounding the vocals, but the Stax vocals are quite forward and focused, while the YH-5K moves those vocals further back. The Stax and the Yamaha also have similarities in their FR measurements in the upper midrange with that 1.5K peak followed by the 2-4K recession. And both seem to slightly benefit from amps with higher distortion, perhaps because distortion products of the 1.5K peak help fill in the 2-3K region.
All in all, I want to hear more of the YH-5000. It's probably the strangest, yet most interesting, headphone tuning I've encountered in a long time. It's certainly distinctive, and since I'm maintaining a collection (as opposed to finding the one headphone to rule them all), I find it strangely compelling.