I've come to the sad realization that I prefer headphones for enjoying music...

(Think it's partly a studio-psychological-damage. I listen to speakers, I move my head, and my mind goes "Whooooooa, what happened to the phase there, and why did the tonality suddenly shift...")

Pretty much. 800S is my favourite pair of cans for listening to music, and that's all genres, a lot of it being in the synth/electronic category. They sound great (to my ears/liking, not the flattest, not the most of anything, but they reproduce music with great detail and an effortless presentation), they are comfortable to wear, they are relatively lightweight, they are easy to drive well.

Might still get another pair of higher end cans for studio work (use 660/280pro). But for enjoying music, unless I run into something I find much more enjoyable to listen to music in than my 800s, I'm not actively searching. There's definitely "different", and I wouldn't mind a pair of STAX:es, and a pair of great planars, but so far I haven't found them overall more enjoyable than the 800S. (Hell, if music is less than stellar production-wise, I sometimes even would pick the HD660s over pretty much everything else...)

Binaural doesn’t translate well to speakers. (And it’s slightly flawed in that it still doesn’t model your specific head/ears so without significant wizzardry it won’t be 100%)

But yes, as more and more music is listened to through headphones, I’m kind of wondering this as well…

There are a lot going on with stuff like Atmos though, and virtual studio simulations for mixing in headphones, so we might see interesting things in the future.

Re repriduction of different kinds of music I think the challenge is exactly the same. The challenge is to reproduce a signal with flat frequency response, correct transient and phase response, with low distorsion. As long as you do that, you reproduce every kind of music well. There may be tradeoffs that ate more important in some kinds of music. Soundstage/positioning perhaps being more important than sub bass in jazz/chamber music but (perhaps) not in EDM for instance. But I still think the challenge is the same. Have a good enough system and everything will sound good on it. Or be correctly reproduced atleast, which may not be what sounds the best, but that’s a whole other can of worms.

That can of worms may also be part of it, because, we might for the most part not actually want the recorded music to sound like the real thing, we want larger than life. Have an anecdote from a former colleague who's worked a lot on recording classical music for radio and TV. He had this story where he worked on an audiophile symphonic recording. They used a small set of really high quality microphones, set up as a blumlein pair as main source + some more microphones to capture and be able to adjust tonal balance, width and room in the final result. Apart from slight corrective EQ, they weren't supposed to process the sound at all, it should be all natural. Compression or artificial reverb were strictly prohibited. They did several different mixes with different mic balances, but the producer weren't satisfied and still thought it sounded unnatural. So, without telling anyone, they sent the mix through a Manley VariMU compressor, just compressing a few dB:s at most. That's the version that got released...

> For orchestral sound like classic music, jazz music, the challenge is higher as the goal to reproduce original sound

I think this is incorrect. It's not any more difficult to correctly reproduce a recording of an orchestra than of pop/rock or even entirely DSP-generated sound. The challenge is probably rather in recording the music in the first place.

If you would record an orchestra with a high quality binaural rig, run some 3d-scanning + DSP to correct for the shape of your head and ears, and listen through high quality headphones, you would probably not be able to tell the difference (atleast not if we assume we could achieve a truly blind comparison, that is, you would not see whether it was the orchestra playing or not, and that you wouldn't feel if you had headphones on or not...).

You only have two ears after all, and they just react to sound pressure level changes in a very small volume of air.

It’s a question of amplitude. In theory, a single driver is optimal, since it’s a point source. Creating a driver capable of loudspeaker sound pressure levels, that can reproduce the entire human hearing frequency range with high resolution and low distorsion is tricky (almost impossible). Doing the same at headphone audio levels is much much more doable, probably even easier than overcoming the problems with multiple drivers and extremely short listening distances. With multiple drivers you also have crossovers, with their problems and added distorsions and phase issues.

Also, you have technologies as planar magnetics or electrostatics, that are very difficult to use well at loudspeaker volumes, but extremely viable at headphone levels. (Electrostatic designs are also an example of a seriously high-performing single driver loudspeaker, albeit one that requires BIG speakers if you want full range bass reproduction at somewhat respectable levels.

In the loudspeaker world, some companies are jumping through a substantial amount of hoops just to get speakers to behave like a single point source even though constrained by the need to use multiple drivers, like Genelec TheOnes (three-way coaxial point-source design with substantial audio and DSP trickery...).

Most (all?) high end headphones are single-driver designs. If there would be substantial advantages to instead using multiple drivers I'm sure many high-end headphone designs would, but they don't...