Please bear with me, I'd consider myself still a noob when it comes to desktop DAC/Amps. Although I've heard a few DAC Amps in quite a few public headphone festivals, most aren't heard with my headphones nor my source tracks of music. And I'm still mostly learning from the gurus here in this forum and from Japan Head-Fiers.
So this initial impressions will be somewhat unstructured and a lot of it will be based on a DAC/Amp I'm familiar with, the Benchmark DAC1 Pre. I won't get into the specs of the HP-A8 as you can find the info easily yourself. Or please go here for its specs :-
A little bit of background...
of my pursuit of this DAC Amp. I've owned the Benchmark DAC1 Pre for about 7 months and although it's a great DAC Amp, I was getting somewhat bored of it's 2D sound - both as as DAC to my Stax setup, and as a DAC Amp for my Fostex TH-900 headphones. Furthermore, since I took my Benchmark DAC1 Pre & Stax setup to the Fujiya Spring Headphone Festival 2012, other chaps from Head-Fi brought in other DACs (Yamamoto, Eximus, etc.).
Last month my wife & I went to Fujiya Avic to demo the HP-A8 but in somewhat more "controlled" environment. I brought my Benchmark DAC1 Pre, my DX-100 with optical cable, and my Fostex TH-900. My wife (who's not technically inclined but has as sharp ear) was my "other pair of ears". Spending about 30-40 mins there with Fujiya's HP-A8 demo, we listened to a few tracks rather carefully and concluded then the HP-A8 was more detailed, more 3D and basically a leap improvement over the Benchmark DAc1 Pre.
As such I went back the next day and bought the last HP-A8 they had on the shelf.
To my disappointment, at first listening, the new HP-A8 didn't sound the same as the demo unit. It sounded almost the same as the Benchmark DAC1 Pre. And so the burn-in began. Although SQ did change somewhat after 8 hrs, it didn't change much. A few days later and 100 hrs burn-in later, the SQ did improve incrementally but still not the vast difference both of us heard at the store. At that point, I felt the purchase wasn't worthy of an upgrade that I was looking for. However after talking to Currawong, whom advised burning in to 350 hrs at least, and we talked about power conditioners, strips and cord, I further ended up picking up the VH Audio Flavor 4 power cord, and Oyaide MTB-4 power strip.
The result after 250 hrs of burn-in and power source upgrade for the HP-A8?
DX-100 with 16/44.1 -> 24/192 tracks of 80's pop, classical, jazz, and country
Headphones/earphones used include the Fostex TH-900 and FitEar TO GO! 334
Switching between the Benchmark DAC1 Pre and HP-A8 involves the headphone jack naturally, and optical cable from the DX-100.
Again, as mentioned before, this has a lot to do with comparing it against the Benchmark DAC1 Pre that this DAC Amp is supposed to replace.
Despite the long in and power source upgrade, for most of the 80's pop music (Michael Jackson's Thriller, Footloose, and Top Gun all in 24/192) it was very difficult to differentiate between the two DAC Amps. Even Michael Buble's It's Time (16/44.1) with a fair decent amount of vocals and acoustic instruments, the HP-A8 and Benchmark were very similar. At times it was difficult to tell the difference between the two. This was tested with both headphones/earphones and confirmed by my wife.
The difference between the two was much clearer with Isaac Stern's Four Seasons (24/192), Lana Del Rey's Born to Die (24/44.1), Buena Vista Social Club (24/96), and Anne Murray's The Best...So Far (16/44.1). The presentation of the HP-A8 sounded more 3D, smooth yet detailed, and more immersive. The difference isn't night and day, but at the same time isn't so subtle - it's obvious enough to pick up after critical listening. Where in pop music or modern vocal jazz (Michael Buble) frequency response sounded pretty much similar across both DAC Amps, for tracks with strong vocals, or natural acoustic instruments the bass seem to extend a little more deeply and trebles sound more clear and transparent. But it's more the immersive 3D presentation that captured my attention initially.
Going by memory of the Eximus DP1, although wasn't as detailed HP-A8, I would place in between the Benchmark DAC1 Pre and the DP1 in my DAC Amp scale for specific genre and type of music. But for the 80's pop tracks I had, as you can guess I'd treat the Benchmark and HP-A8 equally.
Another note I'd like to add though, although the HP-A8 and Benchmark sounded rather similar, the HP-A8 was just a tad more mellow than the Benchmark. The way I felt was the Benchmark had more of a slam-in-the-face impact but somewhat 2D, whereas the HP-A8 was more 3D, it was also somewhat more subtle and mellow in impact. For example, with some pieces of Vivaldi's Four Seasons, I was anticipating the next (chord?)<whatever!> to come in with a wham (at least with the Benchmark) but with the HP-A8, it just creeped subtly.
However, on the other side of the coin, with the HP-A8's 3Dness, found I could hear some instruments in some classical pieces more clearly. Whereas with the Benchmark's 2D slam, that instrument drowned in the slam and everything sounded more congested. This is more apparent in lower volumes.
Finally, I also did test the DSD capability of the HP-A8. After some fiddling with filename lengths and DSD formats (only DSF accepted, not DFF!!), the HP-A8 does play DSD files very nicely. Across the board on clarity and transparency, the same DSF file sounded better via the SD card than via Audirvana Plus (which does convert to PCM in real time before output via USB). I'm quite satisfied with the SQ of DSD playability.
Non-SQ Aspects of the HP-A8
Although DSD sounded nice, getting it to play it is a real pain. Firstly, only DSF is accepted, not DFF. Luckily my SACDs are in ISO as such I just have to re-extract into DSF. Then the next problem is a 30 character limit in the file name. For some strange reason even for some files with less than 30 characters, the HP-A8 wouldn't recognise it. Also all the DSF files have to sit in one folder. So really the DSD capability is really more a proof of concept than for practical use. Fostex is supposedly working on a DSD over USB driver but they're having some issues. Once they've released it, I'd be much more satisfied with my HP-A8.
Also having only unbalanced RCA output is a little tedious. I do wish Fostex included balanced XLR which I'd feed into my Stax setup. But for now, I have to use unbalanced.
Overall, I think the HP-A8 is a decent desktop DAC Amp in the similar grade to the Benchmark DAC1 Pre. As to whether the HP-A8 is worth it's price, I think it really depends on where you live. In US, the HP-A8 is USD$2000, whereas the Benchmark DAC1 Pre is $1400. In that scenario, I personally don't feel the HP-A8 is worth the $600 difference of my money.
However, as I'm in Japan, where the HP-A8 is approx USD$1300, and the Benchmark DAC1 Pre is sold here for USD$1730, the HP-A8 worth it's value. If Fostex gets the DSD over USB driver working, the HP-A8 will show it's true capability.
P.S. Apologies for the very scattered and unstructured review. I may come back later on to clean it up.