Summary:
Well, read the pros/cons, but if you want a straight line FR response and low distortion, that's what benchmark DAC promises and delivers along with an array of I/O.
If you want a DAC+HP amp combo, DAC1 USB may not necessarily be what you're looking for.
Do note that you may feel that non-rolled off highs are making it harsh and unpleasant on some systems. You may also interpret that as benchmark DAC being highs-centric and mids being somewhat recessed.
You may also end up with a combination of straight FR for all the following: DAC, AMP, HEADPHONES which your ears might not like. Or DAC-1's HP amp and your headphones.
What's in the box:
There's the dac, the manual, the power cable and a coax->optical adapter...Oddly, no usb cable. Although both the DAC1 people state their belief in a to spec non-magical USB cable, and, well, everyone has one...
The manual gets a special mention. Interesting read... Although some of it has distinctly marketing feel that (IMHO) should be left to brochures (more on that below).
Build:
The DAC1 case is rock solid and very heavy. Looks/feels like cast iron? Under the cover it's all tiny surface mount components (non-user-servicable-land). This seems to go well in hand with the design philosophy for DAC1 (transparent sound/min noise). The DAC1 board looks like a pc motherboard
Sound:
I think people often mistake treble amount for detail and resolution. Well, DAC1 is trebbly (if I'm reading the graph correctly, 0-20khz @ -0.5DB . To me - Very treble-centric. Incidentally, since treble is so prominent (not rolled off) - I can't even turn it up enough before my ears start hurting from the treble to hear the mids. Consequently, because of the masking effect, it sounds to me like there are recessed mids/not enough mids detail.
Features
Lots of I/O, Switchable 220/100v psu, a few jumpers and switches:
You can set level as variable or calibrated or off for line out/xlr, turn analog stage off on connecting hp and +10db gain for HP out.
The manual talks of nearly 0 jitter, but neglects to quote a figure...http://www.stereophile.com/digitalprocessors/108bench/index6.html does show the DAC1 jitter measurements in PS. Interestingly, they're not any lower than the stated AQVOX ones. AQVOX( 110ps over all Toslink - peak to peak 40ps over all TOSLINK - RMS, 3ps rms internal).
No power switch... It's actually a feature (monitors inputs for signal/suspends). Except, when I suspend the PC, the usb lights are on...and stayed on until the morning. I contacted Benchmark and they were nice enough to point out that if I flick the input switch it will go to standby.
No LCD screen either. There's a table similar to the bios post error table to troubleshoot things (5 quick flashes = X, 25 quick flashes =Y). I guess, arguably - no LCD - probably less noise, since the LCD backlight will have to share power with digital/analog stages? SR leds would be nice.
USB does 96k/24b without drivers. Having had poor experience with Musiland's buggy custom drivers, this probably is a good idea. However, 96k/24b is not uncommon or particularly impressive nowadays. Still, the 'sweetspot argument' and ASRC to 110khz make this a moot point
Now um, there's one more thing to note about the manual and marketing. They both quote DAC CHIP features... E.g. the jitter immunity thing and de-emphasis filter. Neither really seem to be unique to DAC1. Have a look at the datasheets below. Most DAC chips to de-emphasis. Now for the jitter immunity part: The Yulong D100 DAC section (well, the USB path) is nearly the same as DAC1 (pretty sure someone said it's the same as the m902 or Stello (can't remember,but can't be bothered googling at the moment - spent too long cooking). In any case, according to the DAC1 manual, it seems like most of the anti-jitter benefits on the DAC1 come from the USB->I2S->AD1896(mostly here actualy) ASRCto110k->AD1853.
So it looks like D100 will do well there too, since it goes USB->IS via TSA1020B->AD1896 ASRC to 110k->AD1955 (pretty similar to DAC1 there, except the newer AD1955, that seems to have marginally better THD+N, and SNR and a whole heap of circuits relating to things we don't need (DSD, etc))
I'm a bit confused about the 24b/192k stickers on both DAC1 and D100. I really thought that internally it'll all be ASCRed to 110k regardless of the input SR and input type. Even if your input is 192k via, say coax. The D100 review/benchmark do mention that is the case/110k max is for both the stability (AD1896, not AD1955 as the first page says as per http://www.benchmarkmedia.com/discuss/feedback/newsletter/2010/07/1/asynchronous-upsampling-110-khz ?) and DAC section performance sweet-spot being 96khz. I certainly believe them about the sweet spot being 96k
http://www.analog.com/static/imported-files/data_sheets/AD1955.pdf
http://www.analog.com/static/imported-files/data_sheets/AD1853.pdf
Presuming it's (fft 1k 0db for 48,96,192k sr) they (benchmark media) look to be correct.