Benchmark is a professional audio, not audiophile company, which makes it somewhat unusual that the audiophile press noticed and promoted it, but to their credit they did. Truth should win. Particularly impressive is the measured performance in the Stereophile review, and the comparative and subjective comments in The Absolute Sound.
As a pro audio company, Benchmark's design seems generally orthodox, but very carefully done and not too tweaky like an audiophile design. Opamps are Texas Instruments outstanding LME49860 with a pair of TI-sourced 5532 opamps possibly on the headphone or volume control circuit. The headphone amplifier is based on another Texas Instruments chip, the LME49600. DAC chip and Asynchronous Sample Rate Converter are Analog Devices AD1853 and AD1896 respectively. USB interface chip is TI TAS1020B. Digital Audio Transceiver is an AKM AK4114. All the digital and analog audio chips are very high performance.
Benchmark claims near perfect USB jitter rejection, but I can clearly hear a small but noticeable difference between the same audio data over USB from PC compared to dedicated disc player over S/PDIF. The difference sounds like greater jitter via the computer. To be fair, if there is greater jitter from the computer signal path, it may or may not be the fault of the USB interface. There are many possible ways jitter could be introduced from storage, to ripping, to playback, etc. Whatever the cause, Benchmark does not appear to have practically eliminated jitter over USB. This is somewhat disappointing, since it would be far more convenient to computerize music than play from physical optical discs. To be clear, over USB the Benchmark's sound is outstanding, but not as good as direct from disc, which is cleaner, less hashy, has more realistic harmonics and instrument voice sounds, etc. The difference may be inaudible to non-audiophiles, but why would non-audiophiles use high-performance audio gear?
A casual glance at the circuit board suggests some room for improvement. There aren't nearly as many power supply sections as a more typical audiophile approach. This means that many of the circuit sections must share the relatively few standard three terminal voltage regulators, which could increase power supply modulation distortion. All the signal chips do have very high conventional power supply rejection specifications, which counteracts power supply modulation issues, but more regulators would further improve power supply rejection as we learned many years ago from Ben Duncan. Though he was writing about analog circuits, power supply rejection is at least as important in digital audio circuits since digital signals impose large, high-frequency current swings on power supplies, which like digital curcuits are ultimately analog. Digital signals traverse between their high and low signal levels rapidly and often, and very much more stressfully to the power supply than analog signals.
The AC bridge rectifier is a solid but cheap Japanese unit that could be improved with a soft recovery replacement. The transformer is a nice toroidal. Not having the AC enter the chassis at all would be better, but less practical for it's intended use in professional sound monitoring. As an aside, the DAC1 runs noticeably hot and uses about 15 Watts at idle and when operating. It doesn't really have a sleep, low power or off mode, but mutes the outputs when switched to standby. This is wasteful in terms of home energy use, but important for having consistent performance in studio use.
My Benchmark DAC1 PRE model, which includes an analog input, but no remote volume control is no longer made as of October 2012. The PRE is intended to be used as a preamp/control amp with multiple digital inputs, an analog input, an input selector switch and output volume control. It's replaced by the slightly more expensive DAC1 HDR that adds an optional remote volume control. It's perhaps worth mentioning that the Benchmark DAC1 has gone through multiple versions over time, but not an excessive number. For example the version in the 2008 Stereophile review is probably several revisions behind the different model I got in 2012. All the models should have common central circuits and general similarity. Most of the relatively minor differences are in features, inputs, outputs, etc.
It would be interesting to hear whether the improved jitter performance in the DAC2 audibly betters the effects of jitter coming from computer audio. And it may be worth checking out whether jitter performance can be improved over USB using professional audio personal computer card-based digital audio outputs such as the Lynx AES16e.
The latter uses the professional AES interface, which has separate clock and data signals, elminating a major source of jitter caused by multiplexing clock and data over a single signal, as happens in the common consumer S/PDIF interface. In addition, the purpose-built professional AES card may have better signal timing integrity than motherboard USB or S/PDIF interfaces.