To be perfectly candid, I had my doubts at first when I was offered the chance to review the 212se digital preamplifier from DiDit High-End. On the one hand, I was intrigued by a special edition of the well-received 212 DAC/pre. But then again the 212 is built around ESS Technology’s Sabre ES9018 sigma-delta 32-bit DAC. It was inevitable that after some 30 years of searching for a musical CD sound that certain notions would crystallize for me as far as DAC chip types, upsampling, and digital filters. Eventually, my ears pointed me toward R2R DACs operated in non-oversampling (NOS) mode. The Philips TDA1541 and TD1543 made it into millions of CD players over the time frame from about the mid-1980s to the mid-1990s. Actually, I didn’t think too much about these chips until I heard them much later in non-oversampling mode and without the crappy op-amps that accompanied them as buffer stages in those days. In the immortal words of Schoolhouse Rock, “Zero is my hero,” at least when it comes to oversampling. Unfortunately for me, sigma-delta modulator (SDM) DACs have come to dominate the audio scene in the past 20 years. They are everywhere; literally millions of them reside in iPods around the world doing their thing with MP3 music files. An SDM DAC may be easier to manufacture, and it might make economic sense, but I’ve never before warmed up to its sonic footprint.
And so I agreed to give the 212se a listen before committing to a review. To say that I was in for a surprise is to put it mildly. This was the first SDM-based DAC that to my ears sounded non-digital in the best sense of the word. Hey, the 212se was so good that I was willing to put my bias aside and embrace it for the long haul. You may be wondering who exactly DiDit High-End is and how it pulled off this trick. The name is a clever acronym for Different in Design, Different in Technology. This Dutch company was founded in 2007 by a core team of four individuals, Patrick Schoon, Rients Steenbeck, Roy van der Hulst, and Sebastiaan de Vries, all of whom are passionate about music and bring complementary skill sets to bear on audio design and manufacturing.
Be forewarned that the obligatory overview of features and technology will be longer than usual—there’s just a lot to say about this unit.
The 212se is a special edition of the DAC212RCA and is fully differential from input to output, meaning that it is the difference between inverting and non-inverting signals that is being amplified. There’s more to implementing a balanced design than simply adding an XLR output connector. The trick is to take advantage of the common mode noise rejection offered by balanced operation while avoiding increased distortion. DiDit has paid this close attention, including use of current mirrors and routing the inverting and non-inverting signals on the PCB as a tightly coupled pair. The advantage of such a PCB layout is that the adjoining two magnetic fields partially cancel each other out, which results in a lower inductance circuit and purer signal transmission. During the review process I was sent the latest USB board and a comprehensive set of installation instructions for accessing and replacing the old board. It was a fairly painless process and gave me a chance to examine and admire the guts of the unit.
Only XLR analog connectors are provided, which means that those of us with single-ended amplifiers or preamps will need to use XLR-to-RCA adapters. In this case, it is recommended that the DAC be set to unbalanced mode via the internal software. This switches circuit relays for optimal unbalanced performance. I initially tried cheap generic adapters but those were no match for the Purist Audio Design adapters that Santy Oropel, the U.S. distributor, sent me. The Cardas adapters are very good, as well.
Digital connectivity is prolific and includes the usual suspects of AES/EBU, two coaxial, one optical, and USB. In addition, an I2S input via HDMI is accommodated for compatible devices such as PS Audio PerfectWave Memory Player, and Wyred4Sound and Cocktail Audio X50 music servers. If desired, the coaxial inputs could be changed to dual BNC to allow for DSD over BNC, one BNC per channel. And if you’re into this sort of thing (I’m not), Bluetooth streaming data can be received via the aptX codec or in AAC or MP3 format. I should also mention in passing the headphone jack on the front panel, which is not part of this review.
All of the unit controls are accessible via a cigar-shaped remote control. Power, source select, mute, and volume buttons provide the primary user interface. However, internal software menus and sub-menus allow the user the ability to tweak all performance aspects. These are accessible by holding down the mute button for about three seconds. The unit’s very cool dot-matrix display is your visual guide as you scroll through a host of main menus and select the one of interest. The one menu you should definitely become familiar with is titled “DAC,” and its sub-menus allow the user to tweak all of the ES9018’s settings. In particular, two settings definitely worth experimenting with are the digital filter roll-off (Fast or Slow) and the quantizer bit depth. The ES9018 is a multi-bit sigma-delta DAC, and although a quantizer setting of 6 is considered to be default, settings of 7, 8, and 9 are also available. (I was told that Qsize of 9 is being discontinued as an option, and it certainly did not work on my unit.) Theoretically, the main benefit of increasing the quantizer size is that of reduced high-frequency noise in the audible bandwidth. Strangely enough, a Qsize setting of 8 turned out to be quite noisy, so the real choice was between 6 and 7. I quickly zeroed in on 7 as my favorite, which sounded livelier, with better rhythmic drive. For the record, most of my critical listening was conducted with the Slow filter setting and a Qsize of 7. Out of curiosity, I asked the factory about any testing they may have done to identify measurable differences between these two settings. The Audio Precision tests they undertook, including input-output linearity, failed to identify any clear differences. Perhaps not surprising, since measurements rarely capture perceptual nuances experienced by audiophiles.