We put together the dCS972+Elgar and the Wadia 27ix and made a comparison based on the Wadia 270 transport, Cello Encore preamp, 2 sets of Encore power amp bridged for 200W, Transparent speaker cables, cello string interconnects and the Watt/Puppy 5.1
The Wadia was very 'HiFi' while the dCS combo just sound REAL. After the test, everybody in the room can easily tell what is "digital sound". The dCS sound like a super quite LP with the dynamic and clarity of CD!
The winner is clear. The dCS owner left with a smile on his face while Wadia owner was very fustrated....
Every audiophile and music lover should hear about and listen to this product. It makes CDs sound as liquid, pellucid and flowing as LP. It transforms the sound of your entire CD collection. And if used with the equalization features of Cello's Palette preamp, as I do, it makes virtually any CD, no matter how badly recorded, listenable, if not enjoyable.
The main benefits: added air, the music flows much more easily, you relax and get drawn into the music, the tension and chopped-up cardboard sound characteristics inherent in CD sound disappears, gobs more inner detail and harmonics, just sounds right.
The best analogy I can think of is CGA versus EGA versus VGA versus Super VGA or 1024x768 or 1280x1024 screens. The higher you go, the more resolution you get, the easier it is on the eyes, the sharper the image, etc., even if it still the same source.
As mentioned elsewhere, it's a shame this product has not been highlighted by the press more. No product in my view (not my Wadia, not my Mark Levinson, not my Audio Synthesis DAC) elevates the sound of CD more than an upgrade to the Purcell, used with the Elgar or Delius. 10 on a 1-5 rating.
I use the dCS Purcell, the commercial version of this professional product. The Purcell is cheaper and looks nicer but differs in that it hasn’t got as informative a front screen and doesn’t (currently) convert PCM audio to DSD, as used in Super Audio CD players, which the 972 will do. This is unfortunate since I’d dearly like to hear what DSD conversion sounds like! Apparently there is no consumer interface yet defined so there isn’t one on the Purcell. However, if one is ever ratified there’ll be one available as a hardware and software upgrade.
But it will upsample 16-bit 44.1KHz audio to 24-bit 192KHz and what a difference that makes!
I use mine with a dCS Elgar DAC, Jeff Rowland pre/power, Wilson WATT/PUPPY 5.1 and Transparent cables. The sonic effect: more detail, more realism to instruments that have complex harmonics – you can hear “more” of the strings on a guitar or the richness of a trumpet, particularly the decay into silence which enhances overall realism. Smoother and richer in the midrange, less “bloat” in the bass, less harshness in the treble. In short, more natural and less digital, although the ultimate sound quality is (still) not up to the best in analogue. Having said that, the advantages of vanishingly low background noise, plus the sheer convenience and robustness of digital sound are hard to beat.
OK there has to be something unusual going on here. I have some theories but I’m not an engineer! Most of the so-called explanations for higher-sample rate audio sounding better talk about rolling off some sort of high-pass filter that needs to block the sound of half the sample rate, or 22KHz, which is close to the limit of human hearing. So higher sample rates have a smoother roll-off because half their sample rates are well away from the audible range and the roll-off filter slope can be less acute.
Certainly this might be a valid reason for some of the effects in the sound, but as a fundamental explanation I think it is flawed. Most of this theory is based on the much-publicised “Nyquist Theorem”, which states that to sample any single frequency all you need is a sample rate of twice that frequency. In other words, you sample the peak and the trough of the sine wave and join the dots at the other end to produce the relevant output again.
Of course, that works fine for a single frequency, but a musical instrument consists of a fundamental frequency plus several overtones, and also “beats” between those overtones that give slight changes in amplitude at certain frequencies and provide much of the character of the instrument. My conjecture is that you really need to sample at a rate appropriate for each of those as well, resulting in (more-or-less) a sample rate that’s twice the sum of all of the frequencies you want to sample. Mind you, not all of us can detect the smallest of sounds which correspond to the least significant overtones and beat frequencies… Hence we can get away with generally increased sample rates (e.g. 96KHz) to give most of the space needed for capturing the significant tones and discarding the least significant that are masked by louder sounds.
That’s my explanation for higher-sample-rate digital audio sounding better, in general, but not for why the Purcell or 972 sound better. After all, they have to be adding dummy data, there’s no extra information. So there’s got to be an algorithm inside that’s looking at the overall picture as data streams through it and adding extra bits to suit. In particular the conversion from 44.1KHz to 192Khz is not an exact one so only a few of the “points on the graph” match every now and again, and the vast majority of the data points converted to analogue are, in fact, derived.
Hence the extra richness on the trumpet is an approximation based on the major tones and the occasional bits of data that give the algorithm “clues” about the overtones and beat frequencies. It’s good enough to fool me! Obviously upsampled 16-bit 44.1KHz data is not as good as “true” 24-bit 96KHz recordings, of which I have exactly one (Chesky sampler), but it’s a lot better than regular CD!
Downsides: the Purcell takes a short while to lock to a CD or DVD, and the Elgar takes another short while to lock again to the Purcell after it has done it. It’s only a second or so each, but if you stick a CD in and hit “play” there’s a chance you’ll miss the first two or three seconds of the first track you play. I get into the habit of loading a CD and waiting until the dCS kit has settled down, then hit play after that.
Finally, there was once a glitch in the Purcell and it switched back to simply passing 44.1KHz data through without any upsampling at all. I didn’t notice because the display was switched off and thought something else was broken! I spent an evening of tweaking: unplugging/rearranging signal and power cables and moving the loudspeakers around to remove a steely, digital sound, until I discovered the fault and switched upsampling back on. Music restored!
It's strange there's not been as much of a buzz on this product as there should be. In my opinion, it's the first real breakthrough in CD playback in a decade. The dCS 972 is a digital to digital converter that upsamples 16/44.1 to 192/24. Of course, you'll need a dCS Elgar to convert it to analogue, adding another $12,000 to the cost, AND, you'll need a good transport (like a C.E.C TL 1) Having said that, it'll probably costin the region of a Wadia 7 and 9 combo or a Linn Sondek CD12, and it really is a whole lot better! The upsampling will not add any more detail or resolution (if it's not there, it's not there) to the recording, but obviously, it gives the music more smoothness, lushness and transparancy. It's just so, so very musical. If you're a midrange freak, you'll swear you were listening the best possible analogue playback, minus the ticks and pops and inner groove distortion. Until some other technological breakthrough comes along, the dCS 972/Elgar combo swim right on top of the foodchain.