 |
|
|
|
|
|
 |
 |
|
|||
 |
 |
HOME | |
||
 |
|
||||
       |
|
||||||||||||||||||||||
| Chord
Electronics DAC64 Issue
14 - November 2001 |
 |
|||||
|
Just when it seemed as though CD had reached the end of the road, up pops a product to challenge old beliefs and confound prejudices. I mean, we all think we know how it is - right? CD's limited 16bit 44.kHz specification is what stops it from achieving the highest possible standards; so you only have to up the sampling rate to 96kHz and go to 24bit resolution to reach Audio Nirvana. Well, maybe not. If the above encapsulates your views on matters digital, think again. Chord's DAC64 addresses the real problem with digital audio - a problem 24/96 won't solve! Their solution makes CD sound better than you could ever have imagined. It's not inexpensive; breakthroughs never are. But it's probably cheaper than replacing your entire CD collection... One thing needs underlining from the start. The special filtering/ buffering in Chord's DAC64 is not an enhancement device. We're not talking about circuits that superimpose a feel-good factor onto the music making it sound 'better'. Rather, the WAC filter and RAM buffer act as liberating devices - allowing you to hear your recordings as they were meant to sound. |
 |
|||||
|
The RAM Buffer has three settings; off, minimum buffering and maximum buffering. It sequentially takes all the data, re-times it, and then sends it out. This ensures jitter-free local clock operation without needing to send back a clock signal to the digital source. With maximum buffering there's a delay of about four seconds before you hear music; with minimum buffering it's about one second. Maximum buffering really does wonderful things to the sound, correcting the timing errors that cause poor sound quality. This doesn't mean that as a result Chord's DAC64 is totally uninfluenced by the absolute quality of the CD/DVD transport used. But it's probably fair to say it's much less sensitive than usual. I used a cheapie Marantz CD-6000 OSE with stunning results. I had interfacing problems using the DAC64 with a Pioneer DVD-545; crackling interference occurred and could not be eliminated. Going to a Denon DVD-2500 cured this, but there was a problem playing 24/96 music DVDs; the opening couple of seconds were lost when a new track was selected. Pressing Play, then Pause, and waiting a few seconds for the start of the track to cue, solved this. |
||||||
 |
With the DAC64 finally up and running, I started going through my CD-collection, playing all sorts of material, new and old. Naturally, I sampled many of my best recordings, and was pleased to hear them sounding better than ever before. Whether from digital or analogue masters, the beauty and freshness of the sound impressed time and time again. At the same time I also dug out discs that for one reason or another had always disappointed. One of the first I tried was Claudio Abbado's DG set of Mahler's Sixth symphony - an analogue recording first released on vinyl in 1980. I bought the LPs when they came out, and was distinctly underwhelmed by what I heard. |
|||||
|
Unlike DG's excellent 1978 Karajan recording of the same work, Abbado's sounded cramped and boxed in - as though the original multi-track tapes had been mixed down to a ferric-oxide compact cassette. The sound lacked range and depth, with little impression of the vast orchestral forces used in this powerful symphony. The CD, released six or seven years later, seemed much the same... Played on the DAC64 with fours second buffer, Abbado's Mahler 6 miraculously acquired superb depth, space, scale, and separation the very qualities that seemed to be totally absent. Suddenly, hall ambience was apparent. Soundstaging was 3D holographic rather than pancake-flat, and the stereo was vivid and highly detailed. Yet the tonal balance was rich, warm, and smooth - the increased clarity wasn't being artificially enhanced by frequency response shifts. It was as though the true qualities of this recording were being revealed for the first time. I've known this recording for more than twenty years and played it on lots of different systems. It's never sounded any good: Until now. Even more remarkable was the transformation of the classic Pretenders track `Brass in Pocket'. My late '80s CD pressing sounds thin, hissy dynamically flat, and colourless a text book example of how CD highlights everything that's wrong with a recording, and misses all the things that are right. Played on the DAC64 without buffering, the sound was clear and detailed, but still rather `flat'. Switching in the four-second RAM buffer changed things completely. Now, the music sounded vivid and three-dimensional. Chrissie Hynde's voice was always recessed and well-back in the mix; now it was projecting powerfully into the room. The whole sound was bigger and musically more engaging. Rhythmically, there was much greater forward momentum and drive. Going from the four-second buffer to the one-second buffer, the soundstage collapsed to a degree. Guitars and drums were less projected, and the vocals lost their much of their superb holographic out-of-the-box forwardness and immediacy. Not a bad sound, and scads better than no buffer. But the magic was no longer quite as apparent. Solo Piano showed the benefits of buffering very clearly. There's a greater sense of notes starting from nothing, quickly reaching their peak, then dying away. You can definitely follow the decay of the notes more easily - the balance between attack and decay is much more even. That `broken bottles' brittle top-end you sometimes get with CD piano was completely absent. Low frequency clarity improved too. Via the DAC64, piano bass notes are more resolutely projected. Suddenly, small details like trills and subtle dynamic inflections tell with far greater authority. Instead of being vague and shadowy, left-hand detail is clearly articulated. The notes making up a chord are more precisely defined so that harmony is enriched. |
||||||
 |
Result? Fuller more focussed piano tone. No longer do you have to play the music loudly to create a big room-filling sound. Subjectively, this increases the signal to noise ratio. Tape hiss can be very pronounced on CD because (unconsciously) one often turns up the volume to achieve a sense of presence and immediacy. With the DAC64 this isn't as necessary. |
|||||
|
Deep bass definitely seems fuller and weightier, especially when the four-second buffer is engaged. Bass drums show noticeably greater power and solidity, and I consistently noticed lots of subtle low frequency disturbances on classical CDs (rostrum thumps, traffic noise, air conditioning) that usually pass unheard. Good bass helps create a sense of scale, plus a feeling of expansion in climaxes. |
||||||
|
CD, for all its superficial sharpness and attack, often lacks power and weight the very qualities that subjectively create a big sound. One of my favourite 12in singles is Randy Crawford's `Street Life' - it sounds awesomely huge and exciting on vinyl, but disappointingly tame on CD. It's not a question of loudness, but scale and presence; no matter how loud the CD is played, it never really projects. Now, had you asked me a few weeks ago, I'd have said it was absolutely impossible for CD to offer impact comparable to a good 45rpm 12" single like `Street Life'. But the DAC64 in four-second buffer mode via its balanced audio outputs, produced levels of scale and power that almost equalled the 12" single played on a top-class turntable with state-of-the-art moving-coil cartridge. One of analogue's greatest attributes is its ability to produce results that exceed than the sum of its parts. Digital is often the opposite despite boasting excellent specifications, results may be sonically less good than you'd expect. Just because there's no background noise and low peak level distortion, it doesn't mean reproduction will be musically engaging. From the recording engineer's standpoint, digital is regarded as truthful but unforgiving-an exacting medium to work in. Analogue is said to be kinder and more forgiving. People say analogue alters the sound in ways the ear finds pleasant, whereas digital gives you the truth pure and unvarnished. Really? I strongly disagree. Chord's DAC64 surely explodes both these old chestnuts. What impressed time and again was the way Chord's DAC64 (maximum buffer) consistently revealed the strengths of a recording, not its weaknesses. You heard what was right with a track, rather than what's wrong with it. A perfect example was Otis Spann's 'Goin' Down Slow' from the Fi/Analogue Productions sampler CD -track 4. Technically, it's not a great recording. Tape hiss is high; and there's some background hum. But just listen to the way the piano leaps out of the speakers. Listen to the dynamic quality of the voice; the live feel to the performance. This is exactly the sort of track that normally sounds spellbinding on good analogue. And an absolute dog's dinner on CD... |
||||||
|
The soundstage is huge. Yet it's not brute-force dynamics; more a subjective feeling of power and expansion as things get loud. The soundstage diminishes noticeably when you go from the four-second buffer to one-second. And when the buffer is switched out altogether the sound becomes very `flat' indeed. Subjectively, it's like singer and piano have moved back from the microphone. |
 |
|||||
|
Most tracks on the Fi sampler demonstrate the enormous benefits produced by the DAC 64, especially when used with its four-second buffer. Previously, when I listened to this disc using an ordinary CD player, only Janis Ian's `Breaking Silence' (track 9) impressed. This is superbly engineered, with low background noise and clean climaxes. It's the sort of production that sounds good on most systems. Other tracks on the disc are not so impeccably engineered, but this is deceptive. Many have qualities that transcend their superficial limitations, producing results greater than the sum of the parts. Interestingly, `Breaking Silence' was less improved by the four second buffer than other supposedly `poorer' tracks. The DAC64 reveals detail in recordings that ordinary CD players often obscure. Making a choice between the one and four second buffers is essentially down to personal taste. The four second option gives the greatest depth, space, and separation. On suitable material you hear a marvellously refined sumptuous sound: One that's rich, weighty, fullbodied, exquisitely subtle and holographically dynamic. Yet this too is only part of the story. |
||||||
 |
Playing Jethro Tull's Songs from the Wood, the sharpness and detail of the sound was fantastic. With the buffer on four seconds, the many tricksy tempo changes were fantastically crisp. The production is not particularly sophisticated, but the instrumentation is complex with large numbers of over-dubs. Every detail was audible; clarity and cleanness were something to marvel at. I definitely prefer the four second buffer. But for times when a leaner more forward balance is required, the one-second option is useful. On certain discs, the increased sharpness is definitely preferable. Bass can sometimes be a little too rich and warm when set to four-seconds. If this is the case, switching to the one -second buffer produces a drier leaner sonic impression. |
|||||
|
Whether or not the four second buffer was helping to disguise the timing limitations of the Marantz CD-6000 OSE player is not clear. Alas, my heavily re-clocked Pioneer PD-75 is still away being repaired I think it may've had it's chips literally! Actually, when this review was originally mooted, it was supposed to be about DVD and the potential of its 24bit /96kHz technology. However, the Pioneer DVD player I was going to use didn't interface properly with the DAC64, and in any case the improvement with the 24/96 discs (replayed on a Denon machine) over CD seemed fairly minor. Encounters with 24/96 have so far left me feeling mildly impressed and slightly disappointed. Yes, there's a difference. But it's less than I'd have hoped. Frankly, if I couldn't live with CD, I couldn't live with DVD either- it's not significantly better. If Chord are right about the causes of disappointing digital sound, then it's hardly surprising 24/96 only sounds slightly better than 16/44. What this implies for the future is hard to predict. There's no guarantee that `better' formats like SACD and DVD-A will survive in the long term. You can, however, invest in a Chord DAC 64 and in doing so upgrade your entire CD collection at a stroke, as well as buying a DAC capable of accepting any of the new PCM derived `high- resolution' technologies (although not SACD for the present). Speaking as someone with a big CD collection, that route looks far more interesting than worrying about future systems that may or may not deliver the goods sonically. The breadth of music available on CD is quite awesome. And its cheap! Ordinary CD has far more to offer than most of us ever realised. Using the DAC64, I've been hearing my CD collection as never before. It's not often a product leaves me gobsmacked, but Chord's DAC64 has done just that. I don't think I can live without one. CD was 18 years old in March 2001. With a little help from Chord's DAC64 it may finally have come of age! During use, the unit gets quite warm. Operation is fairly simple, but the toggle switches for input selection and buffering are not marked. Nor is there an illustration in the handbook to say which is which. The illuminated porthole in the top gives the unit a fantastic appearance, and nearly everyone who's seen a DAC 64 wants to buy one - on looks alone. Nice to see people have their priorities right! |
||||||
|
Technical
description Therefore, to resolve timing of one microsecond an impossibly high sampling rate of 1mHz is needed. Luckily, 44.1kHz sampling can be made to accurately resolve transients by means of digital filtering. Unfortunately, to do this the filter needs infinitely long tap lengths. Chord say all present reconstruction filters offer relatively short tap lengths - the longest commercially available device has about 256 taps. The combination of short tap lengths, and the filter algorithm employed, generates transient timing errors - errors that are very detrimental to the final sound. Apparently, going from 256 taps to 1024 taps gave what Chord describe as massively improved sound quality -much smoother and better focussed, with soundstaging of increased depth and precision. Chord's experiments initially seemed to indicate that infinitely-long tap lengths were necessary going from 1024 taps to 2048 gave a further significant improvement. At this stage a new type of algorithm, the WTA filter, was developed. It was designed to minimise transient errors from the outset, thereby reducing the need for very long tap lengths. Chord claim their WTA filter (using 256 taps) outperformed all other commercial filters - even when the latter used 1024 taps. However, WTA filters still benefit from long tap lengths, and currently Chord are using 1024 taps. The filters are implemented in Field Programmable Gate Arrays using a specially-designed 64bit DSP core. Fourth-generation Pulse Array DACs are used, employing 64bit 7th order noise shaping and 2048 times oversampling. Chord claim this gives improved measured performance, better resolution of detail, and smoother more focussed sound quality. The unit will run from 30V to 300V AC without adjustment, incidentally. |
||||||
|
TECHNICAL SPECIFICATIONS Chord
Electronics DAC64 Manufacturer:
Back to Last Page | Top | Forward to Next Page | Technical Review index
|
||||||
 |
|
|
|
|
|
|