November 1997
One of those frequent audiophile dreams is to find the infamous giant-killer product. Such a beast would give performance well beyond its price, earning accolades that it sounds like it costs several times what it actually does. I personally think this whole view of the world is seriously warped. Fact is, products should be priced based on how much they cost to produce. And designers shouldn't spend serious money on things that don't have a positive impact on sound quality. I can understand why huge high-end amplifiers cost serious bucks; just the metalwork for heat dissipation alone can drive the price to the stratosphere. And expensive speakers often feature incredibly complicated cabinets, outrageous drivers, and premium crossover components. What I just don't see is how anyone can justify spending multiple kilobucks on a stereo D/A converter. Digital technology just doesn't cost that much. Frankly, I'm hard pressed to see any reason why a normal DAC should cost more than $1000 under any circumstances. If you dig into the pricing for the digital parts used, from companies like Crystal or Burr-Brown, you'll see what I mean. One fallacy with that line of reasoning is failing to consider the R&D costs and the overhead of running a small production line. Companies like Audio Alchemy successfully (for a while) dealt with that problem by planning from the outset to sell a lot of units. After all, if you make something less expensive, you are very likely to have more buyers pick one up. Camelot Technologies seems to have learned this lesson as well. Their Arthur V3.0 DAC provides just about everything you could want for stereo listening in a compact $899 package.
While it's nice to get inexpensive products, it makes you feel better if they don't look it. The Arthur is available with a black or silver faceplate, and both are very attractive units. There's no power switch, and the digital lock light is usually the only thing that shows the DAC is turned on. Three switches with a solid feel are your only controls, and unlike many inexpensive products these make no noise whatsoever when you flip them. The standard V3.0 includes a huge array of digital inputs on the back:
In addition, you can add an ST Fiber optical input as an optional upgrade ($179). Essentially, you can hook up just about anything to this DAC. One three-position switch on the front switches between the coaxial, AES/EBU, and optical inputs. The ST upgrade replaces the Toslink as the optical choice. The other switch selects one of the I2S inputs, so unless you're using that connection method it stays in the off position. One additional switch reverses polarity to give a 180 degree phase shift. This works effectively, and the manual gives a good description of what to listen for.
I tried a few other sub-$1000 DACs before Camelot's box arrived. After hearing what those units were capable of and looking at what you got for the money, I was unimpressed. I bought the $2000 Lexicon DC-1 instead, which justified the higher price tag with a huge array of features. When I asked, Camelot's Howard Schilling was certain the Arthur V3.0 would sound better than the DC-1 does. After all, he pointed out, Lexicon is putting eight D/A converters inside their box, while the Arthur only has two. Accordingly, they can use better quality DACs and associated circuitry, along with a cleaner board topology. Seems reasonable, but I certainly don't believe every manufacturer claim I hear.
To see how the Arthur V3.0 stacked up, I went through a number of writing contortions. The two digital sources used were the Rotel RCC-940AX (Toslink optical output, Monster Toslink cable) and the Rotel RCD-955AX (coaxial output, DH Labs digital cable). Preamp configuration varied, but they always went to my Warner Imaging VTE-401S power amplifier with JPS Labs Superconductor interconnect. Speakers were the NEAR 50Me, and speaker cable from was also from DH Labs.
To get a general feel for how the Arthur sounded, I pulled my Rotel RSP-960AX preamp from retirement and connected it to the DAC with the DH Labs BL-1 interconnect. Listening to that configuration for a month left me very impressed; this was the first time I'd heard my analog preamp driven so well that the combination was competitive with the DC-1. After that warm-up, the first order of business was to determine how to configure the V3.0. After removing eight screws and popping the cover, there are two internal jumpers that you can use to optimize the DAC's performance. The first controls whether dithering is performed. Dither adds a bit of noise to help make the overall output of a DAC more linear, and should normally be used. Turning dither off makes for strange listening at times. Some noisy tracks, like "Pressure Points" from Camel's Stationary Traveller, were less grainy during the noise with dither disabled. The louder portion where the noise was masked seemed a bit less clear. Looking for a pattern, I listened to the opening of "Pretty Maids All in a Row" from the DCC release of the Eagles Hotel California. The piano slowly rising from the background noise here sounded cleaner and more transparent with dithering on. While there were some tracks that I preferred with dither disabled, overall I found the sound quality rougher, which is never a good thing in digital reproduction. Dither stayed enabled for the rest of my listening.
The Arthur V3.0 uses Pacific Microsystems' HDCD filter chip. One of the technical details associated with using this filter is a 6dB attenuation of non-HDCD discs during playback. Using any sort of volume decrease isn't likely to improve sound quality, so Camelot offers another jumper to turn off the gain reduction. Experimenting with this setting always favored the configuration where gain wasn't reduced. Adding the attenuation dulled the sheen around cymbals, giving a less natural edge to their envelope, and some of the ambiance in complicated recordings was lost. One possible explanation of the magnitude of the improvement is that the unattenuated configuration outputs 4.4V, while the regular one produces a mere 2.2V. A higher voltage going to the preamp is bound to improve dynamics and keep the noise floor lower. A warning, though; your preamp may not be compatible with such a high input. Also, this will reduce the useful range of your volume control, so if you already find you don't have enough fine control at low volumes this will just aggravate that problem. On the positive side, this high-voltage mode makes it an ideal DAC for those using a passive preamplifier. Gain reduction stayed off after I determined it was a step backward in my system, and I didn't miss it a bit. Note that there are also a set of jumpers on the board labeled as controlling the filter configuration. You should not change any of these settings; they are used to match the HDCD filter to the DAC chips used, and there's nothing to usefully alter there.
The next compatibility concern was with the digital inputs. I switched between the Toslink optical and coaxial inputs to see which offered better sound for me. The results were different from what I normally notice. While, as usual, the bass was fuller and high frequencies more detailed with the coaxial connection, there was something strange going on in the midrange with the optical hookup. In my system, using the Toslink inputs with the Arthur pushed all the vocals backwards in the soundstage, and the timing seemed very weird in spots. While the rest of the frequency spectrum was rendered very well, this was enough of a problem that I really didn't find the overall result acceptable. With the coaxial input, everything was fine, so that's what I used for testing after noticing the differences. Note that Toslink outputs vary in quality enough that you may get totally different results from the Arthur. Based on my recent listing to inexpensive DACs and transports, I'd have to say that using a Toslink connection with a $899 DAC is a waste of money anyway. You really should upgrade to a transport with at least a coaxial output before buying a DAC this expensive.
Now that I knew I was getting the best sound I could get out of the Arthur, the inevitable showdown with the DC-1 was almost ready to proceed. During an informal comparison, the Camelot unit was obviously far louder than a DAC normally is. After all, the standard CD output level is 2V, so even the attenuated 2.2V output of the Arthur is 0.82dB louder than normal. Watch out for this if you're trying to directly compare the V3.0 to other DACs. The full 4.4V of the Arthur with gain reduction disabled is another 6dB higher still. Since the DC-1 is actually a preamp, too, it's easy enough to turn it up to match levels as closely as possible. Pulling out the voltmeter and testing at the loudspeakers revealed that the Lexicon had to be run at +10dB to come close to the Camelot's output; it was off -0.35dB there. The +11dB setting was obviously +0.65dB. Since one was a little two low and the other a bit too high, I switched between +10dB and +11dB when doing the comparisons to try and factor out differences from the volume change.
Those ugly technical details out of the way, I worked my way through my usual testing to get a feel for what the Arthur did better or worse than the DC-1. On the Alan Parson Project's "Too Late" (from Gaudi), the Camelot DAC portrayed a very warm bass line with good separation between bass guitar and drum. The vocals seemed a little bit recessed compared with the Lexicon, but the Arthur gave a wider soundstage. Aggressively recorded cymbals on recordings like Peter Gabriel's "Red Rain" (from So) really sliced through the air with extremely sharp transients on Camelot's DAC.
Since it showed so much difference in the dither modes on the Camelot, I returned to "Pretty Maids" for this comparison. The tape hiss in the otherwise quiet opening was much louder on the Camelot than on the Lexicon, even when I was playing the DC-1 a little higher overall. This is not necessarily correct (can't tell without the master to check against) or better, but it's certainly different. If you want to hide from noise in the recordings you listen to, the Arthur V3.0 is not the DAC for you. While I was appreciating DCC recordings, I tried "Crazy on You" from their version of Heart's Dreamboat Annie. The acoustic guitar opening lets you hear that the edge of the strings are better on the Arthur, and the drums were a bit more crisp. Across all the recordings I listened to, it was obvious that in this configuration the Arthur really was the winner at giving the maximum detail. The Lexicon had a slight advantage in presenting really clear vocals, but it's hard to tell if that's not just because the more defined treble in the Camelot DAC made the vocals sound less present by comparison. Performance at the bottom end was similar, with the Lexicon leaning more toward a tight, controlled presentation while the Arthur was a touch fuller.
Now, this isn't exactly a fair comparison. Sound from the Lexicon is hobbled by going through a preamp and one cable more than it needs to. It's much more appropriate to compare the DC-1 ($1995) by itself to the Arthur V3.0/DH Labs Interconnect/Rotel preamp combination (~$1600). This made for a much tighter contest. While the Arthur was the clear winner in presenting a wide and well defined soundstage in my previous tests, the fully unleashed DC-1 was now its equal. The differences in treble definition also closed considerably, with the Arthur still having a small advantage. It was certainly still the once and future king of throwing recording hiss at you. On recordings where I had a preference for one setup over the other, the results for which combination I liked better were pretty evenly distributed.
Overall, the Arthur gives a very clean presentation with loads of detail without becoming harsh or unpleasant. Still, this is certainly not the DAC for someone who wants a really relaxed or tube-like quality to their digital sound. The deep bass playback may not be the ultimate available under all circumstances, but it's very good, and exact performance is going to depend on what preamp you're using and the quality of the cables used for the connection.
The Arthur V3.0 has potential beyond what I've mentioned so far. You've always got the option to use one of its seemingly endless alternate digital inputs, and of course there's the digital cable to experiment with (Camelot's Excalibur line starts at $149 for 1M). There are a number of alternate power supply options available as well. You can replace the IEC power cord with something better; Camelot sells the Sir Percival ($149/1.5M) and Sir Bors ($349/1.5M) cords as potential upgrades. Why fool with that, though, when you can use a different form of power altogether? The Charm II ($299) lets you run the Arthur for several hours off of battery power, using the DC power input on the back. Removing the need to interface with the ugliness that is our power lines and avoiding the whole rectification process can't help but reduce noise and improve power stability. Camelot also sells CD transports (starting at $1395) that utilize the I2S interface, which transmits the digital clock and data lines in a fashion that highly reduces the potential for jitter based timing errors. Looking at what you get for your money, I personally find it hard to believe any transport should cost more than $1000, but I've been accused of being cheap before.
The technically minded may wonder just what Arthur's got hiding inside his armor. The clean and neatly laid out circuit board houses enough circuits to cover even a round table. All S/PDIF inputs are conditioned with an RS-422 data receiver before heading to a Crystal 8412 input receiver. I2S signals bypass this whole section and go right to the next stop, the Pacific Microsonics PMD 100 filter. This interpolates the data to 20 bits and performs out-of-band digital filtering. Dither and gain attenuation, if enabled, also happen in this stage. Two Burr-Brown PCM 1702 20-bit D/A converters provide the first view of the signal in analog form. This analog output needs to be properly amplified and buffered before it's ready to see the real world, and Camelot's "charge pump" design is claimed to provide a unique yet simple approach that yields superior frequency and phase response. Powering all these pieces are no less than seven independent power supplies optimized for the device they are dealing with.
I've been rather outspoken on my preferences in digital technology. I still am quite convinced that we're quickly moving to the day where a DAC that doesn't implement additional features like DSP crossover technology is going to be at a huge disadvantage. Plus, I expect surround formats to become increasingly important for music applications. So for me, a unit like Camelot's just doesn't pack a big enough digital feature list to be a long-term purchase. But given that the vast majority of music people want to listen to right now is still stereo, I can certainly understand that not everyone wants to follow that path. There are two overlapping groups of potential customers I see this DAC being perfect for. One is the set of audiophiles who don't give a hoot about home theater, subwoofers, or the rest of that scene. This includes people who are still dedicated to analog sources and rightfully want to stay with an analog preamp. The other potential buyer for an Arthur V3.0 is the audiophile working within a controlled budget, someone who wants excellent digital sound in a product that integrates with whatever components they buy without spending a fortune.
For $899, you can't get a better DAC right now than the Arthur V3.0. Nothing else I know of combines so many features and options at this level of quality in such an affordable package. While there are, as always, some issues with whether its presentation will exactly match the qualities you value most, anyone searching for a sub-$1000 converter to buy should consider adding Camelot's unit to their shopping list. And those looking to spend more should certainly consider an Arthur that's fully decked out in power and cable upgrades. I predict that you'll see a lot of recommendations for this unit in the upcoming months. Odds are good at least one reviewer will call it a "giant-killer." You won't hear that phrase from me, but I'll certainly state there's a whole lot of expensive DACs out there that look rather obsolete now that the Arthur V3.0 is available.
.....GS (gsmith@westnet.com)
Arthur V3.0 Price as configured: $899 USD Camelot
Technology Web-Site: www.camelot-tech.com Camelot Technology Responds: My father and I would like to express our sincere thanks for your review of the Arthur v3.0. The v3.0 reflects Camelot Technology's commitment to affordable digital audio and video excellence. We intend to build upon this pledge in the coming year with our newest product offerings, which include the Merlin PRO CD Transport at $1595, The Dragon v2.0 Jitter Reducer at $699, the Morganna CD Player at $1395, and, last but not least, the Crystal Vision VPS-1 2D Adaptive Digital Comb Filter at $599, the redesigned version of the groundbreaking Audio Alchemy VRE v1.0. We consider it a privilege to be included in your magazine, and we support you 100% in your efforts. The Internet is a medium that receives a lot of respect and attention in our company, as its users rank among the most qualified and informed buyers of our types of products. Thank you again for all of your time and consideration. Sincerely, Howard Schilling E-mail: camelot@camelot-tech.com |