Wadia 850 CD player Page 3
"All well and good," I'm sure you're saying to yourself, "but can you stop gushing and let me know whether there were any downsides to the Wadia's sound?"
Well, it took me a while to get a handle on this, because it seemed music-related, but I occasionally got the feeling that there was some mid-treble brightness that had me reaching for the volume control. However, as soon as I got my hands on the remote, the feeling had passed. Strange, particularly as this was more prevalent with a preamplifier in the chain.
The optional digital input board ($995) allowed me to test the Wadia 850 as a separate D/A unit using both a Mark Levinson No.31.5 transport to play CDs via an ST-optical link, and the Sonic Solutions Sonic System hard-disk editing system to play back 24-bit material. (For the latter, I used a Digital Domain VSP/S to convert the Sonic's TosLink output to ST-optical.) The 850 acquitted itself well in this mode. In fact, my own high-definition recordings were reproduced about as well as I have heard them, with a noticeable superiority compared with the noise-shaped 16-bit versions we've released on CD. This suggests that Wadia's claims of high resolution may well be justified.
As suggested by a reader in this issue's "Letters," it would be interesting to compare the reduced ringing of Wadia's filter on 44.1kHz recordings with the reduced sinc-function ringing of a regular filter when it handles 96kHz-sampled data. Unfortunately, I ran out of time to do such a comparison during the preparation of this review. I may well do it in a future issue.
Against the BAT
The analog output levels of the $4950 Wadia 850 and the $4500 Balanced Audio Technology VK-D5 (reviewed elsewhere in this issue by Jonathan Scull) matched to within 2mV, so no additional level matching was necessary. All comparisons were done using the Levinson No.380S preamplifier. As well as side-by-side comparisons involving disc swapping from one player to the other, I used the BAT to drive one of the Wadia's coaxial data inputs via a 0.5m length of Mod Squad Wonderlink I fitted with BNCs. Interconnects were balanced: initially, the softer-balanced CZ-Gel for the Wadia and the more upfront AudioQuest Diamond for the BAT; then the reverse.
Given the convergent evolution of state-of-the-art CD players, the VK-D5 has a surprisingly different presentation from the Wadia 850. As J-10 correctly noted, the tubed player has quite a midbass bloom to its sound that can be very seductive. By contrast, the Wadia sounded drier in the bass, with a greater emphasis on the leading edges of notes. There was still plenty of low-frequency body to its character, and which one I preferred was dependent on the music playing. If the recorded balance was itself on the lean side, I preferred the BAT; if not, then the greater delineation of the notes led me to prefer the Wadia overall.
There was a similar story at the high end. The slightly more laid-back treble of the BAT gave recorded strings a slightly more mellow nature, which was very welcome on typical fizzed-up classical recordings. By contrast, the 850 was definitely more upfront, with a very slight emphasis on the metallic nature of violin strings. But with naturally balanced recordings this became inconsequential, allowing me to appreciate the wealth of recorded detail, the Wadia presenting a clearer picture into the soundstage. On Anthony Michaelson of Musical Fidelity's excellent new recording of the Mozart Clarinet Concerto, for example, issued with and thoroughly documented in the April 1998 issue of the English magazine Hi-Fi News & Record Review, you can more clearly hear the supportive acoustic of London's Henry Wood Hall on the Wadia, even though there is more bloom overall from the VK-D5.