Reviewer: Paul Candy
Source: Rotel RCD-971 CD player, Sony SCD-XE670 SACD player (both modified with IEC jacks to eliminate captive AC cords); Apex AD-1100 DVD player
Preamp/Integrated: Bryston B60 integrated, Audio Zone AMP-1, Audio Zone PRE-T1 preamp (in for review), Houston Mini-2 integrated (on loan).
Amp: Audio Zone AMP-2 monoblocks (in for review).
Speakers: Meadowlark Kestrel 2
Cables: DH Labs Q10 loudspeaker cables, DH Labs Revelation and Air Matrix interconnects, DH Labs D-75 digital cable, DH Labs Power Plus power cords, Wireworld Aurora III and Stratus power cords
Stands: Premier three-tier, filled with sand
Powerline conditioning: Blue Circle BC86 MkII Power Line Pillow.
Sundry accessories: Caig Pro Gold contact enhancer; Walker Audio Vivid CD treatment; Black Diamond Racing Cones Mk3; AudioPrism Isobearings; dedicated AC line with Hubbell outlets
Room size: 13' x17' x 8', long-wall setup
Review Component Retail: TubeDac+ $475.00, 3Xac Power Supply $115.00

Redbook stinks. At least that's what the folks behind SACD and DVD-A would have you believe. I don't disagree that there are flaws with 16-bit/44.1kHz. But while I've found much promise in at least the SACD format, the fact remains, CDs continue to sound better with each passing year and will continue to be the data carrier of choice for the foreseeable future. Plus, I'm not about to replace my music collection yet again. While I enjoy SACD, there is still a paucity of titles, most are overpriced and the great 'unwashed' masses couldn't care less. Furthermore, the sales staff members at the big chain music stores (at least in Toronto) are absolutely clueless about these new formats. This leads me to believe that the major record companies don't completely back or promote SACD and DVD-A either. Personally, I think both DVD-A and SACD are destined for the dust bin. I'm awaiting the day when I can download uncompressed music at my choice of resolution via a high-speed wireless connection, complete with album graphics, liner notes, video etc. Until then, Redbook is fine with me.

Many criticisms directed at the 16-bit/44.1kHz format are a result of the use of so-called steep brickwall filters which cause nasty phase shift artifacts to creep their way into audible frequency ranges. Current high-speed DACs and digital manipulation processes such as upsampling and oversampling attempt to minimize these problems by pushing anti-aliasing artifacts into the ultrasonic range. As I mentioned in my review of the Music Hall Maverick:

"Upsampling or sample rate conversion is an attempt to eliminate the sonic shortcomings of standard 16/44 conversion by increasing, via interpolation, the sampling frequency from 44kHz to 96kHz or higher. An additional oversampling filter (or sample rate converter) is series-connected to the existing oversampling filter, which, in many cases, may actually reside inside the DAC chip itself. The concept is to upshift anti-aliasing noise into extreme ultrasonic regions where a low-order, gentle reconstruction filter avoids the higher-order, drastic "brickwall" artifacts of gross phase errors and bright, edgy, fatiguing sound. While upsampling clearly doesn't increase resolution, music playback should, in theory and as a function of minimized conversion damage, sound smoother, less brittle and edgy. If all of this rings a bit like oversampling, you might be correct. There's a bit of verbal controversy as to what distinguishes up- from oversampling. I suspect a fair degree of marketing hype at work but in the end, the proof is in the pudding".

Thus far, my experience with three different upsampling players/DACs has left me unconvinced. With upsampling switched in, I detect smoother grain-free playback with enhanced air and space around instruments and voices. But it also sounds somewhat artificial and unnatural to me, with images shifted about and the inherent flow of music impeded. Indeed, the entire aural landscape seems subtly altered. Perhaps I haven't yet heard a proper implementation of this technology. Be that as it may, I am doubly suspicious of upsampling DACs and players that do not allow this feature to be switched on and off on the fly to make meaningful comparisons.

Not surprisingly, there is a growing underground movement in the digital world that contends that these new DACs, fancy filters and hi-tech processes actually do more harm than good. They possibly subject the waveform to phase shifts and ringing that allegedly distorts the music hiding in those pits of your favorite CDs. Their argument? You are hearing an altered version of the recording encoded on that disc. Think of Elvis circa 1968 as opposed to the endless hordes of Elvis impersonators currently roaming the planet's casinos and side shows. No matter how convincing these guys can be, brothers and sisters -- or so this argument has it -- they are not the leather-clad King who shaked, rattled and rolled on that famous 1968 television special. These digital malcontents contend that if done right, a simpler approach without over- or upsampling will offer improved resolution and musical satisfaction. Srajan's review of the Zanden Model 5000 MkII DAC is a case in point. 47 Labs, Audio Note, iLungo and others like the subject of today's review, the Scott Nixon TubeDac+, are examples of companies who subscribe to that seemingly Luddite and retrograde philosophy. I'm not an audio engineer. I can't tell you who's right or wrong. But I do have a decent pair of ears attached to my skull and believe that there are several ways to skin the proverbial digital cat. I must admit that the theory does appeal to me. Of late, I have discovered a preference for audio gear that utilizes a minimal number of components and attempts to preserve original waveforms. The Audio Zone AMP-1 and -2 and my Meadowlark Kestrel 2s are prime examples of said philosophy.

The $475 TubeDac+ is a tube-buffered version of Nixon's DIY Dackit housed in a cute 3"x4"x1.5" black ABS box with a lone 6922 poking out the enclosure. The overall signal path appears to be no more than a few inches. The input receiver is the ubiquitous CS 8412 while number crunching is performed by the venerable Philips TDA 1543 DAC. Digital and analog filters are conspicuous by their absence. The signal taps directly off the DAC and routes into the single 6922 triode with passive current to voltage conversion. In this configuration, the TubeDac+ inverts absolute phase. Scott Nixon suggests that some listeners will not notice the difference unless directly comparing his with other DACs - in which case he recommends reversing speaker cable leads. Even without comparing other digital gear, I preferred playback with my speaker leads reversed. I think most people would notice a difference, too. Just reverse your connections and fuhgeddaboudit. There is no de-emphasis filter. If you own mid-80s CDs especially from Denon, they may sound bright. There were altogether very few discs ever pressed with this pre-emphasis curve and I could not find a single one in my 400+ CD collection. However, Scott Nixon does offer DIY info on how to construct a simple RCR filter to manually switch de-emphasis via a small toggle.

Output impedance is 100 ohms and signal output is 1.75 volts, just slightly lower than the 2V industry norm. The stock TubeDac uses a single 12 volt AC supply. My review sample came with an upgraded 3-amp toroidal power supply called the 3Xac, housed in a wee black ABS box that retails for $115.00 shipped. The TubeDac+ differs from the standard non-plussed TubeDac by substituting premium parts in key positions, including "Black Gate non-polars for output caps, Oscons in the post regulator supplies, Hexfreds, an upgraded 4-amp desktop transformer, as well as a simple passive switch-able sinc filter". On the back of the TubeDac+ are a pair of analog outputs, a single coax digital input, the sinc filter switch and power inlet. On the front are two LEDs, one for signal lock with a valid digital source, the other to indicate a pre-emphasis disc.

The power supply has a standard IEC power inlet on the rear as well as the power umbilical for the TubeDac+. While Scott offers varying configu-rations as DIY kits, the TubeDac+ is sold assembled only. All in all, these are cute, well built little boxes that will fit into the palm of your hand. Parts quality and build seemed excellent. I did not encounter any difficulties with the TubeDac+ at all and it performed flawlessly with my Rotel and Apex disc spinners.

During the review process, I asked Scott a couple of questions:

Could you go into more detail about the sinc filter? I really don't think I could notice any difference one way or the other.
It's a passive filter that is now switchable. All non-oversampled DACs are -3dB down at 22kHz due to not having a digital filter. The sinc filter passively fills in the 22K level to 0dB. This is very useful in some systems but can be a mild irritant in others. Unless a setup is missing 'air', I prefer to operate without it.
What got you interested in pursuing non-oversampling, no digital filter design?
I was simply bored with standard digital and had read about filterless DACs in late 90s' Japanese magazines. After a major surgery in late 01, I decided this avenue would be more interesting. I had built CD players with tubed outputs from 1985 to '89 and DACs with tubes from 89 to 01 under the brand name Anodyne. To 'switch' to non-oversampled DACs with tubes was a natural segue for my business. I decided to do it low cost, sell direct and bypass the dealer markup - at least to start out with. 6 months after post surgery, I was basically starting from scratch: Plain Jane looks, but "va va voom Betty" performance.
How do you reply to those who argue that omitting digital filtering will expose the partnering preamp to excessive levels of noise and distortion? It seems to be a bizarre claim to make as I certainly found your DAC extremely easy to listen to.
With most program material, there won't be a problem since the out-of-band noise is 40-50dB or better below the reference level. If you had a piece of "music" with a continuous 10kHz sine wave, then yes, there would be some quite apparent problems with out-of-band noise. But I've heard virtually no ill effects with any equipment hooked into my DAC. I'm sure some ultra wide-band gear exists that would have issues with the HF trash, but our ears can't hear it. Because my sinc filter needs it, I do employ a gentle low-pass filter set at around 65kHz with or without the filter engaged. That lays to rest the ultrasonic issue in the TubeDac. It's 6db down at 120K and 12dB down at 240K - not that there's any signal there anyway...