Theta Data Basic II CD transport + DS Pro Basic IIIa D/A processor

Home | Audio Magazine | Stereo Review magazine | Good Sound | Troubleshooting

(This article adapted from a 1996 review in Stereophile)

--- Data Basic II CD transport:

Outputs: one RCA coaxial S/PDIF, one AES/EBU.

Jitter: <20 picoseconds typical.

Dimensions: 19W by 3.5" H by 15.5" D.

Weight 20.5 lbs.

Price: $1995 (option. AT&T ST output adds $300: optional

Single LaserLinque optical input adds $800, plus $250 for the cable).

--- DS Pro Basic IIIa D/A converter:

Frequency range: DC-2lkHz (no tolerance given).

S/N ratio: >108dB.

THD at full output: 0.002%.

Balanced out level: 4.2V RMS.

Output impedance: 120 ohms (balanced); 60 ohms (single-ended)

Digital inputs: one AES/EBU, two coaxial or RCA jacks, TosLink optical.

Dig. input loops: one, on RCA jacks.

Digital filter: 8x-oversampling, custom-software-controlled.

Computing power: 86 effective MIPS

Internal mathematical resolution: 56 bits.

Output data resolution: 20 bits.

Linearity 20 bits.

Dimensions: 19W by 2.24" H by 13" D.

Weight 19 lbs

Price: $2695 (optional HDCD module adds $459: optional AT&T S-type input adds $300: optional Single-Mode LaserLinque optical input adds $800).

-- --- ---


NOTE:

In recenty years, 's chief digital designer, Mike Moffat, has been invoved in a follow-up company, .

develops much lower-priced products -- gear that also provides functionality, ergonomics and even portability not technically possible only a few years back...


My relationship with digital has been a stormy one. I was working in a record-pressing plant when CDs were first introduced, and I think I know how buggy-whip manufacturers felt upon spotting the nascent automobile. I bought one of the first-generation Magnavox machines and took it straight back to the store — surely it wasn’t meant to sound like that? And I worked at a large NYC record store during the mid-’80s, which meant that every day I had to figure out how to put more CDs on the shelves and which LPs to send back to their manufacturers —jobs that stuck in my craw.

As both the recordings and the technology that plays them have improved, however, I’ve made my peace with the format. Perhaps it wasn’t totally by choice, but let that pass. One of the first digital pieces that really impressed mc was ’s DS Pro Prime, which I thought did a terrific job of sorting out harmonic detail and rhythmic pace. So when JA suggested I audition the combination of the Data Basic II and the DS Pro Basic IIIa, I was more than tolerant of the idea I was enthusiastic.

This enthusiasm seems family-wide—even my cat approves of the Pro Basic IIIa. This is highly unusual: Turntables fascinate her, but she generally ignores digital gear. Yet she has wiled away many an hour on top of the Basic II taking comfort in its warmth and solid construction. Me too, now that we’re on the subject.

DESCRIPTION:

Both pieces are built—if you know what I mean. They feel solid and kinda chunky, and their weight belies their slim lines. The Data Basic II is a basic CD transport. The fascia sports a power button on the left and the disc drawer in the center. Under the display on the tight, eight dome-shaped buttons control the drawer and the play options. The rear panel accepts a modular IEC power-cable plug as well as RCA S/PDIF, AES/EBU, and TosLink data outputs. AT&T glass or ’s own Single-Mode LaserLinque optical out puts arc available as options. There is also a rocker-type power switch that powers-off the unit—the front panel switch only de-activates some of the dig ital circuitry. The unit rests upon four compliant feet; there is also a spike that mechanically “grounds” the transport to the surface upon which it sits.

The original Basic transport was in production for over two years and was considered by some to set the standard for transports costing less than $3000.

The Basic II’s improvements center around the power supply — along with the isolation of what calls the “crucial” digital sections. In addition to working on power regulation and isolation, has rethought the transformers used, substituting low-flux models said to bring improvements to low-level resolution and dynamic contrast

The DS Pro Basic IIIa D/A processor offers a variety of input options, including AES/EBU, two RCA S/PDIF, tape in and out (in the digital domain), TosLink, and (again) the option of ST or Single-Mode glass. It has both single- ended and balanced analog outputs. The front panel boasts an input selector, a tape/source switch, indicators for each of the inputs, a polarity switch, and an on/off switch. HDCD capability is available as an option, and when HDCD-encoded discs arc played, a red LED indicates that the circuitry has been engaged.

describes the Pro Basic II topology as follows: “Input selection is performed by a field-programmable gate array which replaces numerous logic ICs and the attendant circuit- board layout problems. Following the input selection, the signal is routed to a Crystal CS8412 decoding IC which decodes the S/PDIF datastream into bit [clock] word [ clock] and data. An advanced phase-locked loop circuit then performs jitter reduction. The bit, word, and data signals are sent to the heart of the digital section: two Motorola DSP 56001 microprocessors implementing ’s proprietary algorithm. The 56001s perform eight times oversampling to 20-bit precision.

“The oversampled datastream is routed to four 20-bit Burr-Brown PCM1702K digital-to-analog converters. Each DAC converts one phase of the signal for fully differential operation. Following each DAC is a high speed current-to-voltage section based on the Analog Devices AD841 . . . At this point, each phase of the balanced signal is sent to a buffer for the balanced output and, in addition, is sent to a summing net work which does a balanced-to-single- ended conversion using both halves of the balanced signal. This summing stage gives the single-ended outputs a 6dB S/N ratio advantage, compared with simply using the plus phases of the balanced signal as the single-ended signal. After the summing stage, the now single-ended signals are buffered and routed to the output jacks.”

uses custom-wound low-flux transformers in the IIIa — one for the digital section and one solely for the analog. This, they claim, prevents the digital section from compromising the analog section through the transformers. The microprocessors and input ICs have their own regulated power supply, separate from the phase-locked loop (PLL) section. The analog section employs six individually regulated power supplies.

The Pro Basic II differs from the III in three major areas. The PLL circuit, which generates a clock signal in sync with the transport’s while rejecting jitter or other spuriae, was found to have significant impact upon the sound. Experimentation with different PLL filters revealed that they all affected the sound in different ways, so now optimizes them for each model. In addition, has installed a new filter program that adds dither to the signal below the threshold of audibility. Finally, they have implemented a new analog filter; which in turn required careful matching of capacitors to the circuit. The IIIa employs Wima capacitors, to which attributes increased transparency and the elimination of unnatural sibilance (interview with Neil Sinclair and Jim White).

ASSOC. EQUIPMENT USED:

I auditioned the gear in the con text of several systems, but most of my critical evaluation stems from teaming it with an Audio Research LS22/VT 130SE/Aerial Acoustics 10T setup. Straight Wire Virtuoso Gold balanced interconnect and speaker cable connected everything. My Lion LPI2/Naim Armageddon/Naim ARO/Sumiko Transfiguration Temper/Naim Prefix/Supercap front-end served as the analog comparison. An assortment of accessories tuned the rig: VersaLabs Red Rollers, Flat Rollers, Ground Block, and Wood Blocks; Audio Power Industries Power Wedge 112; MIT Z-Series power cables; Highwire Audio Power Wrap (on components with non-replaceable power cables); Shakti Stones (on pre- and power amps); Bedim Ultra Clarifier; Townshend Seismic Sink (under turntable); The Shelf by Black Diamond Racing (under the amp); and Golden Sound DH Cones (under everything but the Linn). I also listened straight into the Stax Omega/SRM-T1S—-a hi-rez minimalist headphone system that was like putting an electron microscope on the combo.

SMALL BEER:

Let me address the small stuff first.

Small -- like the buttons on each unit’s faceplate, and the discreet gold lettering that identifies them, which is virtually unreadable if the gear is not precisely at eye level. Is this a big deal? Maybe not, but even after using the Data Basic II for several months I found the location of most of these buttons confusing, and my wife has yet to discover the logic behind putting the stop command next to the drawer open/close command. (She told me to say that.)

In the context of day-to-day use, I wasn’t severely inconvenienced. I’d open the drawer; pop in a disc, and do everything else with the remote from my listening position. However, when preparing for WCES, I made several CD-Rs with the Meridian CDR inserted into the tape loop of the Pro Basic II and I found that I cycled through the input options of the unit by “braille” — I just kept hitting the selector until I heard something. Most folks probably won’t notice or care about such things, but if you’re one who might check this feature carefully.

Here in desert-dry Santa Fe, we are static electricity connoisseurs; we each perform our own little grounding rituals before we touch our audio gear. I found the Pro Basic II more sensitive to static than most of the processors I’ve used. I’m not talking about scrambling the DAC with a spark from my finger tips, but periodically I heard chh-chh-chh noises faintly coming out of one speaker or the other. The fix was simple —all I had to do was power down the system and then power down the Pro Basic IIIa. When I turned the system back on, the sound would be gone. The knows about this and (rightly) addresses it in their owner’s manual.

As to the rest, I was impressed by how carefully thought-out the units were. Many of my favorite tweaks made no improvement on their sound — which I took as a sign that had done their homework and built well- balanced products in the first place. Power cables were one area that did bring audible enhancements, so I used MIT’s Z-Cords. Confounding conventional wisdom, these digital units benefited greatly from being placed on high- quality platforms such as Black Diamond Racing’s The Shelf, Barry Kohan’s Bright Star Bases, or Townshend’s Seismic Sink. Go figure.

I used the Data II and the Pro Basic III as a single unit most of the time, although I did try each with other components. I fed the Pro Basic IIIa with the McCormack SST-i as well as with the AES/EBU outputs of the Audio Research CD-i. The Data II clearly mated well with the Pro Basic l in these comparisons. (I matched input type, of course.) The Data II/Pro Basic l combination revealed more low- level detail I could hear further into the music. I don’t just mean that I heard more space, air, and all of that audiophile stuff— although that was certainly true. I also heard more of the telling musical details that so distinguish one performance from another.

But the Single-Mode LaserLinque connection system (see sidebar) was the main reason I used the two units together nearly exclusively. If I heard more detail, and hence less grain and hash overlaying the musical information when combining two components, then changing over to ’s proprietary glass system put that change into warp drive. Mama! It wasn’t subtle. I just didn’t want to forfeit that level of clarity and nuance, so. . . I didn’t.

LISTENING TESTS:

’s gear has garnered the reputation of being a trifle, shall we say, relentless in its presentation of music’s top end. Not shrill-sounding exactly, but hyper-detailed and aggressive. Perhaps this was true at one time, but I found these two components to sound naturally balanced, although brimming with detail. There was an abundance of air surrounding the performers and an unusually well-developed portrayal of the performance space itself— neither of these being qualities generally associated with CD playback in my experience.


Fig. 1 Steve Reich’s Drumming basic rhythmic element [Reich originally recorded the work on LP in 1973 for DG (2740 106), but this 1987 release is an better realization, enhanced by a brilliant recording by Judirh Shennan.]

A case in point is Calamus’s The Splendor of Al-Andaluz (MA Recordings M026A), which was recorded in an ancient Spanish abbey, the stone walls of which impart a profound reverberant signature to the proceedings. “Modaddar ‘Va Mushniix” is a deceptively simple piece that has as its heartbeat a pulse established by a large, loose-skinned frame drum, a small clay drum called a dumbek, and some metal shakers. Over this rhythm sail two melodies, one taken by a piercingly pure-toned cane flute, the other by a mellow-sounding portative organ. To hear the intertwined melodies over the hypnotic propulsion of the percussion is to be pulled into a dance — or perhaps trance — state. But as easy as it is to describe, it’s a bear to unravel.

The two melody voices are distinct, but can blur together through some players. So, too, the frame drum:

Indeed, as much as I have listened to this disc, it was not until I played it on the combo that I clearly heard that there are, in fact, two frame drums playing —one of them a tar, with the metal rattles, the other a bender, which is much larger. I’d never heard the sound of the tar’s membrane as distinct from that of the larger drum before. Then, also, there’s the character of the room the musicians are playing in: a large stone chamber that has a mesmerizing decay — fast and clean, but incredibly extended in the finish. ‘With the , the decay seems to last forever; on most digital gear of my acquaintance, it sounds like someone’s riding the gain with a “click-stop” pot.

Another CD that tests the Pro Basic II ability to unravel complex information is Steve Reich’s Drumming (Nonesuch 79170 I would not be guilty of over-statement if I claimed that encountering Drumming in 1976 changed the way I heard music—certainly it shook my convictions about something as basic as rhythm and forced me to reevaluate the power of the beat The work is divided into four parts played without a pause. Part one is scored for four pairs of tuned bongos; part two for three marimbas, played by nine players and accompanied by two women’s voices; part three for three glockenspiels, played by four players, and accompanied by whistling and piccolo; while the fourth section com bines the instrumental threes from all three preceding sections. The 57- minute work has only one rhythmic pattern (see fig.1). It starts by constructing the basic pattern, with only one beat played in each cycle of 12 beats. Gradually the other beats in the pattern are added, and for the rest of the work every musician plays that core rhythm, or some portion of it. The pattern changes phase position, pitch, and timbre constantly, but never key—which is the conventional method of maintaining interest in a prolonged musical composition, at least in Western music.

Sounds like it would be dull, doesn’t it? But it’s not; it’s subtly hypnotic. As the basic beat begins to drift among the players, new and fascinating rhythmic patterns form; these clusters come to the forefront, break up, and reform into differing patterns made from the same materials. There’s no melodic progression — hell, there’s no melody! — but changes in pitch (and instrumentation) keep the ear/mind involved.

Drumming gives the critical audiophile a cruel tool for evaluating low- level resolution, pace, and subtle dynamic shifts at both ends of the spectrum. It was recorded in RCA’s Studio A, and one can hear the large room — not overly reverberant, but certainly not dead —in the decay of the individual drum beats. This is most noticeable in the early passages, where the rests out number the beats, but I’ll be damned if it didn’t inform even the busiest pas sages through the . As the drum- beats shift emphasis, and cluster and re cluster in changing rhythmic patterns, the combo kept the drive and swing intrinsic to the piece alive. The complexity of the patterns is starkly set against the acoustic, and rye heard processors go mad trying to reproduce the contrast: the beats become blurred; the emphasis sounds wrong; the decay of the instruments no longer dies a natural death, but is cut off abruptly. Shabby treatment for a work of art capable of shattering preconceptions.

The III presented all of the little details that make the overall effect of the work so powerful. In fact, I was reminded of the first time I saw Steve Reich and Musicians; having come to his works through the recordings on DG and ECM, I was unprepared for the showed rich stew of overtones that hovered over the stage, clanging and clashing against one another. Those discs, recorded in arid studios, didn’t offer a clue as to that dimension of his sound. But in his later recordings, captured by Ms. Sherman, it’s all there to be revealed —if your system’s up to it. And the is.

I would be remiss if I didn’t give special mention to the amount of bass the DS Pro Basic extracts from discs. It’s impressive. It doesn’t sound hyped-up because it is well differentiated in pitch. Articulate, too. This baby’s got bottom. I know that bass is supposed to be one of the glories of digital, but I’ve never found the bass from CDs to be particularly convincing. ’s on to some thing, though, because the IIIa keeps everything lithe and focused. And deep — did I mention that?

The illusion of layered depth was very well-constructed with the . While my current listening room doesn’t sustain this quality as well as my previous one did, the DS Pro IIIa/Data Basic II combo managed to distinguish themselves in this area. Is this merely an artifact of the recording process, as some have claimed, or does it have musical relevance as well?

Personally, I would never sacrifice timbral accuracy or relaxed articulation for it, but I appreciate the hell out of it when it done well, as it is here.

SUBJECTIVE TEST SUMMARY:

So a rave? Well, yeah. Okay, I’m not a measurement engineer, who has more experience with cutting-edge digital products than almost any other man on the planet, nor am I John Atkinson, who goes home every night to the mighty Mark Levinson No.30.5/31 combo. The Data Basic II transport and DS Pro Basic IIIa D/A processor are pretty much the most expensive and ambitious digital gear I’ve auditioned for any length of time. But they offer performance commensurate with their cost. I thoroughly enjoyed them in my system and consistently discovered new levels of musical enjoyment in my CDs — which is saying a lot. As long as we have the 16-bit CD standard there will be cause to complain about CD sound, but with the Data Basic II/DS Pro Basic IIIa, I really didn’t suffer much at all. Nope, hardly a single bit.

Not!

= = = BASIC PREMISES: ’s Neil Sinclair and Jim White provide some thoughts = = =

Neil Sinclair had come to Santa Fe to deliver some equipment for testing and had brought Jim White— his right-hand man at — because the airlines were offering a “friends fly free” deal. I couldn’t pass up the opportunity to get them together for a talk about , the DS Pro Basic Lila and Data Basic II, and the state of digital itself. It was a gray February afternoon, perfect for settling into a relaxing conversation — but I gotta tell you, these guys were hard to keep up with. They passed the conversational ball back and forth rapidly once they warmed to the subject, so I just held on and listened intently. We began talking about records and music when Jim got on the subject of “digital resistance”

Jim White: I think consumers have a lot of misunderstandings about what digital means. They don’t understand the full implications of it—not that it’s complex, but it’s not within the scope of most people’s understanding. So any technical explanation as to why they’re having trouble is often disregarded as technobabble.

Neil Sinclair: Even those who are somewhat technical have gotten used to living in an analog world and feel uncomfortable because they really are just different worlds.

Wes WP: So what are the new ground rues?

Sinclair: That all depends on where you’re trying to go. For example, in analog you have generation loss. In digital, you don’t necessarily—you may even have generation gain. Volume controls are also different, for example, if you use a really good resistor in analog, you don’t lose much at all. If you control volume through a DAC, you lose resolution. You lose the way a note fades out, you lose the room — the DAC starts sounding gritty. You have to design power supplies differently for analog and digital. Grounding is crucial in both, but it’s applied differently. Amplifiers have to be designed with great attention to magnetic fields, but not much is given to RFI.

WP: Tell me about ’s Basic line

Sinclair: Our goal with the Basic series has always been to see how close to our benchmark pieces we can come, while making them affordable. Someone to whom music and sound are important could afford to buy our Gen.V and the Data III, perhaps, if they were willing to move heaven and earth. But that’s a $10,000 combination, and we know that it’s out of reach for most people. The Basic products, on the other hand, cost under $2500 each.

How close can we come? We think of these as catch-up pieces — they represent technology that was available only in our state-of-the-art products not all that long ago. But it takes us a while longer to deter mine how to do it at a lower Cost. It’s exactly like the computer industry, where you can buy the latest processor and the largest hard drive early on and pay a hefty price, or you can wait a while and see those features offered as standard not much later.

For example, we sell a re-clocking device —the TLC — that strips the jitter from the signal and offers TosLink inputs in addition to coaxial for just under $200. A few years back, you couldn’t have bought anything like that for the money. Now it’s priced so that anyone with an aging CD player could buy one and a D/A converter for a relatively small amount of money and garner a huge increase in sound quality Yet, it’s good enough that until I began using our Single-Mode option, I had one in between my Gen.V and Data III — so it also represents a lot of technology for the money.

WP: Sounds like it’s time to ask about Single-Mode. What’s it do?

White: Basically, with AT&T glass cables you have two pieces of glass with differing numerical apertures that give you an effect called total internal reflection, where the light never leaves the first piece of glass because of the angle that it strikes it at. If the light were incident at a certain angle, it would go through, but it doesn’t. The aperture, or the diameter, of AT&T glass cable is 62.5 microns [um] which is rather large for the wavelength of the light used. The light itself has many paths that it can take as it travels down the fiber, and then the connectors themselves have flat ends. Flat isn’t good, because as the light hits, some of it bounces back and only some of it passes through. And whatever bounces back ends up as jitter.

In Single-Mode, the diameter of the glass is only 83 um—an eighth as wide as the AT&T—which means that, given the wavelength of the light used, there’s only one path it can take down the fiber. The connecting surfaces are dome-shaped, which means they make contact when joined, so you get very little back-reflection.

WP: Is that also why recommends Optigue -- because it standardizes the optical impedance at the connection?

White: That’s right, and because it eliminates any air in the connections.

Sinclair: Even though it’s formulated for use with glass, we find it improves the transfer for TosLink as well —it’s still a closer impedance/refractive index match to the plastic than air!

White: One other advantage to Single-Mode — one that it shares with AT&T optical -- is you eliminate the ground connection between the transport and the processor.

Sinclair: That’s a pathway for garbage to travel along!

WP: Single-Mode is offered as an option on all your units. How does it add to the cost?

White: It $800 on each end and $250 for the cable.

WP: Should someone spend their extra $1000 buying a better processor or adding Single-Mode to Pro Basic IIIa?

White: That’s a tough one. It would depend upon the system.

Sinclair: You could buy a single-ended Gen.V for the extra money if your system is single-ended, that’s what I’d do.

White: But if your system is balanced, I’d get the Basic IIIa, because you get a lot of benefit from running its output balanced.

Sinclair: You get balanced DACs, and you get greater linearity and drive ability in the analog stage.

White: This is the kind of question that customers ask us all the time:

“Should I get a Basic III with a Data Basic, or should I get a Gen.V and keep my current transport?” It’s really hard to answer because it all depends on the system and the person using it.

WP: Since the ultimate answer so system- and user-dependent, why not just confine ourselves to a description of the Pro Basic IIIa and the Data Basic II?

White: We’ve done a lot of work on the transport, but we haven’t really discussed all of it —just bits and pieces in various pieces of literature. We start with the power supply. Basically, the II’s an all-digital machine. All of those clocks generate very high frequencies and they all interact. The primary route for this interaction is through the power supply. In the Data Basic II, the main thrust for improving it came through work on the power supply. The more we address questions such as: what kinds of power supplies work; what kind of regulation works what kind of transformers or grounding schemes or regulation or power-sup ply capacitors sound best—the more we find that it has a tremendous impact on the end output.

Sinclair: That’s really the key.

White: There are additional things we do to the signal itself, as far as risetime. We have a very fast transformer-coupled output that maintains the integrity of the signal while giving the proper output impedance. We could probably design a better interface for just equipment, but in order to work well with a really wide range of equipment, we were forced to optimize the industry standard.

Sinclair: Our first prototype, which we called the Frankenstein, had a separate cable between the processor and the transport for the clock — which was really neat, but you couldn’t hook them up to anything else.

White: Now you’re starting to see systems like that from other companies.

WP: is there anything else about the transport you’d like to mention? I was impressed by how substantial it is.

White: We’ve done a lot of work on the physical package, including vibrational behavior — you noticed the “grounding” spike, for example, obviously, there are benefits from working there, but we found the differences from improving the power sup ply to be so significant that we decided to focus on that—and on re clocking the digital signals. But there’s more work to be done there in the future. Let’s just say that that’s an area where we are engaged in ongoing research.

WP: Onto the Pro Basic IIIa…

Sinclair: Its digital processing section is similar to the Gen.V. That’s one of the keys to what a is; really, because we use DSP for reconstruction filters. The Basic III uses two thirds the computer processing power of the Generation V. as well as a similar algorithm —just not as complex.

We don’t tend to use the same parts that the mass-market companies do, nor do we use similar topologies —open up a lot of CD players and you d notice how they all have the same parts and very similar circuits, with the differences mainly falling in the analog circuits. We do use some standard elements —we use a Crystal [ chip] like many other people do — but we do unique things with our phase-locked loop, and to control jitter. We’ve done a lot of research into the audibility of jitter and we’ve experimented a lot with filters to reduce it. Our reconstruction filters are unique.

White: One of the biggest changes in the Basic IIIa, as compared to the III, is in the software program itself. We found that adding certain amounts of dither at select frequencies improved the linearity of the DAC—which means the listener hears an amazing amount of low level detail and decay.

Sinclair: And cleanliness.

White: Exactly—all from noise that is supposedly below the level of audibility. For instance, hearing the natural decay of a bell is something that digital supposedly doesn’t allow you to do but using dither and our DSP reconstruction you can hear that clearly.

Sinclair: You’ve heard digital play back turn decay into a process where it steps down, rather than smoothly fades out — or just truncates it. We’ve been working for a decade now just to get that right But in music, you don’t have that easy a task — usually you have complex sounds, with each fading off with different timing...

White: That’s something that analog naturally does right. It can’t do it wrong. It has all kinds of noise and other distortions occurring but it does that right because you can hear down into the noise-floor—easily. The noise may be at a very high level, but at any one frequency it’s actually at a very low level. The ear can hear an event occurring way down into that noise. So while analog has its problems it gets that part right. For the first time were approaching that with digital.

The PLL filter is another area where has done a lot of work. It has major sonic implications in imaging. Things are much more clearly defined, and the spaces between them are blacker. What the circuit does is take an incoming dock signal and compare it to the outgoing signal. If the two aren’t the same, the error artifact can pollute the clock that drives the DAC. The PLL filter eliminates that effect—that’s a very tricky business, especially to accomplish without degrading other areas

WP: The Pro Basic IIIa also has HDCD available as an option...

Sinclair: We didn’t want to make people buy it. For normal discs — which is still most of them—we feel pretty secure about our own filters. When we ye done tests with HDCD discs — some of which axe excellent —our results are not so conclusive Some people prefer our filter on some HDCD discs, others prefer the HDCD filter. People don t even agree as to which discs are better.

After all this time, there are still only about 25 HDCD-encoded discs. We get calls from people all the time asking us to guarantee that HDCD is the wave of the future, and we just can’t do that. We offer it if people want it if they don t fine — or if they want to wait and see if it takes off, that’s fine too. There’s no price difference between purchasing it as part of a new machine or adding it to one you have. We did feel that we should make it available, though, to give customers a choice.

WP: Any other features?

Sinclair: The DS Pros flexible having multiple inputs: optical, AES/ EBU, three S/PDIF...

WP: Then you happy with S/PDIF as an interface.

White: Let’s just call it a necessary evil. Ideally, you’d have three separate cables: data, word, and clock. But then you d have three separate interfaces and you’d be paying for three different interconnecting cables. It would cost consumers a lot.

Sinclair: Plus, you’ve already got too many standards — TosLink, AES/ EBU, S/PDIF, AT&T/ST. and, of course, we’ve added Single-Mode. It’d be hard to get consensus to create another standard.

White: Not to mention all the sub-standards of coax. You’ve got BNC, SMC, and all the different connectors that coax has. Digital is already confusing enough to most people.

WP: Isn’t this where we [audio journalists] came in?

= = = = =

OBJECTIVE MEASUREMENTS:

As with other Digital products I’ve measured, the DS Pro Basic IIIa had a very high output level, 3.75V at the single-ended outputs and 7.5V at the balanced jacks. maintains that this high output level improves their products’ musical performance. When driving a normal high-gain preamplifier with the DS Pro Basic IIIa, however, you’ll end up making very small adjustments to the volume control at the bottom of its range (particularly when using the balanced outputs), where its interchannel balance is most likely to be awry

The output impedance was an extremely low 14 ohms (single-ended) and 28 ohms (balanced). This combination of high output level and low output impedance makes the DS Pro Basic III ideal for driving a power amplifier through a passive attenuator. The DS Pro III’s DC offset levels were a moderate 4.5mV (left channel) and 7mV (right channel). Although the processor locked to 32kHz and 48kHz sampling frequencies, it produced a distorted waveform when processing 32kHz data. This is a moot point for most users because the digital audio broadcasting frequency of 32kHz isn’t used in North America. The DS Pro Basic III doesn’t invert absolute polarity

Fig 2 is the DS Pro Basic III’s frequency response and dc-emphasis error. The response is ruler-flat, with only a slight rolloff at 20kHz. This response, determined by the Basic III’s custom digital filter is flatter than that of most processors using off-the-shelf digital filters. Crosstalk (fig.3) was excellent, measuring more than 120dB of channel separation at 1kHz.

A spectral analysis of the DS Pro Basic III’s output when decoding a 1kHz, —90dB dithered sinewave (flg.4) showed superb performance. The noise level is low the trace of second harmonic distortion seen in many processors in this test is almost non-existent, and only a hint of power-supply noise can be seen in the audio signal (the tiny bump at 120Hz). A wideband spectral analysis (fig.5) confirms the Basic III’s low noise, along with the effect of the output filter above the audioband.


Fig.2 DS Pro Basic IIIa, frequency response (top) and dc-emphasis error (bottom) (right channel dashed, 0.5dB/vertical div.).


Fig.3 DS Pro Basic IIIa, crosstalk (R—L. channel dashed, 10dB/vertical div.)

Fig.4 DS Pro Basic IIIa, spectrum of dithered I kHz tone at —90.31 dBFS, with noise and spuriae (20-bit data, 1/3 octave analysis, right channel dashed).

Fig.6 is the DS Pro Basic III’s linearity, which shows that the unit’s PCM 1702 DACs are performing well to about —110dB, suggesting performance approaching 20-bit resolution. Looking next at the ’s reproduction of a 1kHz, -90.31dB undithered sinewave (fig.7), we can see the good performance of the DACs, along with a low level of audioband noise overlaying the wave form. The ’s reproduction of the same signal with 20-bit resolution (fig.8) shows fairly uniform step size.

The DS Pro Basic III’s noise modulation plot (fig.9) was excellent; suggesting the unit’s noise floor doesn’t shift in level or spectral content as the input signal changes amplitude. An FFT-derived spectral analysis of the DS Pro Basic III out put when processing data representing a full-scale mix of 19kHz and 20kHz tones (fig.10) showed low levels of intermodulation distortion. The 1kHz difference component; here buried well below -100dB, usually protrudes into the -95dB range.

I next measured the DS Pro Basic III’s clock jitter with the Meitner LIM Detector. The signal source was a PS Audio Lambda playing test tones from the CBS Test Disc.

Fig.11 shows the jitter spectrum when the DS Pro Basic III was processing a 1kHz full-scale sinewave. Note the absence of signal- correlated jitter components, which would be seen as spikes in the trace.

Instead, the plot is extremely clean, indicating the unit’s clock jitter is primarily the more sonically benign “white” jitter rather than periodic jitter. Although free from periodic jitter components, the RMS jitter level measured at the 8x clock pin was a highish 500ps.

The RMS level dropped to 450ps when the unit was processing a signal of all zeros, but rose back to 500ps with a 1kHz, —90dB sinewave input. The three traces all looked similar, suggesting that the DS Pm Basic III has a good ability to reject incoming jitter; although its intrinsic jitter is moderately high. There’s a lot of speculation about the audibility of jitter, particularly with regard to whether the jitter is random or periodic. Higher levels of jitter are tolerable if that jitter is random rather than having a specific frequency.

I was impressed by the DS Pro Basic III’s bench performance, particularly the high channel separation, low noise, and good low-level performance.


Fig.5 DS Pro Basic III a, spectrum of digital silence (20-bit data, s-octave analysis, right channel dashed).

Fig.6 DS Pro Basic IIIa, departure from linearity (right channel dashed, 2dB/vertical div.)


Fig.7 DS Pro Basic IIIa, waveform of undithered 1 kHz sinewave at —90.31 dBFS (16-bit data).

Fig.8 DS Pro Basic IIIa, waveform of undithered 1 kHz sinewave at —90.31 dBFS (20-bit data).


Fig.9 DS Pro Basic IIIa, noise modulation, —60 to —100dBFS (10dB/vertical div.).

Fig. 10 DS Pro Basic IIIa, HF intermodulation spectrum, DC—24kHz, I 9+20kHz at 0dBFS (linear frequency scale, 20dB/vertical div.).

Fig. 11. DS Pro Basic IIIa, word-clock jitter spectrum, DC—20kHz, when processing 1 kHz sinewave at 0dBFS; PS Audio Lambda transport (linear frequency scale, 10dB/ vertical div., 0dB/ns).

== == ==

(adapted from review in Stereophile, May 1995)

== ==

ALSO SEE: Theta Digital, official site

Re-creating the Musical Event-- Digital’s Mike Moffat discusses digital

== == ==

Prev. | Next

Top of Page   All Related Articles    Home

Updated: Tuesday, 2015-04-21 17:19 PST