Jitterbuggin' Page 3

I'm always suspicious when I hear more detail; it's easy to mistake an etched and analytical character for increased resolution. I have no such reservations about the DTI Pro; not only was it more resolving of detail, but the presentation was actually smoother and more liquid. In particular, the treble was softer and gentler, with a greater liquidity and freedom from hash. Similarly, the mids were less hard and strident, giving the presentation a greater sense of ease. I also heard more air and bloom around the treble; cymbals seemed to hang in space, surrounded by air.

That's not all. The DTI Pro's effect on the bass was profound. Low frequencies became tauter, quicker, punchier, more extended, and had better pitch definition—the entire bottom end was more dynamic, alive, and exciting. Bass guitar had greater dynamics, better-defined pitch, and more of the inner detail that tells you you're hearing a bass guitar and not just a low-frequency sound. These improvements greatly added to the music's pace and drive.

These impressions were consistent whether I used the Mark Levinson No.31, Counterpoint DA-11, or NAD transports, or the Sonic Frontiers SFD-2 or Adcom GDA-600 processors. The differences were, however, greatest with the NAD and the Adcom.

After this assessment, I tried the DTI Pro with the Mark Levinson No.31 Reference CD transport and No.30.5 Reference digital processor—arguably the best-sounding digital front-end on the planet. To my surprise, the DTI Pro improved this combination significantly. The difference wasn't as dramatic as with the Counterpoint or NAD transports driving the Adcom processor, but it was enough of an improvement to make me want to put the DTI Pro back in the system immediately.

Because the No.30.5's digital filter can accept 20-bit words, I tried the DTI Pro in its various output word lengths. Comparing the 16-bit output to the 18-bit output, I heard a big increase in low-level detail, space and depth, transparency, and the other qualities described earlier. Going up to 20-bit output words improved the presentation only marginally over the 18-bit output. Based on this auditioning, I must conclude that Audio Alchemy's resolution enhancement does indeed work, and that the DTI Pro's jitter reduction is only partially responsible for the unit's large effect on the presentation.

In short, the DTI Pro was a musical revelation. These differences were dramatic—an across-the-board elevation of the musical presentation, not the kind you hear only after extensive comparisons.

Listening to the Ultrajitterbug
The UltrajitterBug had a less profound effect on the sound than the DTI Pro, but the presentation was nonetheless markedly improved—I could even hear the UJB's positive effect between the Mark Levinson No.31 transport and No.30.5 processor. When used between the Counterpoint or NAD transport and the Adcom processor, the difference rendered by the UJB was significant, and well worth the cost of the unit. The following comments reflect my auditioning of the UJB with the Adcom processor with the No.31, NAD 502, or Counterpoint DA-11 transports.

The UJB produced improvements very similar to those gotten with the DTI Pro: a more transparent, holographic soundstage, tighter bass, more precise image focus, and lots more space and depth. In addition, the UJB made the treble sound smoother, more refined, and less "chalky," the very top end more open, airy, and extended. Overall, the treble was better balanced and sounded more natural.

I'll go back to the UJB's effect on the soundstage—this was perhaps the biggest improvement rendered. Not only was the soundstage more transparent, but the UJB produced so much more space between instrumental outlines. That common digital phenomenon of making the presentation sound like one big flat wall of sound was replaced by a tangible sense of bloom around instrumental and vocal images. The sense of soundstage layering was also enhanced, with the UJB allowing the processor to resolve many more layers of depth.

The UJB also greatly improved the bass, tightening up the bottom end, improving pitch definition, and adding dynamic impact. Listen, for example, to how bassist John Patitucci doubles the melody with pianist Kei Akagi on the title track of Akagi's excellent Playroom CD (Bluemoon/Moo R2 79342). Adding the UJB made it so much easier to hear the pitch of the acoustic bass and the interaction with the piano; without it, bass pitch was less defined. The UJB's effect on the bass was less than that of the DTI Pro, but its improvement was nonetheless significant and welcome.

Just for kicks, I cascaded two UJBs in front of the Adcom processor. The sound improved again with the second UJB, although the difference was not as great as that heard by adding the first. Because the UJB produces a 6dB/octave rolloff in jitter energy starting at 1kHz, two UJBs would roll off the jitter at 12dB/octave.

Overall, the UltrajitterBug made a big improvement to the system—an improvement well worth its $699 price.

Listening to the VSP
Putting the VSP between these same transports and processors produced sonic results similar, but not identical, to those produced by the other two jitter-reduction units. The VSP had a greater effect on the presentation than did the UJB, but less than that of the DTI Pro.

The VSP's most salient characteristic was its transformation of the bass presentation. Putting the VSP in the chain snapped the bass into tight focus—like pulling taut a slack trampoline. Bass guitar suddenly had more dynamics, better pitch definition, and more detail. I could more clearly hear the attack and decay of each note—a character which made the music more rhythmically involving. Dave LaRue's terrific bass playing on the Dixie Dregs' new Full Circle CD (Capricorn 42021-2) was much more interesting and involving with the VSP in the chain. The Kei Akagi disc really highlighted the VSP's cleaning up of the bass and better resolution of low-frequency pitch.

A side effect of this increased low-frequency definition was a thinning of the presentation—the midbass was leaner, tighter, and had less weight. Adding the VSP was very much like switching from an underdamped to an overdamped loudspeaker—the underdamped unit has more bass, but it's of lower quality; the overdamped loudspeaker is leaner and less weighty, but presents much more detail and dynamics.

Of the three jitter attenuators reviewed, the VSP had the greatest and most beneficial effect on the bass; it was a close call compared to the DTI Pro, but the VSP was clearly more effective in improving the bass than the UJB.

The VSP's effect on the soundstage was also dramatic, and similar to the improvements I heard with the DTI Pro and the UJB: increased transparency, greater focus, more depth, and a sharper, more vivid spatial perspective. It's interesting how reducing jitter allows the processor to reveal depth, present a sense of space between instruments, and generally make the digital presentation more analog-like. I didn't, however, hear from the VSP the same degree of detail resolution I got from the DTI Pro.

Similarly, the VSP didn't have the same smoothing effect on the mids and treble heard with the DTI Pro and UJB. John McLaughlin's acoustic guitar on his Qué Alegria CD (Verve 837 280-2) had a little more edge than it did with the DTI Pro or UJB.

Overall, I liked what the VSP did for my system—despite the fact that the AD1890 chip does rewrite the audio data.

Each of the three jitter attenuators reviewed here improved the musical presentation of my system. This was true whether I used a $299 CD player as a transport with a $750 digital processor, or when the jitter attenuator was put between the Mark Levinson No.31 transport and No.30.5 digital processor. The beneficial musical effects of these devices, over a wide range of transports and processors, suggests that all of them will improve any digital front end. You should, however, audition them in your own system before buying.

Of the three, the Audio Alchemy DTI Pro rendered the greatest improvement on the presentation. Not only did it provide the benefits heard from the other two, but it went the extra step by increasing low-level detail, removing edge and glare, and dramatically deepening the soundstage. The DTI Pro is a breakthrough in digital audio reproduction.

The Sonic Frontiers UltrajitterBug's effect was similar to that of the DTI Pro, but of lesser magnitude. It tightened the bass, smoothed the treble, and greatly increased soundstage transparency. What you don't get with the UJB that the DTI Pro provides is the improvement in low-level resolution, and the DTI's midrange and treble liquidity.

At $1295, the DTI Pro is considerably more expensive than the UJB, but its effect on the music is well worth the price. Putting the DTI Pro in front of the Adcom GDA-600 made the $750 Adcom sound like a $4000 processor.

The UJB was the bargain here. At $699, it offered many of the DTI Pro's benefits, but at a much lower price. Although on an ultimate sonic basis the UJB fell short of the DTI Pro, the former provided much more than $699 worth of improvement to my system.

The Digital Domain VSP Digital Audio Control Center is the odd man out of this group. Its $1495 price reflects the unit's other functions (sample-rate conversion, multiple inputs and outputs) which are more of use to professionals than to hi-fi consumers—if you need these features, I highly recommend the VSP. Although I liked the VSP's effect on the music, the DTI Pro offered a greater musical improvement for $200 less.

I urge you to try one of these jitter attenuators in your system. They provided sonic improvements that far exceeded their asking prices, regardless of the quality of the digital front-end. In fact, a jitter attenuator should be considered an essential part of any high-end system.

hollowman's picture

It's good that now-classic gear is being re-measured with modern tools (and evolved skill sets).
I hope JA can get around to measuring digital gear that is STILL sought after. Such as DACs and CDPs with classic Philips chipsets: TDA1541, SAA7220, etc.

John Atkinson's picture
hollowman wrote:
I hope JA can get around to measuring digital gear that is STILL sought after. Such as DACs and CDPs with classic Philips chipsets: TDA1541, SAA7220, etc.

To that end, I dug out the sample of the Magnavox (Philips) CDB472 that I bought in 1988 to run some modern tests on it. Unfortunately, while the transport still worked, the display just said "ERR" (for error) with every CD I tried. It looks as if the laser pickup had died in the 30 years since I last used it. :-(

John Atkinson
Technical Editor, Stereophile

hollowman's picture

I used to own that model. It was an improvement over the famous CDB650, while still retaining the 650's Hall-effect turntable motor (later transports at that price range used a cheaper "toy" DC motor).
About your issue:
Well, it could be a dead laser. You may be able to know for sure if you remove the cover, hold your Smartphone camera over the laser and press PLAY. (The phone camera will allow you to safely see the laser redness w/o hazard to direct eyesight).
If the laser is okay, some of those electro caps may be completely dried out.
FWIW, I've owned over a dozen (now) vintage CDPs and have yet to encounter a bad laser.

There are plenty of vintage audio refurb/repair houses in big cities. I recall a Stereophile YouTube post featuring one in NY (Leeds Radio???).

In any case, it's worth having device of this vintage in good working cond ... you never know when science my re-require its utility ;)

John Atkinson's picture
hollowman wrote:
Well, it could be a dead laser. You may be able to know for sure if you remove the cover, hold your Smartphone camera over the laser and press PLAY. (The phone camera will allow you to safely see the laser redness w/o hazard to direct eyesight). If the laser is okay, some of those electro caps may be completely dried out.

Will try this. Thanks.

John Atkinson
Technical Editor, Stereophile

JRT's picture
John_Atkinson wrote:

"...the Magnavox (Philips) CDB472 that I bought in 1988..."

My first CDP was a Magnavox (Philips) CDB-473. Not sure now, but I think I bought that in 1987.

hollowman's picture

I do think measuring legacy digital gear is important (preferably equip that's been refurbished-- replace dried caps -- but NOT modified).
I don't think such a project has been undertaken by any other major publication -- specifically one with a standardized lab and routines (test methodology) as Stereophile.

It would be good to gain insight into issues, such as certain audiophiles' preference of classic multi-bit sound, or non-oversampling.

Having a database of revised metrics would be beneficial to both manufactures and audiophiles. It's not only test gear and test methods that have improved. But the list of test parameters has also increased (linearity, jitter, Dirac pulse, etc).
And there is room for growth to this list as well . For example, back in the mid-80s, Bob Carver published some measurements (in Audio mag), regarding his Digital Lens technology. Carver introduced 'scope measurements via Lissajous pattern showing (L-R)/ (L + R) ratio from an LP record vs. CD. They looked different:

Carver 1985

I don't think manufs and journalists paid too much attention. There may very well be many more objective metrics lost in the journal archives.

AudioIdiot63's picture

Hi John,
Very interesting again after a long time. Around 2000 I designed the Assemblage DAC3.1 and "jitterbug" D2D-1. I replaced the AES21 that was getting obsolete and couldn't doe 96k with a discrete dual PLL using a VCXO. According to the Clock Jitter analyzer designed by Dr Remy Fourre it had an intrinsic jitter of 1,5ps and a much lower jitter attenuation frequency. If you ever come across one of these I would be very interested how it performs on this test. Nice article thanks.