What does Jitter sound like?

At its worst, jitter is revealed as a hard, flat, fatiguing glare across the midrange & treble. A low-jitter clock relieves strain and adds suppleness & body. Further reductions in jitter get closer to analog-like treble delicacy, tighter bass control, and overall ease. Sounds a little like a good massage...

Dpac996, these are my personal observations going from stock oscillator in my SCD-1, through three generations of Audiocom Superclock up to SC4. In each modification, only the clock (or the clock power supply) were changed. I installed each clock myself & made the comparison within hours of deinstalling the previous clock. There was immediate dramatic improvement, and further improvement over up to 200 hours of break-in. If this is just "going for it" then so be it. There is enough controversy about measured jitter that I'll go with the evidence of my ears on this one.

The SCD-1 is different from some CDPs, insofar as a Synchronous Time Accuracy Controller (S-TACT) chip that is located next to DAC in the analog section of the player, isolates master-clock generation and signal voltage-pulse generation in one chip at a remove from noise propagated by high-speed operations in the RF & digital sections. The stock quartz oscillator chip is located in the analog section 1" from the S-TACT. In a clock upgrade the stock oscillator chip is removed & replaced by a lead-out to a 45Mhz clock PCB. To eliminate the possibility of noise induced by capacitive effects across the PCB, I lifted Xin pin of the S-TACT from the PCB and made short ptp connections between Xin & clock PCB. So in view of Sony's set-up that theoretically isolates master clock from digital noise, I am inclined to consider that audible improvement may be largely due to accuracy of the replacement clock.

I've heard audible improvements upgrading power to clock, and also after upgrading power and various passive SMD components around DSPs, PLLs, and motor/servo in the digital section. Whether this is due to reduced jitter or to reduction of other noise I cannot say. But I have no doubt that improvements in the digital domain other than clock are audible as well.

Esoteric sells an optional out-board clock for its top player that's accurate to a half part per billion; magazine reviewers claimed to hear a difference. I know a fellow in Japan who makes similar claims using an external pro signal generator to clock his player.

All this might sound like an exercise in diminishing returns but my ears tell me otherwise.

Shadorne, Feb. Stereophile is one of the better issues in recent memory, with entertaining assaults by Atkinson & Fremer on the "MP3 as indistinguishable from CD" fiction propagated in the mainstream press. At the opposite extreme there are (finally) several less-than-superlative reviews of megabuck products (e.g. Escalante Fremont & BAT Rex.)

Of course you won't believe it, but RBCD on my tweaked CDP and vinyl on my tweaked TT run neck & neck. Making comparisons across analog & digital media raises the bar & can be helpful in measuring progress. MP3 sounds great doing an 8 minute mile on the treadmill.

Like some deleterious affects in audio, it may be easier to appreciate & to describe in its absence. Sometimes you don't know what you got till its gone. But dancing angels it most definitely is not.

Audioengr,

From either a theoretical or experiential perspective, do you think the phomenenon of speed instability in a turntable is similar to digital jitter?

The reason I mention comparison between speed fluctuations in TT & digital jitter in CDP is because I experience these phenomena quite differently. As the speed stability of my TT has improved through various upgrades, I hear improved dimensionality, soundstage, bass control, dynamics, liveliness, focus, etc., even though pitch remains audibly imperfect. On this narrow point of pitch stability the CDP surpasses and I suspect will always surpass the TT (as evidenced particularly with piano music.) But somehow the ear is forgiving of even quite audible fluctuations in timing that originate in the analog domain. Perhaps this is because the physical locus of the stylus ensures that timing variations are applied uniformly across all spectra at each point in the LP groove.

In contrast, digital jitter seems to smear timing quite differently and more objectionably so. Having progressed through four generations of clock in my CDP I can say that most of what listeners think is synthetic & irritating about RBCD relates to jitter.

Amen Mrtennis.

My listening tests relating to the perception of jitter were gathered through four generations of clock upgrades within one CDP-- eliminating all other variables (except for cognitive dissonance), and disposing of all controversies regarding the quantitative measurement of jitter. A separate set of observations was formed through successive upgrades of DC power to clock, and to the digital sections of the player including motor/servo, PLLs, and LSI DSPs. This exercise led from simple stock 3-pin regulators, to sophisticated discrete regulator circuits, to batteries, to larger low-impedance batteries, to batteries with capacitance, to batteries with massive capacitance. To those entrenched in the "data is data" camp, I suggest disposing of theory & gambling a few hundred dollars on a simple clock upgrade and report back to the forum.

In the course of five years of developing the CDP, I've corresponded with perhaps a dozen amateur and professional modders who were all impressed(some to amazement) by the improvements associated with clock upgrades in otherwise respectable CDPs. There has been no criticism of the results of a clock upgrade.

Anyone seeking a more professional approach should visit Steve N's site, take a look at his Pace-Car product, note which clocks he uses, how he uses them, and how they are powered.

If you think that RBCD sounds synthetic or fatiguing or that the resolution of RBCD can't approach the musicality of vinyl, then you're probably hearing jitter.

Shadorne, I suggest speaking with Kyle at Reference Audio Mods 951-780-2869 about a Superclock for one of your devices. The SC4 retails for $300 and there are probably cheaper used or old-stock SC3s. I have no experience with budget players, but reportedly he has had decent results with Oppo and others under $500.