[Geardaddy]

What is this "Digital Lens"?  I am not familiar with that device....

[Rx8man]

Geardaddy,

The Digital Lens was designed by Paul McGowan of Genesis at that time (mid 90's) for around $1800.00

It was an earlier attempt at reducing the subject of our topic here called jitter.

An interface device for reclocking the digital signal, it was inserted (yes I know, extra cables etc.) between your Player/Transport and Dac.

We're still looking for the solution............ 

[Geardaddy]

What I am honestly waiting for is a wireless DAC/pre-amp that does not need mods.  As one software engineer put it on a Sonos modification thread on Audiogon recently:

"....if you just turned it (Sonos) into a digital file 'pass through' and let the DAC clock do the heavy lifting? It just seems ridiculous to me that there isn't an easier way to transfer the file into the DAC cleanly without worries of transport jitter.

What am I missing here? "

I know the Slim Device TP is an attempt at this, but it still needs further mods to make the grade (a la ModWright or Paychev).  The Nova Physics Memory player is another variation of this concept (sans networking technology) but I do not have $17,500 to burn on that....

[darrenyeats]

I know the Slim Device TP is an attempt at this, but it still needs further mods to make the grade (a la ModWright or Paychev).  The Nova Physics Memory player is another variation of this concept (sans networking technology) but I do not have $17,500 to burn on that....

Then there's the Linn Klimax DS for 20k etc etc.

The fact is the SB3 and Transporter are exactly the kind of device you seek. The TP measures amazingly well.

I've already posted earlier in the Part 1 my blind test experience. In that post you may have noticed that my preference was exactly the reverse of my (short and long term) preference sighted. This has convinced me that subjective listening on subtly different components such as digital sources might be ok informally. As for deciding on purchases or conducting product reviews, it is total bunkum. Total and utter waste of time (according to my personal experience).

All these debates come down to the blind vs sighted debate in the end. Here is where a bunch of people jump in and say "well this is a conversation killer, so don't go there Mr Yeats". But so many conversations in audio are a house of cards built on the foundation of sighted listening. It's a totally broken model for public evaluation yet people repeatedly trumpet that "their ears" are what is most important. The ears aren't the problem (you use your ears in a blind test, well don't you?)...it's the brain that is the issue and this is what is taken out of the equation in blind tests.

I can't put it any simpler: NO DEVICE will EVER be safe from modders who BELIEVE their mod is better. Someone will always pipe up about how this or that mod has transformed the sound. Doesn't mean a thing to me unless they evaluated it blind.

I realise this issue is extremely polarised, and the majority on audiocircle are in the other camp, but this is my view and I state it unashamedly. And it IS relevant. This is the underlying issue that causes the polarisation in ALL these types of audio debates. This is why there is often not a meeting of minds. This is why so often people agree to disagree in the end with these conversations. This is why no-one can agree over the importance of interconnects. Or whether there is a difference between SACD and CD. Or the audible threshold of jitter. Or, or, or. Blind vs sighted is THE issue in audio, so I beg that everyone - even if they don't agree with my particular answer - ponder upon the importance of the question.
Darren

[BrianM]

Darren, I for one completely agree with you in principle.  They do say people who go blind somewhere along the way undergo a significant transformation in how their brain processes what they hear.  But, since we pretty much all listen and shop with eyes open, we're kinda stuck with what the eyes and all their psychological trappings feed to us, sometimes in competition with our ears.  Is this bad?  As for components never being "safe" from modders, I don't know whether that notion needs a positive or negative value assigned to it; tinkering under the hood and listening for changes is doubtless fun and absorbing. Same reason some guys work on their cars (or used to, before they became phenomenally sophisticated precision machines that no layperson could expect to improve on ).

[darrenyeats]

Thanks Brian.

In terms of Geardaddy's question, a device that doesn't need mods, the answer is two-fold one for each world-view:

1. In terms of all measurements and any blind listening reports I know about, the Transporter player is his answer and it doesn't need mods to sound great. If anyone is able to produce evidence, reports, shootouts etc to the contrary I would be very interested to read them, and I mean that very genuinely.

2. If you're into sighted listening then there will never be any device that can't be improved significantly by modding.
Darren

[Geardaddy]

Darren, I totally agree with you.  The whole "psychoacoustic" realm (we in biomedical sciences call the placebo phenomenon) is incredibly powerful.  I have always done my best to perform blind testing with multiple participants when I am evaluating gear, but as BrianM says....its hard.  A friend of mine and I are planning a blind shootout between a Sonos + Exemplar USB tube DAC vs Boulder modded SB3 and my Granite Audio 657.  Should be interesting.  I am tired of pontificating about this crap.  I need to do some listening...  

[darrenyeats]

Darren, I totally agree with you.  The whole "psychoacoustic" realm (we in biomedical sciences call the placebo phenomenon) is incredibly powerful.  I have always done my best to perform blind testing with multiple participants when I am evaluating gear, but as BrianM says....its hard.

Geardaddy, I agree totally - it's very difficult in terms of finding the time and other people to assist etc. It turns into a long day to do some simple comparisons if you want to arrive at a real conclusion.

IMO however, the actual blind listening bit is quite fun and enjoyable. Plus the results tend to be quite interesting. Plus...other people tend to be quite interested about your findings too, more so than usual. Of course the result of a blind test is still based on your taste and your hearing faculties. The only thing filtered out is subconscious bias.
Darren

[darrenyeats]

A friend of mine and I are planning a blind shootout between a Sonos + Exemplar USB tube DAC vs Boulder modded SB3 and my Granite Audio 657.  Should be interesting.  I am tired of pontificating about this crap.  I need to do some listening... 

Geardaddy, can we expect a report in this forum? I for one would be very interested to read this. If you could include a stock Transporter too, that would be grand from my POV!
Darren

PS: I'm sure you know this already but don't forget to match the levels of each player with an SPL meter. I use noise or multi-band signal to do this. Being a bit louder is an unfair advantage.

[Geardaddy]

Agreed.  Thanks for the reminder.  I believe my partner in crime has a meter....

[BobM]

It sounds like the source of jitter in all of these discussions seems to be the internal clock. And the methods proposed to reduce jitter within the unit itself (not on the disk) is to treat, replace or intercept the clock's signal and make it more precise.

So, in summary, what tweaks are available to us, outside of treating the disk)? I've tried to summarize the discussions below:

(1) treating the existing clock crystal (e.g. Jon Risch's sand bag tweak, or applying Moretite or some other kind of vibration reducing compound)
 - effectiveness, slight
 - cost, cheap
 - ability to do ones-self, easily done

(2) replacing the clocking mechanism with something more precise (e.g. Superclock or some other internal clock replacement)
 - effectiveness, noticeable
 - cost, @$200-400 parts plus labor
 - ability to do ones-self, need specialized knowledge

(3) external reclocker
 - effectiveness, about the same as replacing the internal clock (I'm guessing here so please step in if I am mistaken)
 - cost, not inexpensive
 - ability to do ones-self, plug it in (may be dependent on the ability of the player to interface with such a device)

(4) what else?

Please corret any assumptions I may have made above that you perceive to be incorrect, based on your experience (not your opinions, please).

Enjoy,
Bob

[darrenyeats]

Back on topic, how about Benchmark (and other manufacturers) claiming their DACs are jitter-proof. Can I recap the conversation so far for myself and other non-engineering types, and if I'm inaccurate or ambiguous I'm sure one of the engineering types will correct me gently.

To recap, we have established that any signal even a jittered one can be expressed as a sum of sine waves. This is one way of measuring jitter...feed in a clean sine signal then jitter that signal. Then we look at the jittered signal in the "frequency domain" e.g. expressed in a graph with a loudness for each pure frequency. The output graph will show the input pure frequency loudest, but with new different frequencies added that are much quieter. These are referred to as sidebands in the graphs opaqueice posted.

This is why some people were talking about adding noise to the signal as an equivalent test for jitter, because when viewed in the frequency domain this is what jitter does.

The issue with jitter is that these noise products may - in part - be tightly correlated to the input signal. Such signal-correlated noise is, as I understand it due to psychoacoustic reasons, more noticeable i.e. there is a lower threshold of audibility for humans.

Nevertheless, looking at these sidebands if they are vanishingly small or - much more significantly - don't increase as input jitter is increased, then there is some validity to claiming the device is immune to jitter. Music isn't a pure sine tone but as discussed any signal, including music, can be expressed as a sum of sine tones, so these measurements are a valid indicator.

So...AFAIK the Benchmark boys (and others too plausibly) may be on to something. And on this thread, since opaqueice posted the Benchmark graphs nothing has been written to change this impression...any takers?
Darren

[woodsyi]

I like hard scientific explanations as much as most anyone, but I don't want to make the mistake of reducing everything to known equations either.  While I was getting my masters on religious studies, I did a lot of study on modern science (quantum physics and biology) and religion.  What I got out of it is that what we think is "hard" science is "only" a human construct to describe the natural phenomena.  At all times, however, we should always remember that it's only a map and not the real thing.  Music IMHO may not be reduced to just the sine waves.  Frequency, time and amplitude are quantifiable units we have defined to understand the music phenomenon.  Let's not get carried away looking at the shadow on the wall and thinking it is the real. 

[art]

It sounds like the source of jitter in all of these discussions seems to be the internal clock. And the methods proposed to reduce jitter within the unit itself (not on the disk) is to treat, replace or intercept the clock's signal and make it more precise.

So, in summary, what tweaks are available to us, outside of treating the disk)? I've tried to summarize the discussions below:

(1) treating the existing clock crystal (e.g. Jon Risch's sand bag tweak, or applying Moretite or some other kind of vibration reducing compound)
 - effectiveness, slight
 - cost, cheap
 - ability to do ones-self, easily done

And how much will it reduce jitter? It won't. It might isolate it from microphonics. Not as much a problem as you might want to believe.

(2) replacing the clocking mechanism with something more precise (e.g. Superclock or some other internal clock replacement)
 - effectiveness, noticeable
 - cost, @$200-400 parts plus labor
 - ability to do ones-self, need specialized knowledge

"Precise" has nothing to do with jitter. Clocks that come out of a certain corner of the globe seem to think so. They are wrong. Period.

Won't do diddly-squat unless you have a really clean power supply, and good distribution chain inside the unit. Which most don't have to begin with. Problem best addressed by fixing existing problems first.

(3) external reclocker
 - effectiveness, about the same as replacing the internal clock (I'm guessing here so please step in if I am mistaken)
 - cost, not inexpensive
 - ability to do ones-self, plug it in (may be dependent on the ability of the player to interface with such a device)

Only works for SPDIF. Which is the only application that jitter needs to be addressed. (I already showed in the other thread that jitter in a stand-alone player is not much of a problem. Stop worrying about it unless you have lots of time and money. Mostly the latter.)

(4) what else?

Please corret any assumptions I may have made above that you perceive to be incorrect, based on your experience (not your opinions, please).

Enjoy,
Bob

Ok, I did! Based on designing and building this stuff for >20 years.

Pat

[BrianM]

Only works for SPDIF. Which is the only application that jitter needs to be addressed. (I already showed in the other thread that jitter in a stand-alone player is not much of a problem. Stop worrying about it unless you have lots of time and money. Mostly the latter.)

I probably missed it somewhere, but to what extent do you feel the SPDIF problem is ameliorated by a so-called optimized digital cable length?

[Geardaddy]

So Pat, based on your experience (and forgive me for not remembering the particulars of the last thread ), what is/are the optimal ways to deal with jitter? 

You mention a new clock being worthless without a good power supply...how so?

[AphileEarlyAdopter]

It sounds like the source of jitter in all of these discussions seems to be the internal clock. And the methods proposed to reduce jitter within the unit itself (not on the disk) is to treat, replace or intercept the clock's signal and make it more precise.

So, in summary, what tweaks are available to us, outside of treating the disk)? I've tried to summarize the discussions below:

(1) treating the existing clock crystal (e.g. Jon Risch's sand bag tweak, or applying Moretite or some other kind of vibration reducing compound)
 - effectiveness, slight
 - cost, cheap
 - ability to do ones-self, easily done

And how much will it reduce jitter? It won't. It might isolate it from microphonics. Not as much a problem as you might want to believe.

Well, going by the popularity of footers and ball-bearing devices I think you are underestimating this issue. I have had very good improvement even with the Vibrapod cones under my Oppo.

(2) replacing the clocking mechanism with something more precise (e.g. Superclock or some other internal clock replacement)
 - effectiveness, noticeable
 - cost, @$200-400 parts plus labor
 - ability to do ones-self, need specialized knowledge

"Precise" has nothing to do with jitter. Clocks that come out of a certain corner of the globe seem to think so. They are wrong. Period.

Won't do diddly-squat unless you have a really clean power supply, and good distribution chain inside the unit. Which most don't have to begin with. Problem best addressed by fixing existing problems first.

I agree the power supply has to be good, but the clock also should be of good quality. For eg. when I got my Oppo modded, I just went with the power supply mods alone. If I had the money, I would have gone for a better clock too.

(3) external reclocker
 - effectiveness, about the same as replacing the internal clock (I'm guessing here so please step in if I am mistaken)
 - cost, not inexpensive
 - ability to do ones-self, plug it in (may be dependent on the ability of the player to interface with such a device)

Only works for SPDIF. Which is the only application that jitter needs to be addressed. (I already showed in the other thread that jitter in a stand-alone player is not much of a problem. Stop worrying about it unless you have lots of time and money. Mostly the latter.)

..

Well..not sure you are being precise here. There are two sources of jitter, when you have a transport +DAC(or receiver). The system or clock jitter in the transport and the interface jitter caused by the SPDIF. Lets say some how, you take care of the SPDIF interface. Still the system jitter is there and needs to be minimized. Reclockers like the ones from Empirical Audio, address both jitter - there is a master clock input (into SB3 for example) and I2S connection to your DAC (if it has that input) to take SPDIF out of the equation.

The Genesis Digital Lens is a reclocker getting the data into a 500KB memory and clocks it out, but then again you will have a SPDIF interface jitter to contend with.
The Apogee Big Ben is a 2 PLL reclocker, which again has the SPDIF out jitter, but can be reduced if you use the BNC connection and your DAC or receiver has a proper 75ohm BNC input.

(Yes, SPDIF jitter happens because in consumer products the digital inputs are not true 75ohm (terminated with a transformer like in pro products) and the cable/terminations may not also be true 75 ohm. I speculate with audiophile digital cables people are trying different cable/termination impedences with their equipment and when there is a match you get good performance. There is also the issue of throughput, which I am not sure how far it affects the jitter).

[*Scotty*]

AphileEarlyAdopter, for the purposes of this discussion a DVD/CD player is a black box.You put a footer under it and the sound may change but there is no way of knowing why it changed.
Noise on the digital supply rails can cause jitter or make it worse the two are inextricably linked.
Jitter occurs with the SPDIF interface because the clock is embedded in the data stream and cannot be recovered properly, period. Impedance mismatches and reflections as well as deformation of the squarewave the data train consists of make the problem even worse. SPDIF cannot be fixed or cured because it is inherently flawed.
Scotty

[BrianM]

SPDIF cannot be fixed or cured because it is inherently flawed.

That may well be, but like a lot of dogmatic statements it might just suffer from a narrow perspective.  The transducers in your speakers can't be "fixed" or "cured" either, and are inherently flawed.  (So is your listening room.)  Are we losing more sleep over these facts, or is it more about the jitter?  Which is worse, an awesome DAC + SPDIF or an awesome CDP which is probably compromised in other ways?  I'm sure nobody's against a better solution than SPDIF being more widely accepted, but the question (I thought) is whether SPDIF can be ameliorated to below the point at which other factors become more relevant to improving the sound.  Or, perhaps, whether it is already below such concerns.

[darrenyeats]

I agree S/PDIF is compromised, the reason being the timing is part of the data signal.

The DAC manufacturers who claim jitter-proofness use the general approach, one way or another, of isolating the clock of the DAC from the S/PDIF timing. I can imagine some implementations using buffers although these have additional challenges.

Whatever the implementation, it seems some manufacturers claim they have achieved the goal of clock isolation. IMO this negates the downside of S/PDIF. Don't get me wrong - I don't like S/PDIF still because I'd rather things be designed well from the ground up. But I have confidence enough in the technical argument and the measurements to use S/PDIF in the right DAC (and Benchmark, as discussed, have published such measurements). Until someone convinces me otherwise of course...
Darren