[BrianM]

Steps to reduce jitter -

• Reduce vibrations to your component. Use a rack with spikes and place each shelf on spikes. Lots of shelves, vibration reducing components available.
• Use a power conditioner or make sure power is clean of spikes/noise etc. Balanced power conditioner makes quite a bit of difference
• A component with good power supply makes the best of the oscillator/clock in the device. 50% of the cost of the pricey Halcro amp goes into designing a power factor corrected power supply for it. After market power cords help improve dynamic demand of current of the device. These dont have to be pricey. Simple DIY designs make quite a bit of difference
• In the case of CD/DVD players, a clean CD with some anti-glare spray helps the laser track efficiently reducing too much servo tracking and in turn putting more demands on the power supply, which will affect the whole system and jitter.
• With computer based play back, it is much easier to isolate the audio output device than to use the builtin soundcard. Examples would be USB/Firewire/Network(like Squeezebox) sound cards/devices. It will be easier to get a power supply to these devices than to the whole computer.

Here is my system -

Cryoed Hubbell wall outlet -> Empirical Audio power cord -> Furman Power Factor Pro -> Black Sand Violet power cord -> Blue Circel MR1200 Balanced Power conditioner -> Modded SB3 with modded Elpac power supply -> Zu Ash -> Behringer SRC2496 (with Vh Audio power cord)  -> Sound Professionals Glass Toslink -> Panasonic XR55 --- (bi amp) ---> Silverline Sonatina speakers.
My system is all digital till the speakers. Any change in the above setup I immediately notice a difference. So I am used to hearing the effects of jitter.

I don't mean any offense, but am I the only person who finds it strange to see so much $ in quality power conditioning (which apparently also removes jitter) leading up to a ~$250 HT receiver?  I understand the XR55 is a very nice receiver, but...I mean...is it a hi-fi component?  I thought jitter was kind of a hi-fi concern?

[marklivia]

I kinda have to agree. Mind you, I do have a Panny 55 in my own modest HT  setup and i love it,  but it would not make the cut in my 2 channel rig. It next to my Red Wine Sig 30.2 ........ sorry, not even remotely close in the clean signal dept.

[darrenyeats]

ok, that means you can hear the effects of jitter. You are an audiophile with golden ears

Is that an insult? 

FWIW I have done a double blind test with transports and it indicated jitter is audible.

It was an SB3 vs a Monarchy CLD-M401 CD transport, both going via coax S/PDIF into an old, not particularly jitter busting DAC. The SB3 was chosen as better 4 times out 4, despite sighted impressions (short or long term) seeming to be the opposite.

The negative is that we only did 4 rounds, and there is 1 in 8 chance of getting a 4/4 (one way or the other) by pure chance anyway. The positive is that the differences sounded pretty clear blind and choosing didn't seem difficult or agonising. The coax cable, DAC and rest of system was identical, and the DAC had a switchable input so quick A/B was possible. The person doing the switching didn't know which transport corresponded to which switch, he didn't let me know with which input he started each round and I stated my preference as "the first one" or "the second one" based on his starting choice each round. I could switch back and forth as much as I liked until I made a decision. Other than the number of listens, it was a good double blind test.

I hope this report is of help to the discussion. The DAC used was venerable so it was a good jitter test.

It doesn't get us closer to what level of jitter is audible though. It doesn't really speak about the claims of modern DAC manufacturers that their DACs are in effect immune from jitter. I'm open-minded about that.
Darren

[Ethan Winer]

I say jitter is more than noise, ergo adding noise to an otherwise "clean" file won't answer the question. Others in this thread have echoed these thoughts.

On what do you base that opinion? Forty thousand people could echo those thoughts and, logically speaking, it doesn't make those thoughts correct. I'm sure you agree with that! So again, what is your better proposed test? It's not fair to stand on the sidelines and throw darts. Ya gotta at least offer an alternate test proposal. It would also help if you'd explain why you believe jitter is more than noise.

Ethan, I notice a change or softening of your stance. Now you need to talk about jitter in the actual terms it is measured i.e peak-to-peak variation in time/picoseconds  (just ignoring the spectrum for simplicity sake).

No softening, though I hope nobody takes this as anything more than a technical discussion that aims to get to the bottom of what is audible and what is not. If I have an axe to grind, it's only my disappointment over seeing people repeatedly avoid the issues I keep raising.

If you use this parameter, most CD/DVD players under $1K have very high jitter.

Yet again you have failed to identify what is high, and how high it has to be before it's audible. Do you have an answer, or should I just stop asking?

If you do not recognise these, you are doing a disservice or misleading audiophiles here.

From my perspective, I'm the one who's actually serving audiophiles by trying to separate truth from fiction. Nobody wants to squander money on something that doesn't matter, so it's in the public interest to find what's worth buying and what's not.

Here is my system - Cryoed Hubbell wall outlet

Understood fully.

Peace folks, really.

--Ethan

[miklorsmith]

I wholly agree that # of votes does not equate to degree of correctness.  This is one of the most difficult lessons in our pursuit and one of the sociological phenomena referenced above.

I base my opinion on my experience, what I've heard, what is so commonly wrong with digital sound, and how good my two sources are now.  Yes I know there are other design concerns.  What I've heard wrong with so much digital before is cleaned up so much that it seems there has to be more to it than signal/noise ratios.  Bad sounding digital can have very low noise levels.  Could *something else* be the culprit?  Sure, maybe "jitter" is the catch-all scapegoat and there's some other pixie dust we're all missing.

Apparently it is possible to record a jittered sound, that would be a very good test.  I will report when I've had time to check out the Stereophile CD.

[TheChairGuy]

Steps to reduce jitter -

• Reduce vibrations to your component. Use a rack with spikes and place each shelf on spikes. Lots of shelves, vibration reducing components available.
• Use a power conditioner or make sure power is clean of spikes/noise etc. Balanced power conditioner makes quite a bit of difference
• A component with good power supply makes the best of the oscillator/clock in the device. 50% of the cost of the pricey Halcro amp goes into designing a power factor corrected power supply for it. After market power cords help improve dynamic demand of current of the device. These dont have to be pricey. Simple DIY designs make quite a bit of difference
• In the case of CD/DVD players, a clean CD with some anti-glare spray helps the laser track efficiently reducing too much servo tracking and in turn putting more demands on the power supply, which will affect the whole system and jitter.
• With computer based play back, it is much easier to isolate the audio output device than to use the builtin soundcard. Examples would be USB/Firewire/Network(like Squeezebox) sound cards/devices. It will be easier to get a power supply to these devices than to the whole computer.

My steps are quite different:

1.  Buy a decent sounding, $49 Coby AC/DC one-box CDP with a top loading mechanism (not a drawer - less mechanization and lower failure rates can likely be good things)

2.  Toss the lousy switch-mode power supply and replace with Radio Shack (regulated) linear one  ($30)

3.  Put 4 squishy/dead isolation puds (LAT International) on the bottom, and a 6 x 6" wedge of 0.125" EAR Isodamp on top to provide excellent vibrational control ($40)

4. Run thru $40 Isolation transformer so nasty digital has doesn't travel back to the other components

5.  Add a Dakiom Feedback Stabilizer to the outputs ($59)

6.  Run signal from 2-channel analog outs via Monster Cable of choice to preamp inputs.  Use volume control judiciously.

7.  Run preamp to tube amp(s) to drown out Redbook nasties with pleasing distortion (be it jitter or inadequate sampling or whatever  )

Viola! I can listen to CD's now.  In fact, it's decent now - as good as anything I have heard at shows or homes...and maybe $220 spent all in all 

John

[Daygloworange]

I wholly agree that # of votes does not equate to degree of correctness.  This is one of the most difficult lessons in our pursuit and one of the sociological phenomena referenced above.

I base my opinion on my experience, what I've heard, what is so commonly wrong with digital sound, and how good my two sources are now.  Yes I know there are other design concerns.  What I've heard wrong with so much digital before is cleaned up so much that it seems there has to be more to it than signal/noise ratios.  Bad sounding digital can have very low noise levels.  Could *something else* be the culprit?  Sure, maybe "jitter" is the catch-all scapegoat and there's some other pixie dust we're all missing.

I totally agree with Mike here. I'm not sure what makes the big difference in transports and DAC's.

At RMAF 2007 I heard a CDP used solely as a transport feeding into a DAC sound absolutely lifeless. It was during setup on Thusday night in the GR Research room. It was just being hooked up when I walked in. I had never heard Danny's system before. It was the dullest sound I've ever heard period. We thought something was wrong with the amps. We found another CDP, this one (IIRC) was a pro machine, with a stable platter system, hooked it using the digital outs to the DAC, and it was a whole different ballgame. Could I have picked out the difference in these 2 transports blind?? There's no doubt I could have with 100% reliability picked out the difference.

FWIW, the upgrades that have really astonished me, have been to the power supply section of my Bolder analog modded SB3. My setup is a minimalist one, modded SB3 feeding the analog outs directly into a pair of ATI 1202 stereo power amps with solid silver IC's. No preamp. I've done a number of A/B's with 4 different power supplies, and the results are clear to me. Blacker background, better resolution, better timbre, better imaging. The power supply simply makes it sound like a different player all together.

Mind you, the modded power supply alone costs more than some CDP's, but I have no reservations about my decision.

Since this, I've sent out the power supply for my studio digital recorder and a few other pieces to be gutted and upgraded.

Cheers

[AphileEarlyAdopter]

I kinda have to agree. Mind you, I do have a Panny 55 in my own modest HT  setup and i love it,  but it would not make the cut in my 2 channel rig. It next to my Red Wine Sig 30.2 ........ sorry, not even remotely close in the clean signal dept.

Right, it probably is the weakest one in the chain. But my system reflects any change upstream. So I am thinking the Panny must be transparent. Especially in biamp mode with a good aftermarket power cord it sounds fine to me (used with a C7-IEC adapter). My previous amp was a Audio Refinement Complete Int. Amp. I still get the same smoothness, but better bass and dynamics now (but midrange is a little recessed). Maybe with SPDIF the Panny runs with as good a clock as it is in the source. I dont know but it sounds very good. Just now before getting to bed I heard Dave Brubeck's Take Five track. The saxophone sounded so very palpable. Anybody in SF Bay Area feel free to send me an email and stop by anytime. 

[AphileEarlyAdopter]

Ethan,
Please read this article, if you have not done so already -
http://www.stereophile.com/features/368/

Here is the jitter specs of various players in the same article/series.
http://www.stereophile.com/features/368/index4.html

This is an old article so the players are not current. Also, Stereophile changed the measurement mechanism. So these numbers are not comparable. But all player reviews in Stereophile have jitter measurements. So you can do some comparisons probably in the last year or two. Also, note that the spectrum throws another wrench into the works. Two players, with same peak-to-peak jitter may sound different based on the spectrum. But the POINT is you want this peak-to-peak jitter as low as possible.
Jitter is a very common issue, without addressing this you do not have high-end sound.

[AphileEarlyAdopter]

ok, that means you can hear the effects of jitter. You are an audiophile with golden ears

Is that an insult? 

FWIW I have done a double blind test with transports and it indicated jitter is audible.

It was an SB3 vs a Monarchy CLD-M401 CD transport, both going via coax S/PDIF into an old, not particularly jitter busting DAC. The SB3 was chosen as better 4 times out 4, despite sighted impressions (short or long term) seeming to be the opposite.

The negative is that we only did 4 rounds, and there is 1 in 8 chance of getting a 4/4 (one way or the other) by pure chance anyway. The positive is that the differences sounded pretty clear blind and choosing didn't seem difficult or agonising. The coax cable, DAC and rest of system was identical, and the DAC had a switchable input so quick A/B was possible. The person doing the switching didn't know which transport corresponded to which switch, he didn't let me know with which input he started each round and I stated my preference as "the first one" or "the second one" based on his starting choice each round. I could switch back and forth as much as I liked until I made a decision. Other than the number of listens, it was a good double blind test.

I hope this report is of help to the discussion. The DAC used was venerable so it was a good jitter test.

It doesn't get us closer to what level of jitter is audible though. It doesn't really speak about the claims of modern DAC manufacturers that their DACs are in effect immune from jitter. I'm open-minded about that.
Darren

Hi Darren,
Come on, I like to be called an audiophile too
Seriously, no offence meant. It just that it is good to see people concur on this issue.
Somehow I did not like the stock Squeezebox sound. It did not have the edginess of a player, but it was 'soft-sounding' to my taste. I first added the modded ELPAC power supply and could not notice much change. Still I was not getting the clarity I wanted. So I sent it in for mods to boldercables.com, btw, this was the first time, I was getting anything modded. After the digital mods, I got the clarity I was looking for. But ultimately, I want to go with the Empirical Audio mods/design. I am hesitating because if I do that, i will have to go with a DAC and an analog power amp. This will cost quite a bit of money in my case (for a I2S DAC and an integrated amp ). It all depends on how our economy is going to be soon

[opaqueice]

One simple test to establish a threshold of jitter audibility is as follows (I apologize if someone suggested this earlier, I haven't read the whole thread):

start with pure tone (say a 1 kHz sine wave).  Now add side bands to that, i.e. take the original signal and add a tone at 1,100 Hz and a symmetric one at 900 Hz, with some small amplitude.  That's precisely what jitter with a spectrum sharply peaked at 100 Hz would do, and the relation between the amplitude of those sidebands and the peak-peak or rms amplitude of the jitter is simple to work out. 

Having done that you can just increase the amplitude of the sidebands (starting with them very low) until they become audible, and use the relation between jitter spectrum and sideband amplitude to convert that into a threshold for jitter audibility.  This whole experiment could be done in an hour using free software (Audacity for the tones, and foobar with the ABX plugin to find the threshold).

The test is limited in that it simulates only a very specific type of jitter - jitter with a spectrum concentrated at one frequency (100 Hz in the example I gave).  That's supposed to be easier to hear than the effects of than broad-spectrum jitter.  It's also highly data correlated (in a kind of trivial sense) - that should also make it easier to hear.  Picking a test-tone that's near the most sensitive part of the human hearing range will enhance audibility, and in general effects like this are much easier to hear with pure tones (because all the complex stuff going on in music tends to mask these things).  So there are three effects all going in the direction of making the jitter more audible, and so this test might significantly overestimate the audibility of normal jitter - but it still seems worth doing since it's so simple.

Let me note that professional audio researchers have investigated jitter audibility already, using more sophisticated methods - there's some literature on it, and on what the jitter audibility thresholds are.  But it's more fun to do your own experiments.

[Daygloworange]

The test is limited in that it simulates only a very specific type of jitter - jitter with a spectrum concentrated at one frequency (100 Hz in the example I gave).  That's supposed to be easier to hear than the effects of than broad-spectrum jitter.  It's also highly data correlated (in a kind of trivial sense) - that should also make it easier to hear.  Picking a test-tone that's near the most sensitive part of the human hearing range will enhance audibility, and in general effects like this are much easier to hear with pure tones (because all the complex stuff going on in music tends to mask these things).  So there are three effects all going in the direction of making the jitter more audible, and so this test might significantly overestimate the audibility of normal jitter - but it still seems worth doing since it's so simple.

I'll reserve final judgement till I hear jitter properly isolated. I don't doubt that it could have a significance. But having said that, I suspect it's effect is not severely profound and detrimental to audio playback. I'm looking forward to hearing any proper demonstration that correctly isolates jitter. A sine wave test would be cool, but hearing a music playback with varying levels of jitter would be very interesting.

There are shortcomings and tradeoffs to everything. I don't know why, but very often, digital IMO is overly scrutinized as compared to analog, which is well known to be plagued with high noise and non linearities.

Despite that, analog does have a signature, and a character that many have grown to like, myself included (for certain things).

The examples I give earlier about self noise and analog tape non linearities are quite pronounced and very easy to isolate. Anyone of those non linearities highlighted on it's own would be disconcerting, but because they are low in amplitude in relation to the audio, they mostly go unnoticed.

I've never heard an audiophile complain about the effects of aliasing, tape machine rumble, print through, crosstalk, etc....
I've never heard one audiophile ever mention hearing tape splices, which make a distinctive pop when they go over the playback head on a tape machine, or the pitch difference between the 2 tape sections spliced together from different areas of the reel.

Cheers

[BrianM]

DGO, there must be something about the nature of the old analog recording techniques that gives audiophiles that romantic feeling they're, at bottom, looking for in this hobby; and something about the nature of digital recording that makes them obsess about and pine for absolute perfection.  When we're watching "On the Waterfront" we don't care at all about amazingly perfected hi-def computer graphics special effects, and when we're watching Star Wars Episode III we suddenly do.  It's a great reminder of how our enjoyment level can't be always tethered to seeking perfection in all things.

[Daygloworange]

Brian,

There is no doubt about it. Analog is musical. Tape saturation is very, very cool. But it's also blurring. It also adda low mid emphasis and soften's high frequencies and compresses very nicely. A lot of large diaphragm condensor mic's and tube mics have accentuated top ends, and with the combination of accentuation of the mic, and the roll off of tape, it can create a nice soft shimmer.

It sounds nice, but it's non linear. It's kind of like sugar. It can be too much. Ask any studio engineer. But most engineers have lost the sweet tooth for it from overexposure, and only want it at certain times, not all time.

Half the charm from a cranked 100 watt Marshall guitar amp is from cone break up from the speakers, not just the tube compression and distortion. A properly distorted tube amp sings.

It's the same thing.

You couldn't pry my vintage Marshall's out of my cold dead hands. 

Cheers

[Joey B]

If the Stereophile test cd that has the jitter example on it also has the same track without jitter , why try and create a jitter test ? Just use those 2 examples .

Why complicate things needlessly ?

Joey B

[Ethan Winer]

Please read this article, if you have not done so already -
http://www.stereophile.com/features/368/

What I see there makes no sense. The two graphs that show jitter - Figures 2 and 3 - show the sidebands at very high levels around -60 and -50 dB respectively. If the jitter really were that high it would be clearly audible and highly objectionable. I cannot accept that any CD player or any other digital system could have such high levels of artifacts, even very early players from the 1980s. At -50 that would be worse than the hiss from a pre-recorded cassette, and we all know that CDs are infinitely quieter than any cassette.

The data in the Stereophile article is also wrong because the stated jitter levels for those graphs are listed as 145 picoseconds and 561 picoseconds. Contrast that with Figure 4.28 in Ken Pohlmann's book Principles of Digital Audio (4th ed.) where he shows the spectrum for 2 nanoseconds of jitter as being below -120 dB. Now, Pohlmann's example is 2 ns peak to peak, where Stereophile uses RMS. But even still, Pohlmann's jitter is four to 14 times higher than Stereophile's, yet the artifacts are 60 to 70 dB lower. So ether Pohlmann doesn't know how to measure jitter properly, or Stereopile doesn't know how to measure jitter properly. Guess who my money is on.

--Ethan

[Ethan Winer]

One simple test to establish a threshold of jitter audibility is as follows ... start with pure tone (say a 1 kHz sine wave).  Now add side bands to that

Yes, and this is exactly what I proposed about 137 pages back.

It is well known that jitter manifests as non-harmonic sidebands, so the real issue is how far below the music such artifacts must be before we can consider them insignificant. This is trivial to test, and I'm willing to go one better (in favor of jitter believers) than your proposal to add correlated side bands. I propose adding the nastiest sounding treble-weighted noise I can concoct at varying levels below music and low frequency test tones. This will make the artifacts as obvious as possible, and then we'll all know how soft is soft enough. If nobody can hear noise that sounds like gears grinding when it's 80-odd dB below the music, then we'll have disproved an entire category of beliefs such as the audibility of capacitor distortion, jitter, dither and ultra-high bit depths, and more.

--Ethan

[miklorsmith]

We're saying different things, Ethan.  I don't disagree about your idea on listening tests though.  I simply don't see the point in simulating what jitter theoretically should sound like to decide absolutely where the audibility threshold is.

Rather, we should focus on jittered tracks themselves with actual jitter levels.  This seems like a sensible way to go about it.  Your proposed methodology leaves open the possibility that other audible, jitter-related problems exist which were not part of the testing protocol.

On the ubiquitous audiophile compliment "transparency", I'm sure growing to hate that word.  It's just about completely meaningless.

[opaqueice]

One simple test to establish a threshold of jitter audibility is as follows ... start with pure tone (say a 1 kHz sine wave).  Now add side bands to that

Yes, and this is exactly what I proposed about 137 pages back.

 

It is well known that jitter manifests as non-harmonic sidebands, so the real issue is how far below the music such artifacts must be before we can consider them insignificant. This is trivial to test, and I'm willing to go one better (in favor of jitter believers) than your proposal to add correlated side bands. I propose adding the nastiest sounding treble-weighted noise I can concoct at varying levels below music and low frequency test tones. This will make the artifacts as obvious as possible, and then we'll all know how soft is soft enough. If nobody can hear noise that sounds like gears grinding when it's 80-odd dB below the music, then we'll have disproved an entire category of beliefs such as the audibility of capacitor distortion, jitter, dither and ultra-high bit depths, and more.

Well, I don't think most hard core "audiophiles" are going to accept this, no matter what.  But for the rest of us it will be interesting.  As for adding nasty noise, it's nice to have a clean and precise relation between the sidebands/noise and the jitter spectrum and amplitude that would have produced it.  If you can find a way to do that, it sounds like a good idea.  But I suspect simple anharmonic sidebands like I suggested are going to sound pretty nasty as it is.  Another option is to increase the frequency of everything - I forget where human hearing is the most acute, but wherever that is, that's where to put one of the sidebands and/or the center frequency.

[miklorsmith]

So, your opinion is jitter manifests solely as noise artifacts without any effect on primary signal?  Is this scientifically accepted?  I don't know the answer.