Jump to content

Directional Digital Cable


Recommended Posts

35 minutes ago, AussieMick said:

but I haven’t experimented with digital cable.

What audio signal would you be sending through the coaxial cable e.g.  44.1kHz 16 bit stereo from a CD transport, 96kHz 24 bit stereo from the internet, 5.1 channel Dolby surround from a DVD player?

 

Most importantly, can you tell us what device you'd be using to receive the digital stream?  [If it's an advanced DAC, it should thoroughly "recondition" the signal.]

Edited by MLXXX
Link to comment
Share on other sites



39 minutes ago, Darren69 said:

Sometimes the crap is consumer driven. Permeate free milk anyone?

Yes I can't help feeling that some audiophiles will never be satisfied that jitter through a SPDIF cable can be dealt with, or that streamed audio can be received bit-perfect, buffered, and then reclocked with a rock steady clock. And there are other audiophile preoccupations over extremely minor differences.

 

Meanwhile we routinely watch movies and U.S. sitcoms on our 50Hz television system in Australia with the sound running 4% fast, and yet most viewers seem to be completely unaware of the anomaly. (I notice the anomaly more these days with so much of my viewing being at correct speed via Netflix or from Blu-ray discs. The "chipmunk" edge to dialogue sped up by 4% annoys me.)

  • Like 1
Link to comment
Share on other sites

4 hours ago, AussieMick said:

Of course, if the chip has to “work”, it produces more noise which can be injected right into system.

Not really. The buffering and reclocking algorithms remain active even if the incoming stream is pristine. 

 

As for noise, if you look at the specifications for standalone DACs on the market you'll find the analogue audio that emerges has a signal to noise ratio that is more than sufficient for the program material being carried by the digital stream.  I note that 16 bit CD sound has an inherent noise from using only 16 bits. However you'd only hear that at a very high gain setting.

 

As for 24 bit sound, that precision doesn't really exist for the least significant bits, as room noise and other noises in the recording process necessarily occupy the least significant bits.

Edited by MLXXX
Link to comment
Share on other sites



9 hours ago, AussieMick said:

Of course, if the chip has to “work”, it produces more noise which can be injected right into system.

Current draw you are quite correct is different for certain integrated circuits vs others, the

cheap fix usually seen, is a resistor to provide a modicum of separation between devices.

The better approach would be to supply from separate high impedance current regulated

supplies, to reduce adjacent noise between devices. 

 

Voltage regulators too factor into this, typified by the cheap 78xx and 79xx types,

The problem with these sharing circuits doing different functions,  is their low

impedance capability,  which even without a device being connected,  loads a circuit

constantly at 10ma of current between their output and ground. 

 

Seen as an attribute in lesser circuits  ,to in some cases relieve circuits of output capacitance, their widespread use in digital circuitry, is indicative of simple price cutting.      

Edited by stereo coffee
Link to comment
Share on other sites

  • 1 month later...
Not really. The buffering and reclocking algorithms remain active even if the incoming stream is pristine.    As for noise, if you look at the specifications for standalone DACs on the market you'll find the analogue audio that emerges has a signal to noise ratio that is more than sufficient for the program material being carried by the digital stream.  I note that 16 bit CD sound has an inherent noise from using only 16 bits. However you'd only hear that at a very high gain setting.   As for 24 bit sound, that precision doesn't really exist for the least significant bits, as room noise and other noises in the recording process necessarily occupy the least significant bits.     

 

 

Here is PS audios engineer talking about jitter and clocks.My opinion is use specced cables, doesn't need to be expensive. 

 

  • Like 1
Link to comment
Share on other sites

On 11/07/2019 at 10:21 PM, AussieMick said:

Anecdotally:    There are fields of science where the signals they are working with are a lot (!!!!) more sensitive than what is put through a digital audio cable.... and where there is a lot more money involved... and where the risk of getting it wrong is a lot greater.

 

When we look at those fields... we see that do not have "directional cables"..... if the direction was important for some reason, then they would know, and use it to their advantage.

 

Empirically:    When we look at science/physics, etc..... there is no reason to expect that a well made and designed cable would be directional.   Any of the potential explanations for why it could happen are too "small" to have an a real effect (and we can measure things many many times smaller than can practically affect the digital signal).

 

 

That is not to say that you will never hear a (real) difference between digital cables.   Only that when you do, that the reasons often provided are not the real reasons.

 

 

Paradoxically.... one of the things which leads to a good quality high frequency cable, is highly toleranced manufacturing processes.....  and those with such manufacturing processes are making a lot of cable .... which in turn drives down the price of the product.     That doesn't mean expensive cables won't be good. ...... but it doesn't mean that sensibly priced cables will necessarily be inferior.

Link to comment
Share on other sites

On 12/07/2019 at 8:53 AM, scumbag said:

 I was not involved in any debate or discussion about cable directivity.

Sure you were....  Your contribution was (allow me to paraphrase)  "I tend to avoid discussions about (for example) cable directivity"

Quote

name calling and brow beating

I tend to put people like this on ignore.....  it's almost never about .... 

 

Quote

I guess that's how you guys like it.

....  "ad hominem".

 

I just like to get my point across....  like you're doing right now.   See?!   Everyone's happy :) 
Live and let live.... scroll on.

 

 

PS:  Don't put us (or anyone) in "camps".   History shows that things don't work out well once people get put in camps.

Edited by davewantsmoore
  • Like 1
Link to comment
Share on other sites



You just have to love these 'discussions' about cable directions and so forth - it keeps life interesting - it seems that some people can hear differences and some can't - that's pretty simple, eh!  

If, to you, they sound better in one direction than the other, just use them this way - if they don't change with direction, this is just one less decision you need to make about your system. 

 

I did try this and I thought that changing direction on a well 'run-in' cable did seem to make a difference for the first couple of hours but after that, the sound seemed no different - as we all know, the idea of self-delusion/confusion/remnance/system reboot/etc is common and could quite easily create this 'apparent effect'

 

  I don't know why anyone would get so 'heated' and argumentative over such a thing.

 

Now, a seriously aggravating subject to get stuck into is the notion/idea of cable 'break-in' - this must surely create enough 'hot air' to keep us warm for the rest of winter!

 I do find that 'breaking -in' cables is an extremely irritating fact of life yet many people find this a total waste of space even talking about it and others just like having a good old 'ear-bash' about it. 

And yes, you can get a more critical result with good headphones/amp in this regard, just to add my view on this, if there's any doubt!

 

Curiously, I sent off to the Hagermann people for one of their accelerated break-in' units (not cheap either) and was really disappointed that it didn't work at all on the cables that I tried it on  - aah, such is life!

Link to comment
Share on other sites

Some cable companies have a theory that the drawn copper changes sound.

Who cares, just connect the arrows and sleep well.

McGowan himself says he is sceptical.

Audioquest has a demo visiting room where you self can test the difference..

I don't bother about arrows it's worth to note that some Digital cables of the one meantioned had shielding and Arc that only works correct one way.

Link to comment
Share on other sites

On 04/09/2019 at 2:51 AM, Dankeshon said:

Here is PS audios engineer talking about jitter and clocks.My opinion is use specced cables, doesn't need to be expensive. 

I've listened to such YouTube videos by Ted Smith before. He appears to know his subject matter very well, and he explains things clearly and concisely. He appears to have a genuine belief that he can hear jitter at the level of 0.2 picoseconds.  He doesn't however mention use of any objective verification process for his personal listening tests. (Nor does he mention setting up objective listening tests for others.) He appears not to have carried out an objective verification of his listening ability for jitter, unless he reports having done that in some other video, or in some article or blog on the net. 

 

So at the end of the day we have someone who is very articulate and knowledgeable as regards relevant technology, sharing his subjective impressions as to what he has perceived.  He has as much credibility in relation to sharing his subjective impressions as you or I, or anyone else on this forum. We can believe that he is sincere, but we cannot know whether the differences he thought he heard in various situations were actually because of jitter at a low or infinitesimal level, or not.    

Link to comment
Share on other sites

22 hours ago, HdB said:

If, to you, they sound better in one direction than the other, just use them this way - if they don't change with direction, this is just one less decision you need to make about your system. 

Well of course.   (Duh!) .... I think any monkey can figure that out.

 

I assume that when people want to talk about it .... they want to talk about "why".

 

How could this happen?

Is it something to do with the device? the cable? my hearing?

Why do I hear it?

Why are some cables so expensive?

 

The answers/discussion of those things are not always quite as simple or straight forwards as "you're imagining things"...  and "just pick your poison and be happy" addresses none of the above things.

 

Link to comment
Share on other sites

1 hour ago, MLXXX said:

He appears to have a genuine belief that he can hear jitter at the level of 0.2 picoseconds.

It isn't as implausible as it sounds once we realise that 0.2ps .... 1ps....  500ps..... 1ns .... are all dramatic over implications of what is actually happening, and can't just be compared to each other.

 

"How much 'ps' jitter can you hear" is a completely flawed question.

 

1 hour ago, MLXXX said:

He has as much credibility in relation to sharing his subjective impressions as you or I, or anyone else on this forum.

I think you are potentially quite mistaken about this.    You are correct the is not necessarily more subjectively qualified..... however when he sets up  tests .... he has much more ability than you do to create specific, known, differing levels of jitter - using a decent test setup (ie. automated, and instant switching ... variable controlled) to compare the differences.

 

Yet, when you sit at home and compare A vs B, what are you actually comparing (do you have any real idea?) .... and how are you comparing it  (I would guess not in a way where you are able to instant switch between the two examples, or in a way which investigates and controls all other variables).

 

 

The arguments against subjective testing are not simply (or only) that "human hearing and perception is not reliable" .....  but it is that when testing is approached without any objectivity or rigour.... there are often massive uncontrolled variables.... even down to an uncertainty in what you are actually comparing.

 

It isn't to say that we can't listen and report the difference ..... It's that if we do that, but we also don't know if what we were listening to was actually different, or by how much ... or whether there were other uncontrolled variables affecting the result ..... then we can't reliably say very much about our subjective result due to those uncertainties.

 

 

 

 

1 hour ago, MLXXX said:

we cannot know whether the differences he thought he heard in various situations were actually because of jitter at a low or infinitesimal level, or not.    

However if we take the (very safe) assumption that his tests were designed and executed orders of magnitudes better(understatement) than yours or mine ...... then, that does improve the validity.

ie. he is more sure (than us) that he really tested X vs Y jitter ... and that what he was listening to was free from uncontrolled variables, and that the test was executed well (eg. instant switch, etc. etc.)

Link to comment
Share on other sites



2 hours ago, davewantsmoore said:

ie. he is more sure (than us) that he really tested X vs Y jitter ... and that what he was listening to was free from uncontrolled variables, and that the test was executed well (eg. instant switch, etc. etc.)

That wasn't apparent to me in the three videos.  I can recall no mention of instant switching etc.  It was more a case of "with this setup I heard jitter", and  "with this other setup I heard less jitter".

 

2 hours ago, davewantsmoore said:

"How much 'ps' jitter can you hear" is a completely flawed question.

I was merely quoting a figure the speaker himself mentioned of 0.2 picoseconds. 

 

Whatever kind of jitter is involved (be it rapid, slow, repetitive, or random) a deviation of 0.2pS is tiny as a percentage of the time between digital samples of PCM audio.   For example if we think of  a [non-oversampled] 96kHz sample rate, each sample comes every 0.00001041666 seconds, or 10.416 microseconds, or 10,416,666 picoseconds.   So we are looking at at a ratio of 50 million to 1.   If it is  44.1kHz sampling, a sample is required every 0.00000226757 seconds, or 2.267 microseconds. Relative to 0.2 picoseconds that is a ratio of 113 million to 1.

 

In dry air at 20 degrees Celsius, sound travels at about 343 metres per second. In 0.2 trillionth of a second, sound will travel  0.0000686 micrometres.  The diameter of a human hair ranges from about 17 to 181 micrometres.  The wavelength of a 20kHz sound wave is about  1.716cm, or 1,716mm, or 1,716,000 micrometres. So 0.2 pico seconds would involve an extremely small portion of even a high audio frequency wave. 

 

To be frank, the claim of being able to hear jitter at the amplitude of 0.2pS (whether it be rapid, slow, repetitive, random, or a mixture of these) strains credulity.  It's the sort of extreme claim one would want to see backed up with an objective test of the claimant's hearing ability.

 

I have set this video to start where the claim is made about 0.2 picosecond jitter. It appears that the speaker used the manufacturer's spec sheet as the way to ascertain that the jitter was at that infinitesimal level:-

 

 

 

Edited by MLXXX
Link to comment
Share on other sites

I was merely quoting a figure the speaker himself mentioned of 0.2 picoseconds. 
 
Whatever kind of jitter is involved (be it rapid, slow, repetitive, or random) a deviation of 0.2pS is tiny as a percentage of the time between digital samples of PCM audio.   For example if we think of  a [non-oversampled] 96kHz sample rate, each sample comes every 0.00001041666 seconds, or 10.416 microseconds, or 10,416,666 picoseconds.   So we are looking at at a ratio of 50 million to 1.   If it is  44.1kHz sampling, a sample is required every 0.00000226757 seconds, or 2.267 microseconds. Relative to 0.2 picoseconds that is a ratio of 113 million to 1.
 
In dry air at 20 degrees Celsius, sound travels at about 343 metres per second. In 0.2 trillionth of a second, sound will travel  0.0000686 micrometres.  The diameter of a human hair ranges from about 17 to 181 micrometres.  The wavelength of a 20kHz sound wave is about  1.716cm, or 1,716mm, or 1,716,000 micrometres. So 0.2 pico seconds would involve an extremely small portion of even a high audio frequency wave. 
 
To be frank, the claim of being able to hear jitter at the amplitude of 0.2pS (whether it be rapid, slow, repetitive, random, or a mixture of these) strains credulity.  It's the sort of extreme claim one would want to see backed up with an objective test of the claimant's hearing ability.
 
I have set this video to start where the claim is made about 0.2 picosecond jitter. It appears that the speaker used the manufacturer's spec sheet as the way to ascertain that the jitter was at that infinitesimal level:-
 
 
 
His work is jitter, and he has meeters to check this easily.
There is another link from a professor I believe further up.

Ted is more practical and explain it easy in lamens terms the one above from the university of Essex where his student published in Stereophile has more information.

The problem with sceptics is that they believe jitter is not a problem.
I guess you need to do more Abx testing.. [emoji38]
Link to comment
Share on other sites

1 hour ago, Dankeshon said:

His work is jitter, and he has meeters to check this easily.

If you listen to the part of the video I quoted he makes no reference to measuring the jitter himself.

 

The fact that 12 months later the manufacturer lengthened the time constant of the PLL suggests to me that with certain types of input the phase-locked loop may have been hunting or overshooting.

 

1 hour ago, Dankeshon said:

The problem with sceptics is that they believe jitter is not a problem.
I guess you need to do more Abx testing.. emoji38.png

It doesn't appear that Ted did any ABX testing whatsoever.  He simply reported what he thought he heard.

 

As I've said, anyone on this forum can do that.  

 

1 hour ago, Dankeshon said:

Ted is more practical and explain it easy in lamens terms

Yes, he explains things very clearly. He seems to have an in-depth knowledge and the ability to explain technical matters well.

Edited by MLXXX
Link to comment
Share on other sites

4 hours ago, MLXXX said:

It was more a case of "with this setup I heard jitter", and  "with this other setup I heard less jitter".

In that case, then he's unlikely to be "testing properly".

 

Like we know a lot about how to measure an electrical signal with a volt-meter .... we also know a lot about how to listen to audio for differences.

 

Quote

I was merely quoting a figure the speaker himself mentioned of 0.2 picoseconds. 

Yes, sure... I understand that.

Quote

Whatever kind of jitter is involved (be it rapid, slow, repetitive, or random) a deviation of 0.2pS is tiny as a percentage of the time between digital samples of PCM audio.

Yes, it is ... but don't think because it is small, that it is less relevant.... that's no how jitter works ;) 

 

Quote

In dry air at 20 degrees Celsius, sound travels at about 343 metres per second. In 0.2 trillionth of a second, sound will travel  0.0000686 micrometres.  The diameter of a human hair ranges from about 17 to 181 micrometres.  The wavelength of a 20kHz sound wave is about  1.716cm, or 1,716mm, or 1,716,000 micrometres. So 0.2 pico seconds would involve an extremely small portion of even a high audio frequency wave. 

This is not the right way to think about jitter..... what you have written here tells me that you do not understand what is going on.

 

It's why in the other thread I told you to stop "mixing them up" ... ie. stop comparing sampling jitter to other forms of "time distortion" that was being discussed.    It is not the same effect....  and nothing to do with how "much of a portion of the high frequency wave is displaced"   (as you noted it is infinitesimally small .... millions of times smaller than a time distortion of the waveform that we could hear in our timing mechanism).

 

 

What is happening (in short, at the risk of over simplifying) is as individual samples are repeatedly* displaced in time, relative their neighbouring samples, this creates intermodulation distortion components in the output.    Not time distortion like you said ..... but harmonic (and in-harmonic) distortion.    ie. a raised and dirty noise floor.      The repetition (See: * above) conspires to produce further worsening and intermodulation  (which is where the frequency spectrum of jitter, ie. its periodicity, becomes important).

 

 

Quote

To be frank, the claim of being able to hear jitter at the amplitude of 0.2pS (whether it be rapid, slow, repetitive, random, or a mixture of these) strains credulity.

Given what you have said above....  I can appreciate why you would say that  ;) 

 

 

EDIT:   Although to be very clear.... I am not saying it IS.    I think the audibility of these things are often overstated..... my only beef here is the understanding of "how it works", as the industry often gets it wrong.

 

Edited by davewantsmoore
Link to comment
Share on other sites

9 minutes ago, davewantsmoore said:

What is happening (in short, at the risk of over simplifying) is as individual samples are repeatedly* displaced in time, relative their neighbouring samples, this creates intermodulation distortion components in the output.    Not time distortion like you said ..... but harmonic (and in-harmonic) distortion.    ie. a raised and dirty noise floor.      The repetition (See: * above) conspires to produce further worsening and intermodulation  (which is where the frequency spectrum of jitter, ie. its periodicity, becomes important).

The IMD components would be extremely small, and with 0.2pS jitter, quite possibly immeasurable directly as IMD.  What test equipment would you suggest to use to measure directly the IMD produced by 0.2pS of jitter?  It should be possible to do a theoretical calculation of the vanishingly small IMD involved.   

 

What you have written in the paragraph I've quoted above could apply, at least broadly, with very large amounts of actually audible jitter.

 

An example would be a decades old  receiver plugged into a CD player using an optical fibre cable,  where the PLL circuitry in the receiver has gone out of adjustment (e.g. through severe component value drift) making the PLL's locking unnecessarily volatile. This is  resulting in continual clocking variations falling short of an actual loss of sync or a bit reading error.  In this example, jitter timing displacements over a long period would tend to average out to zero. However to the extent they didn't over a particular interval of time they would create a net time distortion forwards or backwards. 

 

This is generally the case with jitter. It will tend to net out to zero over time, but in the short term there will generally be very minor timing and phase changes advancing or retarding the waveforms, or parts of waveforms, compared with rock-solid perfect timing. 

Link to comment
Share on other sites



2 hours ago, MLXXX said:

The IMD components would be extremely small

That depends.   Typically, yes.

 

2 hours ago, MLXXX said:

It should be possible to do a theoretical calculation of the vanishingly small IMD involved.   

No.  There are infinite possibilities in how the periodicity could shift.

 

2 hours ago, MLXXX said:

In this example, jitter timing displacements over a long period would tend to average out to zero. However to the extent they didn't over a particular interval of time they would create a net time distortion forwards or backwards. 

This is generally the case with jitter. It will tend to net out to zero over time, but in the short term there will generally be very minor timing and phase changes advancing or retarding the waveforms, or parts of waveforms, compared with rock-solid perfect timing. 

I am not sure why I am still surprised ....

Link to comment
Share on other sites

1 minute ago, davewantsmoore said:

I am not sure why I am still surprised ....

I have no idea why you should be surprised at all at what I've written.  Jitter does result in timing changes as well as phase changes.  And 0.2pS of jitter would result in challenging to measure spurious products.

 

Anyway...

 

 

Could you please answer this question I raised:-

 

2 hours ago, MLXXX said:

The IMD components would be extremely small, and with 0.2pS jitter, quite possibly immeasurable directly as IMD.  What test equipment would you suggest to use to measure directly the IMD produced by 0.2pS of jitter?

 

Link to comment
Share on other sites

I'm happy that my YouTube links was so popular.

 

But really?? Commenting about an Dac electro engineer that hear 0,2ps of jitter???

 

Who cares, he says he can hear it.. That's a statement

 

Why don't the null-tester Ethan Winer make a Dac??

 

"-Of course everything sounds the same so a 10$ Dac with a freebee clock does the job.."

 

-Jitter is unhearable.

 

But here they are not supported by science...

 

They are supported by the -everything sounds the same clan.

 

 

Link to comment
Share on other sites

11 hours ago, Dankeshon said:

-Jitter is unhearable.

 

But here they are not supported by science...

Gross amounts of jitter (not typically encountered in well designed equipment) can indeed be audible.  In a particular case (a particular example of jitter, and a particular panel of listeners) that could be confirmed with a scientific listening test. On the other hand, very small amounts of jitter typically wouldn't be audible. In a particular case, that could also be  confirmed with a scientific listening test.

 

I don't think anyone suggests that jitter is never audible. The usual position taken is that jitter would need to be of a nature and size sufficient for it to have a noticeable impact for the listener.

 

Exactly where the threshold lies would vary with the individual.  A scientific way to investigate that would to allow a listener to hear two versions of a passage of music, one with a certain type and amplitude of jitter and one with much a much lower amplitude of the same type of jitter. If it turned out that the listener couldn't distinguish between the two versions, then the jitter would appear to have been below that individual's threshold of perception, at least for the particular passage of music played to them.

Link to comment
Share on other sites

15 hours ago, MLXXX said:

Jitter does result in timing changes

Which are millions of times below the threshold of audibility.

 

15 hours ago, MLXXX said:

And 0.2pS of jitter would result in challenging to measure spurious products.

Measurement is simple (so that is no a challenge) .....  but how large they are, or what what frequencies they congregate ... that depends.

 

15 hours ago, MLXXX said:

Could you please answer this question I raised

I thought it was a typo or some sort of trick question....   If you need to ask this question.... then, that should be telling for you to reconsider how sure you are about other things  ;)

 

https://en.wikipedia.org/wiki/Audio_analyzer

 

Link to comment
Share on other sites



  • Recently Browsing   0 members

    • No registered users viewing this page.




×
×
  • Create New...
To Top