What’s the difference between jitter and noise?

Posted by: Consciousmess on 04 April 2018

As I’ve recently realised I’ve been suppressing what my system does by not using ‘good enough’ interconnects, it’s led me to gradually making each of the other component connections far better - at a comparably cheap and affordable cost considering the electronics.

My next change will be the digital interconnect, which presently is a Clearaudio £180 one. Satisfied yes, but now I know this must mask what the nDAC outputs, including adding jitter and noise.  Because it adds jitter and noise, the rest of the chain must progressively increase this and amplify. Firstly, am I correct with this hypothesis?

Secondly, I am certain a commensurate digital interconnect (i.e. to nDAC + 555PS DR + 552 DR) will improve the sound tremendously, but I’m hazy as to whether this digital cable will let more signal through and restrict the jitter, or let more noise and jitter through? Perhaps highlighting flaws in the source??

Please advise!

Posted on: 04 April 2018 by Huge

To answer the question in the title...

Both are unwanted variances of the signal.

Jitter is variance in time.
Noise is variance in amplitude.

Posted on: 04 April 2018 by Ardbeg10y

Is there a way to detect Jitter or Noise?

I'm a happy listener to my various setups, and have no idea how to detect this.

Posted on: 04 April 2018 by Consciousmess

Aren’t these the two things that make a signal sound worse? If it is high definition I point out.

A cable should minimise them and make the sound far better. A cheap DAC becoming an expensive DAC just because of this.

Posted on: 04 April 2018 by TOBYJUG

Wasn't sure if a cable could make jitter and noise worse, but then remembered an article about Shunyata Research cables.

http://shunyata.com/2017/06/13/medical-applications/ 

 Not sure how this would manifest in listening and what artefacts are noticed.    NDac has a buffer thats meant to minimise jitter.

Posted on: 04 April 2018 by ChrisSU
Ardbeg10y posted:

Is there a way to detect Jitter or Noise?

I'm a happy listener to my various setups, and have no idea how to detect this.

If ever there was a case of 'ignorance is bliss' surely this is it? Pour yourself another dram and enjoy the tunes!!!

Posted on: 04 April 2018 by Simon-in-Suffolk

To the OP, jitter is noise - or specifically a type of noise. There are many sorts of noise - which typically describes unwanted variance to the signal - and with jitter there is noise jitter - random variations, and periodic jitter - periodic variation such as mains hum modulating a clock

The interesting thing with jitter is that is refers to the noise with clocks waveforms - that is it is the variance of the  clocking of a  data stream. This adds, as Huge said above, to the noise in the time domain. Now if this clocked stream contains timed samples that represent the frequencies that make up a continuous signal - as in digital audio -  this noise passes from the time domain to the frequency domain when the sampled data is reconstructed into an analogue signal. The interesting thing is that the power of noise between domains remains constant so - the jitter time domain noise energy  gets exactly converted to frequency domain noise energy.

Now what does jitter sound like?  -  With increasing jitter the reconstructed clocked sample stream  will  sound more and more grainy and unnatural. In nature we don't really experience time domain noise - so to our hearing increasing time domain jitter makes the original clocked reconstructed sample data sound more and more unnatural and grainy.

As far as your cables are concerned - they won't affect the jitter from the clock - but  they  can cause phase variation of the clocked signal (jitter) through slewing, reflections and other physical interactions of the cable, connector and clocked signal. Now most DACs these days don't rely on the clock signal accuracy to reconstruct the sampled data - they use their own clocks instead. However noise in the clocked signal can couple into the receivers electronics and cause system coupled noise of its own. So you will hear in many systems differences between different SPDIF cables.

FWIW I believe this is the same affect we notice with ethernet patch leads and different ethernet switches - it its the receivers system coupling of the noise in the clocked ethernet signal - in this case the the physical ethernet clock (which is not connected with audio sample clocks)

 

 

Posted on: 04 April 2018 by Consciousmess

Thanks Simon in Suffolk, very educating  and knowledgeable!

Posted on: 05 April 2018 by Rich 1

Gosh Simon, your concise explanation should be in one of the Hi Fi journals. An easy to understand explanation for the layman. Rich 

Posted on: 05 April 2018 by Ardbeg10y
ChrisSU posted:
Ardbeg10y posted:

Is there a way to detect Jitter or Noise?

I'm a happy listener to my various setups, and have no idea how to detect this.

If ever there was a case of 'ignorance is bliss' surely this is it? Pour yourself another dram and enjoy the tunes!!!

My wife had different plans.

Posted on: 05 April 2018 by Innocent Bystander

Some noises give me the jitters...

Posted on: 05 April 2018 by Dave***t
Simon-in-Suffolk posted:

 As far as your cables are concerned - they won't affect the jitter from the clock - but  they  can cause phase variation of the clocked signal (jitter) through slewing, reflections and other physical interactions of the cable, connector and clocked signal. Now most DACs these days don't rely on the clock signal accuracy to reconstruct the sampled data - they use their own clocks instead. However noise in the clocked signal can couple into the receivers electronics and cause system coupled noise of its own. So you will hear in many systems differences between different SPDIF cables.

Simon, is the coupling you mention the same kind of thing which is absent with optical connections (and maybe ameliorated by galvanic isolation)?

If so, does this not suggest, in concert with the fact that that DACs use their own clocks, that optical connections should be best? I ask because I'd read that the downside of optical connections was increased jitter - this was one of the particular benefits of Hugo etc, was it not, to be largely immune to optical jitter?