RFI and optical cables

Coax is generally preferred over optical because of better jitter performance (the more complex optical interface is more prone to jitter). For a good discussion of jitter in audio see this article  http://www.positive-feedback.com/Issue43/jitter.htm. Optical is adequate for TV to a dac, but for bluray I have optical to a Bravo 96kHz digital interface which reclocks to reduce jitter and then a coaxial output to the dac. This arrangement is better sounding than optical straight to the dac. 

Mike-B posted:

Bhliain nua sasta SJB,   

Agus leat fein. 

Go mbeirimid go leir beo ar an am seo arís.

SJB

johnG posted:

Coax is generally preferred over optical because of better jitter performance (the more complex optical interface is more prone to jitter). For a good discussion of jitter in audio see this article  http://www.positive-feedback.com/Issue43/jitter.htm. Optical is adequate for TV to a dac, but for bluray I have optical to a Bravo 96kHz digital interface which reclocks to reduce jitter and then a coaxial output to the dac. This arrangement is better sounding than optical straight to the dac. 

I understood the Naim DAC recloccked using a RAM buffer or some such to eliminate incoming  jitter?

SJB 

The issue with cheaper optical cable and transceivers is that they ate bandwidth limited.. This can effect the changing of the transport wave form timing with respect to the shape of string of data values (Manchester encoding is used) and physical vibration of the fibre cable. Now although transport timing doesn't these days affect sample timing or jitter, the process of receiving and decoding this data with subtle timing variations causes digital cross talk. This digital cross talk can affect coupled digital clocks and audio circuitry.

Electrical transmission has a more stable transport frequency.. But as you say can conduct common mode RF currents.  In the end you choose your poison, and many choose electrical transmission with good quality send and receive electronics.

Simon

Sloop John B posted:

I understood the Naim DAC recloccked using a RAM buffer or some such to eliminate incoming  jitter?

SJB 

Yes it does but I guess it can never be completely eliminated, and as Simon points out the timing variations will have an effect via digital crosstalk.

An optical cable has one advantage - there is no galvanic connection between two components. Thus any chance of inducing hum (e.g. from power supply) between the two is gone.

Also the opto-electric converters are not always of best quality and introduce noise again at the receiving end.

Optical has never sounded better than a good coax SPDIF in my case.

Adam Zielinski posted:

An optical cable has one advantage - there is no galvanic connection between two components. Thus any chance of inducing hum (e.g. from power supply) between the two is gone.

True, but all quality electrical SPDIF units I have seen incorporate galvanic isolation. Most if not all Naim digital lines are galvanically isolated.

Simon

Simon-in-Suffolk posted:

The issue with cheaper optical cable and transceivers is that they ate bandwidth limited.. This can effect the changing of the transport wave form timing with respect to the shape of string of data values (Manchester encoding is used) and physical vibration of the fibre cable. Now although transport timing doesn't these days affect sample timing or jitter, the process of receiving and decoding this data with subtle timing variations causes digital cross talk. This digital cross talk can affect coupled digital clocks and audio circuitry.

Electrical transmission has a more stable transport frequency.. But as you say can conduct common mode RF currents.  In the end you choose your poison, and many choose electrical transmission with good quality send and receive electronics.

Simon

 Can it  be inferred by the above post that better made optical cables will not have the limitation quoted above for cheaper optical cables and transducers ?

SJB

Not really, but...

If you read NDS's description it actually says that they use an optical internal link to reduce system noise. 

So much for science and theories ))

Optical cables aren't as good with Gaelic. Something to do with the accents not being aligned with Salisbury.

Tá mé ag magadh faoi na cáblaí

In Polish they get even more clumsy 

Sloop John B posted:

 Can it  be inferred by the above post that better made optical cables will not have the limitation quoted above for cheaper optical cables and transducers ?

I guess better cables will have less limitations and possibly could effectively remove them.. I have yet to use optical with a very high quality transport source..with more modest sources I find optical totally transparent (!)

Simon

Hi guys if you go over to head fi,under chord dave and read the last few posts,i think you'll find it rather interesting. 

Cannot find the thread....

Sorry ,its in head-fi under CHORD ELECTRONICS DAVE 

Not just the cables affecting the transfer rate but the Toslink Audio specification itself. Old Toslink Audio modules are limited to 3Mbps and since this is the standard for audio so far, there hasn't been a huge move to move away to better units suitable for other data transmissions. After all, if one end is Toslink Audio and the other is Toslink Data, they are not going to talk to each other. As a manufacturer you would need a good reason to use a standard that might cause problems.

Much multi channel audio is moving towards a combination of HDMI with serperate coax cable to carry the clocking signal. For stereo, coax itself should be sufficient for all hi-res streams.

I use Toslink optical between my source and my DAC and it works fine up to 96/24.  Not sure I buy all this noise/jitter problem stuff as most of it is just assertion by the protagonist.  But at this point I have to come out as a BAB (bits are bits) fundamentalist.