[Don_S]

Ted,
Boot me if I am in the wrong place or point me in the right direction if my search failed me. I could not find what I needed.

Circle members, On purely technical specs how would you rank the three for performance (first, second, third) as inputs to a DAC? Please don't confuse the issue with toslink or USB.  Those are not options I have available. I am going to have two digital outputs added to a music server that currently only has analog outputs. I get one from column "A" (AES/EBU) and one from column "B" (S/PDIF i.e. BNC or RCA).

I own AES/EBU and RCA digital cables but no BNC digital cables.  I could get a BNC cable if that is voted best. Right now my concern is which S/PDIF is better to have added and should I even worry about which S/PDIF is better if I have AES/EBU out. I do not have a DAC yet and have not done my homework.  How common are DACs with BNC input?  A few years very few companies (Wadia) used it. 

Thanks for any help.

[ted_b]

Why not USB??

[Speedskater]

For a rather short cable, let's say about 15 feet, I would rate them no difference.

[jarcher]

There are theoretically better mediums, and then's there's the reality of how they are implemented in the device(s) you are using......

E.g optical - whether toslink or ST - should theoretically be the best, but often in implementation they aren't, whether it's on the source-side (the media server) or the DAC side. 

My vote based on my limited experience to date would be a good USB board on the media server feeding a good asynchronous USB DAC on the other side preferably using a USB cable that's only passing along the signal & not the power from the media server.  DAC's w/ good USB inputs often claim to offer good circuitry to isolate it from the PC / media server's bad / dirty power, but if you can avoid passing that along w/ the aforementioned USB cable, that would seem to be even better.  I suppose a good USB / SPDIF and/or ST optical would be another way to approach it.  That's what I did for my media server.  I.e. USB - convertor - SPDIF RCA to try to leave the grunge behind on the media server + because my DAC is not asynchronous.

Besides that, I suppose "theoretically" the next best options would be ST optical, BNC, then AESBU (balanced), then SPDIF RCA (unbalanced).

[werd]

Go with rca over bnc. Use the bnc connect adaptor if you want but there is far more buy and sell market for rca.

[sts9fan]

Why not USB??

Well that needs a computer. Maybe he's not using one.

I think engineers will say BNC.

[wakibaki]

AES/EBU cables use an XLR connector and 110 ohm three-conductor balanced twisted pair.

The S/PDIF standard is defined to use 75 ohm coax. It also commonly uses RCA connectors, but these are not available with a controlled impedance (although some manufacturers may claim to produce such). 75 ohm BNC connectors may be used for S/PDIF and are generally considered superior to RCA, but in practice, little difference in performance is likely to be encountered, particularly with shorter cables. If long cables are used, or 24 bit samples are being transmitted, as is possible in some implementations, and particularly where sample rates greater than 48 kilosamples/sec are employed, then BNC would be the preferred solution being technically superior.

If 16/44k1 is being used there is almost certainly no disadvantage to using RCAs on a cable of any length likely to be encountered in an domestic environment.

Some disagreement about this subject has occurred in the past, it being suggested that reflections in the cable give rise to jitter since the clock is extracted from the transmitted data, RF attenuators having been proposed as offering a way to improve performance, but while these do have some impact on reflected edges, they also attenuate the wanted signal which is likely to cause problems of itself. There is no evidence based on analysis of reproduced sound quality to support the use of attenuators, but as with many 'tweaks' subjective reports of improvements persist.

Some S/PDIF receivers have a purely resistive termination, others use transformers. In general, transformer types offer the better performance, but again, if the situation is not particularly demanding (a consumer application as opposed to studio use) the added expense is almost certainly not justified by any perceived difference in sound quality.

There will of course be those who disagree with the views expressed here, but they are reasonably representative of those widely held by the engineering (objective) branch of the audio community.

w

[john dozier]

Both RCA and XLR are poor choices for digital audio. Neither was ever designed for the speeds at which digital operates. A well designed BNC is the only choice. Be sure to get the 75ohm version as there are also 50ohm BNCs used in RF transmission. It also makes a more secure connection than RCA and again, imho, better than an XLR. Regards

[Speedskater]

quote
AES/EBU cables use an XLR connector and 110 ohm three-conductor balanced twisted pair.

We have a typo or an oxymoron here. twisted pair is 2 conductors.

quote
Both RCA and XLR are poor choices for digital audio. Neither was ever designed for the speeds at which digital operates.

While the RCA connector was designed long before S/PDIF audio, S/PDIF audio is at a very low frequency.  RCA connectors do very well with higher frequency video signals.

It's April 1st so the myth that S/PDIF connectors are important goes on.

[wakibaki]

We have a typo or an oxymoron here. twisted pair is 2 conductors.

Neither.

Second line below 'hardware connections':

IEC 60958 Type I Balanced – 3-conductor, 110-ohm twisted pair cabling with an XLR connector, used in professional installations (AES3 standard)

http://en.wikipedia.org/wiki/AES3

...so I presume the balanced pair are twisted and the 3rd conductor left to make it's own way willy-nilly in the insulating sheath, or it could be in the form of a braided sheath enclosing the pair inside the insulating sheath.

Otherwise I confess to having looked it up and depending on what I found.

w

[john dozier]

There is no myth. The RCA and the XLR are both lousy for digital audio. Ask any competent engineer and you will get the same opinion.

[Speedskater]

They are counting the optional shield as the third conductor.  A rather unorthodox way of counting conductors.

[Speedskater]

Other than nobody loves RCA connectors for mechanical reasons, I don't recall any of the experts writing that they are 'lousy for digital audio'.
Some experts:
Henry W. Ott
Keith Armstrong
Jim Brown
Bill Whitlock
Stephen Lampen
Ralph Morrison
Neil Muncy
Dick Pierce
Bruno Putzeys
Tony Waldron

So can you help me out?

[john dozier]

Ask Guido Tent, who also has a professional audio company, Grimm Labs. Regards