Digital cables, USB, Coax, Toslink?

Not looking to start an argument but has anyone got a technical reference to why the sq should be affected by the different types?

I have no idea, I do however expect that popcorn might be needed for this thread.

I want to avoid that. I am interested in any technical explanations as to why there might be differences though. My simple guess with no knowledge at all, is that it will be more about the end connections than the transport media, but what do I know.
I am currently running a daily swap ( to let my ears adjust) between co ax and usb using a radio stream to avoid nas or ripped source errors creeping in. The USB does sound different, why?

Because datatransfer in this case (spdif) is not buffered. It is real time,meaning that biterrors can not be corrected with a retransmission. So a “0” can easily turn into a “1” and that can cause audible distortion in the digital to analog conversion. So yes,in this case quality of cable matters​:wink::grin:
Now of for popcorn… :grin:

1 Like

That makes sense though so no popcorn.

Might aswell close the thread now then!

In the context of your question about sound quality it doesn’t make sense. Cables don’t lose bits unless they are defective, with a bad connection in a plug or a tight kink for example. If there are differences in sound quality that’s going to be down to noise shaping causing a subjectively better or worse change to the audio signal.

Hence my assumption re connections.

I guess my question really resolves around the different processing requirements for usb vs spdif transmission then?

I dare say the implementation in the box you are connecting is the significant factor here. Naim tend to concentrate on optimising their Ethernet and electrical SPDIF inputs and outputs, and generally recommend that you use these where available, instead of WiFi, Toslink or USB.
Other manufacturers may, of course, take different approaches.


They don’t “lose” bits but the integrity of the data can not be guaranteed. As said,at the receiving end a bit transmitted as a “0” can easily be interpreted as a “1” due to bad cable characteristics. In a buffered configuration this is checked with the CRC and if necessary data is retransmitted, Not so when using e.g. a coax cable. Data is then directly fed to the DAC,real time, so no opportunity transmit a block of data again. And then I am not even considering clocking aspects.

Interesting info, thanks. I am using an all Auralic system with an Aries streamer usb connected to a Sirius pre processor and resampler, finally into their Vega dac via their proprietary cable, hdmi type.
The main processing seems to take place in the Sirius so maybe the usb vs coax from the Aries is less significant.

Over a local network this would be rare and it would indicate a broken cable that needs to be replaced. I can’t say with any authority how common it is over the likes of USB or Ethernet, but I suspect not that common.

1 Like

Data integrity can be safely assumed for the most part, if bits are being lost or misinterpreted at any meaningful rate due to a bad cable, you’ll quickly end up with distorted audio due to breakup. It is timing and noise that needs to be placated at transmission level.

The diferent technologies have their own benefits and weaknesses, and this is at the physical level alone, the so called layer 1 which is always analog. In practice it becomes considerably more complex:

• Toslink provides perfect galvanic isolation, but limited bandwidth. Still, a well implemented Toslink (that is good quality fiber, connectors and optical transceivers) can reliably do 24/192, and sound really good while at it.

• Coax is higher bandwidth and simpler to implement and control, but will create an electrical *bridge" potentially carrying RF, EMI and ground loops all the way across to the DAC analog circuitry.

• USB has even higher bandwidth but is electrically noisy, and being two-way (duplex) plus carrying voltage across even more complex and difficult to implement. Even the Amanero interfaces many high end audio manufacturers prefer are known to be noisy, one of the reasons Chord like dual coax better.

@ChrisSU hit the nail on the head IMO, it is the quality of the implementation that will make the biggest difference in SQ in the end. Experiment with the different types of connections and let your ears decide.

1 Like

So far I have discarded toslink in favour of 75ohm coax. I am now thinking I am hearing more bass with the usb connection that the coax. On the other hand it sounds as though the co ax is a little more open , spacious in the mid range. Am I hearing just what I want to hear?
I am now listening to usb and upsampling to dsd 256 (x44). I think I like it but only an hour or so in so far.

Whatever you say guys/girls,fine with me.
I have the popcorn ready so bring it on. Let the show begin! :grin:

We are getting near to that point.

I know which I prefer but like a dog with a bone I have to know why. :grinning:

That likely means the difference is quite subtle, kind of expected IME.

It could be but I think it is more than that. There is something that just feels right when it all comes together. I am just playing a playlist of acoustic rock covers. Black Hole Sun, Chris Cornell, a live recording. Hairs on the back of the neck, if Mrs Bruss wasn’t watching I think a tired and emotionally tear may have appeared. It puts you in the audience with all the atmosphere of the time it was recorded.

1 Like

Interesting… can you point me towards the album please Bruss. Superunknown will be getting a play tonight but I’d like to listen to the track you mention.

Amazon playlist, Acoustic Rock. 90 songs and outstanding quality. Now Nothing Else Matters, Miley Cyrus Elton John, et al, you wouldn’t know.

1 Like