Coaxial Vs. Optical digital S/PDIF connection, which sounds better?

Hello and welcome to whatishifi blog; since you discovered it, why not having a look around!

Note: Unfortunately, back in 2011 when we did the following test I did not keep photographs and notes of all the details, but I remember very well all the important things and conclusions. Still, let me apologize for not remembering every detail - I was not aware back then that one day, I would have a blog.

For over a long time, I am using a CD transport and a dedicated D/A converter in order to listen to my CD’s. Like many of you, I was wondering if S/PDIF (Sony/Philips Digital Interface Format) coaxial digital cable connection sounds different from the S/PDIF optical TosLink fiber.

There is a trend in audiophiles in favor of the coaxial digital cable connection. Optical fiber is supposed to be inferior because:

• Optical fiber has supposedly smaller bandwidth than the digital coaxial cable.
• Optical fiber supposedly introduces jitter.
• The actual digital signal is electrical, meaning that if it travels through an optical cable, it has to “suffer” two conversions: it has to be converted from an electric signal to light at the transport and then it has to be converted from light to an electrical signal at the DAC. These conversions are a compromise and they are suppose to be bad.

But as I have repeatedly learned, Hi-Fi culture includes a lot of logical looking assumptions that are not necessarily true. So I had on my mind the idea to make a proper coaxial Vs. optical S/PDIF digital connection test. Soon, one evening, the opportunity for this test arrived with the help of my good friend Panos: he had borrowed a Marantz DV6500 multi format DVD SACD player (that unlike our Pioneer transports, it had both digital coaxial and optical outs) as well as a Kimber digital cable, a Van Den Hul digital coaxial cable and a QED optical fiber. We also had an economic optical fiber and a cheap looking digital coaxial cable.

Marantz DV6500 digital outputs.

Having already an excellent Parasound D/AC-1500 DAC with multiple digital inputs meant that we could do a reliable A/B blind test, instantly changing the digital inputs at our convenience and comparing the sound on the fly.


Pano's Parasound D/AC-1500 has multiple digital inputs.

So what we did was to connect the Marantz with the DAC simultaneously with a digital coaxial and an optical fiber and play a good CD recording. One of us would listen, the other would stand in front of the DAC and would change the digital input from coaxial to optical and vice versa. The test was blind, meaning that the listener would not know which input he listens to, neither he could see the input indications lamps of the DAC.

Each listening round would consist of 3 listening tests. The listener would choose each time his favorite input without knowing which type of input he listens to. In order for the listening test results to be granted as valid, the listener had to favor all 3 times the same thing. Then, we would change positions and we would repeat the test. In order to accept that our results are valid, both of us had to favor the same input all 3 times. Also, we were not in a hurry; the listener would ask for the change whenever he felt like it.

The differences that we heard had mostly to do with the 3D soundstage of the music. When we say that connection A is better than B we mean that music was livelier, better focused and with a greater soundstage that would make you feel like the artist is standing in front of you. The rest of the system back then was a Parasound PLD-1100 pre amp, a Parasound HCA-2200II power amp and Audio Spectrum Hermes speakers - excellent floor standing speakers with SEAS drivers, not widely known outside Greece.

Here are the results:

• No name cheap digital coaxial cable (the ones with thin cable and yellow RCA) Vs. no name 15€ optical fiber: optical was better.
• Kimber digital cable Vs. no name 15€ optical fiber: optical was better.
• Van Den Hul coaxial digital cable Vs. no name 15€ optical fiber: they both sounded identically good.
• Van Den Hul coaxial cable Vs. QED optical fiber: they both sounded identically good. (Here we also realized that the no name fiber sounded the same with the QED).

Unfortunately I do not remember what exact models were the Kimber, the Van Den Hul and the QED. I remember that the Kimber digital cable was not coaxial but a braided design. I also remember that the Van Den Hul was a very thick cable and it had cost around 200€.

So, our conclusion was that both digital coaxial and optical can sound equally good, it just looks like that it is easier to achieve top performance with an optical fiber. I mean, we needed an expensive Van Den Hul to make the digital coaxial achieve top performance. But we could achieve same top results with an inexpensive, humble optical fiber. This means that it is much easier to get it wrong the coaxial way. So, even if both connections can achieve the same high level of performance, our verdict was the optical fiber. Sorry die hard audiophiles!

Let me add a few more things. For digital or video signals, you need a cable with a very effective shield and 75Ω (Ohm) characteristic impedance. Ideally, you also need 75 Ohm characteristic impedance on the connectors. (Please note that characteristic impedance does not mean that if you measure the cable resistance with a multimeter you will measure 75 Ohms! If you want to learn more about 75 Ohm characterisitc impedance you can easily find a lot of info on the web).

Any variation from this 75 Ohm characteristic impedance will make the cable to behave not as the ideal transmission line intended, causing degradation of the signal it carries. Unfortunately not all cable manufacturers pay attention to the 75 Ohm characteristic impedance – not to mention what happens in the connectors, where ideally the digital cable should be crimped and not soldered!

Also, since the coaxial connection is an electrical one, it can reveal possible ground loop problems of the system.

On the other hand, a decent optical fiber with proper connection ends that lock securely will do the job.

Back then, just due to syndrome of optical inferiority, I bought the most expensive quartz optical cable.  In reality, all it offers me is pleasure of ownership.

It is not that the digital coaxial connection is inferior, it is that it is easier to get it wrong and it will not forgive possible design flaws. Both connection types when applied correctly can lead to top performance.

Happy listening as always!

Chris

P.S. Here is another article in defence of TOSLINK capabilities.