Sound quality differences between NAS units

[DavidHB]

This issue does surface from time to time; I seem to recall an attempt on the Lejonklou forum a while back to rank NASs for sound quality. Whatever you think of that, there seem to be enough independent observations of differences between NASs to suggest that the observers were not deceiving themselves. At the same time, it is reasonable to suppose that the laws of physics have not been suspended for their or our benefit. The observed behaviour has to have a cause, and that cause (whatever it is) has to be able to travel through the network and affect the sound quality of the output of the player.

I can think of 3 candidates for consideration. The first is that the bits and bytes of the digital stream are somehow changed. Personally, I think that the workings of TCP/IP (chopping up the stream into packets at the sending end, routing those packets through the network, and then reassembling the stream, using error checking, at the receiving end) make data corruption a fairly unlikely possibility. The next possibility is that different devices introduce different, but always minuscule, degrees of timing error ('jitter') as they handle the data. As most quality players have highly accurate clocks to which they lock the data stream at their end, it seems to me that the concept of a NAS (or any other network component) introducing jitter which the players cannot handle is somewhat implausible. But, with my limited knowledge of the subject, I could be wrong about that.

This leaves the third candidate - noise. Noise is all-pervasive in electrical and electronic systems, and the digital domain (which is typically only interested to know whether a transmitted value is a 0 or a 1, so off or on) is tolerant of high levels of noise. If (or rather when) that noise percolates into the analogue domain, the story is very different; one of the markers of top quality systems is a high (these days, very high) signal to noise ratio, and a correspondingly low noise floor. I believe that, as time goes on and technology develops, the elimination (or, more likely, the control) of noise at all stages in the reproduction chain, both digital and analogue, will become even more important than it has been up to now.

Coming back to the difference between NASs, if noise is the issue here, where does it come from? Well lots of places probably, but there is a strong body of opinion that points the finger at power supplies. Over on the Linn forum, for example, you will find plenty of references to experiments, both in conditioning mains power (a favourite topic in the US and Canada, it seems) and in replacing or modifying the power bricks and wall-warts that come with most of the devices we buy these days. Would power supplies be the issue in the present case? I don't know, but a combination of improving the power supplies and network noise isolation would be the first possible remedy I would try.

The point here, of course, is that it is probably misleading to think of a NAS in isolation as having a sound, good or bad. What it (or, more probably, its power supply and network interface(s)) does is affect sound quality in particular installations. It is not surprising that, in a given context, different NASs will behave differently. Whether they exhibit the same differences in a different context is, of course, another question.

David