Sound quality differences between NAS units

[lyapounov]

(19-08-2016 10:01)DavidHB Wrote:  I can think of 3 candidates for consideration...

I was to write the same thing roughly, but DavidHB did it brillantly !

Though I have a question, and I think Simon has the answer : how is the content transferred ?

UPnP says it sits on top of TCP but only for the discovery and other stuff (like the former channel D in the old telecommunication times), but when it is about sending the content, it seems to be based on UDP using unicast protocol.

Simon : can you confirm ?

If so, then it explains a lot : TCP guaranty that the content is exact when received, while UDP does not. This is why TCP is not a real time protocol, because if the receiver is not happy, it asks the sender to resend the same content; incompatible with streaming. Therefor, it could be that the communication is more or less corrupted depending on how the extremities really cooperate, or the quality of the copper in between, or other external problems, like load on the network, interferences on the copper (people would be suprised how much electronic pollution there is on earth...) wifi transmission etc.

As an exemple : wifi quality degrades a lot when humidity is high. Reason being that the wifi frequency is an open one, the one of the microwave oven, which is the absorption frequency of the water molecule. I remember talking to the city of Amiens in France, 8 years ago. They had built a wifi network in the city which worked perfectly; except that they built it in winter. And when spring arrived, the tree leaves grew, and water in the leaves absorbed all the wifi signal power, and the network was poor quality. You can even check this at home, under very rainy days.

So degradation of the signal means more error correction, more interpolation, more CPU power, jitter, and eventually less accuracy in the data stream itself. Again, TCP manages, but not UDP. Therefor the concept of QoS in UPnP AV.

I am not an expert in electronic. But for sure, the better is your sound system, the more you will hear / feel such differences.

The only scientific experiment I would see would be to record the data which are transmitted, and check if they are exactly the same or not depending upon different factors such as time of the day, different servers, etc...

Just my two cents.