roderickvd
100+ Head-Fier
No, it is not a must. I understand it might have come across when I said “problem solved” but it is necessary only when you have a ground loop.
I had no ground loop, but bought one to satisfy my audiophile curiosity. Measurements show that there is a good 20 dB of noise attenuation and harmonics even in the absence of a ground loop. Though measurable, I admit that I cannot hear it — only logical since the Ares’ own noise has been measured well below the threshold of human hearing.
Expectation bias is with everyone.
I would say the semantics are that all transmission methods have issues. Which is where I find the beauty in the audio hobby: it’s all about compromises.
This is sound advice (no pun intended). Implementations differ and so do personal preferences. As always, I just try to contribute by sharing engineering knowledge, then to each to pass his or her own judgement.
You are right on the voltage levels to make it ones or zeros, but otherwise USB works differently from what you describe. There is no sort of lossy error correction. USB packets contain a CRC (a “check” value calculated from the actual payload, to see if the packet was received correctly). The USB client sends the host a confirmation when a packet was received successfully with a matching CRC. In the absence of such an acknowledgement, the host retransmits the package. If the cable or connection is poor, and too many packets are dropped, you will hear skipping. Packets are either received with full integrity, or they are not, they are not mutated.
Indeed battery powered devices are by nature free from ground and mains noise so they are perfect in that regard. Of course like you say always test your system, because switching circuitry in them can and will still inject noise on connections (particularly ICs). But less so than noisy PCs and such, that’s for sure!
Although I don’t find much in the argument that USB audio cannot or will never be good because USB was not invented for it — we could make a list of the greatest inventions that were intended for a different purpose — I am totally with you that I2S with a great master clock really is the best you can get. Unfortunately all Denafrips DACs reclock I2S signals too and so are limited by the quality of their own clock.
I’m usually more diplomatic but let me say that this is nonsense. I will go out on a limb and say that electronics can be microphonic in the presence of loudspeakers, which is the basis for all matters anti-vibration, but tuning a chassis is just nonsense.
I think that @sajunky provided a good factual description of how things work. And again everyone looks through the glasses of their own choosing.
Almost any: at least those with good shielding, preferably with ferrite beads, and good fitting of connectors. This guarantees high noise rejection.
As do I. I hold no grudges against coax or Toslink, and encourage anyone to choose what they love the music with most. Just want to be transparent on what’s the deal with USB.
