Can This Really Matter?

I hear it again and again, everything matters in the digital chain. So I recently unboxed my new Bryston BDP 3 renderer and it’s connected to my Ayre QB9 Twenty USB DAC. I have a SSD inside a generic enclosure (with a fan) that will power up via bus. I have a twenty cent USB cable running from the SSD to the Bryston. So here’s the question, does the enclosure matter and does the cheap USB running between between the SSD and into the Bryston really matter? If so, it wouldn’t seem practical to spend much money on upgrading these accessories. And if so, what do you suggest?


A bad (not to spec, defective) cable can cause trouble even in digital paths, but a solid cable will be no different from a really expensive one.  Bits are, in fact bits unless there is a serious defect (impedance, shielding, etc.)


Now, before people pop, there ARE issues int he digital path.  At the DAC chip itself (not the box) you must have timing. I explain this int he sonogyresearch blog -- but that only applies to SPDIF, not USB which is asynchronous.  A much bigger issue is ground and power noise, which is why sources, LPSs etc can affect sound.  I power all my sources with a custom low noise LPS, (linear power supply) and wherever possible I also use a bridge (e.g.: dedicated ethernet --> USB box) to further isolate the DAC from the noise of a computer. All streamers are computers, although they may be embedded in a fancy box.

You would not believe how much testing i have done on these various issues, both as me and through the company.  Many things surprise me and do matter.  Digital cables, past the point of engineering correctness, are not one of them.


Beware the placebo effect. One thing i can say, for sure, is every system sounds better with a glass of good wine.  really. And that tells you a a lot.  The wine (etc) variable must be accounted for in real listening if apples are to be compared to apples, which, when in this world, is my business.


The digital streaming argument is 1’s and 0’s are just that and you can’t change them. Yes this is correct but there are a lot of hitchhikers that will jump on the 1’s and 0’s in the form of noise. Noise in the digital signal path is not noticed until it is gone. One of the most cost effective upgrades you can do is to add a fmc into the upstream signal path for the 1’s and 0’s. You want to create and separate a dirty side a clean side for the digital signal. Yes all cables do matter in they way they introduce and eliminate noise. Enjoy the music!

You have to assemble the 1s and 0s correctly to produce an accurate waveform to correctly reproduce the music. That is where almost every product, be it hardware, software, or both, will differ somewhat. How different and how much it matters case by case is another story. Just like all the other fine details that go into good sound for discerning ears.

to sgreg1 and mapman

1. Yes digital noise is the enemy - actually ANALOG noise on digital paths, created by high frequency digital circuits and switching power supplies. Cables will do little to fix this. Filters and isolation components like bridges do a lot.

2. The reconstruction of the 1s and 0s into PDM or PAM, and the reconstruction of that into a continuous waveform, is, of course, the entire issue.The question posed is quite different: do USB cables in particular, which carry an asynchronous stream, of 1s and 0s, have much effect? As noted above they can't really block much noise unless they are far more than cables. And unless they really suck, they cannot create bit errors.



oh, i should add to the above to finish the logic...and the digitally generated noise impacts a) the reconstruction process and b) the analog stage during and after reconstruction.  Some posit that it can affect the quantized PAM/PDM --> analog, by affecting zero crossings, but i think that is unlikely given how well proven digital transmission intergaity is (witness: gigabyte internet into your house...amazing)