Article: Review | Innuos PhoenixUSB Reclocker

[barrows]

Rajiv, thanks for the review and especially for sharing the technical info from Innuous.  Respectfully, I would suggest though, that you might refrain from such definitive language as this:

 

"The reason DIY tinkerers like me end up with spaghetti digital chains is not because we love the clutter, but because manufacturers have not (yet) offered components that deliver the SQ we are able to achieve"

 

Unless you have personally tested every USB output device available for audio delivery to a DAC, you are not well informed enough to make such definitive statements.  The review would be more accurate if you included a statement such as: "within the realm of my experience" or something to that effect.  I would also suggest that even if you had tested every such component, your only metric for evaluation is subjective, and as such, also not definitive.  The "quality" of USB output can be measured definitively through the eye pattern diagram combined with a spectral analysis of the USB power rail: these are the only relevant factors for USB transmission and subsequent audio performance (covering both the ultimate signal integrity and the noise factors).  I have no problem with subjective evaluation, but when an evaluation is entirely subjective, such should be mentioned as well, rather than making definitive statements about a thing being "better".  Statements such as, "in my system i found this sounded better to my ears" are often used as caveats by reviewers to make sure the readers understand the limitations of the subjective evaluation process.

 

 

[barrows]

5 hours ago, Superdad said:

 

 

Hi Barrows. Please take this in the playful Friday spirit I intend: You might do well to take your own advise regarding the above.

Because both an eye-pattern and VBUS spectral analysis will in fact not tell you all.  Proof?

Right there for you to check with two original Sonore microRendus, a v1.3 board and a v1.4 board, wherein the only difference between them is the latter's use of a low-phase-noise Crystek CCHD-575. The better clock is clearly heard, yes? But I challenge you to spot the difference in an eye-pattern test.

 

Sure, improved signal integrity common to both through the Rendu-series use of a USB hub chip at the output (and common to all USB "regenerators"--including the Phoenix--since JS started this trend back in 2014) can certainly be seen.  But the low-phase noise clock is not going to present in an eye-pattern; to understand its benefits one has to dig deeper (our 'white paper' points the direction).

 

Likewise, the 5VBUS lines (really the whole USB "ground" connection)  are a pathway for common-mode leakage currents which travel through all audio systems, so looking just at the noise level/spectrum on the +5VDC line will not expose those factors, which again can propagate ground-plane noise and increase receiver chip jitter by modulating the reference voltage used to determine clock thresholds (inside the DAC stuff, and I refer again to our paper).

 

Hope your Colorado weather is lovely this weekend and that you have something fun planned.

Cheers,

--Alex C.

Hi Alex, not sure what/how you mean about spectrum analysis not showing the noise?  And indeed, I would expect a highly resolved eye pattern measurement to show the difference of phase noise for the USB clock.  After all, phase noise is essentially jitter, one way or another, and jitter is derived from the eye pattern.  What I would admit is that one would need a very highly resolved eye pattern test to see the difference.
As far as my criticism goes of Rajiv’s definitive statements, I stand by that, as a reviewer, one should be very careful about their language.  A reviewer, published in a very widely read forum such as AS, needs to hold themselves to a very high standard, as their words carry a high degree of gravitas.