Innuos Phoenix USB Reclocker Review

[MingChops]

Hehe. Quantum tunneling isn't my straw man -- that was a Synergistic Research claim for their speaker cables. What I am opposed to is asking others (Amir) to test every quack idea a fertile audiophile manufacturer's mind can come up with that has little to do with audio.

Oh those guys, the ones who did those carbon discs! They are stop the world I want to get off levels of crazy!

I mean it could be even argued to an extent as to why this product was tested, given the notion of USB transfer being async, it was always a foregone conclusion this product would do nothing to the resultant audio output. But given its astronomical costs then it is good to see nonetheless - and without repeating old ground, the tests amrim did run prove that, and were necessary, alas I just for completeness would have liked that extra test running too.

Anyway if nothing else, Innuos is now a manufacture I would never consider because of this gimmickry. If they can fleece customers so gratuitously in one area, they are not immune to doing it in equal measure elsewhere.

 

[voodooless]

But given its astronomical costs then it is good to see nonetheless - and without repeating old ground, the tests amrim did run prove that, and were necessary, alas I just for completeness would have liked that extra test running too.

I understand your viewpoint. But why ask @amirm for this, especially if it mandates tools that he doesn't have and costs an arm and a leg (and possibly a kidney )? What about the manufacturer? Should they not put their money where their mouth is and show themselves that what they claim is actually valid?

 

[BeeKay]

costs an arm and a leg

I assumed this would be a typical German idiom …

 

[voodooless]

I assumed this would be a typical German idiom …

It's definitely not at all a German idiom

 

[MingChops]

I understand your viewpoint. But why ask @amirm for this, especially if it mandates tools that he doesn't have and costs an arm and a leg (and possibly a kidney )? What about the manufacturer? Should they not put their money where their mouth is and show themselves that what they claim is actually valid?

Well if the Modi test only required a small amount of additional time, it would in my view have been a warranted inclusion as then neither time nor money are a factor. If the test required 100k of additional gear however, then absolutely no it would not be warranted - no singular component would warrant that expense. But my understanding is the Modi test gives us enough of a basic surface level test to quickly identify if it does what it says on the tin or not, so far as a re-generated jitter free signal, acting in place of a master clock.

I would assume the aforementioned 100k device goes some orders of magnitude unnecessarily deeper in testing? - certainly such a device is beyond my skill set and would yield results even more detailed and obtuse than perhaps the one I am already seeking! But fundamentally I'm trying to tackle this from a logic point of view, whereby I see merit in a quick test that would set out to control a variable and an assumption, before then running the tests that were run.

And yes, any manufacture selling products especially in this area of audio should be providing some level of detailed metrics to support the claims; switches, clocks, and frankly cables and the like too. But that's never going to happen for obvious reasons!

 

[KL....]

I assumed this would be a typical German idiom …

Off topic but a reply to clarify....
'costs an arm and a leg'.... it can be suggested that this saying is English and originated about the time of 'Jack the Ripper', why?.... because at the time there was a great (black market) demand for body parts, especially Limbs. Initially, they were digging up dead bodys for experimental medical purposes but they found that those parts were not fresh enough. Highly likely, at that time, people also began requiring amputations (for minor injurys or afflictions), and not just being killed/murdered then being immediately dug up for experimental medical purposes.... hopefully that is sufficiently concise and no need to be (further) precise.

That was then and (now) there is an expression, 'become an Organ donor at your own risk'.... hopefully that is sufficiently concise and no need to be precise.

 

[Da Capo]

Impressive looking PCBs and components. Wouldn't it be amusing if the I/O board is multilayer, with hidden traces that simply run the USB input lines to the outputs?

View attachment 430698

Thanks for showing these products are basically useless.


View attachment 430695

View attachment 430696

View attachment 430697


JSmith

 

[voodooless]

But my understanding is the Modi test gives us enough of a basic surface level test to quickly identify if it does what it says on the tin or not, so far as a re-generated jitter free signal, acting in place of a master clock.

I very much doubt this. The reason for its poor jitter performance is most likely the lack of additional clocks into the CMedia USB receiver. It only has a 12 MHz run PLL to generate the needed 48 MHz USB clock, but also make whatever is needed for the audio sample rates. Multiples of 48 kHz will be fine, multiples of 44.1 kHz are not. A USB reclocker will do nothing to combat this. The Modi 3 PCB clearly shows the additional clocks though, partly explaining the improved jitter performance vs the multibit.

 

[dustinyo]

Clearly it didn't work because you fed it dirty power. You'll need to upgrade to a solid gold $20,000 power cable. /s

 

[MingChops]

I very much doubt this. The reason for its poor jitter performance is most likely the lack of additional clocks into the CMedia USB receiver. It only has a 12 MHz run PLL to generate the needed 48 MHz USB clock, but also make whatever is needed for the audio sample rates. Multiples of 48 kHz will be fine, multiples of 44.1 kHz are not. A USB reclocker will do nothing to combat this. The Modi 3 PCB clearly shows the additional clocks though, partly explaining the improved jitter performance vs the multibit.

That was the whole reason to suggest the Modi (the old Modi 2 specifically), because it has been used to show even over USB, variances dependant on the incoming USB signal:

ifi iPurifier3 USB Filter Review

This is a review and detailed measurements of the ifi iPurifier 3 "USB Audio and Data Signal Filter." It is on kind loan from a member and costs US $149. The iPurifier 3 comes in a gorgeous and nice to feel ceramic (?) enclosure. The enclosure give is enough weight to feel substantial but not...

www.audiosciencereview.com

It's an anomaly in the DAC world, and an unrealistic use case and pairing for the Innuos - this I'm not disputing. But it is a device that even despite the async nature of USB, shows us if the upstream device is actually doing anything for not. Something I think would have been interesting to see proven, rather than take at face value that it is.

 

[voodooless]

That was the whole reason to suggest the Modi (the old Modi 2 specifically), because it has been used to show even over USB, variances dependant on the incoming USB signal:

From the teardown, it seems to use the same 12Mhz clock only as well, and from the review you posted on the iPurifier3, it will clean up some noise, but I would expect the jitter spectrum to still be relatively bad, looking at the sidebands of the 1kHz dashboard.

 

[MingChops]

From the teardown, it seems to use the same 12Mhz clock only as well, and from the review you posted on the iPurifier3, it will clean up some noise, but I would expect the jitter spectrum to still be relatively bad, looking at the sidebands of the 1kHz dashboard.

Indeed I may have inadvertently latched on to jitter as being the metric to focus on here, whereas yes sideband noise is perhaps the correct metric I should have been harping on about - that may have confused a few no doubt and I can only apologies, but not being an engineer I was always trying to approach this from a logic standpoint, whereby that is to control a variable/remove an assumption.

The saliant point remains however, that the Modi test is able to detect and differentiate its output based on the incoming USB quality (which the D70s is notably immune to), and thus why I think the test would have been useful to show the Innuos even does the basics of what it says it can do at an implementation level. The suspicious mind such as I have, wouldn't be surprised if it did sweet FA to the signal! If amirm still had the Modi 2 and it wasn't much time to run the test, would have been nice to see that result. - But yes I think you're right we would indeed be looking at noise and not jitter as the resulting distinction (my bad).

 

[JustJones]

Thanks for the review and it is nice to see the occasional nonsense component tossed on the bench but please don't waste time doing more tests with this thing. Speakers are waiting.

 

[JSmith]

Video review is up;

Innuos Phoenix USB Reclocker Review [Video]

Video review of the Innuos Phoenix USB Reclocker Review is out: Please subscribe to the channel if you have not done so. :) Text review at: https://www.audiosciencereview.com/forum/index.php?threads/innuos-phoenix-usb-reclocker-review.61000/

www.audiosciencereview.com

JSmith

 

[Moonray64]

Reclocking USB to USB is a nonsense, since all modern USB interfaces are now asynchronous and buffer/reclock themselves. Better save money to buy a better DAC.

 

[mcdn]

Impressive looking PCBs and components. Wouldn't it be amusing if the I/O board is multilayer, with hidden traces that simply run the USB input lines to the outputs?

Snap! That was my first thought too. Probably best to actually use the power supply otherwise it would look suspicious when everything worked with the Reclocker turned off though. But the rest? Might as well not be there. Incredible reliability.

 

[AndreaT]

There is no doubt in my mind that this Innuos product belongs the The Emperor’s New Clothes category, and it is useful to redistribute wealth better than Socialism. Kudos to Amir for allowing us to discern what matters in Audio Reproduction and what doesn’t, and to help me allocate my hard earned disposable income to better products.

 

[mhardy6647]

It's always refreshing to come on ASR and read a proper review of one of these products rather than the "I heard a difference that was not subtle. Clearer bass, less digital hash in the mids and a palpably lower noise floor" or similar nonsense one sees everywhere else.

Seems like an apt time to note that such quotes nearly always* include an unbiased third-party assessment from the kitchen.

If it doesn’t turn back time it’s not a proper reclocker.

Isn't that a declocker?

What i learned from this review is that @amirm could sell his Frankenstein usb cable for 5 K and be at peace with himself.
I'm on pre-order.

I don't even think that quantum mechanics has mathematical language that can accurately describe pre-ordering a reclocker. Plus, I am sure there's no grammatical tense that is suitable. OK, make that pretty sure. Maybe Latin has one.

Sorry I am always so late to parties like this!
______________
* ahem -- yes, this is these are anecdata; I haven't collected any statistics on the occurrence of such testimonials... but I may.

 

[JSmith]

declocker

I had one of those, was great to declock all my clocks so I could then at a later time reclock my clocks. Even my ex-clock was affected due a strange quirk of space-time, so I had to declock my ex-clock and then reclock it... which then became my present clock. It became confusing as to whether I should or would push meetings forward or back and what time they actually were. Matt O'Dowd did a video on it... time dilation of reclockers, check it out.


JSmith

 

[mhardy6647]

When I was in a first-year student in grad school (and this is absolutely true) -- I spent a semester "techniques in biochemistry" session working in the lab of a rather esteemed physical biochemist noted for his expertise in analytical ultracentrifugation*. Without going into too much detail, subjecting interesting macromolecules (proteins, nucleic acids... those kinds of things) to intense centripetal acceleration can tell one interesting things about their size, shape, and hydrodynamic properties. The technique(s) can be used, e.g., to study associating systems (which is what I was doing at the time). Some of these studies are performed at equilibrium; others are kinetic, measuring, e.g, sedimentation velocity as a function of... whatever.

Where this gets interesting is that the experiments are done at very high rotation rates: 50,000 to 100,000 rpm.

With a totally straight face, I asked a postdoc in the lab whether the tangential velocity at the outer diameter of the centrifuge rotor (where the samples are located) was close enough to c that allowance was required for time dilation when calculating velocities.
The postdoc laughed.
______________
* In my day, this still meant using a Beckman Model E
I enjoyed the heck out of running these instruments (the lab had two of them)!