Topping announce a new USB (high speed) isolating device. For those directly connecting computers (such as NUCs, Nuclei, or similar) to the DACs and suffer degradation in SQ maybe a nice solution to eliminate ground loops and other ‘noise’. Not crazily expensive too.
https://twitter.com/topping2008/status/1622777314688831491?s=46&t=tiWMz9aE3aoPV12N9STUXg
Will be interesting to see the ASR take on this considering brand and device type.
Downside: Might double the costs of esoteric USB cables 
double+. As you need 2 cables and the device
What’s in that box that wouldn’t already be in a Topping (or any competently engineered) DAC?
It may compensate for weirdly engineered high-end cables and boutique DACs. Not necessarily needed for Topping gear 
Ouch, what will Amir do now? He always praises Topping, but also never has anything good to say about devices like this.
And no mention of the oscillator for the USB protocol (25Mhz if I recall correctly), which does make a difference.
Good point
I fully expect he’ll say it will only help only those with serious noise or ground loop issues and will have otherwise absolutely no benefit to sound quality. And that folks should probably sort out their ground loops and noise in a fashion that doesn’t involve purchasing a device such as this.
He might. I’m very anti-snake oil, but in the (quite specific) case of USB from a computer – where the output is invariably noisy – I see no harm in fixing up the USB waveform as early as possible using a relatively inexpensive device. At the very least, as long as the device is competently designed, no harm should result.
Although I don’t use USB regularly, when I do, I have a Holo Audio Titanis for such clean-up purposes, which is similarly priced to this Topping unit. Jeff Zhu at Holo Audio knows what he’s doing.
Note that @Archimago has posted tests of both the Topping HS01 and the Intona 7055-C USB 3.0 SuperSpeed Isolator and concluded that they both do work. Probably fair to assume that the HS02 will work as well as the HS01, if you’ve got a noisy USB hub.
Though he does also add, “I don’t worry about “noise” in the USB system itself when we add a good modern DACs in typical sound systems.” I think it’s kind of foolish to worry about that, too, unless you hear something, or perhaps if your DAC is USB-powered.
When I used USB directly from computer, I got a ticking noise whenever the computer did some work. An iFi Micro USB cleaner/reclocker fixed that (I assume this topping device would also fix that). So there are some cases where the benefits are very big and easy to hear.
It gets harder once you talk about the effect of USB protocol clock. I experienced this myself when I upgraded my clock in a normal microRendu to a Crystek clock (25Mhz). It really did a difference (I had a RME ADU-2 fs DAC back then).
And then you add linear power into the equation (or any clean DC power), which will improve the clock among other things, and we reach the point where Amir tells it does not matter, but my ear tells something different.
I have not found USB output to be “invariably noisy”. If the USB output to my RME is noisier than optical it certainly is below the audible threshold and inserting noise reduction devices has provided no apparent benefit.
If you suffer from audibly noisy USB I would think using optical input makes a lot more sense than throwing money at stacks of dubious hardware fixes (unless you imagine only DSD can deliver desirable SQ).
Well, reviews are for individual pieces of hardware, not manufacturers. I’ve seen both good and bad reviews for same manufacturer. I for one think isolators are useless and find it disappointing coming from Topping, but hey, they’ve been making DACs with M*A support for a while…
Other than perhaps believing in magic. And going down the path of adding linear PSUs, re-clockers and other things that could do no harm.
USB clock is irrelevant if using async transfer.
Even if these devices work, it’s not clear there’s any point to them. For instance, take this example with the D10 Balanced USB-powered Topping DAC:
Clearly a cleaner signal, but inaudibly so.
The differences in the noise seem to come from suppressing power supply harmonics (multiples of 60Hz), which were below -130dB. They are only visible because of the very low noise floor.
No, although that’s a very common misconception. But it’s not clocking for audio frequencies (44.1, 48, and multipliers), which would be irrelevant for async USB. Instead, it’s the clocking of the underlying USB protocol at 24Mhz, which has its own problems with jitter and re-assembly at the end (XMOS etc). The better the signal reaches the USB receiver, the less it has to work, which leads to less internal noise generated, which in turn means a more accurate oscillator inside the DAC (oscillators are extremely sensitive to electronic noise and voltage fluctuations).
Better transports/streamers have quality clocks for USB protocol for a reason, it really does make a difference. Remember that USB receivers are small computers in themselves with CPU and memory.
Here is my own hack in my microRendu (which btw sounds excellent with the better clock and a good LPS, definitely better than the RPI4 is replaced):
The misconception goes the other way, i.e. from a lack of understanding of the USB audio protocol.
Yes, and their only role in an async transport is to ensure those memory buffers are neither over- nor under-flowing. It’s no different than any digital transport, and async digital transports’ clocks have no influence on the DAC’s clock.
