Galvanic isolation absolutely makes sense if there is a ground loop that causes hum.
No other reason exists to go that path.

In your case the opposite is true, the USB connection supplies ADI-2 DAC with a ground reference through the computer mains/PSU connection.
This can avoid disturbances under certain circumstances, e.g. if you use AUDEZE headphones that can unintentionally transform a static electricity field into sound.

Yeah. I absolutely agree with you. The problem is to differentiate audio voodoo with real stuff. Marketing bullshit doesn't make things easier.

KaiS & Curt, thank You for reply. Cheers!

If you do not hear any hum, then you have no hum problem that needs to be fixed.

The device does not lead to any sound improvement, digital audio data are sound-neutral and it does not matter whether the digital audio transport was via ADAT, SPDIF or USB.

The important thing is the lossless transmission in the digital domain, which everyone can prove quickly and easily with the ADI-2 DAC/Pro thanks to the "Bit test" feature.
Also Jitter is no issue because Steadyclock FS technology removes any clock jitter, the clock signal is being refreshed and the final D/A conversion is being performed with the local FS (femto second) clock in highest quality.

It sounds far better than the RME's USB input.

It is obvious that mediummanting does not know what the bit test provides him. Otherwise, he would have been able to validate long ago that the bit test is passed with and without the additional device.

It would then be clear that in both cases it is a lossless transfer of digital audio data and that the final D/A conversion is performed in both cases with identical digital audio data which leads to the same analog wave form after the conversion.

Whoever is still talking about differences at this point has unfortunately not understood anything..

Furthermore, I believe that mediummanting does not use the correct test methodology to exclude psychoacoustic phenomena, which must mislead him here.

What I would at least expect from him is, to simply run the bit test with and without this device.
It should - if the test was performed correctly (i.e. correct choice of player, volume at 100%, best use of the ASIO driver) - always come to the same result, that the bit test was passed.

Well, and then there can be no sound differences. Again this is clear, obvious and no matter of a discussion.

You know the story of Don Quijote, don’t you?!
This is how it feels fighting against all this Snake Oil.

Maybe just let it be, let them spend their hard earned money, someone else enjoys getting it.

You are right, do not fix things until broken.....

My LAMY multipen stuck, finally gone, cannot repair, no spare parts available even from Germany, it is more than 30 years old ... luckily, they will send me a replacement pen free of charge to make sure I do not live without a LAMY pen.

WOW, 30 years plus of warranty and goodwill, impressive !!

Summing things up: don't bother until you have ground loop problem…

My subject question came from considerations if there is a place for any improvement. Since ‘I don't know what I don't know’ I had to ask.

Btw. On my other system iFi iDefender has mitigated buzz problem below audible threshold. It was a life saver.

Thanks for all of You for sharing thoughts! :-)

I have an iFi Zen Dac before I bought the RME Dac, still using it for my computer, noise found, noise source is from USB power fed from computer (Mac mini) to Zen Dac .... some friends recommend the iDefender, borrowed one, tried but not successful, then I use a USB cable with barrel plug, get 5V from either a powerbank or iPhone original charger, problem solved, clean.

Not to say iDefender is not good enough, but share the experience, every case can be different.

Another case is RME .... I stream Apple Music to RME dac via Apple CCK3, if I feed power to the CCK3 from computer, noise will appear at phones output when I connect it to a soundbar via 3.5mm cable, the solution is iPhone original charger, clean.

It works for my setup/connection, I will not say it is a global solution. But Bit test is the best tool to judge where the problem at.

For me it's working for many years without any problem. Source of ground loop issue was related to CD player which RCA connector mass is connected to AC ground pin. Connect computer via USB and loop is ready. This is not real galvanic isolation and differences of potential between devices may occur. I suspect it can be dangerous.

If this is the case I pull back my recommendation and delete it in my posting above.

To be honest I just read that one page and it's not completely clear to me which device is meant in MC's quote, nor what you'd have to do with exactly to break things. So just adding a word of warning should be sufficient?

Yes, of course in this case USB galvanic isolation is not advisable at all!

It‘s the worst case scenario:
Picking the mains stray current from the unearth device and feeding it through your body into the earthed one.

There should be an earth in a system somewhere, doesn’t matter where if it’s only a single one.
I do in fact use a separate cable with a special blind plug to achieve this:

https://www.conrad.de/de/p/wolfgang-war … 37382.html

No, there’s no analog audio running through USB.

This USB ground inter-connection functions as link from ground (devices) to earth (mother earth, really).
It’s equalizing the electric potentials and preventing adverse effects cause from floating ground potentials

We had some cases where headphones started to produce hum caused by such floating ground potentials.
AUDEZE headphones are are quite prone to this effect, and there are others to, even my nice Jecklin Float Electrostats.

Sieht aus wie ein Scherz, ist aber ein reales Angebot von Conrad Electronic.
https://www.conrad.de/de/p/wolfgang-war … 37382.html

Natürlich kann man sowas auch mit Resten aus der Wühlbox für 0 € basteln.
Würde ich hier aber einem Laien auf keinen Fall empfehlen immerhin findet sich in dem Stecker die lebensgefährliche Netz-Spannung.
Insofern gehen die 22 € eigentlich OK.