DirectStream: USB Sound Quality better than Bridge II

I recently upgraded to the DS, and needless to say, I’m am thrilled with it, but I am surprised that I am getting decidedly better sound quality from USB. I have hundreds of hour of burn-in with music and burn-in tones from Track 8 of the Chesky Test and Burn-in CD using both the USB and Bridge inputs.

I am using Uptone Audio’s USB ISO Regen and Ultracap LPS-1 galvanic isolation power supply, which has greatly improved the performance of USB. At worst I would have expected the Bridge II to sound equally good, but it is clearly not to my ears. Less bass, thinner, harsher high frequencies. The Bridge II volume setting needs to be lowered to 94 to equal to spl of the USB input.

I am running JRiver MC 24, so all variables may not be the same because JRiver requires use of DLNA to play to the Bridge II, while the USB connection is direct.

Can others confirm or contradict my results or offer any thoughts on why I’m getting this result?


What is your Ethernet source?

Agree usb sounds much better. I think it’s due to rfi and other nasties on your local network that is not / can not be? blocked out by the Bridge. I certainly got improvements when moving from BTHome Hub router to a google Wi-fi one for example (before I discovered usb)

I am using a TP-Link N300 router and generic ethernet CAT 6 cable direct connection.

Your observations make complete sense. Routers are not clean and the power supplies are not audiophile grade at all. You are passing tons of noise down the Ethernet cable.

Everything I’ve read seems to declare the superiority of ethernet over usb because of all the noise associated to usb. Based on your comments, it seems that ethernet suffers the same issues.

I’ve found USB even without galvanic isolation to be more alive vs. Ethernet into Bridge II. With an isolator, USB is much better for me.

Ethernet absolutely passes noise especially when typical network devices are used that are very noisy. This is why people observe differences powering network devices with linear power supplies and using FMC’s.

The Uptone EtherRegen is probably going to be a big step up for Ethernet. It’s basically an audiophile Switch designed from the ground up to be clean for audio.

Whaaaat? Check your settings. Something is off with JRiver settings if your USB sounds better than the B2.


Hmm…running from (Pink Faun galvanic isolated) Melco N1ZS10 to DSsr with both Audioquest Diamond RJ/E and USB streaming Qobuz, Tidal and on the Melco stored files sounds GREAT over BridgeII, but is defeated by USB (with exception of the Tidal Masters…).
My PWT is defeated by both :slightly_smiling_face:

Well, I guess there is no consensus on which is better.

I am going to look into the Uptone EtheRegen when it hits the market. The LPS 1.2 is getting a lot of
positive feedback for its improved sound over the LPS-1. I have great respect for designer John Swenson.

Just two weeks ago I measured the DSjr output level in the JRiver DLNA mode. Finally, I found that DLNA lost 2.5dB of Vrms compared to USB direct connection. -2.5dB is approximately equal to a 30% reduction (2.5Vrms to 1.8Vrms), which is the 5 scale volume in DS/DSjr.

Therefore, objectively speaking, it is recommended to retest at the same output level - to avoid the psychological gap induced by loudness differences, the human auditory system is very susceptible to this.

The USB has always sounded better to me in my rig than the B2. In part it may have to do with which you make the effort on with regard to noise. I had gotten an ethernet USB galvanic isolation solution going prior to Paul coming out with the LANRover. So the USB was more optimized early on, and I didn’t particularly have a need for the B2, so haven’t been motivated to optimize that input chain. Some folks also like brighter sounding options, have a need/want to use the B2 (direct streaming from a service, MQA playback, etc.)

1 Like

I did some additional testing after reviewing the replies to this post. Woot’s comments reinforced my thinking that I might not have set JRiver correctly, so I went back into JRiver DLNA server settings. Changing the output mode from “Original” to “Specified output format only when necessary”, which allowed me to access the DSP studio and set the DSP output settings to match what I had for the USB input.

And Eureka! both USB and Bridge II sound so much alike that I can’t hear a difference, and the both sound fantastic! SPL output levels for both were also now equal.


You should see an uptick in performance by utilizing an LPS on your switch and using aftermarket ethernet cables like the SOtM DBL-CAT7 or even something like the Supra Cat 8’s.

It’s been so long since I set up my JRiver that I can’t remember how anymore. Short from browsing through the JRiver forums, are there any instructions/tutorials on the PS forum that show how to accomplish this?

Here is one link from the PS Audio “How To” link, but it’s a bit out of date: (

I also got help with Bridge ii setup with JRiver from the Support team, and they offer excellent help, advice, and patience.

1 Like

Much appreciated!!

Thank you . . .

usb better for me also ( i use nas ans jplaystreamer). But the choice of the ethernet cable on the pc is important also…

New poster here, had a Digital Link III as my main DAC for about a decade and loved it. Recently acquired a DS Jr. and so far so good, with a few usability issues (pertaining to the crummy MControl app).

I understand that it’s impossible to account for every variable in a given user/commenter’s system, so it wouldn’t be prudent for anyone to say which SHOULD sound better between the Bridge II and USB, but I really wish someone from PS Audio would step in here and clear a few things up:

  1. For a given setup and isolating as many variables as possible, what MIGHT or COULD sound better, or DID sound better during PS Audio testing - for example the same FLAC file stored on a laptop (not NAS) over a user’s Ethernet Bridge II connection vs. the same file direct to the PS USB input - which has the POTENTIAL to sound better? Feel free to frame that within any engineering perspective necessary - does one maintain better signal integrity and does another somehow (as bizarre as it would be) drop significantly more bits/packets or introduce more “noise” into the system (see below)?

  2. Users seem to be confused or something about “noise” via Ethernet and USB. It looks like people are divided down the middle over whether the (or more) “noise gets into” the DAC (or Bridge II) via Ethernet vs. USB. For one thing these are digital signals - so I’m not sure what “noise” is referred to unless somehow you’re talking about a type of jitter. Claiming that “noise” gets into the PS Audio product and that it affects sound quality is basically saying that PS Audio engineers did not do their jobs and their DACs are somehow not rejecting - OR - that the DAC is actually downright processing this “noise” into the audio signal and that it’s degrading sound quality as a result.

As an engineer, albeit not a network or computer engineer, I am pretty sure that “noise” is a meaningless concept in this context and that ONLY jitter matters when we’re talking about the transmission of a digital signal over a conductor/cable which is within spec (resulting in no dropped packets or bits or interference with the checksum function).

Soooooooo could a PS Audio engineer please put to bed the constant talk about “noise” over either USB or Ethernet? And - If anyone who is talking about “noise” actually means “jitter” then please elaborate for the rest of us - up to and including how you might think that “noise” is actually adding to or creating “jitter”.