USB output vs Streaming Output

Using Audirvana 1.14.0 on a Macbook Pro. Can output 2 ways to a DAC: either USB directly off the Macbook into the DAC, or via DLNA/UPnP to DAC via Ethernet cable into the DAC.

Any advantage to either of these methods in terms of sound quality?


The best way to make this determination is to listen to each type of connection for yourself on your system…

:notes: :eye: :headphones: :eye: :notes:

Listening for yourself is indeed probably the best way (despite us not being terribly objective about what we’re hearing). A word about why I think so:

USB is probably going to be more reliable. UPnP/DLNA has the potential for better sound quality depending on a number of factors. So listen to each and see whether you hear anything sufficiently better that would make you want to have what will possibly be the somewhat more fiddly UPnP/DLNA connection.

1 Like

Thanks very much. Agree on the fiddly part of UPnP. Since I’m sitting next to the DAC (literally) I will buy a high-quality USB and see how it sounds.

I should have rephrased my question. It really had to do with the origin of the digital stream from Audirvana. Is the stream sent to the USB port on the laptop and then to a DAC the same as the stream arriving from the router over an Ethernet cable (or possibly WiFi) to a DAC? Put another way, what is the role of my laptop in an Audirvana session which plays music from both a local library (such as a NAS) and, for example TIDAL?

Sorry for the confusion–I guess I’ve never understood how streaming software such as Audirvana or Roon sends the music to a DAC or other UPnP device.

1 Like

The short answer is yes, the bits sent would be the same, because some version of PCM or DSD is the only language DAC chips speak. (“Some version of DSD” is, technically speaking, some version of PDM, but that’s a little beyond the original question.)

@DWR The answer is found in the simplicity of the transmission path… In both cases the Mac manages the Ethernet bus responding to Audirvana request…

The Mac manages the USB bus and that signal is sent via your USB cable, therefore there is little potential for signal distortion from RFI and EMI when using a good quality USB cable designed for audio transmission especially using a short USB interconnect (best).

In the case of Ethernet transmission in your case… These cables are generally not shielded and susceptible to RFI and EMI interference of the signal integrity…

The digital audio packet data sent to the DAC is ischronous and there is no error correction of the signal being received by the Input architecture of your DAC, so signal integrity is imperative for the highest quality playback experience… The only “bits” are those found in the original recorded file, the signals sent through the system are analog electric pulses the system components react upon… So these pulses are susceptible to inducted noise into the component architectures and cabling that precipitate jitter on the clocking employed in digital electronics like your DAC.

:notes: :eye: :headphones: :eye: :notes:

Thanks. The proximity of the DAC to my listening position suggests high-quality USB. The length is 2M at most.

While I’m at it, here’s another question. While I’m sitting at my laptop with an Audirvana session going, and listening randomly to both local (NAS) and Tidal files, to what device does Tidal stream to? My laptop, which then sends it back over the LAN to the UPnP DAC, or does it send it directly to the DAC?


Audirvana is the player of the Tidal file and playlist manager and does all of the processing in the application and this data is sent to the output buses of the Mac…

Thanks very much. :grinning:

1 Like

I have a very nice well shielded inexpensive Ethernet cable from Ghent Audio (their JSSG360 model). Also, Ethernet cables have transformers at either end providing a level of galvanic isolation USB cables lack. Or if you’d like complete galvanic isolation, you can go with optical Ethernet, which is quite inexpensive for some very good compact equipment. (See MikroTik.)

So again, many different options and all just down to your budget and the way you like to do things.

Well… for sure what you are describing will work… In the case of a 2 meter run the better the cable employed here, the potentials for waveform distortions are lowered, however, the shortest cable is best… not all cable designs are equal in their ability to reject noise and maintain the data-packet signal integrity without phase/timing errors… Galvanic isolation is useful if re-clocking is employed… remember we are trying to maintain the integrity of the data-packet signals… less is better, however if there are ground-loop problems from poor playback system power/ground/earthing strategy, galvanic isolation solves ground-loop problems… Many new DACs and streamers employ some form of galvanic isolation in their design topologies.

1 Like

@DWR … If you are so inclined, and would like to get the most from the USB output to your DAC, I recommend using the iFi Audio iPurifier3 USB Audio and Data Signal Filter… This will provide noise-filtering and re-clocking… It also allows you to use a USB 3.0 cable… the USB 3.0 cable separates the data lines from the request signaling lines and power lines for higher throughput/speed and data-packet integrity…

:notes: :eye: :headphones: :eye: :notes:

That’s a great suggestion. I will check it out.


I have a iP3, and tried it with a half a dozen DACs.

What I can report is that a good DAC with a good USB cable does not need this gadget.
Here’s a good example:


I have an Audioquest Jitterbug. Roughly the same device as the Ifi purifier. My experience is that it even made the sound quality worse on some of my DACs (much ‘flatter’ or ‘duller’). On other DACs I could not hear a difference at all. I have no measurements of this, but in this case I trust my ears.

Maybe with a poor quality DAC it will give some improvement, but I even doubt that.

@DWR, trust your ears and do your research before buying ‘esoteric’ devices like this.


I think that the reason for this is that these gadgets reduce the bandwidth.

1 Like

@Cloclo @AndyLubke @Jud @DWR
The rationale for recommending the IP3 is in this particular application where @DWR is considering a 2 meter in length USB cable… Also because he then can use a USB 3.0 cable…

This Audio Science Review post does not specify the length of the USB cable and does not specify if they tested using a USB 3.0 cable… I have experience with a 2 meter in length very good USB 2.0 cable and it did benefit from re-clocking and noise reduction…

My personal current set-up requires a short interconnect scheme, however, I’m employing USB 3.0 from a very highly filtered USB 3.0 Interface installed in my Sonnetech Thunderbolt 3 PCI chassis (Elfidelity AXF-107 Ultra PCI Express 1x/16x Power Supply Filter) that is connected to my MacBook Pro via USB 3.1 because good quality TB3 cables didn’t perform as well as the WireWorld USB 3.1 cable I now use… the output of the USB 3.0 interface installed in the filtered PCIe chassis, feeds iGalvanic 3.0 via USB 3.0 (WireWorld) that is feeding my DAC with a very short USB 2.0 cable (WireWorld)

A simple USB 2.0 interconnection has problems in this scenario… no matter how short the cable, primarily because of the noisy USB/Thunderbolt interface of the Apple computer and the nature of the signaling being shared on the data lines of the USB 2.0 cable protocol.

:notes: :eye: :headphones: :eye: :notes:

I had the same experience with the Jitterbug. And I don’t know that I trust iFi in the noise reduction area, as they sell supposedly low noise power supplies that in fact produce a fair amount of noise.

I have used an UpTone Audio IsoRegen for years and wouldn’t be without it. But all this is getting into tweaking territory beyond the scope of the original question.


Maybe… Why was the question posed initially?

So you could talk about your favorite tweaks? :wink: