Ethernet vs usb

[jon2020]

Hi everyone,

 

What are your thoughts on the following statement regarding ethernet vs usb?

 

Reference-grade Roon with the Antipodes DX Gen 2 | DAR__KO

 

"According (to) our man in New Zealand the argument against Ethernet (and for USB) runs like this: compared to Ethernet, USB has the potential to carry more noise from the server to the DAC but generates less of its own noise inside the DAC (think: receiver chips). Ethernet remains a good strategy with a noisy server. However, it can be bested by USB when a low noise server with a decent clock is present."

[matthias]

AFAIK, this is the statement of the team behind the Sound Galleries Music Server as well.

 

Matt

[Astralark]

I see Merging Revenna protocol has more advantages than USB solutions because of precise clock accuracy.

 

 

Sent from my iPhone using Computer Audiophile

[Astralark]

I think this is the problem of cable. You could not find a high price tag HDMI, USB or Ethernet cable. The connection has no problem. Ethernet cable has less shielding layers (and less research and designs for audiophile applications).

 

 

Sent from my iPhone using Computer Audiophile

[Cebolla]

Of course, all of this is a non issue with network file streaming such as UPnP/DLNA. In that case the network is not used to carry the realtime digital audio signal, since audio file decoding and playback occours at the receiver side of the network, not the sender side.

[jon2020]

AFAIK, this is the statement of the team behind the Sound Galleries Music Server as well.

 

Matt

 

Have checked out their website and the thread here at CA - cool stuff!

[jon2020]

I see Merging Revenna protocol has more advantages than USB solutions because of precise clock accuracy.

 

 

Sent from my iPhone using Computer Audiophile

 

The Merging Ravenna protocol is not quite mainstream yet for home computer audio systems - looks promising.

[jon2020]

Of course, all of this is a non issue with network file streaming such as UPnP/DLNA. In that case the network is not used to carry the realtime digital audio signal, since audio file decoding and playback occours at the receiver side of the network, not the sender side.

 

You are saying that DLNA network file streaming via ethernet/WiFi would be superior to a direct usb connection as it does not involve cabling.

This would bring us to the issue of "a low noise server with a decent clock" being better via usb.

[diamondblack]

The caveat "low-noise server with a decent clock" is a very tall order if you are looking for something really decent.

[jon2020]

The caveat "low-noise server with a decent clock" is a very tall order if you are looking for something really decent.

 

That may be true but we can take baby steps to get there , like isolating usb power from the audio signal, regenerating and reclocking the usb signal before the dac input.

Though my iFi usb 3.0 provides these baby steps in one box, it is nevertheless a great gigantic leap in SQ from before.

Others have reported very satisfying results with the Regen or Intona or other products of such ilk.

 

And specialty music servers with low noise and decent clocking in the market today have produced very, very good results.

[firedog]

I think it is one of these audiophile generalizations that turns to dogma.

 

It all depends on the specific implementation, not the format or type of HW.

[jon2020]

I think it is one of these audiophile generalizations that turns to dogma.

 

It all depends on the specific implementation, not the format or type of HW.

 

Certainly, dogma should be eschewed and it would follow from your statement that a well-implemented usb connection can sound as good as a well-implemented ethernet connection.

 

This would be true in that different hardware manufacturers may recommend one connection over the other for the best SQ in a particular product manual.

 

Just wondering if the theory behind the statement by Antipodes is grounded.

[DavidL]

I think it is one of these audiophile generalizations that turns to dogma.

 

It all depends on the specific implementation, not the format or type of HW.

 

Agreed but my impression is that more design work by the manufacturer and effort with hardware setup by the buyer is necessary to achieve as good SQ via USB as via ethernet.

[jon2020]

Agreed but my impression is that more design work by the manufacturer and effort with hardware setup by the buyer is necessary to achieve as good SQ via USB as via ethernet.

 

Would prefer manufacturers come up with THE complete solution at point of sale so that consumers can dispense wirh the add-ons and extra wires.

With USB, it is easier for hardware manufacturers to control the variables but with ethernet, the home network is much more individualised.

It would appear that a direct usb connection is more predictable than an ethernet set-up for both manufacturers and consumers while consumers have to work much harder with the ethernet option.

[jon2020]

Here is another perspective :-

 

Audio Ramblings - the PS Audio BHK Signature Preamplifier and DirectStream DAC - Positive Feedback

 

"Using Lightning DS via the Aries from Auralic (USB) versus MConnect (a free app to control files via DLNA) to send files to the DS presents USB as being more detailed and resolving as opposed to a richer, warmer, more analog like via network. Which is right, which is better? Both. Can live with either simply on musical terms, ....."

[soundquest]

Would prefer manufacturers come up with THE complete solution at point of sale so that consumers can dispense wirh the add-ons and extra wires.

With USB, it is easier for hardware manufacturers to control the variables but with ethernet, the home network is much more individualised.

It would appear that a direct usb connection is more predictable than an ethernet set-up for both manufacturers and consumers while consumers have to work much harder with the ethernet option.

 

+1

Either way [usb or ethernet], it would be best to have a complete solution with (hopefully) fewer wires and individual components as each one is likely to introduce power supply noise and (potentially) issues in software compatibility. Further, it could offer greater simplicity for consumers. I love that sound quality options for computer audio are evolving, but in some regards I miss the plug and play simplicity of a cd player.

[Cebolla]

Of course, all of this is a non issue with network file streaming such as UPnP/DLNA. In that case the network is not used to carry the realtime digital audio signal, since audio file decoding and playback occours at the receiver side of the network, not the sender side.

You are saying that DLNA network file streaming via ethernet/WiFi would be superior to a direct usb connection as it does not involve cabling.

This would bring us to the issue of "a low noise server with a decent clock" being better via usb.

No, what I'm saying is that network file streaming provides an oportunity to eliminate the need for an external connection between the audio file player and DAC (ie no need at all for the likes of USB audio, network audio via Ravenna, etc).

 

What you do is you build the network audio file player/renderer/streamer into the DAC. You have even linked to an article describing one such device yourself (so seemingly understanding the significance of that "other perspective"):

Here is another perspective :-

 

Audio Ramblings - the PS Audio BHK Signature Preamplifier and DirectStream DAC - Positive Feedback

The PS Audio DirectStream DAC has a direct/internal I2S connection to its network audio file player module, the PS Audio Bridge II.

[jon2020]

No, what I'm saying is that network file streaming provides an oportunity to eliminate the need for an external connection between the audio file player and DAC (eg no need at all for USB audio, network audio via Ravenna, etc).

 

What you do is you build the network audio file player/renderer/streamer into the DAC. You have even linked to an article describing one such device yourself (so seemingly understanding the significance of that "other perspective"):The PS Audio DirectStream DAC has a direct/internal I2S connection to its network audio file player module, the PS Audio Bridge II.

 

Thanks for your explanation.

Now that I understand the more technical perspective, it is interesting to note that from the listener's perspective at PF, either method is good by him in musical terms.

That is quite reassuring in that we dont have to pursue one connection over the other. As has been said many times before in all things audio, sounding different does not necessarily mean better.

[Astralark]

Keep in mind there must be one solution that offers the best SQ. Audio world never have a gold standard and different manufacturers set different standards which didn't help, so we have XLR, BNC, Toshlink, HDMI, USB etc. Oh, please, In order to work on all this connections. An audio device have to prepare all this input/output (to have a good market sell). But, each input/output standard increase the complexity of circuit design that degrade SQ.

 

 

Sent from my iPhone using Computer Audiophile

[plissken]

Hi everyone,

 

What are your thoughts on the following statement regarding ethernet vs usb?

 

Reference-grade Roon with the Antipodes DX Gen 2 | DAR__KO

 

"According (to) our man in New Zealand the argument against Ethernet (and for USB) runs like this: compared to Ethernet, USB has the potential to carry more noise from the server to the DAC but generates less of its own noise inside the DAC (think: receiver chips). Ethernet remains a good strategy with a noisy server. However, it can be bested by USB when a low noise server with a decent clock is present."

 

I think their man in New Zealand sounds confident enough to sit down for a bias controlled test.

 

I know Ethernet is immune to inductive noise coupling up to 30Mhz. This is according to a T.I. paper. And I don't know any creature alive that can hear the base 25Mhz clock Ethernet uses our any of the harmonics even if they could theoretically be measured in our range of hearing. I wouldn't place my money on him.

[plissken]

I think this is the problem of cable. You could not find a high price tag HDMI, USB or Ethernet cable. The connection has no problem. Ethernet cable has less shielding layers (and less research and designs for audiophile applications).

 

 

Sent from my iPhone using Computer Audiophile

 

HDMI, USB, Ethernet are data cables. They can't be designed for Audiophile applications.

[plissken]

You are saying that DLNA network file streaming via ethernet/WiFi would be superior to a direct usb connection as it does not involve cabling.

This would bring us to the issue of "a low noise server with a decent clock" being better via usb.

 

This is testable as I have a PCIe EMU 1212M sound card. Could also connect an 'Audiophile' DAC and do direct comparisons.

[Astralark]

HDMI, USB, Ethernet are data cables. They can't be designed for Audiophile applications.

 

However we has those inputs and even HDMI I2S.

 

 

Sent from my iPhone using Computer Audiophile

[plissken]

That may be true but we can take baby steps to get there , like isolating usb power from the audio signal, regenerating and reclocking the usb signal before the dac input.

 

If your DAC doesn't do that for you then you purchased a poorly designed DAC. There is no audio signal on the USB cable. Only data. The DAC takes the data and reconstructs the audio signal.

[plissken]

Keep in mind there must be one solution that offers the best SQ. Audio world never have a gold standard and different manufacturers set different standards which didn't help, so we have XLR, BNC, Toshlink, HDMI, USB etc. Oh, please, In order to work on all this connections. An audio device have to prepare all this input/output (to have a good market sell). But, each input/output standard increase the complexity of circuit design that degrade SQ.

 

 

Sent from my iPhone using Computer Audiophile

 

They tend to be tools that solve different problems.

 

XLR is good for long run and noise immunity. HDMI,USB, Ethernet are data systems with buffering that are more resistant than SP/DIF, TOSLink, AES/EBU to timing errors.

 

If you have 24/96 or lower audio, in a home environment you need nothing more than TOSLink. The issues of timing really aren't a problem in a direct connected home environment and it gives you EMI/RFI immunity.