Ethernet Cables and Sound

Interesting, thanks for sharing.

Regards.

This is one I use. Cat6 to the Direct Stream DAC. I have found nothing better.

1 Like

Question for people in the know. So if I replace my modem to router network cableā€”router to wall outlet network cableā€”(100 feet CAT5e in wall not replaceable)ā€” then finally replace wall outlet to DSD network cable, is this going to make any difference, given the the CAT5e is in the equation?

Objectively - no, subjectively - depends on you and your environment

It depends on your environment, electrical, RF. If the data is getting there now without too many retries replacing the end patch cables shouldnā€™t make any difference. If you have poor terminations you may see some improvement. I would convert to optical at the last step if you can as this will guarantee true galvanic isolation

I donā€™t use the Bridge II rather I2S into my DS. My streamer is fed by my switch with optical, switch as a SFP slot and my streamer has a fiber PCI board

You prefer Cat6 over the Cat6a offering from BJC (for Audio)?

Both provide dead accurate performance. The Cat6 is already more than sufficient but I see nothing wrong with maybe future proofing. Itā€™s only a few bucks more to go 6A. I simply canā€™t hear a difference.

Thanks - any recommendations on a decent sounding switch - for use after your run of the mill cable modem\router?

For some great homework, go here;

1 Like

BJC ethernet cables are top drawer. No need to look elsewhere.

1 Like

For what it is worth, and I would suggest it is not worth much, I did an experiment the other day. I compared the sound quality of music played in in IRSV room as fed from WIFI vs. CAT 5 and could hear absolutely no difference. That systemā€™s pretty revealing
of just about everything, yet still no difference between the two into the Mac mini setup.

4 Likes

hey rszk, iā€™m no expert but i would be concerned about using 100 feet of cable (of anything) connected into my stereo electronics. stream that stuff man! T

Thanks for the input so far guys. I know for sure that even though Iā€™m running my system this way, the DSD with Bridge II is such an improvement over my previous Bryston BDA-2/BD-P-2 setup, Iā€™m really satisfied. Maybe ignorance is bliss on this subject.

So, I guess Iā€™ll poke a little harderā€¦

What is the consensus regarding network noise in the Hi-Fi networking/streaming context. Intuitively, it makes sense to me to take steps to isolate ā€œnetwork noiseā€ in my signal path. The current details about the digital highway my 1ā€™s and 0ā€™s travel on are noted in my member details.

Galenā€™s recent posts in this thread and related comments by others have prompted me to reconsider the effectiveness and need of such measures.

I think I am going to undertake some critical evaluation of some of my noise-reduction pieces and parts by systematically removing them and noting any perceived differences, when I have some time to set aside.

What say you, PSAers?

Cheers.

hey RsZkā€¦based on my previous experience with the Bryston BDA2 DAC, and now with a DSD, I would suggest the improvements you are pleasantly experiencing are a direct result of the PS Audio Direct Stream DAC replacing the BDA2. Although I havenā€™t had the enjoyment of trying the BridgeII, my sense is the BDP2 would hold itā€™s own, or maybe even better, against the BridgeII - and the DAC is the real deal maker.

1 Like

I have reviewed my back-and-forths with Galen, and my best guess is that he was talking about ā€œnoiseā€ from the perspective of the digital signal only -even though I initially mentioned the affect of ā€œnoiseā€ on the analog circuitry of the DAC. I am absolutely convinced that using isolation techniques improves the sound quality of Ethernet -> BridgeII (with no effect whatsoever on the bits themselves). As a matter of fact, reading posts of an SQ improvement of, say the Matrix, over the Bridge are irrelevant unless the comparison involves a galvanically isolated ethernet feed.

I am using a 100 Mbps T-Link Media Converter (because it can go directly into the Bridge and there is some talk about the Gbps ones being noisier). There was a noticeable improvement when this was inserted. It really took off, though, when I changed out the wall-wart for a 9 volt battery.

Note: I verified bit perfect transfer prior to listening tests under all scenarios.

1 Like

Interesting takeā€¦I too am pretty convinced of the benefits (my system has never been ā€œcleanerā€ or more resolving ā€“ in a good way). However, I think it would be informative to take the time and listen to a couple of system configuration variants with and without some of my ā€œadd onsā€ just to see how things sound taking a step or two backwards.

Thanks for the reply.

Scott

1 Like

+1 on the media converters. I run mine between the router/modem and a passive switch (they donā€™t work directly into Bridge II). A Pink Faun LAN Isolator is inserted in front of a WireWorld Ethernet cable feeding the Bridge, for good measure.

I donā€™t try to optimize my streaming experience. I accept it for what it is.

Here is where noise can be a problem, it isnt the effect on the Ethernet accuracy, that remains bit perfect except under the worst of sitations, it is the effect on the system clock. The system clock controls the digital interpolation coming out of the DA block. A bad clock introduces jitter and non linear quantization noise in the X and Y axis. This moves the closest, and most accurate, value with respect to time and amplitude.

This clock error is what impacts the sound. It is feasible that different Ethernet cables COULD introduce noise that MIGHT effect the system clock. A galvanic isolator might mitigate the issue if several things are true; the cable even has noise on it, it gets past the filters in the NIC card and it gets into the system clock circuit blockā€¦past the built-in noise safeguards in that critical block.

This isnā€™t the ā€œsoundā€ of an Ethernet cable. It, if ā€œitā€ is really happening, is the sound of the system clocks response to noise induced quantization errors. And the noise is assumed to be there, and introduced by the Ethernet block.

This could be proven with noise injection and detection on the system clock. Are we really after a real problem, or failing to define that we have one?

The most critical item in digital are clocks, and I doubt that these are left to the wolves and with no significant amount of noise shielding. That jitter value? The system clock is where it all begins.

Galen

1 Like

It would be interesting to hear what @tedsmith thinks about all this.
I use Sboosters on my Nucleus, HDD and Asus router, and wireworld Ethernet cables everywhere, then a Lumin with a chord aes into the DSD.
Are the Sboosters and wireworld Ethernet cables overkill ?