XMOS XU208 USB BRIDGES - THE LATEST GEN HAS ARRIVED!
Apr 29, 2016 at 10:49 AM Post #931 of 3,865
   
Yes, saw it. It's my opinion, he has his own. Different people different experiences, different environments and rigs.
But there is a lot of devices on the LAN all must be with GI. From NIC in PC server, through swithes, to NIC in SB.
But if one need could always:
1. Unplug the ethernet cable from SB (it's still playing)
2. Use wifi
3. Use medical ethernet isolation
and then compare the sound :)
I haven't found differences

Well I'll be try all those later today - I have 4 different CAT5e/CAT6/CAT7 cables coming.  One STP the other UTP.
 
   
Not to mention, twisted pair cable is, by design, built to reject noise and interference, otherwise it wouldn't be able to handle the bandwidth that it does for network use.

Yes - hence the good advice to avoid the 'flat' CAT cables.
 
Apr 29, 2016 at 11:06 AM Post #932 of 3,865
  Well I'll be try all those later today - I have 4 different CAT5e/CAT6/CAT7 cables coming.  One STP the other UTP.
 
Yes - hence the good advice to avoid the 'flat' CAT cables.

 
Great.
 
Theory: STP cable should be better than UTP in terms of canceling interferences.
But to use STP in LAN all cables have to be STP and all devices have to be properly grounded otherwise shielding will be acting as antenna. 
In consumer market network, with consumer market LAN devices it is hard to achieve. So better one uses UTP as mhamel said.
BTW: Do you know you were mentioned in stereotimes article?
 
Apr 29, 2016 at 11:11 AM Post #934 of 3,865
I was just posting my findings, you can put ferrites on ethernet cables, what happens ? the treble is reduced, should be impossible, but there you go.
 
I also tuned the tcp/ip settings using this http://www.speedguide.net/downloads.php, it also made a difference.
 
Apr 29, 2016 at 11:30 AM Post #935 of 3,865
   
Great.
 
Theory: STP cable should be better than UTP in terms of canceling interferences.
But to use STP in LAN all cables have to be STP and all devices have to be properly grounded otherwise shielding will be acting as antenna. 
In consumer market network, with consumer market LAN devices it is hard to achieve. So better one uses UTP as mhamel said.
BTW: Do you know you were mentioned in stereotimes article?

Yes - and he recommended thr PUC2 Lite...it was a good review.  Thanks for the explaination on the STP issue.
 
  BTW: one of our polish web magazines tested and reviewed yellowtec PUC Lite back in January. Found it better than BADA usb

I had seen some posts on CA by folks who had both - they prefered the PUC2 lite as well.
 
  I was just posting my findings, you can put ferrites on ethernet cables, what happens ? the treble is reduced, should be impossible, but there you go.
 
I also tuned the tcp/ip settings using this http://www.speedguide.net/downloads.php, it also made a difference.

Well in audio nothing surprises me - theory and test bench measurements are great - but in the end I trust listening...I see SR and AQ have $1000 CAT cables!  You may be right subsituting one set of problems for another.
 
I will open two new threads over the weekend - one on the 'PS Audio' LAN ethernet USB extender phenom and the other on AES67 Audio over IP.  The later actually really interests me alot...
 
http://www.musicdirect.com/p-58154-audioquest-rje-diamond-ethernet-cable.aspx?utm_source=googlepepla&utm_medium=adwords&id=91894008788&gclid=CPytqceUtMwCFc5hfgodggAPag
 
"Just as we discovered many years ago with S/PDIF cables, it is easy to damage the information package. To avoid corruption (jitter), AudioQuest offers new models of high-performance Ethernet cables featuring better metals, critical signal-pair geometry, Dielectric Bias-System (DBS), Noise-Dissipation System—all proven techniques for delivering superior digital audio."
 
And around and around we go....
blink.gif

 
Apr 29, 2016 at 12:40 PM Post #939 of 3,865
 
He took my order on Ebay April 8th - one of the first then said he had none!  Here was his message on shipping April 14th:
It is now the April 26th no F-1.  Worse took two weeks of badgering him to get the driver link and passcode.  Freeshoppingchina.com it took 12 hours.
 
Shenzen is a disaster!  Avoid!  Now I hear the boards he sends do not have the i2s pins attached!

 
Hi. Is this that seller or a better one?  
 
http://www.aliexpress.com/store/product/SINGXER-F-1-XMOS-USB-digital-interface-board-XU208-chip-High-end-U8-upgraded-version/601461_32648384788.html
 
Apr 29, 2016 at 12:43 PM Post #940 of 3,865
  Yes - and he recommended thr PUC2 Lite...it was a good review.  Thanks for the explaination on the STP issue.
 
I had seen some posts on CA by folks who had both - they prefered the PUC2 lite as well.
 
Well in audio nothing surprises me - theory and test bench measurements are great - but in the end I trust listening...I see SR and AQ have $1000 CAT cables!  You may be right subsituting one set of problems for another.
 
I will open two new threads over the weekend - one on the 'PS Audio' LAN ethernet USB extender phenom and the other on AES67 Audio over IP.  The later actually really interests me alot...
 
http://www.musicdirect.com/p-58154-audioquest-rje-diamond-ethernet-cable.aspx?utm_source=googlepepla&utm_medium=adwords&id=91894008788&gclid=CPytqceUtMwCFc5hfgodggAPag
 
"Just as we discovered many years ago with S/PDIF cables, it is easy to damage the information package. To avoid corruption (jitter), AudioQuest offers new models of high-performance Ethernet cables featuring better metals, critical signal-pair geometry, Dielectric Bias-System (DBS), Noise-Dissipation System—all proven techniques for delivering superior digital audio."
 
And around and around we go....
blink.gif

 
 
Without trying to get into a debate about cables or subjective/objective... As for expensive LAN cables - I think that if you get a tested/certified Cat6a cable it will be a better option than dropping large amounts of money on the AudioQuest cable.  (http://arstechnica.com/gadgets/2015/07/340-audiophile-ethernet-cable-gets-a-marginal-pass-on-the-test-bench/)  
 
While the cheap generic cables definitely don't meet spec, the AudioQuest cable barely meets it, while a properly constructed and tested/certified Cat6a cable will outperform it. A 10ft Cat6a rated/tested ethernet cable from BJC is $21. They even include the test report to show the cable's performance. Cat7 is un-necessary if you're not using 10GbE.
 
We have to remember here that the audio data is being encapsulated into network packets using a protocol that is designed to be as error-free as possible, including checksums, retransmits, etc. If you have a switch that can be monitored, it's easy to determine if there is packet loss. You can also check stats on the source. Diagnostics/reporting on the end device will vary based on the type of device/capabilities. This underlying structure is part of what makes Ethernet & TCP/IP great transports for audio data and why so many channels of bit perfect uncompressed audio can be handled simultaneously in both directions, even at high bitrates.
 
That alone shows that the "directional" designation on the AudioQuest cable is complete nonsense. Ethernet is bidirectional. The TCP packets do not care what they contain in terms of audio data.  If there was some kind of signal degradation or loss based on the direction of the signal, it would not be suitable for audio use, especially in a pro audio scenario where something is being mastered. Even more so when it is handling potentially hundreds of high bit rate streams in both directions to multiple devices.
 
Apr 29, 2016 at 12:49 PM Post #944 of 3,865
   
 
Without trying to get into a debate about cables or subjective/objective... As for expensive LAN cables - I think that if you get a tested/certified Cat6a cable it will be a better option than dropping large amounts of money on the AudioQuest cable.  (http://arstechnica.com/gadgets/2015/07/340-audiophile-ethernet-cable-gets-a-marginal-pass-on-the-test-bench/)  
 
While the cheap generic cables definitely don't meet spec, the AudioQuest cable barely meets it, while a properly constructed and tested/certified Cat6a cable will outperform it. A 10ft Cat6a rated/tested ethernet cable from BJC is $21. They even include the test report to show the cable's performance. Cat7 is un-necessary if you're not using 10GbE.
 
We have to remember here that the audio data is being encapsulated into network packets using a protocol that is designed to be as error-free as possible, including checksums, retransmits, etc. If you have a switch that can be monitored, it's easy to determine if there is packet loss. You can also check stats on the source. Diagnostics/reporting on the end device will vary based on the type of device/capabilities. This underlying structure is part of what makes Ethernet & TCP/IP great transports for audio data and why so many channels of bit perfect uncompressed audio can be handled simultaneously in both directions, even at high bitrates.
 
That alone shows that the "directional" designation on the AudioQuest cable is complete nonsense. Ethernet is bidirectional. The TCP packets do not care what they contain in terms of audio data.  If there was some kind of signal degradation or loss based on the direction of the signal, it would not be suitable for audio use, especially in a pro audio scenario where something is being mastered. Even more so when it is handling potentially hundreds of high bit rate streams in both directions to multiple devices.


Many thanks - completely agree!
 
Do you have a link for that cable.  I don't really trust these Amazon cable sellers.  Bought this one from Lowes yesterday (just in case the other's don't arrive on time), you'd think for $479 Startech could provide a quality tested CAT6 cable...geeze.
 
This was $3.98 for a CAT5e 7ft at Lowes
http://www.lowes.com/pd_737251-1571-AC3507-WH-V1_2z8vi__?productId=999930134&pl=1&Ntt=cat5e+cable
 

Users who are viewing this thread

Back
Top