What I am saying is noise can’t enter the DAC through the CAT cable. So if the ethernet cable is making an audible difference its likely not from noise entering the DAC from the cable. As for power cables, I made some cables to test shielding and quickly discovered that shielding did more harm than good. Apparently the shield clamps the AC magnetic field which in turn clamps the voltage, such that the cable cannot deliver proper voltage.
1 Like
I’m not sure it’s possible to make an argument that noise ingress is absolutely impossible.
There are too many experts on this topic that say this is an issue for this to be simply dismissed (especially if our ears say something else). Meridian went on record to say that they didn’t realise the influence ethernet had on audio sound quality and designed a replacement input card for their processors as a result.
As any true debate should hear both sides of the argument, I’ll quote from @Alchemist’s post where he references a comment based on Swenson’s thoughts on this topic (who designs high-end DAC’s for a living):
Here’s a link to a copy of Swenson’s full paper: https://cdn.shopify.com/s/files/1/0660/6121/files/UpTone-J.Swenson_EtherREGEN_white_paper.pdf?v=1583429386
Sorry Jeremy, but there are so many gaping holes in Swenson’s paper! I posted about some of it awhile ago (before the thread got shut down 
)
In that case, absolutely no ethernet switch can make a difference and there is absolutely no point in using an optical rather than copper network cable. This is quite an assertion and well out of line with the experience of many here.
Or there is something else going on. An inferior cable or switch or wiring can cause packet lose. At some point the lose will be audible. Consider streaming video and the picture goes bad or buffers, this is packet lose (or in the case of Paramount Plus bad service). Another possibility is some audio grade cables use a small PCB to attach the connector to the wires, doing this removes the noise immunity built into the ethernet standard by eliminating the benefits of twisted pair wires and differential drivers. The extra weight of a heavy cable may make a less than ideal connection which can cause packet loose - keep in mind these RJ45 receptacles are designed for standard cable and were not designed to support heavy cables. When I used to build my own cables for our PC network many would pass a continuity test ok but not work - the connections just weren’t good enough for CAT 6 speeds. I believe there is a very good reason the people at DCS recommend a simple CAT 6 cable that passes CAT 6 tests.
indeed. Perhaps we should read this (again):
2 Likes
@jandersonhill Jeremy, I presume the response you posted in the other thread was meant for this?
In any case, while it might appear that Swenson knows what he’s talking about, as I mentioned above, his white paper has huge gaping holes in it (which will be obvious to anyone with even basic Ethernet Switch design & engineering knowledge).
I’m not really interested to debate this on their sponsored sub-forum at Audiophilestyles. Besides, Audiophilestyles is really a joke these days, full of subjective mumbo-jumbo, it’s only good for a laugh. 
That’s OK, I feel the exact same way about Audioholics or, worst of all, Audio Science Review.
I’m often convinced no one at ASR has an audio system at all, they just test components and imagine what they sound like based upon the measurements because, after all, your ears are not reliable. 
3 Likes
Dodgy cables are an entirely different and separate topic. They may interfere with the 0s and 1s as you suggest, through packet (frame) loss or whatever. This is an exceptional use case and should not drive decisions about using optical vs copper or using a switch or not.
CAT 6 cable is digitally reliable.
That is the point I am making. High end audio cat cables that can’t pass a CAT 6 certification test may be the problem. They may drop packets and force the DAC to use error correction, which could be audible. So instead of spending more on flighty audio-grade CAT cables, perhaps spend $10 on an IT grade certified CAT cable. Independent labs (sorry I don’t have the link) have proved that some very expensive audio grade cables fail certification tests. I have used flaky cat cables on PCs and the effects are strange - work sometimes, slow other times, and sometimes don’t connect at all. Keep in mind just because audio doesn’t use a lot of bandwidth, the bit speed and timing does not change. So a poorly designed cable that fails testing has a very good change of forcing the switch/router/dac to resend packets. If these resends are too often the DAC will not be able to fully create the audio. I assume (but do not know) that the DAC will use error correction to guess what the lost data is. The guess will likely be wrong, the music still will play, but it will be wrong and you may hear the difference.
To summarize my comments on this thread, I don’t think noise is causing cat cables to make an audible difference, I think its packet error.
Or pseudo effects for some of us. We hear what we want to hear and I believe we are all guilty of selective hearing. And if your over the age of 50, and like me, our hearing changes from day to day. For me its sinus issues - some days my ears are clear, some days they are noticeably clogged and some days I don’t know. I am quite aware that my hearing response changes day to day, sometimes for better, and sometimes for worse. Oh it would have been nice to have the system I have now when I was 30.
You’re suggesting that some boutique (fancy) cables might be causing packet error? Wow, that’s quite a claim. Packet error would introduce random sonic effects rather than anything a manufacturer could control, and ethernet protocols would re-transmit the packet if anything went wrong so the packet error wouldn’t get to the streamer.
You’re going to have to name brands/models to back that up.
I presume you don’t actually mean CAT cables anyway as they’re standards-compliant by definition. You might mean “network cables” which is what some boutique guys call their non-CAT-compliant offerings.
Noise remains far more likely as a cause of any sonic differences.
I found this testing. While it is out of date, I think the claims speak for themselves. I would expect a $350 ethernet cable to pass with flying colors, not squeak by. What concerns me with this product is the use of a PCB for the wire-to-RJ45 connection. The point of twisted pair it to keep air born noise out. But exposed PCB traces defeats the purpose. My design experience with RF radios is that it only takes a small opening to create big problems. We had to completely cover a RF radio circuit with a shielded can and position the antenna as far from critical circuits as possible.
Keep in mind we are dealing with 2 type of noise entering a cable - noise placed on the wires by the electronics - the DAC, streamer, router or switch. I would expect IT grade equipment to not have an issue as they are designed to transmit over 100 meters. I’d avoid cheap switches and routers. The other source of noise is air induced noise radiated from our DACs, TVs, RF signals, etc. The twisted pair with differential drivers is what stops this noise. But, as I stated above, the PCB would negate the advantage of the twisted pair by creating an opening for noise to enter. Its just bad engineering. When it comes to 10GBS and 100GBS transmit speeds I trust the big guys who make 10s of millions of cables per year and not a small audio company who make a few dozen.
As for the listening test mentioned in this article, those can be dismissed as immaterial. We don’t know what equipment was used nor the experience of the listener. But I trust the Fluke test as if they weren’t true I suspect this would of brought on the attorneys who would of had this site taken down.
Again, I am not saying people are not hearing real differences in ethernet cables, what I am saying is its likely not caused by noise getting from the cable into the DAC. Its caused by something else.
3 Likes
Review:
“Then we took the cable we didn’t use on stage and gutted it, exposing its innards. We found an interesting mix of high craftsmanship (a thick polyethylene sheath, genuine S/FTP construction) and corner cutting (masking tape, unterminated shields).”
I’m not sure this is the right place for this post but here goes anyway.
We just got an independent ISP who has installed fibre cables in our street. I have had the connection installed. Everything works great except my Rossini Player which Mosaic cannot see. It worked fine with the old ISP equipment which was using copper cables.
The Rossini manual says some routers supplied by ISPs cannot handle the dCS connection and recommends an aftermarket router instead. The router supplied by the new ISP is Huawei HG8245Q2. The previous one was a standard BT Hub.
Does anyone have any suggestions.
Thanks.
Some ISP routers don’t support the mDNS protocol, which allows devices on a local network to find other devices on the same network, which could be why Mosaic can’t see the Rossini.
Not knowing your network layout, the best starting point (and possible solution) is to test this using a half-decent ethernet switch after the Huawei router (connecting all the local ports into this switch and connecting the switch to the router using an additional ethernet cable).
If it worked before the change in router, it’ll probably work fine again after making this change 
1 Like
Unless you have already tried, try rebooting the router and then restarting the Rossini.
2 Likes
Simon, just to clarify the advice from @hifimckinney you need to get the Rossini to reset. To do this switch off the unit from the rear panel. Using the front panel button alone will not achieve this. Wait a minute or so before switching back on.
If this does not work I suspect that as @jandersonhill suggests the Huwei router may be incompatible and not support the mDNS protocol. Please let us know how you get on as this helps build our knowledge base.
Simon, modern ONTs (like your Huawei) can be quite complicated depending on your ISP’s default configuration of the devices are. Trying to figure out what exactly is going wrong could be tricky, so, you should start with the basics;
-
I agree with the recommendation to try a rebooting the ONT to start with
-
Is your platform running Mosaic connected over WiFi directly to the Huawei ONT? If not and there’s a separate Access Point involved, you might have a NAT/different IP segment issue going on that’s preventing Mosaic from seeing your Rossini
-
If both the Rossini and the Mosaic platform are connected directly to the ONT, then plug a spare Laptop to an empty port on the ONT and check if you can ping both the Rossini and your Mosaic platform’s IP addresses - this will verify that they’re all within the same IP segment and can see each other (FYI ONTs can heave each Ethernet port on a separate VLAN for multiple service from your ISP thus preventing them from “talking” to each other).
-
Verify that your Mosaic connected port on the ONT has IGMP Snooping turn OFF
There’re probably more thing you can try, but start with those.
3 Likes
Thanks for the suggestions and here’s an update from me.
I did a full factory reset on the Rossini, three times, but no progress. I contacted the ISP but they insist their equipment does not stop any traffic and is not preventing the Rossini from working. They suspect that the dealer setup the Rossini with a fixed IP Address which is outside the range of their router and recommend returning to a DHCP setting. I’ve asked the dealer who insists it should have a fixed IP Address and blames the ISP for providing rubbish kit. Each blames the other.
I have tried typing the IP Address shown on the back of the router into a browser connected to it to investigate the options for a fixed IP Address but it says the server is not responding. I’ve emailed the ISP asking for help with that but no reply yet. I’ve called and get the answering service which says the mailbox is full!
So we are not making much progress really.
However, the real mystery was last night when I left the house for three hours with the Rossini powered up. When I got back I noticed that cycling through the screen using the remote control it showed a NETWORK input. I fired up mosaic on the iPad but got the little blue circle spinning round (I’m used to seeing that anyway) but I left if for about 30 minutes and then it connected. It worked.
I tried again this morning and it also took a long time with the message trying to connect to Rossini, but eventually it did connect.
But it is still totally unsatisfactory and I will persevere with the new fixed IP Address or DHCP option if possible.
I will update again if I make progress, but might put the Rossini up for sale. I can’t believe it’s so complicated.
Sorry to say your dealer is absolutely wrong. 
I suggest you do a factory reset of the Rossini, and it should pretty much be plug & play after.
3 Likes