|
|
Post by MikeMusic on Sept 5, 2024 8:56:14 GMT
The magic bullet we are offered so often doesn't work !
We have to work harder or a lot harder to realise the potential
I remember listening to my first CD on a new to the country Sony CD player when first launched
It was utterly horrid. Nails across blackboard much more listenable
Put me off CDs for 20 or 30 years.
Big collection of CDs now, starting in the mid 00s
Streaming comes along. Pointless. Missing most of the albums I want - that changed a bit too
|
|
|
|
Post by assisi on Sept 7, 2024 6:51:07 GMT
Right, it gets complicated when we stack several noise reducing methods on top of each other to gradually reduce the noise on the network. I was sort of thinking that a moat is part of the "gear implementation"...(even if that is one gear or 30 stacked gears). At the end of the day it is the noise that goes out of that last ethernet cable into the streamer that matters, from a network noise perspective. (how you got there is sort of irrelevant for the outcome) But i do agree it is a topic of its own, how you can lower network domain noise. My experience and perspective are that everything before the end of the digital network is relevant with contributions to the mitigation of noise. I consider that there is a potential for noise everywhere in a network. No just the final cable into the streamer. The quality and the synergy between all the components and the connections etc in the network can provide a beneficial outcome.
My network is reasonably complex. There is a router, plus 7 switches and finally a passive filter. Not connected now is Paul Pang Quad switch. There are also various isolation feet etc. By choosing, well each a part of my network makes a contribution to noise mitigation that enhances the streaming outcome that I have achieved.
John
|
|
|
|
Post by MartinT on Sept 7, 2024 7:12:39 GMT
My experience and perspective are that everything before the end of the digital network is relevant with contributions to the mitigation of noise. I consider that there is a potential for noise everywhere in a network. No just the final cable into the streamer. I've been saying the same for a while now. Even changes to the router at the most upstream level affects sound quality. Every step is important. |
|
|
|
Post by HD Music & Test on Sept 7, 2024 7:23:19 GMT
If you use streaming as your main scource the IP inforstructure (from entry into the property to exit to your streaming device imho as important as the devices you use if not more so, as the IP pipe is essential your sound quality restriction
|
|
Tobias
Rank: Quartet
Posts: 320
Member is Online
|
Post by Tobias on Sept 7, 2024 7:33:18 GMT
This is why i think air gapping (decoupling from your network) can be very effective since it will reset the noise level and start over again, from that receiver device.
Then one can tailor a solution to target the exact noise level/profile that that single device has, which will be almost the same for everyone (or very similar), if using the same power supply, for example.
This is the more systematic approach that has been adopted in the "throttle choke" concept, for example.
Having said this, I don´t know how good this solution stack up against other solutions, since i haven´t compared myself.
|
|
|
|
Post by MartinT on Sept 7, 2024 8:03:56 GMT
If you use streaming as your main scource the IP inforstructure (from entry into the property to exit to your streaming device imho as important as the devices you use if not more so, as the IP pipe is essential your sound quality restriction One of the reasons why I use a dedicated 4G+ streamer with nothing else connected to it for streaming. |
|
|
|
Post by MartinT on Sept 7, 2024 8:08:07 GMT
This is why i think air gapping (decoupling from your network) Most air gapping (optical fibre or a moat inside a suitable component) will still allow some noise through. It is not a complete reset. This is why I use a different router - the ultimate air gap. A throttle choke isn't any kind of isolation except filtering - of the noise and the precious signal itself. |
|
Tobias
Rank: Quartet
Posts: 320
Member is Online
|
Post by Tobias on Sept 7, 2024 8:14:43 GMT
Here we disagree.
When it comes to WiFi (for example) then the noise is not transported over the air.
The noise will however be created on the other side again, and it is a receiving antenna..., but at least we now have a better baseline.
Now we need to bother about that receiver and everything after it but it won´t matter what you do before it, from what i hear. (and it doesn´t make sense that it could)
For optical, it is still physically connected and that is something else. Noise seem to find its way when things are coupled together, according to engineers.
|
|
Tobias
Rank: Quartet
Posts: 320
Member is Online
|
Post by Tobias on Sept 7, 2024 8:25:46 GMT
I forgot to mention that the throttle choke concept is not the ferrit solution in isolation. It is a solution that is developed together with a wifi solution as a complete solution.
|
|
|
|
Post by MartinT on Sept 7, 2024 8:42:24 GMT
When it comes to WiFi (for example) then the noise is not transported over the air. Sorry to tell you (and I am an IT engineer) that wi-fi contains an awful lot of packet jitter. So in some ways it's good isolation, but in others it introduces problems of its own. |
|
Tobias
Rank: Quartet
Posts: 320
Member is Online
|
Post by Tobias on Sept 7, 2024 8:50:08 GMT
I´m also in IT and that is usually the problem :-)
As you know, don´t ask an Network engineer about streaming audio.
Network jitter is a problem if we can´t read the ones and zeros but it is not something that travels on the network and causes downstream effects later.
The jitter that we talk about in the DAC gets created in the DAC, due to the noise injected in the DAC, that causes this jitter locally. The jitter does not travel into the DAC.
That is what is explained by Hans Beekhuyzen.
|
|
|
|
Post by MartinT on Sept 7, 2024 9:44:40 GMT
My understanding is that jitter (time-based) and phase noise (frequency-based) can both present as noise into the DAC. In any case, it's safer to assume that some noise gets through and take care to treat it along the path. Grounding boxes to remove noise from the ground plane of components is very effective.
I know you like the wi-fi + throttle cable solution. I have taken care not to transmit the music data over wi-fi, only control signals. They are both valid viewpoints and we have both reached a sound quality we like.
|
|
Tobias
Rank: Quartet
Posts: 320
Member is Online
|
Post by Tobias on Sept 7, 2024 10:08:51 GMT
I agree and i realize that you have A LOT more actual experience in this area, with several methods, and probable an amazingly low digital noise floor, as confirmed by many.
As you said, all i wanted to share was a different way of reducing noise and try to help share the view that it is much less complicated than we tend to think, from a logical standpoint, even if that simple noise problem is actually very hard to "solve"...
But just knowing that it is a noise problem mostly and that noise travel on everything (it is a non IT problem) physically connected. This is something that engineers usually can attest to within areas where noise is a big problem, that is not IT related.
In my mind, we can just forget about digital and IT in audio, because that works flawlessly. This has nothing to do with that, unless you want it to.
|
|
|
|
Post by HD Music & Test on Sept 7, 2024 13:01:51 GMT
Here we disagree. When it comes to WiFi (for example) then the noise is not transported over the air. The noise will however be created on the other side again, and it is a receiving antenna..., but at least we now have a better baseline. Now we need to bother about that receiver and everything after it but it won´t matter what you do before it, from what i hear. (and it doesn´t make sense that it could) For optical, it is still physically connected and that is something else. Noise seem to find its way when things are coupled together, according to engineers. Noise ingress is from ALL sides of the incoming pipe line, hence why you require a full gamut of supression & killling at source techniquies to FULLY tale care of the IP pipe coducted, radiated and imunity issues RIGHT up until the entry of the device, then some form of quality opto coupling barrier (both power & signal) after the socket on the streamer/PC what ever you use.
We see tis everyday in our non audio high speed digital communications work, I'm with Martin mobile router eliminates a lot of the issues, HOWEVER power isolation, signal isolation, and protection from external RF ingress and internal RF control would be a bare minimum chaps.
|
|
Tobias
Rank: Quartet
Posts: 320
Member is Online
|
Post by Tobias on Sept 7, 2024 13:15:57 GMT
Here we disagree. When it comes to WiFi (for example) then the noise is not transported over the air. The noise will however be created on the other side again, and it is a receiving antenna..., but at least we now have a better baseline. Now we need to bother about that receiver and everything after it but it won´t matter what you do before it, from what i hear. (and it doesn´t make sense that it could) For optical, it is still physically connected and that is something else. Noise seem to find its way when things are coupled together, according to engineers. Noise ingress is from ALL sides of the incoming pipe line, hence why you require a full gamut of supression & killling at source techniquies to FULLY tale care of the IP pipe coducted, radiated and imunity issues RIGHT up until the entry of the device, then some form of quality opto coupling barrier (both power & signal) after the socket on the streamer/PC what ever you use.
We see tis everyday in our non audio high speed digital communications work, I'm with Martin mobile router eliminates a lot of the issues, HOWEVER power isolation, signal isolation, and protection from external RF ingress and internal RF control would be a bare minimum chaps.
Yes, I agree on that as well. As soon as we connect something prior to the DAC, regardless of what it is, it needs full attention to noise from all aspects. The ethernet signal just being one of them. |
|
