advertisement


Audiophile Network Switches for Streaming ... really ?

Billsie said:
Long post alert! If you fall asleep reading it…. sweet dreams!
Original system: Naim UnitiServe - NDX - 202 - 200 - Linn Ninkas morphing into :-
Present system: Euphony 2 Box (Intel NUC) server/streamer (with Sean Jacobs LPS) running Roon Core to Euphony Stylus EP (connected to home network via DX Engineering RF filter) - Teddy DAC - Teddy PR1 - TP MB100s - Revel F208s

I’ve been following this thread with interest and frankly I’ve become more and more irritated by all the talk of expectation bias. I’ve seen the videos and always spotted the gorilla playing basketball and wandering across the dance hall. I know that adding a Teddy Pardo power supply to my UnitiServe resulted in an immediate and worthwhile improvement to the sound of my streaming setup. My findings echoed various reviews. I didn’t try removing the PS and reverting to the original Naim PS (why would I, it sounded better and I enjoyed it). Adding the Sean Jacobs power supply to my server/streamer didn’t immediately improve the sound but one day (probably some weeks later) things sounded much better. My thoughts were that it must be the new PS (I’d expected to have to give the PS some time to “burn in”). I didn’t take the PS away and revert to the original power blocks supplied with the NUCs to check…why would I, it sounded better and I enjoyed it. A year or so ago, after reading about them on the Audiophile Style website, I bought a couple of DX Engineering RF filters (about £50). Wow, for the money, they effected a clear improvement to the sound of my system. I didn’t remove them to check if the sound deteriorated without them..why would I?

A few years ago, I went speaker shopping for the day to Hull. I’d read tons of reviews and knew that Focal Sopra 1s would be perfect for me and if per chance they weren’t then PMC Twenty5 24s would be. Well the Sopras were so bright, they were almost painful to listen to and the PMCs were so dull and boring I left the shop shaking my head in disbelief. Clearly, if expectation bias exists I’m immune to it.

A couple of weeks ago I decided to contact HAF with a view to having some convolution filters created to hopefully help the soundstage and calm some perceived brightness. Thierry produced an initial set of filters that I inserted into Roon which rather than improving things, seemed to suck all the life out of the music. (My ears won out over any expectation bias!) A second set of filters sounded much better. I was very chuffed!

Last week, my wife and I had a few days in Scotland (I don’t think they’ll ever need to worry about a water shortage). One morning whilst eating my cornflakes I was treated to “I’d rather go blind” by Beth Hart piped over the hotel sound system. When we got home on Saturday evening I listened to the same track on my HiFi. Bloody hell, was it good. I sat grinning at the fabulous sound, improved so much by the new convolution filters inserted into Roon. I have subsequently played track after track this week transfixed by how good they sound. My daughter humoured me and asked about the filters, so this evening I showed her the Roon DSP window on my iPad only to find that at some stage I’d erased my favoured filter and the system was playing without any filters at all! I’ve now reinserted the filters and it all sounds even better. (Have I finally become susceptible to expectation bias,as clearly something has been going on in my head this week)

So, will I swap out either of the power supplies or get rid of the RF filters? Of course I won’t. My system sounds the best it has ever sounded and I really enjoy it. Will I appraise future changes to my system more thoroughly? Yes, probably. Do I now believe in expectation bias? Yes probably.

Is this hobby fun? YES IT IS!!

Night night
Zzzzzzzz
Click to expand...
TL : DR
 
Billsie said:
Long post alert! If you fall asleep reading it…. sweet dreams!
Original system: Naim UnitiServe - NDX - 202 - 200 - Linn Ninkas morphing into :-
Present system: Euphony 2 Box (Intel NUC) server/streamer (with Sean Jacobs LPS) running Roon Core to Euphony Stylus EP (connected to home network via DX Engineering RF filter) - Teddy DAC - Teddy PR1 - TP MB100s - Revel F208s

I’ve been following this thread with interest and frankly I’ve become more and more irritated by all the talk of expectation bias. I’ve seen the videos and always spotted the gorilla playing basketball and wandering across the dance hall. I know that adding a Teddy Pardo power supply to my UnitiServe resulted in an immediate and worthwhile improvement to the sound of my streaming setup. My findings echoed various reviews. I didn’t try removing the PS and reverting to the original Naim PS (why would I, it sounded better and I enjoyed it). Adding the Sean Jacobs power supply to my server/streamer didn’t immediately improve the sound but one day (probably some weeks later) things sounded much better. My thoughts were that it must be the new PS (I’d expected to have to give the PS some time to “burn in”). I didn’t take the PS away and revert to the original power blocks supplied with the NUCs to check…why would I, it sounded better and I enjoyed it. A year or so ago, after reading about them on the Audiophile Style website, I bought a couple of DX Engineering RF filters (about £50). Wow, for the money, they effected a clear improvement to the sound of my system. I didn’t remove them to check if the sound deteriorated without them..why would I?

A few years ago, I went speaker shopping for the day to Hull. I’d read tons of reviews and knew that Focal Sopra 1s would be perfect for me and if per chance they weren’t then PMC Twenty5 24s would be. Well the Sopras were so bright, they were almost painful to listen to and the PMCs were so dull and boring I left the shop shaking my head in disbelief. Clearly, if expectation bias exists I’m immune to it.

A couple of weeks ago I decided to contact HAF with a view to having some convolution filters created to hopefully help the soundstage and calm some perceived brightness. Thierry produced an initial set of filters that I inserted into Roon which rather than improving things, seemed to suck all the life out of the music. (My ears won out over any expectation bias!) A second set of filters sounded much better. I was very chuffed!

Last week, my wife and I had a few days in Scotland (I don’t think they’ll ever need to worry about a water shortage). One morning whilst eating my cornflakes I was treated to “I’d rather go blind” by Beth Hart piped over the hotel sound system. When we got home on Saturday evening I listened to the same track on my HiFi. Bloody hell, was it good. I sat grinning at the fabulous sound, improved so much by the new convolution filters inserted into Roon. I have subsequently played track after track this week transfixed by how good they sound. My daughter humoured me and asked about the filters, so this evening I showed her the Roon DSP window on my iPad only to find that at some stage I’d erased my favoured filter and the system was playing without any filters at all! I’ve now reinserted the filters and it all sounds even better. (Have I finally become susceptible to expectation bias,as clearly something has been going on in my head this week)

So, will I swap out either of the power supplies or get rid of the RF filters? Of course I won’t. My system sounds the best it has ever sounded and I really enjoy it. Will I appraise future changes to my system more thoroughly? Yes, probably. Do I now believe in expectation bias? Yes probably.

Is this hobby fun? YES IT IS!!

Night night
Zzzzzzzz
Click to expand...


Look, this thread didn't get where it is today by people reporting actual experiences. Stop it, now!
 
Just looks like another "I can't afford it so it must be sh*t" thread to me :)

That being said, I must admit to having absolutely no opinion any whichaway since I've never been interested enough to listen to switches...
I've just plugged my NAS into the RJ45 pass through on my dCS Rossini with a Nordost Heimdall (which I DID listen to before purchase).
 
No way the network transport itself could affect the audio. Especially if there is a buffer on the receiving end.

There is a possible argument that the two grounds get connected and that a crappy ground on the switch leaks to the Audio Streamer.

I say to this argument.
a) If the Audio streamer lets a signal ground from the ethernet segment get to the analog sections its a shit design.
b) If it really was a problem, streamers would offer SFP (fiber optic ports). There is no way in hell ground noise could traverse over light. SFP ports are also not expensive. Most don't offer this so manufacturers don't consider it a problem.
 
David Halliday said:
No way the network transport itself could affect the audio. Especially if there is a buffer on the receiving end.

There is a possible argument that the two grounds get connected and that a crappy ground on the switch leaks to the Audio Streamer.

I say to this argument.
a) If the Audio streamer lets a signal ground from the ethernet segment get to the analog sections its a shit design.
b) If it really was a problem, streamers would offer SFP (fiber optic ports). There is no way in hell ground noise could traverse over light. SFP ports are also not expensive. Most don't offer this so manufacturers don't consider it a problem.
Click to expand...

“No way the network transport itself could affect the digital signal. Especially if there is a buffer on the receiving end.” Is this what you mean? Not seeking to correct you, just to clarify what you’re asserting. Thanks

Agreed re SFP in theory. In practice, SFP tends to be avoided because SFP circuitry tends to be noisy itself, thereby undoing the benefit of transmitting the audio signal as noise-free light. I’m sure there must be better and poorer SFP circuits available but this might be why streamer manufacturers typically avoid incorporating it.
 
David Halliday said:
If it really was a problem, streamers would offer SFP (fiber optic ports). There is no way in hell ground noise could traverse over light. SFP ports are also not expensive. Most don't offer this so manufacturers don't consider it a problem.
Click to expand...

Manufacturers only consider one problem: low sales.
And that is why we are still using S/PDIF and USB...
 
TheFlash said:
SFP tends to be avoided because SFP circuitry tends to be noisy itself, thereby undoing the benefit of transmitting the audio signal as noise-free light
Click to expand...
Actually laser generated light is noisy, due to both the laser and the speckle noise, but tiny compared with the volts level recovered digital signal. As this is the same as the recovered data from a RJ-45 cable, its not important.
What fibre does do is eliminate all mains and power supply related noise
The reason that fibre is not used is cost.
 
SFP tends to be noisy? Whats your source for that please. The fact its one of a couple of worldwide IT optical standards would point to that being bullshit.
 
sq225917 said:
SFP tends to be noisy? Whats your source for that please. The fact its one of a couple of worldwide IT optical standards would point to that being bullshit.
Click to expand...

I think that he is referring to the process of converting data / generating pulses.
 
sq225917 said:
What's the source of the proof, whats the measured outcome.
Click to expand...

That is a good question.
As mentioned earlier Jussi Laako (HQPlayer's developper) has commented extensively at CA/AS on the negative effects noise in the clock and D/A chip. Other digital audio software and hardware designers seem to agree.
I have not seen much measured evidence other than Jussi's ultrasonic IMD and noise plots.
 
sq225917 said:
SFP tends to be noisy? Whats your source for that please. The fact its one of a couple of worldwide IT optical standards would point to that being bullshit.
Click to expand...
Ah, as polite as ever.

It’s the converters. You’re viewing this through your digital spectacles (again), in which case IT optical is perfect. Google is your friend.
 
So that patent shows that optical transceiver circuitry operating at high data rates generates noise, and steps can be taken to mitigate it. Whoop de do - and in other news, bears crap in the woods. Doesn't in any way address the much more relevant (to this thread) question of whether there is any difference in noise generation between optical and copper transceiver circuitry operating at equivalent data rates.
 
I'm not viewing through any lens, I asked for measurements. There's no measurements showing improved jitter, because that's not a thing from any external clock apparently.

But I'll be more than happy to see the addition of an external clock improve the fidelity of the analogue output of a dac by any measure within the audible band; hell let's be generous and include upto a decade past whatever filter frequency is being used in the dac, just to give a fighting chance.
 
You must log in or register to reply here.


advertisement


Back
Top