I tried. I can’t find the DM button. And I don’t want to be the forum cop. That’s why I deleted the message and just commented on his reply to the article.
1 Like
I don’t have the chops to go into the weeds on this, but I am curious: which “clock” are you referring to?
We measured the clock for the Spdif output.
I am reading some comments. It’s clear people didn’t read the article. We do NOT believe in clocks in the switches themselves. That’s absolute BS. A network doesn’t need an audiophile clock. It works fine with normal clocks.
Also: TCP is bullet proof. The data is not the problem. We have known that for decades. But when we started noticing differences - especially with other power supplies on the switches - we wanted to know why… So I started measuring switches.
What I noticed is a big difference in noise. Some switches showed way less noise on the spectral analyzer and Prism. In our (semi) blind test the ‘better’ swithces sounded better (more calm). So… the noise from the switches themselves that get in to your streamer is the issue. That’s measurable (as I said: we did several tests with that).
The next step was to measure the actual influence of that noise (mostly common mode). I decided to measure jitter on the clock in the streamer. The clock for the spdif output that is. And yes: there are clear differences. Measurable, repeatable differences.
3 Likes
Thank you for the clarification. I did read the article, though not the data behind the paywall.
Just for the sake of ending some of the discussion and sceptisism in this thread (I do not mean you by the way):
This is NOT for sharing outside this thread. So please do not share these outside this forum. We invested months and thousands of Euros in this. That’s why we used the Patreon pay-wall. We need to pay rent and eat as well (Yes we are also human beings).
Do realize this is raw data that needs some context. There are also trail measurements in these.
3 Likes
Hi Jaap,
You confuse my original post with somebody else’s. I just translated a small part of your conclusion and then included the link:
I did not post a PDF with the translation of your article.
1 Like
Jaap, please don’t feel you you need to defend yourself, but I would welcome a technical debate on the specific conclusion(s) you make.
→ You’re suggesting that because different Ethernet Switches cause varying Clock jitter/phase-noise on the S/PDIF output specifically on the Metrum Amber Streamer, therefore all Streamer/DACs will have sound quality impact from different Ethernet Switches.
Is that a fair summary of your conclusion? If not, then please correct me. If yes, I’ll have further points to make.
2 Likes
Good to see we’re on the same page on that point. I’m curious though, have you stated this categorically in any of AA’s articles/posts or videos?
And do you feel this justifies spending $thousands on a “Audiophile Ethernet Switch” as opposed to using other far far cheaper methods to mitigate any such noise fro getting into the Streamer/DAC?
2 Likes
I noticed. Sorry for the inconvenience.
1 Like
Good to see we’re on the same page on that point. I’m curious though, have you stated this categorically in any of AA’s articles/posts or videos?
Ow yes…! Several times!
And do you feel this justifies spending $thousands on a “Audiophile Ethernet Switch” as opposed to using other far far cheaper methods to mitigate any such noise fro getting into the Streamer/DAC?
Whether it is worth the extra cost is not up to me. That’s up to the consumer. By the way, I have also measured medical grade insulators and that sometimes works, sometimes doesn’t. The biggest gains can be made with a good, quiet power supply. And as you will also know, those are not cheap.
→ You’re suggesting that because different Ethernet Switches cause varying Clock jitter/phase-noise on the S/PDIF output specifically on the Metrum Amber Streamer, therefore all Streamer/DACs will have sound quality impact from different Ethernet Switches.
Is that a fair summary of your conclusion? If not, then please correct me. If yes, I’ll have further points to make.
No… that is not what I concluded. I CAN influence the streamer and it is something to keep in mind. I have tested the Gimm MU1 and know the design team very well. I know from my listening tests AND from a technical point of view it’s very much immune to noise from the network. It is one of the few products that I know that actually paid extensive attention to that.
@alphaaudio - thanks for your efforts here! Without reading all the details, I think you really measured the differences in switches - and that’s is obiously what I could hear by trying Melco, Innuos, D-Link, Cisco in my chain. Even the Sbooster PSU on Melco was clear better vs Melco without PSU!
Has anyone compared the Melco S100 and S10? Is the latter really worth the significant price difference? In fact, are either better than a good quality data switch?
I listened to both recently and could hear no difference, although it was part of a more extensive demonstration and I may just have been suffering from “comparison fatigue” as it was late in the day. I have suggested to my dealer we try all three: data switch, S10 and S100 in my system at home when it is set up to see if I can hear any difference but interested to learn what others have found.
I know it’s tricky to give some recommendations- but (if possible) listen to Ansuz in your setup as well.
What this switch does in terms of “3D”, space and openness is very special.
Hi, which Ansuz switch did you test, they seem to have a range of 4 switches to choose from? Thanks, Mike.
Mike,
I tested the one for 3000Euro (entry line?) and would have liked to buy it but it only worked with roon in my setup (not with upnp/Mosaic). So if you’re fancy that one please test it in your chain.
1 Like
I can confirm that the Ansuz (D2) switch works great in my system. It works with Roon and upnp/Mosaic.
1 Like
Thanks. And you are welcome.
I started to experiment with a few PSU’s after reading the Alpha Audio website, they all sounded different on my Melco S100, in the end I settled for a well made SMPS. (Ifi Elite) Especially timing and transients improved.
Perhaps something is lost in translation, but your concluding remarks clearly generalises the results;
With your permission, from https://www.alpha-audio.nl/achtergrond/hoe-een-netwerkswitch-de-audioweergave-beinvloedt-een-extreme-deep-dive/ (bold/underlined emphasis mine);
To briefly summarise this long story about the influence of network switches on clock jitter: our measurements have shown that the phase noise in particular decreases with a decent switch that does not transmit much (common mode) low-frequency noise to the streamer. In our measurements we see a direct relationship between low-frequency noise from the network port and phase noise on the clock.
Is that audible? Yes: we conclude that it is audible. As far as we are concerned, the listening tests show that switches with a lot of noise from the port sound less good. And now that we have been able to establish the relationship between this noise (especially low-frequency noise) and phase noise on a clock in a streamer, we think that has also been proven, to be honest.
Also, while you go to the nth degree to objectively measure the digital side of things, no objective diligence was given to the analog side. Instead you simply conclude that it’s audible. 
Considering there are a number of DACs which can tolerate a certain level of jitter on incoming SPDIF interfaces, your conclusion that there’s will be a sonic impact would appear to be quite premature. Wouldn’t you agree?
1 Like