I have a slightly different question that may betray the shallow depth of my technical knowledge. I’ve invested a lot of money to stay the hell away from S/PDIF and USB altogether for a decade now. These tests, as performed so far, would seem to have little bearing on an actual network audio system that sends its asynchronous data into a DAC with a properly designed and executed transceiver. In such a configuration, where is the jitter being dumped into the DAC’s buffer? And further, to Anup’s point, what is the actual effect at the DAC output, i.e., the analog signal we actually hear?
I’m not technical enough to explain why these things make a difference, but I can hear a difference between ethernet switches on my Meridian system (and an even greater difference using my previous Antipodes CX combined with a TT2 for my headphone setup).
Conversely, they make a far smaller difference on my dCS equipment (which I suspect says a lot about their design).
Would I like to know why it makes a difference? Most definitely. Can I do so? Not yet.
PS I generally take equipment back out of my system after a few days to see whether the difference makes it worth me keeping the additional box (or cable) or returning it. The switches I’ve kept were worth keeping.
1 Like
I do see - in this case - a relationship between low frequency noise and an increase of phase noise. Whether a streamer is sensitive to that is up to the design. I don’t have 10 streamers to test it on. Nor do I have the time to do that (I hope you can understand that… this takes a LOT of time).
Hifi is a complex hobby, in that sense that there are a lot of variables. But - my opinion - that doesn’t mean that what we have found, proves the relationship between noisy network switches, increase in jitter / phase noise and the effect of that. Can you make a streamer immune to that? Sure… Grimm proved it with the MU-1. Are all streamers immune? No… definately not. Most of the time I do hear differences in switches. Especially when you swap out the power supply. And our research has proven that the PSU makes the most difference. And that makes sense, since most of the junk has to do with the power supply.
1 Like
That is a good strategy. And well… dCS makes top-notch gear (I love the brand). So, it makes sense that you hear less differences in network-switches.
1 Like
Not sure what you mean. But - depending on the system - there is always a point that digital audio needs to be converted to analog. Async or not… that clock needs to be very stable. Considering the fact that every electrical connection can influence the ground plane, a noisy network can influence everything in a dac, streamer, etc. I do not know a single engineer that denies that. The amount of influence depends on the design. The good brands have a very stable, separate power supply for the clock to prevent issues with noise. But still there could be other issues.
I am not an product designer, nor do I know all designs from all brands. And that’s not my focus as well. I just try to explain things I hear.
The clock in a streamer or clock needs to be stable if optimal sound quality is the goal. I think that is undisputed but, hey, it’s an online forum so you never know!
The clock in a switch merely needs to be fit for its technical purpose. There is no relationship between the stability of a clock in a switch and the sound quality to which it contributes.
I watched your switch comparison video part-live part-catchup. At one point one of you said something like “Have you noticed that the better sounding switches are the ones with better clocks and better power supplies?”. The implication was perhaps that both might contribute to the differences heard - but in reality only the second can.
We agree. 
2 Likes
I can’t imagine I said “better clocks”… I don’t believe in better clocks in switches. I never found any proof it actually improved anything. I definately said “Better power supply”… And I found proof a better PSU improved noise levels.
1 Like
I apologise if I’ve misrepresented you. It would be difficult to find the specific moment in the several hours of footage, so I’m happy to accept this assertion and we can then agree that sonic performance is nothing to do with clocks and a lot to do with power supplies.
1 Like
No… I believe ‘high performance clocks’ in switches are useless and just some piece of marketing BS.
2 Likes
That’s clear then! Fabulous. And agreed. Well it’s either marketing BS or the people making/selling the switches don’t understand the mechanisms at play or a combination of the two.
1 Like
Not sure which of those is worse.
3 Likes
I think, on balance, that I prefer naivety to cynical exploitation but it’s close.
Morally, I am probably with you, but to the blithe audiophile, the outcome is largely the same. The naive vendor might mean no harm, because he or she genuinely believes they’ve got something of value based on, say, a misunderstood engineering principle. But again, it’s the same outcome. Close call indeed. Maybe the naive one is a better pint drinking partner.
1 Like
I like the info, tone, and spirit of that piece, and the closing paragraph. Thanks for posting.
Thanks.
I wrote it.
Banking the compliment!
1 Like
Didn’t know you worked for Reiki.
1 Like
Hi Nigel,
Just to prevent confusion, I hope you have read this in the FAQ of this community:
Members of the industry are welcome here as well, but we respectfully ask that you properly disclose your affiliation. If you represent a manufacturer, distributor, integrator, dealer, or are a member of the press please inform the community by editing your profile as follows:
- Include your real name (first name only is fine)
- Include your company name and position. This can be included in the title field. Please don’t hesitate to contact us if you have issues here.
- Include the website URL of your company (if applicable).
1 Like
I started the company 12 months ago but only finished assembling the first Superswitches a few weeks ago.