I started with a NUC based ROON player/server connected directly to my wifi router via a CAT7 ethernet cable. NUC to DAC via USB, CD Transport to DAC via I2S. CD sound was superior. Using the Ayon Stealth XS DAC and Ayon CD-TII transport.
Added a $25 network switch between router and Roon player. Sound improved. Still not as good as CD via I2S.
Added two FMC’s with LPS’s, removed $25 network switch. Sound improved. Hi res streaming was close to CD via I2S. Redbook streaming vs spinning CD was not quite there.
Bought a LHY SW-8 network switch. Added after the second FMC. Still using basic CAT7 ethernet cables. Quite an improvement in sound. Noise level in songs went down.
Took the optical section out of the path so now just using the SW-8 between wifi router and ROON player/server. Best sound yet. Also connected the Apple TV to the SW-8. Picture improved dramatically. Spent a day running a CAT 7 cable through the house to get to the TV.
Bought silver plated ethernet cables. First between NUC player and switch. Sound improved, less sibilance. Added silver plated ethernet cables from Modem to router to switch to ROON player. Best sound yet. Not using the CD Transport much now.
Bought new DAC- Aeries Cerat Helene. Wow- sound much improved. Suddenly I find the sound with the CD Transport is superior to streaming with the NUC based ROON player. Especially with respect to PRAT. Playing a lot of CDs not so much streaming. LHY switch died. Bought an EE8 switch to get my music back ASAP. LHY helped me troubleshoot and sent me a replacement board for the SW-8. Left the EE8 switch in series with the SW-8. Sounds better that way.
Bought the K50 a month later after getting the new DAC. K50 matches the CD Transport for sound and PRAT. Streaming is wonderful.
To sum up: I have played about 12 hours of vinyl since getting my K50 and Helene DAC. Also playing very little CDs directly on transport. I can hear no difference between CD and FLAC files on the K50 now. No fiber optic in the link from wifi router to LHY switch to EE8 switch to K50. All silver plated (AQ) ethernet cables.
Extremely quiet, detailed enjoyable listening. No fatigue, no cotton ear after listening to moderately loud music.
@Fidas Luckily I didn’t spend 6 mins reading that and instead skipped to the conclusion where I found it had zero to do with sound quality in audio systems.
“Conclusion
In the realm of video conferencing, jitter buffers are essential for mitigating the impact of network-induced timing discrepancies. Platforms like Digital Samba’s SDK/API utilize adaptive jitter buffers that intelligently adjust their size in response to current network conditions. This capability ensures fluid, high-quality video calls, free from interruptions such as choppiness or freezing, even when network stability is less than ideal. By dynamically optimizing their capacity to suit the prevailing network environment, these buffers are fundamental in facilitating uninterrupted, high-calibre video communications, making consistently smooth and lag-free video conferencing a tangible reality.“
Nice try to waste my time though.
I emailed Innuos some time ago when I had a Zenith MK3 streamer which for the money with their new App sounded superb & asked about switches etc, they replied the PhoenixNET was a big seller compared to their USB option. I currently use a SoTM CAT 7 cable which works very well which is probably why the Muon Pro had little effect . I do remember speaking to Elite who also said the PhoenixNet was also very popular.
You’ve also succinctly put it down to a layer 1 issues. If Antipode streaming protocol vs Roon vs Tidal vs Amazon vs Apple vs vs vs vs AND they’ve mucked up something then it’s possible. So which are we talking about?
From a network design perspective there isn’t much that I can do to mitigate ROON and their RAAT protocol if that is what they are still using and know some had complaints about.
The old M$ media player did something nuts like caching in the double digit ms range, so just caching all the time and then feeding that to some horrible USB synchronous DAC. That was absolute rubbish as an architectural approach and longer that way. In that scenario you bet there could be differences.
This is why I like knowing how software behaves. Whether it’s Antipodes on what ever cpu hardware or JRiver on x86 as an example and this includes the endpoint.
If we are talking Streamers? Well those are L3 devices and they 100% behave and can be manipulated like any other L3 device.
I can dust off my DX and literally route it 1000 miles.
Here is what I can tell you: When feeding my RME a 24/96 file from JRiver or from VLC Player and my RME locks to the incoming data stream and reports it as 24/96. No I can’t hear the difference.
Benchmark’s Jon Siau showed his DAC doing perfect data recover with a purposefully built USB cable rife with errors. On the analog out it was indistinguishable.
Yes. The PhoenixNET does indeed seem to benefit the sound of my Antipodes and I am right now basking in the glory of Beethoven’s 7th streamed live from the Proms via the BBC Sounds app to the Squeeze server in my Oladra and then to the Squeezelite player in the Oladra and connected of course via BNC from the Oladra to the Mscaler and Dave. It is a sublime sound.
This is my current network cleanup. I have achieved a very clear sound and transparent performance. I can see deep into the recording. Squeeze Server/Squeeze Player certainly has a hand in this too. I can’t say that the sound is annoying, because everything actually sounds so pure and clear. Roon would round it off a little in terms of sound. But I can’t test it anymore because Roon has been canceled. But I remember Squeeze Auto vs. Roon Auto well. Well, I can’t say whether my network cleanup is top-notch, and I can’t say how high my network noise levels are in the end. I can only hear it. Compared to offline operation (LAN cable unplugged from the K50), I would say that I would like both versions. Offline operation is a little quieter. Others would say more boring compared to my current network cleanup. With the current configuration, I have a very clean, powerful bass. The 2 Muon Pro filters in my system are no disadvantage at all. Will fiber optic help me further - no idea. Is the configuration with a WLAN bridge ideal - no idea. It still sounds very good. I don’t stream, but I tried it and it wasn’t a problem via Qobuz. My network is wired to 1 GB/s all the way to the K50. No throttling to 100 MB/s. Certainly not a cheap network cleanup, but every step was positively audible. But as I said, no idea how much noise is still there…
So, even if Network Acoustics finds or prefers the 1GB/s connection best in terms of sound. Or in other words, the Muon Pro Filter is designed for a 1GB/s connection because it can fully exploit its filter technology, I still have to say that I like the 100 MB throttled speed at the output of the Tempus Switch best! Remembering other switches like the Melco S100 or Innuos Phoenix Net, I always liked 100 MB the best. Innuos only worked with 100 MB, but to my ears it was never a disadvantage that it didn’t have a 1GB/s connection. The K50 doesn’t seem to mind in the least that it is fed with 100MB Ethernet, quite the opposite in fact. 100MB Ethernet speed sounds calmer, more relaxed, simply more relaxed to me. 1GB/s sounds a bit more stressful, crisper you could say, bass a bit more powerful, but the sound is noticeably thinner. I mean 100MB Ethernet gives the music more room to breathe.
You have to remember that as speeds increase:
1: Jitter performance improves
2: Eye patterns improve
3: Time on the line decreases. 100 to 1000? 10X less time on the line, 1000 to 10000? 100X less time on the line than 100mbps.
4: Ethernet PHY’s will actually put the interface in a low power state between transfers. Additionally if you have 1GBE and a switch that supports it, 802.11az will put less current on the line if you are under 40 meters.
10Gbe Ethernet can deliver a complete 16/44.1 album to local cache / buffer under 1 second. The remaining 50 minutes of playback never touch the network connection.
I just don’t see this surviving a setup where you don’t know what’s in play. I also don’t see how these boutique outfits have any experience in this realm vs Juniper, Cisco, Aruba, Arista, Netgear, Extreme etc.
Hello! Everything you wrote may be true, but to me it doesn’t sound like jitter when the 100 MB output on the switch is used. My beloved offline mode doesn’t sound like jitter either. Not in the slightest. I think it makes a difference whether someone is streaming or only needs their network for control. When streaming, I can imagine that 1 GB/s can be better. Depending on the amount of data. Since I only listen to locally stored music, I may not need a speed of 1 GB/s. For me, the 100 MB sounds immediately audible, more relaxed, less dependent on the speakers, more tones, the music has more weight. It also sounds perfect in conjunction with Squeeze Server/Squeeze Player. Squeeze literally takes a deep breath. 1 GB/s doesn’t sound bad, but in comparison it sounds more direct, and to me more stressful, than 100 MB.
Can you clarify local storage and streaming? They are typically mutually exclusive.
From K50 intern SSD local Storage.
BTW many of the talking points I present are from Uptone Audio’s John Swenson and his white paper on the ER. He totally is an electrical isolation and jitter reduction proponent.
Hi, a switch can be very useful. I use a Melco with a linear power supply from Plixir. You should definitely experiment with the power supply units. Apart from Farad and Keces, I found Plixir more exciting. In my test series, I always connected two power supply units, one to the router and one to the switch. This was the best way to make out the differences in sound. I also use a switch because my TV with Apple and my AV receiver also run through it.
I use Supra as the cable for the power supply unit. For direct current, I use the statement from Plixir, which I like best.
The effect with the power supply units on the switch is amazing. A technician once said that the simpler the device, the greater the effect; as with a switch, for example.
Best regards
Alexander
I also have a Plixir LPS powering my router and switch.
I’m very happy with the result!
I also use the Melo S100, previously with an Sbooster power supply, which I later switched to an Antipodes S60. Indeed the power supply makes a significant difference. I found the S60 to be very well suited to my Oladra upgraded EX / CX. If I recall correct they all have essentially the same power supplies.
What do you mean? Does ‘everyone’ have Plixir now?
The new models from Plixir, in particular the Statement, go one better. You can currently buy an almost new one from CM Audio. I recently traded it in when I bought the Oladra😃.
Sure, you should just listen to it in the end…
Apologies, I meant the S60 and Oladra upgraded CX / EX from Antipodes are all essentially the same power supply from Antipodes… [quote=“desbiss, post:296, topic:700”]
I found the S60 to be very well suited to my Oladra upgraded EX / CX. If I recall correct they all have essentially the same power supplies.
[/quote]
All right, I understand! Thanks for the feedback.