jabbr Posted January 31, 2023 Author Share Posted January 31, 2023 8 hours ago, R1200CL said: I’ve been told that someone did a test measuring “noise” from the streaming services. And since they are using different techniques or whatever they are, there is a measurable difference. But this didn’t affect the SQ according to the person I spoke to. Ok what? Explain … but here: Custom room treatments for headphone users. Link to comment
barrows Posted January 31, 2023 Share Posted January 31, 2023 11 hours ago, jabbr said: Excuse me? I said it was designed to conform to 1000base-X. 1000base-X is not designed to reject upstream jitter/noise. What is wrong with that? Is anyone claiming it is designed to do more that the spec? If someone wanted to design a product to reject upstream noise/jitter then using an Ethernet controller designed for the 10GBase-X spec might be a good idea. I’ve seen no literature claiming that the opticalModule rejects upstream jitter, rather acts as a 1Gbe FMC and I’m sure does an excellent job at this. Fiber breaks ground loops and doesn’t transmit EMI. Clearly we seem to be having a misunderstanding here, perhaps this is related to the problem of "tone" in Internet discussions not being properly understood by both of us. Yes, the opticalModule deluxe is designed to be much a much better performing FMC than a bog standard 1000 speed consumer device, and is built and designed to a much higher level in terms of noise rejection, self noise, and Ethernet clock phase noise. To suggest otherwise without any actual knowledge of the device is doing the product a disservice at best. I am not the designer, John is, and I will not speak for him as to the details of the design, but even a quick look at the PCB would make this fairly obvious. SO/ROON/HQPe: DSD 512-Sonore opticalModuleDeluxe-Signature Rendu optical with Well Tempered Clock--DIY DSC-2 DAC with SC Pure Clock--DIY Purifi Amplifier-Focus Audio FS888 speakers-JL E 112 sub-Nordost Tyr USB, DIY EventHorizon AC cables, Iconoclast XLR & speaker cables, Synergistic Purple Fuses, Spacetime system clarifiers. ISOAcoustics Oreas footers. SONORE computer audio | opticalRendu | ultraRendu | microRendu | Signature Rendu SE | Accessories | Software | Link to comment
Miska Posted January 31, 2023 Share Posted January 31, 2023 17 hours ago, jabbr said: This could be something that @Miska etc could be concerned with if he cared about the rate at which packets are issued ... NAA protocol is already designed such way that even if there would be leakage it would not create an issue. I also prefer higher network speeds over lower ones. But already 1 Gbps is sufficient for most use cases. No harm with 2.5G, 10G or similar though, as long as you are either optical or CAT6(a) U/UTP cabling. STP cables are no-go. I use 500 MHz certified CAT6a cables myself, which are perfectly good for up to 10 Gbps. Signalyst - Developer of HQPlayer Pulse & Fidelity - Software Defined Amplifiers Link to comment
Popular Post barrows Posted January 31, 2023 Popular Post Share Posted January 31, 2023 As to streaming from sources like Tidal or Qobuz. It is fairly well accepted amongst many audiophiles that there is a sound quality penalty for doing this. But if true, that sound quality penalty does not seem to be large. I have not heard of anyone really digging into the weeds and putting this theory to a rigorous test. I do not stream here, playing only locally sourced files. My understanding is that Qobuz allows one to download the exact same files as they stream, so this seems the likely way to test for differences-as it is often impossible to know with other services exactly what version of an album one is playing, so having exact same versions of an album to test with can be difficult. There are some conspiracy theorists who claim that there is some lossy compression involved with Tidal/Qobuz to save them money on storage... they claim this is not so, and this could be easily checked. I would like to see someone really do a fair investigation into this. R1200CL and StreamFidelity 1 1 SO/ROON/HQPe: DSD 512-Sonore opticalModuleDeluxe-Signature Rendu optical with Well Tempered Clock--DIY DSC-2 DAC with SC Pure Clock--DIY Purifi Amplifier-Focus Audio FS888 speakers-JL E 112 sub-Nordost Tyr USB, DIY EventHorizon AC cables, Iconoclast XLR & speaker cables, Synergistic Purple Fuses, Spacetime system clarifiers. ISOAcoustics Oreas footers. SONORE computer audio | opticalRendu | ultraRendu | microRendu | Signature Rendu SE | Accessories | Software | Link to comment
dbastin Posted January 31, 2023 Share Posted January 31, 2023 21 hours ago, juniorbudel said: I hadn't thought of that possibility, has anyone reported differences between shielded or unshielded network cable on the “dirty” side of the system? I haven't 'reported' this on AS before, and you might not believe it ... My network is ... broadband service > cable 1 > ER Side b - side A > fibre > Mikrotik CRS 305 (router) > cable 2 > Antipodes server > cable 3 > WAP > wifi > Devialet Nothing else is connected to the ER, it is basically a FMC. Cables 2 and 3 are audiophile circa $1k each, super reveiling. CRS 305 is 10Gbe SFP+. For cable 1 I tried generic Cat 6 stp, Cat 8 stp, Afterdark (optimised for ER - ie sheild disconnected at one end), Wireworld Platinum (cat 8 sheild connected at both ends), commercial grade Cat 6a STP, but felt a commercial grade Cat 6a UTP is most accurate. I could hear differences between all these cable despite the fibre, there's moat, jitter rejection of the CRS 305 and wifi. The differences are small that far removed from the endpoint but still enough to distinguish and judge. But it probably depends what the cable is connecting, so worth experimenting. I will revisit when I get around to improving the SFP transceivers. Maybe of I hack the Cat 6a STP to disconnect the sheild at the downstream end that will be best in this location in my network. Link to comment
audiophilac Posted January 31, 2023 Share Posted January 31, 2023 @dbastin is your devialet connected wirelessly to the network? Link to comment
barrows Posted January 31, 2023 Share Posted January 31, 2023 5 minutes ago, Miska said: NAA protocol is already designed such way that even if there would be leakage it would not create an issue. I also prefer higher network speeds over lower ones. But already 1 Gbps is sufficient for most use cases. No harm with 2.5G, 10G or similar though, as long as you are either optical or CAT6(a) U/UTP cabling. STP cables are no-go. I use 500 MHz certified CAT6a cables myself, which are perfectly good for up to 10 Gbps. Jussi: I know you prefer that users maintain 1 Gb/s speeds for use with NAA, in my experience, for streaming DSD 256, even 100 Mb/s is adequate. There are some audiophiles who even believe streaming at the slower rate results in improved sound quality (lower network noise transmission)-I am not suggesting I believe that though. Is your reason for preferring NAA users have Gb speed networks (or higher) just to have plenty of data headroom to avoid the possibility of any problems arising, or are there other possible reasons? @jabbr of course advocates for using hardware capable of even higher speeds, as that hardware is built to higher standards, but of course one can always use that 10 Gb hardware at 1 Gb, etc... MFJG 1 SO/ROON/HQPe: DSD 512-Sonore opticalModuleDeluxe-Signature Rendu optical with Well Tempered Clock--DIY DSC-2 DAC with SC Pure Clock--DIY Purifi Amplifier-Focus Audio FS888 speakers-JL E 112 sub-Nordost Tyr USB, DIY EventHorizon AC cables, Iconoclast XLR & speaker cables, Synergistic Purple Fuses, Spacetime system clarifiers. ISOAcoustics Oreas footers. SONORE computer audio | opticalRendu | ultraRendu | microRendu | Signature Rendu SE | Accessories | Software | Link to comment
Popular Post Miska Posted January 31, 2023 Popular Post Share Posted January 31, 2023 5 hours ago, barrows said: Jussi: I know you prefer that users maintain 1 Gb/s speeds for use with NAA, in my experience, for streaming DSD 256, even 100 Mb/s is adequate. There are some audiophiles who even believe streaming at the slower rate results in improved sound quality (lower network noise transmission)-I am not suggesting I believe that though. Is your reason for preferring NAA users have Gb speed networks (or higher) just to have plenty of data headroom to avoid the possibility of any problems arising, or are there other possible reasons? And I don't believe that 100 Mbps networks would have lower network noise transimission. Quite the contrary. Typically 100 Mbps networks lack newer features such as 802.3az which is very much relevant for network noise (cable length probing with transmit power adjustment and inactive idle). In addition, faster networks help QoS work better. And helps leave good bandwidth also for input side, such as playing from NAS. Essentially you have more spare time slots and each transmit takes shorter amount of time. At 10 Gbps is only a tiny blip. Since "network noise" is also function of transfer/idle ratio, faster the network, more time it has to idle sleep and lower the average power is. 1 Gbps or faster network can transfer much more data with lower power and shorter time. It is similar to mobile networks and old wired networks. Now across the world 3G mobile networks are being shut down, since they waste so much RF space for so little. 4G network can transfer much more data within much smaller RF space. 5G networks are even better and can do 10x higher speeds it fraction of the power. Or if we start from old 300 or 1200 bps dial-up modem connections up to ADSL and VHDSL, and speeds went up to 100 Mbps on the same wire. This largely up to more efficient signalling algorithms, essentially fancier DSP algorithms to pack bits tighter in the same bandwidth, going for example to bigger QAM constellations. MFJG, barrows, ericuco and 3 others 5 1 Signalyst - Developer of HQPlayer Pulse & Fidelity - Software Defined Amplifiers Link to comment
jabbr Posted February 1, 2023 Author Share Posted February 1, 2023 10 hours ago, barrows said: @jabbr of course advocates for using hardware capable of even higher speeds, as that hardware is built to higher standards, but of course one can always use that 10 Gb hardware at 1 Gb, etc... Yes, and currently my final run to the endpoint is 1Gbe because I like those really low powered small units with a builtin SFP module. Interestingly while the power usage of copper ethernet goes up with speed and distance (distance with 802.3az), with light the power usage goes up with the number of lanes but not really with the speed of each lane. So roughly 1w for an SFP and 4w for a QSFP ... but that's rough. If you look at the power per bit transmitted, it goes down as each generation gets faster (consider 40Gbe as quad 10Gbe, and 100Gbe as quad 25Gbe) the latest generation is 50Gb/lane but that's cost prohibitive for our purposes. Custom room treatments for headphone users. Link to comment
dbastin Posted February 1, 2023 Share Posted February 1, 2023 16 hours ago, audiophilac said: @dbastin is your devialet connected wirelessly to the network? Yes. I haven't tried wired for quite some time. Wifi sounded great with a regular WAP (tp link), better with LPS but I now use a dedicated WAP (Mikrotik mAP lite) powered by a USB battery, with Akiko tuning stick, and each of these made a small improvement. Go figure?! 16 hours ago, barrows said: of course advocates for using hardware capable of even higher speeds, as that hardware is built to higher standards, but of course one can always use that 10 Gb hardware at 1 Gb, etc... I run my CRS 305 10Gbe st 100MB. But I think that may be causing more interruptions when TV and music streaming data download is patchy. My broad band delivers 20G peak ... if the data is there. Link to comment
jabbr Posted February 2, 2023 Author Share Posted February 2, 2023 For people who want to look at 10Gbe low powered ARM endpoints (can be LPS powered) https://www.solid-run.com/embedded-networking/marvell-octeon-tx2-family/clearfog-cn9130/ (I have the original clearfog base and it works great with debian/naa) The Computer Audiophile 1 Custom room treatments for headphone users. Link to comment
Popular Post jabbr Posted February 3, 2023 Author Popular Post Share Posted February 3, 2023 Get these while they are still available ... Intel 100G-CWDM4 100G at <3.5w -- low energy/bit! That's why the pulls are green 😝 muski and MFJG 1 1 Custom room treatments for headphone users. Link to comment
MFJG Posted February 3, 2023 Share Posted February 3, 2023 1 minute ago, jabbr said: Get these while they are still available ... Intel 100G-CWDM4 @jabbrGood Look. Thank you. Link to comment
The Computer Audiophile Posted February 3, 2023 Share Posted February 3, 2023 8 minutes ago, jabbr said: Get these while they are still available ... Intel 100G-CWDM4 100G at <3.5w -- low energy/bit! That's why the pulls are green 😝 Single mode? Founder of Audiophile Style | My Audio Systems Link to comment
jabbr Posted February 3, 2023 Author Share Posted February 3, 2023 44 minutes ago, The Computer Audiophile said: Single mode? Yes, single mode modules are available up to 400/800 Gbe using multiple wavelengths stuffed into each fiber! This is the telltale sign that high bandwidth QSFP modules are single-mode: Custom room treatments for headphone users. Link to comment
muski Posted February 4, 2023 Share Posted February 4, 2023 Intel 100G-CWDM4 @Barrows—I assume these would work in the OMD & opticalRendu? Link to comment
barrows Posted February 4, 2023 Share Posted February 4, 2023 19 minutes ago, muski said: Intel 100G-CWDM4 @Barrows—I assume these would work in the OMD & opticalRendu? I have no idea... One thing to be aware of is the power requirements of these much faster QSFPs, if they require more power, and make more heat, that might be a problem, but it might not. Of course they would only run at 1 GBe in Sonore devices. We do not test with other than 1 Gbe SFPs, but I am interested in hearing of any customer's experiments. muski 1 SO/ROON/HQPe: DSD 512-Sonore opticalModuleDeluxe-Signature Rendu optical with Well Tempered Clock--DIY DSC-2 DAC with SC Pure Clock--DIY Purifi Amplifier-Focus Audio FS888 speakers-JL E 112 sub-Nordost Tyr USB, DIY EventHorizon AC cables, Iconoclast XLR & speaker cables, Synergistic Purple Fuses, Spacetime system clarifiers. ISOAcoustics Oreas footers. SONORE computer audio | opticalRendu | ultraRendu | microRendu | Signature Rendu SE | Accessories | Software | Link to comment
muski Posted February 4, 2023 Share Posted February 4, 2023 8 minutes ago, barrows said: I have no idea... One thing to be aware of is the power requirements of these much faster QSFPs, if they require more power, and make more heat, that might be a problem, but it might not. Of course they would only run at 1 GBe in Sonore devices. We do not test with other than 1 Gbe SFPs, but I am interested in hearing of any customer's experiments. Ok, I just ordered a pair. I'll let you know how it goes. I already have heat sinks on both my OMD and OR. I found them getting a bit hotter running 10GBe SFPs. But it's been months and no issue. Onwards with the SFP journey. I now have a drawerful of them! But I do hear improvements. Thanks, @jabbr for the recommendation. best, muski barrows 1 Link to comment
jabbr Posted February 4, 2023 Author Share Posted February 4, 2023 28 minutes ago, muski said: Ok, I just ordered a pair. I'll let you know how it goes. I already have heat sinks on both my OMD and OR. I found them getting a bit hotter running 10GBe SFPs. But it's been months and no issue. Onwards with the SFP journey. I now have a drawerful of them! But I do hear improvements. Thanks, @jabbr for the recommendation. best, muski Guys these are QSFP28 modules! Btw <3.5W Try them in the Mikrotik CRS504, $699 at Walmart! 🤣🤣🤣 muski 1 Custom room treatments for headphone users. Link to comment
dbastin Posted February 4, 2023 Share Posted February 4, 2023 6 hours ago, jabbr said: Get these while they are still available ... Intel 100G-CWDM4 100G at <3.5w -- low energy/bit! That's why the pulls are green 😝 I am curious how 100Gbe SFP would be an improvement in SQ compared to 10Gbe. Didnt someone say they could not detect a difference in SQ between 10 and 100 Gbe switches? I am also wondering if a 10Gbe SFP is better than 1Gbe SFP for the same kinds of reasons why a 10Gbe switch is better than 1Gbe switch. I am not seeking scientific proof, but are we speculating or making a well informed assumption? Link to comment
Popular Post jabbr Posted February 4, 2023 Author Popular Post Share Posted February 4, 2023 24 minutes ago, dbastin said: I am curious how 100Gbe SFP would be an improvement in SQ compared to 10Gbe. Didnt someone say they could not detect a difference in SQ between 10 and 100 Gbe switches? I am also wondering if a 10Gbe SFP is better than 1Gbe SFP for the same kinds of reasons why a 10Gbe switch is better than 1Gbe switch. I am not seeking scientific proof, but are we speculating or making a well informed assumption? I’ve tried a variety of NICs and switches from 10Gbe, 25Gbe, 40Gbe and 100 Gbe and can’t hear a difference but that’s just me. Of note the phase error at 25/100 Gbe get down to ~50 fs in terms of the clock generators. It is the entire end to end segment that is being tested in 10Gbe (10GBase-X) and the upstream jitter/noise rejection is not done in the SFP(+) module itself. im not speculating, the 10Gbe Ethernet speculation requires upstream jitter/noise rejection ie the signal is required to be cleaned up. There’s a big jump from 1Gbe to 10Gbe but beyond that the residual jitter just gets less. Once it’s inaudible making it less is still inaudible. Folks should realize that an SFP module will physically fit into an SFP(+) and SFP28 port but the QSFP and QSFP28 ports are physically different (wider). The modules I posted are not SFP but QSFP28 muski and Johnnydev 1 1 Custom room treatments for headphone users. Link to comment
Popular Post jabbr Posted February 4, 2023 Author Popular Post Share Posted February 4, 2023 Clarification of terminology: SFP = 1Gbe SFP(+) = 10 Gbe lane quad or QSFP(+) = 40 Gbe SFP28 = 25 Gbe lane QSFP28 = 100 Gbe SFP56 = 50 Gbe QSFP56 = 200 Gbe QSFP-DD = 400 Gbe (8x50) QSFP112 = 400 Gbe (4x100) QSFP112-800 = 800 Gbe (8x100) and that’s all I know 🤷🏻♂️ Johnnydev, Superdad and MFJG 2 1 Custom room treatments for headphone users. Link to comment
TRHH Posted February 5, 2023 Share Posted February 5, 2023 My current setup is based on StarTech single mode: Link to comment
SQFIRST Posted February 9, 2023 Share Posted February 9, 2023 On 1/31/2023 at 9:42 AM, dbastin said: broadband service > cable 1 > ER Side b - side A > fibre > Mikrotik CRS 305 (router) > cable 2 > Antipodes server > cable 3 > WAP > wifi > Devialet @dbastin - just came across your post and the details and wanted to share from my recent experience in case it helps: A single CRS305, connected with your ER SFP(1000base-x), may not be benefiting from the 10gbe protocol as it is most likely using the 1000base-x circuit. Adding a second 10gbe fmc/switch, such as another CRS305, to connect with your current CRS305 Router will then be making a true 10gbe connection and should show effective jitter isolation. You might even find a better position your ER once you see the results. 89reksal 1 Link to comment
dbastin Posted February 10, 2023 Share Posted February 10, 2023 9 hours ago, SQFIRST said: @dbastin - just came across your post and the details and wanted to share from my recent experience in case it helps: A single CRS305, connected with your ER SFP(1000base-x), may not be benefiting from the 10gbe protocol as it is most likely using the 1000base-x circuit. Adding a second 10gbe fmc/switch, such as another CRS305, to connect with your current CRS305 Router will then be making a true 10gbe connection and should show effective jitter isolation. You might even find a better position your ER once you see the results. Thanks, I gathered SFP+/10Gbe spec is the big benefit, not the actual speed ... that is, the spec is the benefit even if the data rate is less than 10Gbe (mine is set to 100MB). Perhaps I am not correct. I can't quite remember how I arrived at my current set up. But for what it's worth in case it prompts someone to suggest something or benefits/inspires someone ... In simple terms, the wifi performed better than 2 x ERs and a Gigafoil (although I since realised some of the cabling between these had shields connected at both ends). Nevertheless, I was distracted/occupied optimising the WAP and learning how to set up CRS 305 as router, I then swapped Ubiquiti Edgerouter X SFP with CRS 305 and realised a good benefit from doing that. At the time I was using a generic FMC (fs.com) and swapped that with 1 ER and realised another good benefit. Then I was experimenting with different cables into that ER. This became somewhat of a distraction that is very satisfying. I am intending to insert a 2nd ER ... Broad band service > ER1 > fibre > CRS 305 > fibre > ER2 > server > WAP --- Devialet ... with a view to using an external clock on ER2, and mindful that ER1 is fantastic as FMC. But perhaps this would be better ... Broad band service > CRS 305 > fibre > ER1 > .. etc ... but this doesn't totally isolate the CRS 305 router with fibre in and fibre out. My guess is the following would be inferior to the above, so I have tried it ... Broad band service > CRS 305 > fibre > CRS 305 > Server > WAP > wifi > --- Devialet Incidentally,,I run fibre from the CRS to the rest of the household network to isolate it from the CRS and hifi. I am happy to consider feedback. Link to comment
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now