Republicoftexas69
Well-Known Member
That’s to bad fiber sounds great in my system.
What ethernet cables are members using?
- Thread starter Lee
- Start date
I guess I'm splitting hairs as the difference is pretty subtle. I might give it another goal and I like the fibre as it's so cheap and flexible.
Republicoftexas69
Well-Known Member
Use quality Finisar SFP using the cheap $3.00 SFP on Amazon can be a crap shoot.
I agree, and if you are using a basic FMC, you also need to upgrade the power supply.
I did use Finisar and it was better than some others I tried as well but putting back my Sablon Ethernet cable made everything sound better.
Republicoftexas69
Well-Known Member
As @audiobomber said "I agree, and if you are using a basic FMC, you also need to upgrade the power supply" and I will say he is more versed in this subject than I. FMC can play a huge roll also. I have also found that different Finisar FSP have different sound and from time to time roll my SFP's
Audiophiling is all about system synergy and personal preference. I found the Sablon 2020 too dark in my system.
With respect to optical fiber implementations, I've been doing this for a while now, and have found every "control factor" with respect to fiber can and does, have an influence on the audio quality. This includes but is not restricted to:
1) The fiber media convertors (FMCs). The better (audio grade) ones, e.g. those used in Sonore OpticalRendus, Uptone EtherREGEN (ER), SOtM products, Lumin, etc, sound better than the el-cheapo, IT- industry grade ones (e.g. TP-Link, etc)
2) The optical transceivers. In my experience, single-mode (e.g, Planet Tech) sound better than generic multi-mode (e.g. TP-link)
3) The optical fiber: For single-mode fiber, I've found that Corning ClearCurve single mode sounds best. For Multi-mode: Tripp-Lite.
4) The power supplies for the FMCs: generic SMPS sound the the worst, LPS sound the best, including those from Sonore, AfterDark, SOtM, etc.
5) Word/Master Clocks for the Ethernet switches or FMCs that have SFP cages. AfterDark makes a nice range of affordable Master Clocks for products e.g. EtherREGEN and SOtM (ER also has a clock input).
6) Clock cables for the Master Clocks for FMCs: better clock cables also improve audio quality.
7) Power distribution for the power supplies for these devices. Audio quality power distributors bring notable improvements.
8) Ground-plane noise reduction: reducing and/or removing ground-plane noise from these devices can provide notable improvment in audio quality. For example, simply connecting a ground cable from EtherREGEN's ground terminal to a Shunyata power distributor GP-NR terminal, Altaira ground hub or Gemini ground hub can provide a notable improvement in audio quality, even "downstream" for components in the main audio rack.
1) The fiber media convertors (FMCs). The better (audio grade) ones, e.g. those used in Sonore OpticalRendus, Uptone EtherREGEN (ER), SOtM products, Lumin, etc, sound better than the el-cheapo, IT- industry grade ones (e.g. TP-Link, etc)
2) The optical transceivers. In my experience, single-mode (e.g, Planet Tech) sound better than generic multi-mode (e.g. TP-link)
3) The optical fiber: For single-mode fiber, I've found that Corning ClearCurve single mode sounds best. For Multi-mode: Tripp-Lite.
4) The power supplies for the FMCs: generic SMPS sound the the worst, LPS sound the best, including those from Sonore, AfterDark, SOtM, etc.
5) Word/Master Clocks for the Ethernet switches or FMCs that have SFP cages. AfterDark makes a nice range of affordable Master Clocks for products e.g. EtherREGEN and SOtM (ER also has a clock input).
6) Clock cables for the Master Clocks for FMCs: better clock cables also improve audio quality.
7) Power distribution for the power supplies for these devices. Audio quality power distributors bring notable improvements.
8) Ground-plane noise reduction: reducing and/or removing ground-plane noise from these devices can provide notable improvment in audio quality. For example, simply connecting a ground cable from EtherREGEN's ground terminal to a Shunyata power distributor GP-NR terminal, Altaira ground hub or Gemini ground hub can provide a notable improvement in audio quality, even "downstream" for components in the main audio rack.
Last edited:
@PumaCat item 5) Expensive/external clocks make no difference in the transmission of data by a network switch in the 1gb or less range. It is possible that they generate less noise which may be transmitted on a copper network cable. item 8) I can understand how connecting a ground cable to EtherREGEN's ground terminal can affect the output on the Ethernet B port. But I don’t understand how that effects the fiber port operation.
Republicoftexas69
Well-Known Member
Every set of measurements I've seen where they connect an external clock to a switch results in higher jitter. I use an EtherRegen with its very good internal clock and have no plans to add external.
The switch with external clock sounds a bit lighter – less rounded – but does allow transients to flow more, which is nice. It is mostly different … but thus not better in all areas we think.
www.alpha-audio.net
Conclusion
It is strange but true: actually the switch with the external clock does not measure as well. We see peaks of the clock reflected in the noise. And with the Dlink, some extra grass shows up.The switch with external clock sounds a bit lighter – less rounded – but does allow transients to flow more, which is nice. It is mostly different … but thus not better in all areas we think.
Alpha Labs - Audiophile Clocks and Network Switches
It is strange but true: actually the switch with the external clock does not measure as well. We see peaks of the clock reflected in the noise.
www.alpha-audio.net
As always, it depends on what is being measured. Most measure jitter, but it depends on wander. This is a low-frequency jitter up to 10 Hz.
What are Jitter and Wander? (technopediasite.com)
UpTone Audio's theory is that the phase noise spectrum of the incoming data is superimposed on the phase noise spectrum of the local clock. This is how phase noise gets from one device to the next, even over optical connections. This phase noise (the same as jitter in the time domain, only expressed in the frequency domain) propagates - both with interfaces such as Ethernet, USB, S/PDIF, I2S, and in the chips on the boards. Even a DAC with a perfect clock sitting right next to the DAC chip will be affected by all the upstream jitter/phase modulation that has taken place in the chips and on the PCB just before it.
Diskussion über "Messungen an Ethernet-Infrastruktur" - open-end-music-professional
Diskussion über "Messungen an Ethernet-Infrastruktur"
www.open-end-music.com
You need an OCXO clock like Mutec REF10 for this. At 10 Hz, the phase noise is a very good -145 dBc/Hz. I use a clock from Afterdark in two Buffalo Switches with the same values. This is not a small but a very big difference in sound! With my system and with my ears: More detail and, above all, all sharpness disappears. The bass is more precise and multi-layered.
You may be right, many audiophiles claim sonic improvements using an external clock, and Uptone now sell Mutec. It's still disturbing to me that measurements clearly deteriorate, even with the Mutec REF10. Seems like there is a tradeoff.As always, it depends on what is being measured. Most measure jitter, but it depends on wander. This is a low-frequency jitter up to 10 Hz.
What are Jitter and Wander? (technopediasite.com)
UpTone Audio's theory is that the phase noise spectrum of the incoming data is superimposed on the phase noise spectrum of the local clock. This is how phase noise gets from one device to the next, even over optical connections. This phase noise (the same as jitter in the time domain, only expressed in the frequency domain) propagates - both with interfaces such as Ethernet, USB, S/PDIF, I2S, and in the chips on the boards. Even a DAC with a perfect clock sitting right next to the DAC chip will be affected by all the upstream jitter/phase modulation that has taken place in the chips and on the PCB just before it.
Diskussion über "Messungen an Ethernet-Infrastruktur" - open-end-music-professional
Diskussion über "Messungen an Ethernet-Infrastruktur"www.open-end-music.com
You need an OCXO clock like Mutec REF10 for this. At 10 Hz, the phase noise is a very good -145 dBc/Hz. I use a clock from Afterdark in two Buffalo Switches with the same values. This is not a small but a very big difference in sound! With my system and with my ears: More detail and, above all, all sharpness disappears. The bass is more precise and multi-layered.
It makes some sense to me to use a master clock, where more than one component is slaved to an external clock. But for just one externally clocked switch, I'm not convinced. Ed Meitner says it is a terrible idea, a clock needs to be a few millimeters from the chip.
JVS: You don't believe in using an external word clock. Why?
EM: Because I think this is the most stupid thing I've ever heard in the audio business. That means you have a precision clock that you have to connect to a wire to connect to a DAC, when the clock should be straight away where it belongs, inside the DAC, beside the DAC chip, if there is such a thing—not through a cable in a different box. This is so idiotic, it's not even funny. It's a money grab.
Meitner on the EMM DV2, DSD, MQA, & Digital Audio
Imagine my surprise while I was preparing my review of the EMM Labs DV2 D/A processor in this issue, EMM Labs' manager of production and social media, Amadeus Meitner, informed me that what I'd thought would be a one-on-one chat with his father, EMM Labs founder and CEO Ed Meitner, would also...
www.stereophile.com
Here are some measurements using a Mutec REF10 and EtherRegen:
"The measurements have shown that it is certainly possible to achieve a better jitter value in the Ethernet signal using an external clock on the EtherRegen. However, the differences are very small. The standard deviation is just less than 1pS. This could be attributed to the inaccuracy of the measuring instruments."
"The differences are much more serious in the common-mode interference. The strength of the interference with a connected clock, with its many peaks across the entire frequency spectrum, is very obvious. There are also interferences in the lower audio frequency range that can have a direct effect here."
Messungen von Ethernet-Infrastruktur - Switches (Nur Lesen) - open-end-music-professional
Eric@Ohrwurm (https://www.open-end-music.com/member/6904-ohrwurm) wird hier seine Untersuchungen zum Netzwerkteil einer audiophilen Streamer-Kette vorstellen. Inhaltverzeichnis: 1. Meßaufbau und -equipment...
www.open-end-music.com
Last edited:
No, these are actual, repeatible, and reproducible results. It's this thing you know, called...science? I should add that I worked as a PhD-level scientist my entire professional career, so I know a thing or do about designing an effective experiment.
I have found the tail end of this thread to be interesting. I have learned new technical details and expanded my thought horizon on network audio. I still have several questions:
1. When using copper cables can someone explain how jitter generates noise which affects the sound created from an Ethernet frame with TCP and IP packets transporting musical data.
2. When using copper cables in the 1gb or less range why do jitter values make a difference other than increasing packet retransmission.
3. The statement “Expensive/external clocks make no difference in the transmission of data by a network switch in the 1gb or less range.” is about the transmission of data over an Ethernet cable and not the generation of sound from that data. Can some one explain why that statement is wrong.
Thank You
1. When using copper cables can someone explain how jitter generates noise which affects the sound created from an Ethernet frame with TCP and IP packets transporting musical data.
2. When using copper cables in the 1gb or less range why do jitter values make a difference other than increasing packet retransmission.
3. The statement “Expensive/external clocks make no difference in the transmission of data by a network switch in the 1gb or less range.” is about the transmission of data over an Ethernet cable and not the generation of sound from that data. Can some one explain why that statement is wrong.
Thank You
Republicoftexas69
Well-Known Member
I will take your word for it. Your qualifications that is. As you did not provide data. And I stand by my statement that these are your experiences and are not reflective of all devices or systems.
I borrowed the following definition from an online source.
Begging the question fallacy is an argument where the conclusion is assumed in one of the premises. It is an attempt to prove something is true while simultaneously taking that same thing for granted. This line of reasoning is fallacious because the assumption is not justified by any evidence.
You beg the question in each of your statements in that you assume the cause is found in one of the data layers of the OSI Model. It's not about the data - it's about the noise that travels with the data as it's transported over the physical layer (Layer 1 of the OSI Model). That noise can make its way down to our DACs and harm the accuracy of the conversion of digital-to-analog. Sound quality can be improved by reducing that harm. It's not just the network that can pass noise - every component along the path can and does add noise, with the computer itself being the worst offender. But once we've taken steps to reduce the harm the computer causes, further gains can be had by addressing the noise our networks can pass.
An external clock that has lower phase noise than the one it replaces tends to pass less noise down to our DAC. When it does that the result is typically an improvement in sound quality. The data is the same regardless, because it has nothing to do with the data or with the packets. The cheap clocks found in consumer switches tend to be noisy enough to harm sound quality. Almost all audiophile switches substitute better clocks that generate less noise. They also tend to be mated to better power supplies that reduce the amount of noise the switch can generate. Again, it's the same data, but less noise is being passed along with the data.
Similar threads
