Ethernet System for Audio: putting it all together

[davewantsmoore]

1 hour ago, dbastin said:

my guess it would be easier to just listen in your system

This is quite a problematic approach to take ..... given these important things:

 

How error prone it is to descriminate between small differences (or even sometimes big ones)

The performance may not be audible.... but still be important for "robust system design"

We don't necessarily know what "noise or no noise" is supposed to sound like (we could assume that it is obvious, but it might not always be that way)

 

1 hour ago, dbastin said:

shield that contribute to the "sound".

Sounds like a big problem to me.

[MrBurns84]

7 hours ago, The Mad Scientist said:

 

Yes, you could route hi-fi data as higher priority via different QoS settings on a Layer 3 managed switch.

 

Definitely won't have any impact on sound quality, and unless you have a serious internal bandwidth issue, it's not really worth implementing.

Tried that - messing around with QoS priorities and tbh, couldnt tell any difference, as I run a 1000 down, 50 up plan, and with only a 4 user household (2 adults 2 kids) i have never encountered any bandwidth limitations even when the kids are watching 4k streams or gaming.

[March Audio]

3 hours ago, dbastin said:

Many makers are probably impregnating materials in their dielectrics, sheaths, etc

Or just cosmetically dressing off the shelf cable with the obligatory woven plastic sheath 

Edited March 29, 2023 by March Audio

[Guest]

43 minutes ago, March Audio said:

Or just cosmetically dressing off the shelf cable with the obligatory woven plastic sheath 

Indeed. I suspect there are few "audio" ethernet cable "manufacturers" with the ability to design and construct a Cat7 or Cat8 compliant cable, and perform the necessary testing to demonstrate conformity to the standards.

[dbastin]

3 hours ago, davewantsmoore said:

Sounds like a big problem to me.

To some extent I agree, and it is because the manufacturers are humans too and may suffer the same ...

 

3 hours ago, davewantsmoore said:

How error prone it is to descriminate between small differences (or even sometimes big ones)

We don't necessarily know what "noise or no noise" is supposed to sound like (we could assume that it is obvious, but it might not always be that way)

 

But they probably do various things as part of creating their brand sound to distinguish from others.  This was covered in the Ethernet Cables thread long ago.

1 hour ago, March Audio said:

Or just cosmetically dressing off the shelf cable with the obligatory woven plastic sheath 

This probably happens too with some brands and their combination of parts, wires, etc to make their sound.

 

1 hour ago, The Mad Scientist said:

Indeed. I suspect there are few "audio" ethernet cable "manufacturers" with the ability to design and construct a Cat7 or Cat8 compliant cable, and perform the necessary testing to demonstrate conformity to the standards.

I reckon there are a few, but perhaps not many, while the rest outsource the manufacturing to those who have expertise and factories that make ethernet cable and specify materials, quality control etc, etc.

 

My supermarket has a bakery but lots of its fresh baked stuff comes from other suppliers.

[davewantsmoore]

11 hours ago, dbastin said:

their brand sound

Any changes to digital cables would, if they make any change to the sound, not change the sound in the same way system to system, equipment to equipment.

[Assisi]

As mentioned above I acquired an AOC cable.  It has now been in for just over 150 hours.  It has settled very nicely.  I do not know whether there is more to come.

 

The start of a note from a guitar or piano is more precise and there is slightly more decay on the note. The bass is more defined and there is more in the treble.  A real musical feel good.   Once again is all about the proverbial less noise floor.

 

As the cable is optical if there is settling it can only be the SFPs and there is not much to them.  Why?

 

I am very happy with the outcome.  Right now I am up stairs listening and I detect the benefit.

John

[March Audio]

1 hour ago, Assisi said:

As the cable is optical if there is settling it can only be the SFPs and there is not much to them.  Why?

 

You might just want to consider the possibility that the SFP has not changed, but it's your perception that has.

Edited April 3, 2023 by March Audio

[dbastin]

Very Interesting John.

 

Just to recap:

• the cable is between SOtM and Melco Swtiches
• you first replaced a copper cable (what specifically?) with a DAC and noticed an improvement
• Then replaced the DAC with AOC and gained another improvement.

Which change was the biggest step up?

 

Where is the SOtM and Melco switches in your network relative to the endpoint?

 

What next?  Your Dream Machine router and both those switches have more vacant SFPs?

 

[frednork]

9 hours ago, March Audio said:

 

You might just want to consider the possibility that the SFP has not changed, but it's your perception that has.

 

And you might want to consider the possibility  that you will never hear (or pay any attention to) any differences that you dont expect to hear, possibly, due to your beliefs and biases.  Perception issues cut both ways.

[Assisi]

8 hours ago, dbastin said:

• the cable is between SOtM and Melco Swtiches
• you first replaced a copper cable (what specifically?) with a DAC and noticed an improvement
• Then replaced the DAC with AOC and gained another improvement.

Which change was the biggest step up?

 

Where is the SOtM and Melco switches in your network relative to the endpoint?

The cables are between the SOtM and the Melco.  The Ethernet cable was an Acoustic Revive Red with Telegartner plugs assembled in Melbourne.  With the DAC cable I considered that initially it may have not been as good as the Red.  After a week I thought that there had been an improvement.  That is why I then considered an AOC that I had read about on another Forum.  It arrived and went in and over five days my perception is that there has been an improvement.  Probably slightly better than the DAC.

 

 

The space for the switches is rather confined and it is a messy process to change cables.  I am not overly enthusiastic to make changes.  I intend to do a comparison when I have time.  I have on order a DAC cable used by Taiko in their latest switch connection for the Taiko extreme.  Interesting to compare.

 

 

The out from the Melco goes via a Red cable to a Waversa Lan filter and then Vertere to the Weiss 502 then XLR to the DAC.

 

John

[March Audio]

2 hours ago, frednork said:

 

And you might want to consider the possibility  that you will never hear (or pay any attention to) any differences that you dont expect to hear, possibly, due to your beliefs and biases.  Perception issues cut both ways.

 

They indeed do.  However, the claim here, that the SFP is changing its performance audibly over 150 hours use, that it " settles in" and improves is extraordinary.

Edited April 4, 2023 by March Audio

[MLXXX]

42 minutes ago, frednork said:

Perception issues cut both ways.

Often said on this forum. However the perception issues do not cut equally, both ways.

 

A small or subtle difference could certainly go unnoticed to a person not expecting to hear any difference.

 

However, a "night and day" difference would be noticed, even by a person not expecting to hear a difference  (assuming they have healthy hearing).

 

So for example if you swap DACs and one DAC is 3dB louder than the other, a difference will be noticed, even by a person who believes that all DACs sound the same.

 

(I know this thread is more for practical tips than theoretical discussion, but I didn't feel I could let frednork's remark pass without comment.)

Edited April 4, 2023 by MLXXX

[frednork]

21 minutes ago, MLXXX said:

Often said on this forum. However the perception issues do not cut equally, both ways.

 

A small or subtle difference could certainly go unnoticed to a person not expecting to hear any difference.

 

However, a "night and day" difference would be noticed, even by a person not expecting to hear a difference  (assuming they have healthy hearing).

 

So for example if you swap DACs and one DAC is 3dB louder than the other, a difference will be noticed, even by a person who believes that all DACs sound the same.

 

(I know this thread is more for practical tips than theoretical discussion, but I didn't feel I could let frednork's remark pass without comment.)

And I was "only" responding to March audio's comment so this is how these types of comments are a big waste of time

[March Audio]

1 hour ago, frednork said:

And I was "only" responding to March audio's comment so this is how these types of comments are a big waste of time

 

Your disagreeing with my point of view does not make it a waste of time.

 

We have 2 scenarios.  One is highly improbable, in fact has no known explanation. It would be seen in general networking errors everywhere if it happened.  

 

The other is highly probable and easily demonstrable.

 

Edited April 4, 2023 by March Audio

[dbastin]

1 hour ago, Assisi said:

Probably slightly better than the DAC.

Hmm, so these are slight changes in the context of a quite elaborate collection of ethernet devices.  Good for you but perhaps not apparent in other systems.  I imagine if you removed the Waversa Lan filter the differences would be more obvious.  Just a thought; the AR Red from your Melco to Lan filter is possibly the next weakest link.

[Assisi]

1 hour ago, MLXXX said:

Often said on this forum. However the perception issues do not cut equally, both ways.

 

A small or subtle difference could certainly go unnoticed to a person not expecting to hear any difference.

 

However, a "night and day" difference would be noticed, even by a person not expecting to hear a difference  (assuming they have healthy hearing).

 

So for example if you swap DACs and one DAC is 3dB louder than the other, a difference will be noticed, even by a person who believes that all DACs sound the same.

 

(I know this thread is more for practical tips than theoretical discussion, but I didn't feel I could let frednork's remark pass without comment.)

Your post is theoretical, not relevant and may indicate a confusion.  If you think that I was saying that there was night and day difference with the cables you are exaggerating.  With the AOC I said:

  “Once again is all about the proverbial less noise floor”

 

The importance of lowered Noise floor is something that you regularly acknowledge that you do not understand.  You are not the only one.  To me, my switch digital front end is all about lowering the noise floor.  It may be a subtle.  It is the difference between just good playback and wonderful.  An impact on lowering noise floor will often be of a minimal benefit.   Nevertheless, it is important and once experienced never forgotten.  The benefit of each implementation can be cumulative.  The search is continual.

 

 

You mentioned DACs.  My mention is of a DAC cable.  In this case DAC stands for Direct Attach Copper.  It is a copper cable with fixed SFPs connected.

 

John

[Guest]

Here we go again. Moderator Approval has been turned on for all replies. Can we not have a respectful discussion?

A reminder of the BIG RED sign next to the reply box:

 

Quote

We are all passionate about audio and the things that also contribute to the accurate reproduction of audio. We all have varying opinions, ideas, beliefs, and above all, we're damned sure we're right!

 

This is the place for discussion of some of the more divisive and controversial audio discussion topics. Volunteer Moderators will only step into a debate if they become personal or severely off topic.

 

All we ask of you, is be polite and courteous to other members, respect their opinion and its difference to yours, and be open minded!

 

Discussion about which members have given which member emoji responses and what they meant is not for this thread (and have been removed). If you want to ask another member something like that, please do it directly via PM!

[MLXXX]

22 minutes ago, Assisi said:

The importance of lowered Noise floor is something that you regularly acknowledge that you do not understand.  You are not the only one.  To me, my switch digital front end is all about lowering the noise floor.  It may be a subtle.

It's not that I don't understand. 

 

It's that in the sense you use it, there are no recordings to illustrate it, and no measurements of it.  I keep asking for even one recording of an audible difference in noise floor from swapping Ethernet cables, etc, and no can has been able to point me to such a recording.

 

(In contrast there are any number of recordings to illustrate poor S/N ratio, high THD, or an uneven frequency response.)

[Assisi]

1 hour ago, MLXXX said:

It's not that I don't understand. 

 

It's that in the sense you use it, there are no recordings to illustrate it, and no measurements of it.  I keep asking for even one recording of an audible difference in noise floor from swapping Ethernet cables, etc, and no can has been able to point me to such a recording.

 

(In contrast there are any number of recordings to illustrate poor S/N ratio, high THD, or an uneven frequency response.)

The impact of lower noise floor is the mitigation of unwanted spurious noise in various forms that should not be present in the play back.  Unfortunately Noise is often present in many or most of our systems.  It subtly intrudes into the listening experience. Until it is lowered it is not easy to appreciate the benefits when it is not present.

 

A reverse analogy would be the final benefit of a special condiment on what is already a delicious meal.  A subtle enhancement.  One is an addition the other is the removal

 

One I do not do recordings.  Secondly even if you had such a recording, I am unable to understand how you would be able to perceive the enhancement in your system.  You need to hear it in another system.  Not your own.  I understand that you have been invited to have such an experience(s) and you have declined.  Your choice.

 

John

[March Audio]

4 hours ago, Assisi said:

The impact of lower noise floor is the mitigation of unwanted spurious noise in various forms that should not be present in the play back.  Unfortunately Noise is often present in many or most of our systems.  It subtly intrudes into the listening experience. Until it is lowered it is not easy to appreciate the benefits when it is not present.

 

So, something that would be easily measurable if present.

 

I mentioned in another thread I was going to do perform precisely this test with various ethernet switches.  I will post the results there.

 

EDIT:  Measurements are up.

 

https://www.stereonet.com/forums/topic/572402-jitter/page/6/#comments

Edited April 4, 2023 by March Audio

[MLXXX]

4 hours ago, Assisi said:

Secondly even if you had such a recording, I am unable to understand how you would be able to perceive the enhancement in your system. 

 

If there is increased background noise because of a noisy Ethernet switch or some other network related issue, all that would be needed to hear it would be to set the gain high during quiet passages when playing back a recording of the system exhibiting the increased noise.

 

The increased background noise would be audible on the most basic system.  And it would be measurable too.

Edited April 4, 2023 by MLXXX

[March Audio]

4 hours ago, MLXXX said:

 

If there is increased background noise because of a noisy Ethernet switch or some other network related issue, all that would be needed to hear it would be to set the gain high during quiet passages when playing back a recording of the system exhibiting the increased noise.

 

The increased background noise would be audible on the most basic system.  And it would be measurable too.

 

Yes.  If noise is of a sufficient level to be heard over the masking of music playing, then it would be obvious if you turned up the volume in a quiet passage.

Edited April 4, 2023 by March Audio

[dbastin]

@March Audio and @MLXXX.

 

The term 'noise' is probably used broadly and imprecisely.  Maybe another way to think of it is interference.

 

Firstly, I gather it's frequency is not audible to humans.  Secondly, it seems to be about how the noise interferes with the proper function of electronics, how that impacts what those electronics is reproducing, and how that impact can be heard/perceived.

 

I reckon there are plenty of other places this subject is discussed and debared more precisely, scientifically, etc.  This thread is not one of them, so please refrain from taking this thread off topic.  You could PM one another or start a thread about noise and see if that takes a similar route to the thread about jitter.

 

In this thread I think the common position (regardless if you agree or not) is noise, jitter and other interferences can/do impact ethernet data transmission via cables, switches, power sources, etc in a manner that can impact how music is reproduced and perceived.

[Guest]

1 hour ago, March Audio said:

 

Yes.  If noise is of a sufficient level to be heard over the masking of music playing, then it would be obvious if you turned up the volume in a quiet passage.

 

I find it perplexing when audiophiles claim that noise is the issue with ethernet, yet the measurements seem to place the noise floor well below the threshold of audibility relative to signal, and even more so when you consider masking etc. Also, when I made a -120 dBFS test tone available, no one else was able to apply enough gain to make it audible. So for most the noise floor is inaudible, but even if they wanted to make it audible, they don't have enough gain to pull it into audibility.....