Audiophile modified switch - improving ND555 or not?

you ask if the cisco 2960, which improves the nds, will improve the nd555 also?

Personally i don’t know but guess it will.

I once tried an audiophile switch, tcxo from paul pang, but was not impressed.   There is also the telegartner switch, for around 6k.  Uptone audio will produce one soon.

Or the aqvox se switch too.

Yes, due to ND555 improvements in clock & buffering. Will it make fancy or 2960 switch or ethernet cables useless or is there still a difference like with NDS. At least Darke Bear was reporting ethernet cable differences but he uses Melco.

Also asking if anyone using “audiophile” switch, meaning better than Cisco 2960. I use HP with better linear power supply + better ethernet cable + NDS and the difference is very clear. Thinking if ND555 is doing this same by the default + naturally some extra improvement but is it really worth of money? Yes, I will test it some day.

I know Paul pang, Telegartner & Aqvox. Uptone was new to me.

yes Gandalf, your question is interesting.   But i doubt we will have an absolute answer.   Some prefer expensive chord ethernet cables, other cheap ones.   But perhaps, as you said, the nd555 will change the things.

No absolute, wondering if 2960 or AQ/Chord cables are broken with ND555? Just cheap basic switch and cheap cable does the trick?

This is even more relevant if you consider that DB considered a posh Ethernet cable actually deteriorated SQ when used on a ND555, preferring a cheap Ethernet cable instead.

Has the superior clocking/jitter control/noise rejection/buffering, or whatever the ND555 does better than the NDS, turned things on their head so that now we no longer need posh Ethernet cables, device power supplies or switches to reject/isolate noise?

I have no idea. Maybe Naim or those with techie knowledge could chime in.

nigelb posted:

This is even more relevant if you consider that DB considered a posh Ethernet cable actually deteriorated SQ when used on a ND555, preferring a cheap Ethernet cable instead.

Has the superior clocking/jitter control/noise rejection/buffering, or whatever the ND555 does better than the NDS, turned things on their head so that now we no longer need posh Ethernet cables, device power supplies or switches to reject/isolate noise?

I have no idea. Maybe Naim or those with techie knowledge could chime in.

but this ethernet posh cable was connected to the melco....so perhaps with a connection to a switch, things could be different.....or not.

nigelb posted:

This is even more relevant if you consider that DB considered a posh Ethernet cable actually deteriorated SQ when used on a ND555, preferring a cheap Ethernet cable instead.

Has the superior clocking/jitter control/noise rejection/buffering, or whatever the ND555 does better than the NDS, turned things on their head so that now we no longer need posh Ethernet cables, device power supplies or switches to reject/isolate noise?

I have no idea. Maybe Naim or those with techie knowledge could chime in.

Yes, that is the question, want to know more...

As i have said before, if you don’t notice a change in the audio over the duration of a track then the network has nothing to do with sq of the replay. Earth loops / circulating earth currents through network patch leads might.

i love the new architecture with its larger spool buffer from Naim because it renders obsolete many of the effects we heard from the first gen streamers and allows quite easy  differentiation  between changes in SQ,... it’s  really simple and easy to test... when playing a track out, is there a SQ change when the Ethernet cable is unplugged...if so is there a difference between start and end of a single track, if not the issue is almost certainly circulating ground currents or common mode currents... simples.. sorry boutique merchandise providers....

Simon-in-Suffolk posted:

As i have said before, if you don’t notice a change in the audio over the duration of a track then the network has nothing to do with sq of the replay. Earth loops / circulating earth currents through network patch leads might.

i love the new architecture with its larger spool buffer from Naim because it renders obsolete many of the effects we heard from the first gen streamers and allows quite easy  differentiation  between changes in SQ,... it’s  really simple and easy to test... when playing a track out, is there a SQ change when the Ethernet cable is unplugged...if so is there a difference between start and end of a single track, if not the issue is almost certainly circulating ground currents or common mode currents... simples.. sorry boutique merchandise providers....

I feel like we're finally getting to a place where it really IS the fact that "1's and 0's sound like 1's and 0's."  I know I'm over simplifying, tremendously, but if the streamer/player has a large buffer out of which it plays the music, the opportunity for a fancy switch or lead to make a difference to said music playback becomes rather minimal perhaps??

To a point.. the Ethernet has always been delivering as you put it ‘1s and Os’. That never really varies. However it’s the processing to receive those 1s and 0s which have caused side effects, as well as RF loading and common mode circulating currents from attached leads. 

Now with the new architecture I believe you can differentiate any processing of network data from the effect of common mode circulating currents, and to a certain extent RF loading.

So if there is no SQ shift between loading the track and then playing from the spool memory then that kind of rules out data network processing interference.

Then when playing the track from the spool and then pulling the Ethernet plug and a shift in SQ occurs.. then this most likely will be because of common mode circulating currents from the lead.

if there is no change evident at all with any of these things the ND555 in that setup has met one of its design objectives and has become isolated from the effects of network delivery and connectivity... so standard commercial 50 pence a metre Ethernet leads can be used and any compliant network equipment can be used with no risk of SQ shift.

Therefore to me that Ethernet lead removal test is so important... it could spare the user from a lot of uneccessary expense.

Simon-in-Suffolk posted:

To a point.. the Ethernet has always been delivering as you put it ‘1s and Os’. That never really varies. However it’s the processing to receive those 1s and 0s which have caused side effects, as well as RF loading and common mode circulating currents from attached leads. 

Now with the new architecture I believe you can differentiate any processing of network data from the effect of common mode circulating currents, and to a certain extent RF loading.

So if there is no SQ shift between loading the track and then playing from the spool memory then that kind of rules out data network processing interference.

Then when playing the track from the spool and then pulling the Ethernet plug and a shift in SQ occurs.. then this most likely will be because of common mode circulating currents from the lead.

if there is no change evident at all with any of these things the ND555 in that setup has met one of its design objectives and has become isolated from the effects of network delivery and connectivity... so standard commercial 50 pence a metre Ethernet leads can be used and any compliant network equipment can be used with no risk of SQ shift.

Therefore to me that Ethernet lead removal test is so important... it could spare the user from a lot of uneccessary expense.

If the new bigger buffer is the saviour, then surely pulling the ethernet is irrelevant as when done the streamer is playing from the buffer that had the ethernet in place.

Obsydian posted:

Simon-in-Suffolk posted:

To a point.. the Ethernet has always been delivering as you put it ‘1s and Os’. That never really varies. However it’s the processing to receive those 1s and 0s which have caused side effects, as well as RF loading and common mode circulating currents from attached leads. 

Now with the new architecture I believe you can differentiate any processing of network data from the effect of common mode circulating currents, and to a certain extent RF loading.

So if there is no SQ shift between loading the track and then playing from the spool memory then that kind of rules out data network processing interference.

Then when playing the track from the spool and then pulling the Ethernet plug and a shift in SQ occurs.. then this most likely will be because of common mode circulating currents from the lead.

if there is no change evident at all with any of these things the ND555 in that setup has met one of its design objectives and has become isolated from the effects of network delivery and connectivity... so standard commercial 50 pence a metre Ethernet leads can be used and any compliant network equipment can be used with no risk of SQ shift.

Therefore to me that Ethernet lead removal test is so important... it could spare the user from a lot of uneccessary expense.

If the new bigger buffer is the saviour, then surely pulling the ethernet is irrelevant as when done the streamer is playing from the buffer that had the ethernet in place.

The bigger buffer does not seem adequate for a full track, but aside from that I think we've discussed in the past that it's got nothing to do with the accuracy of the bits transferred to the buffer but is all to do with electrical 'noise' introduced into the streamer via the ethernet cable (from the network of perhaps RFI) such that different cables may or may not mitigate effects of the connection.  So if unplugging the ethernet when playing from the buffer yields positive results, then different ethernet cables may produce different sounds, if not you either have a good cable to start with (cheap or otherwise), or there's no 'noise' problem.

But once the buffer has depleted then the music ends. My point being it is a flawed suggestion this proves cables make any difference negative or positive.

Obsydian posted:

But once the buffer has depleted then the music ends. My point being it is a flawed suggestion this proves cables make any difference negative or positive.

I don’t see why this is such a difficult idea for the layman to understand. If the Ethernet cable connection is, by whatever means, affecting sound quality in any way, that effect will stop when it is removed. The music that then plays from whatever is left in the buffer must then be unadulterated by such effects. If you hear no difference, there are no such effects. 

Obsydian posted:

But once the buffer has depleted then the music ends. My point being it is a flawed suggestion this proves cables make any difference negative or positive.

I don't follow - the data in the buffer can be assumed to be 'perfect' - it's the playback of the buffer content that is affected by 'noise' which will influence DSP/DAC/and analogue circuitry as this data is processed - if such noise is being introduced by ethernet connections, removing the connection will remove the noise affecting playback of the remainder of the buffered audio.

ChrisSU posted:

Obsydian posted:

But once the buffer has depleted then the music ends. My point being it is a flawed suggestion this proves cables make any difference negative or positive.

I don’t see why this is such a difficult idea for the layman to understand. If the Ethernet cable connection is, by whatever means, affecting sound quality in any way, that effect will stop when it is removed. The music that then plays from whatever is left in the buffer must then be unadulterated by such effects. If you hear no difference, there are no such effects. 

i am not sure it can be an absolute theory.  I can’t explain, but something doesn’t work in this idea.  Just my feeling....

ChrisSU posted:

Obsydian posted:

But once the buffer has depleted then the music ends. My point being it is a flawed suggestion this proves cables make any difference negative or positive.

I don’t see why this is such a difficult idea for the layman to understand. If the Ethernet cable connection is, by whatever means, affecting sound quality in any way, that effect will stop when it is removed. The music that then plays from whatever is left in the buffer must then be unadulterated by such effects. If you hear no difference, there are no such effects. 

My point here is whatever is in the buffer was via the cable and whatever affect, if you then remove the cable what remains in the buffer is due to the cable until the buffer ends.

i tend to think similarly.  I can’t prove or explain it, i have no technical knowledge, but what Obsydian says seems more logical to my eyes.. When you remove the cable, the sound is still influenced by it for a moment.  The second aspect for me is that is difficult to mesure differences for a very short time too.  It would be not enough to show the consequences of the network and cables.    But perhaps i am wrong,  it is just my impression.

Guys maybe I am cross purposes, but MR ROOSTER ???? is on the track I was heading, but was the topic was more any ethernet cable plugged is causing noise, either way the same applies.

Or have I not had enough coffee today ????

Obsydian posted:

ChrisSU posted:

Obsydian posted:

But once the buffer has depleted then the music ends. My point being it is a flawed suggestion this proves cables make any difference negative or positive.

I don’t see why this is such a difficult idea for the layman to understand. If the Ethernet cable connection is, by whatever means, affecting sound quality in any way, that effect will stop when it is removed. The music that then plays from whatever is left in the buffer must then be unadulterated by such effects. If you hear no difference, there are no such effects. 

My point here is whatever is in the buffer was via the cable and whatever affect, if you then remove the cable what remains in the buffer is due to the cable until the buffer ends.

The buffer will be the transferred data, that is all.  That data is almost certainly going to have been entirely accurately transferred from the NAS/wherever, the error correction features of ethernet should ensure this - the electrical noise that is introduced from the LAN via ethernet or RFI effects etc will not affect the integrity of the data itself but will affect circuitry processing the data in the buffer to generate the eventual analogue output - if bog standard ethernet cables were not up to the job of transmitting data accurately across the LAN we would soon know it when accessing photos/video/documents from the LAN from a NAS as they would be far more sensitive to data corruption than audio.

Just for fun a couple of weekends ago I was watching the utilities on my QNAP NAS which show network traffic. I could see that data moves to my Nova at the start of each song and only then - I suspect with longer pieces it reloads the buffer during those pieces as well.

Data was transmitted for 5-10 seconds.

I’m only posting this as an indication of the size of the buffer (spool) - some of the tracks I played  (44.1/16) were around 8 minutes long so the buffer is at least that big. Others can put the technical information more succinctly than me.

My understanding is that you only ever hear data from the buffer, never from the cable (though may only be applicable to the newer streaming solutions).

Obsydian posted:

Guys maybe I am cross purposes, but MR ROOSTER ???? is on the track I was heading, but was the topic was more any ethernet cable plugged is causing noise, either way the same applies.

Or have I not had enough coffee today ????

hey Obsydian, perhaps i have not correctly understood of what was about?  i wanted to say that i don’t believe that, if you remove the ethernet cable for a short time, it can explain that this cable had no influence if the sound is the same.

OBSYDIAN posted:

My point here is whatever is in the buffer was via the cable and whatever affect, if you then remove the cable what remains in the buffer is due to the cable until the buffer ends.

???? And FRENCH ROOSTER also wrote,....I tend to think similarly.I can’t prove or explain it,I have no technical knowledge,but what Obsydian says seems more logical to my eyes..

◾ I agree with you.!!
There is no need,in this case,to have technical knowledge,..."Just listen... As Simply As That".

Now we know that Simon in Suffolk does not believe in,...or in contrast to many,many other,..hear no difference between ehternet cables.
That is,in any case,the view that I have received.

I remember the thread that Hungryhalibut started two years ago about the Cisco 2960-Switch,he heard a difference and recommended Audioquest Vodka Ethernet cable in that thread.
Simon dismissed this,or tried to technically explain that it played a minor role if you used "posh-cables".

But I bought three AQ Vodkas because of Hungryhalibuts recommendations,and anyone who listened to them has heard the difference.

But we are all different,my advice is to trust your own hearing and be very,very carefully when you are testing.....as well as try to have a logical thinking just like Obsydian and even French Rooster above.

I'm not saying that anyone is wrong,..However, it is very easy to make mistakes when testing different products.

It is quite,...or very common among people with great technical know-how,to dismiss this with the influence of cables in debates on various forums.
Such debates have been going on for decades on forums.!

I ignore various technical explanations,...I/we test.
And I've yet to come across someone who hasn't heard the difference on cables,when we've been testing in groups the last few decades.

Some in our group have the last 7 months tested various ehternet cables,they are incredibly knowledgeable and "Attention To Detail".
I would think that this test of ehternet cables is complete if maybe 5-6 months to.

The best so far,is a ehternet cable from Transparent that costs in Sweden € 600:-.

Those in our group who are testing have read this thread,they wrote like this in a comment to me...

◾ He is wrong.! Everything affects of course. That he does not hear the difference says nothing...????

I can not argue in technical English,but still want to try to get my/our opinion on this topic.

And I have the utmost respect for Simon's post,...but here our perception and experience differ.

/Peder ????

Obsydian posted:

ChrisSU posted:

Obsydian posted:

But once the buffer has depleted then the music ends. My point being it is a flawed suggestion this proves cables make any difference negative or positive.

I don’t see why this is such a difficult idea for the layman to understand. If the Ethernet cable connection is, by whatever means, affecting sound quality in any way, that effect will stop when it is removed. The music that then plays from whatever is left in the buffer must then be unadulterated by such effects. If you hear no difference, there are no such effects. 

My point here is whatever is in the buffer was via the cable and whatever affect, if you then remove the cable what remains in the buffer is due to the cable until the buffer ends.

The buffer holds digital data, just 1s and 0s. It cannot retain RFI and the like. 

ChrisSU posted:

Obsydian posted:

ChrisSU posted:

I don’t see why this is such a difficult idea for the layman to understand. If the Ethernet cable connection is, by whatever means, affecting sound quality in any way, that effect will stop when it is removed. The music that then plays from whatever is left in the buffer must then be unadulterated by such effects. If you hear no difference, there are no such effects. 

My point here is whatever is in the buffer was via the cable and whatever affect, if you then remove the cable what remains in the buffer is due to the cable until the buffer ends.

The buffer holds digital data, just 1s and 0s. It cannot retain RFI and the like. 

???? ChrisSU,....Theoretically Yes,....Practical is something else,it's our conclusions after many, many tests for a long time.

/Peder ????