Ethernet cable installation...

Hungryhalibut posted:

ken c posted:

 

...  initially i will run the cable outside the conduit where other cables are - unfortunately including my 100mm sq radial mains -- but such is life!! :-( 

i will make sure that i will be able to isolate the effects to the cable -- will be interesting. i will make sure the system is sounding excellent before trying... so that i will feel really stupid doing it (in the sense of, why bother? )

enjoy

ken

That's seriously big mains. 

Ooops! sorry i meant 1000mm sq  -- thats what active needs :-)

enjoy...

ken

nigelb posted:

Ken, that is a perfectly fine way of connecting your network. Even so, assuming you stream from the Unitiserve and use the NAS for backup duties, then the hierarchy is still as I have indicated above with the NAS to switch last as this is merely for backup.

Remember this is not a proven hard and fast rule but I have heard benefits from putting the best cable on the switch to NDS and would certainly suggest you experiment with loose ethernet cables on this link first.

Cheers

 Nigel

actually this is interesting nigelb -- why? because when i tried the AQ cinnamon between switch and NAS, i felt the sound had changed in a way i didnt like. from what you say, the NAS to switch link is not normally in use while playing music through my NDS, so i am left wondering why i heard a difference? but as i always say, in this hobby of ours, its important to isolate changes in order to make valid conclusions. who knows, maybe something else was causing the exaggerated sound i heard?? or thought i heard? anyhow -- this test put me off the AQ cable -- rightly or wrongly... Oh dear...

enjoy

ken

Mike-B posted:

nigelb posted:

Mike - are you sure you don't want to try AQ Vodka? 

Hi Nigel ........  like I said .......  I'm willing to try if a buddy wants to bring one to test much like I've done with the Cinnamon.  

I've a very open mind but do find it hard to justify - or maybe I should say I can't imagine - the value in such small SQ gains against the cost of these boutique cables.  I've heard positive gains moving from my previous Supra Cat7+ (1300MHz) to MeiCord Cat6 (250MHz),  so it seems big cat & mhz numbers don't mean much.   That said I do understand electricity - I lectured on it - & also knowing the price of manufacturing cable I am amazed by the exotic sales price of this boutique stuff.

I guess I'm just a tight fisted old skinflint       

Yes Mike, I get your scepticism and understand your technical reasoning. I was teasing, but too suspect posh cables are somewhat overpriced.

Still loving the iFi iPower by the way. What a great bit of kit and definitely not overpriced.

Cheers

ken c posted:

nigelb posted:  so i am left wondering why i heard a difference?

Mike-B posted:

ken c posted:

nigelb posted:  so i am left wondering why i heard a difference?

As I understand it .........  The renderer plays from the data stored in the buffer & not via a cable & the data stream is checked & verified as its loaded to the buffer,  so in theory there cannot be any cable influence.  

However whilst the data stream cannot change - bits are bits (etc) - it is carried by Manchester encoded voltage & this is analogue. The cable pairs have resistance & impedance variables & that can change. Sub-optimal connections such as incorrectly orientated pair length geometry, twist pair distortion etc. can result in near end crosstalk, return (reflection) loss, skew delay & reduced bandwidth changes that potentially can cause data transfer anomalies & it could be audible.   ............   only my theory ........   

 

 

Yeah, what Mike said.

Seriously, my comments re Ethernet cable influence is that the streamer will be...err...streaming from the Unitiserve when listening to music but the link between the switch and NAS is purely for back up duties and not for primary serving duties. Hence my prioritisation of Ethernet cables.

As I say try your 'best' cable on the the link to your streamer. Purely an observation (audition) on my part and no technical justification whatsoever.

nigelb posted:

Mike-B posted:

ken c posted:

nigelb posted:  so i am left wondering why i heard a difference?

As I understand it .........  The renderer plays from the data stored in the buffer & not via a cable & the data stream is checked & verified as its loaded to the buffer,  so in theory there cannot be any cable influence.  

However whilst the data stream cannot change - bits are bits (etc) - it is carried by Manchester encoded voltage & this is analogue. The cable pairs have resistance & impedance variables & that can change. Sub-optimal connections such as incorrectly orientated pair length geometry, twist pair distortion etc. can result in near end crosstalk, return (reflection) loss, skew delay & reduced bandwidth changes that potentially can cause data transfer anomalies & it could be audible.   ............   only my theory ........   

 

 

Yeah, what Mike said.

Seriously, my comments re Ethernet cable influence is that the streamer will be...err...streaming from the Unitiserve when listening to music but the link between the switch and NAS is purely for back up duties and not for primary serving duties. Hence my prioritisation of Ethernet cables.

As I say try your 'best' cable on the the link to your streamer. Purely an observation (audition) on my part and no technical justification whatsoever.

Actually I have just realised that Ken has the SSD version of the Unitiserve so maybe the NAS link to the switch does play a part when streaming and this is why Ken does hear a change when he swaps this Ethernet cable. My Unitiserve is a hard disc version so is the only link from the server that is 'operational' during streaming.

Mike, could this be an explanation?

............  well err um ....... if this cable priority theory is correct then yes.

I'm not that familiar with SSD UnitiServe,  but I understand the data flow when playing is NAS-Switch-Netplayer.  

Guys just remember with TCP the data flow is two way, typically in a duplex connection. Both peers have send receive sequence numbers and Windows that are actively being used. The renderer will be sequencing communications to the control point and media server. There will also be regular lower level data queued out of the renderer handling ARPs, Multicast and other broadcast type data.

That network cable will look like a busy two way street, effectively with lots of sample media lorry trailers interspersed with the two way traffic travelling to the renderer.

Simon

Mike-B posted:

Hungryhalibut posted:

That's seriously big mains. 

...........  and something to SERIOUSLY avoid running an ethernet near.

Similar size to your speaker cables then 

Simon-in-Suffolk posted:

Guys just remember with TCP the data flow is two way, typically in a duplex connection. Both peers have send receive sequence numbers and Windows that are actively being used. The renderer will be sequencing communications to the control point and media server. There will also be regular lower level data queued out of the renderer handling ARPs, Multicast and other broadcast type data.

That network cable will look like a busy two way street, effectively with lots of sample media lorry trailers travelling in one direction arriving at the renderer.

Simon

Also don't forget the (electrically) noisy factories on each side of the street to represent sources of RFI (which gets filtered differently by the electrical characteristics of the each specific cable in use).

The streamer has to deal with all these negative influences at the same time; so, if you add all these different effects together, it's not surprising that there isn't a consistently 'best' cable option for all (or even most) circumstances.

Mike-B posted:

OK OK  I know the network players only need 32MHz/100mbps & the Cat6A 500MHz performance means its got headroom reserve in bucket loads,  & yes the proof of the pudding is in the sound quality  ..........  but my engineering head .............

Yeahhhh 

Now Mike, I suggest focussing on the common mode earth currents as well as the cross talk that you are aware of and your engineering head should appreciate why  different 'data cables' ( price has little bearing  here) can cause connected sensitive audio equipment to sound different.. and shielded Ethernet cable can make this more apparent. These are all sound engineering RF / transmission line coupling design principles. That paper I have posted from Texas Instruments ( the manufacturer of the Naim renderer DAC chips) on the subject is a fantastic engineering tutorial on this (IMO).

Simon

I read it Simon,  yes interesting.  Problem is it leads me on to read other TI papers when I should be doing something more domestic or agricultural   

i am none the wiser in all this highly learned banter. 

i repeat the specific question: in a streaming setup consisting of NDS, UnitiServeSSD, Netgear ReadyNAS Duo and Netgear switch, is the cable from NAS to switch relevant when playing the NDS. Yes or No -- with some explanation please.

enjoy...

ken

Mike-B posted:

I read it Simon,  yes interesting.  Problem is it leads me on to read other TI papers when I should be doing something more domestic or agricultural   

Nahh .. It's a case of priorities 

My agricultural was some tree felling and cropping at the weekend... can't do much more until I burn some of it off...

ken c posted:

i am none the wiser in all this highly learned banter. 

i repeat the specific question: in a streaming setup consisting of NDS, UnitiServeSSD, Netgear ReadyNAS Duo and Netgear switch, is the cable from NAS to switch relevant when playing the NDS. Yes or No -- with some explanation please.

enjoy...

ken

Ken it's relevant in as far it has to work. As far as all this chat of EMI and common mode earth currents.. conceivably if you have some sort of unhealthy common mode RF current flow that is not being correctly decoupled by your network switch.. Then yes different cables here may be audible, but that would ring alarm bells to me of issues elsewhere.

ken c posted:

i repeat the specific question: in a streaming setup consisting of NDS, UnitiServeSSD, Netgear ReadyNAS Duo and Netgear switch, is the cable from NAS to switch relevant when playing the NDS. Yes or No -- with some explanation please.

ken c posted:

i am none the wiser in all this highly learned banter. 

i repeat the specific question: in a streaming setup consisting of NDS, UnitiServeSSD, Netgear ReadyNAS Duo and Netgear switch, is the cable from NAS to switch relevant when playing the NDS. Yes or No -- with some explanation please.

enjoy...

ken

Ask three consultants the same question and you'll get five different answers! 

Mike-B posted:

ken c posted:

i repeat the specific question: in a streaming setup consisting of NDS, UnitiServeSSD, Netgear ReadyNAS Duo and Netgear switch, is the cable from NAS to switch relevant when playing the NDS. Yes or No -- with some explanation please.

I refer my learned friend to my earlier post ......

...... if this cable priority theory is correct then yes.

...........  with US/SSD ....... the data flow when playing is NAS-Switch-Netplayer.  

 

In my setup, without a US, the files go from NAS to switch to the 272. In Ken's case, where the US takes the files from the NAS and then serves them, the files must go from NAS to switch to US, back to switch and finally to the NDS. Is that right? 

All of my wires are AQ and all make a difference, for whatever reason. So in Ken's case it would seem to make sense to use decent leads on all connections. But that is simply conjecture. 

Of course if the US is transcoding then 'data' from the NAS will be quite different from the US. However an electrically noisy device connected on the network may be conducting through to other devices unless properly decoupled or filtered by Ethernet cables, perhaps with lower capacitance to ground. For me I prefer to get a quality engineered switch for decoupling and remove some of the variables.

Simon

Hungryhalibut posted:

Mike-B posted:

ken c posted:

i repeat the specific question: in a streaming setup consisting of NDS, UnitiServeSSD, Netgear ReadyNAS Duo and Netgear switch, is the cable from NAS to switch relevant when playing the NDS. Yes or No -- with some explanation please.

I refer my learned friend to my earlier post ......

...... if this cable priority theory is correct then yes.

...........  with US/SSD ....... the data flow when playing is NAS-Switch-Netplayer.  

 

In my setup, without a US, the files go from NAS to switch to the 272. In Ken's case, where the US takes the files from the NAS and then serves them, the files must go from NAS to switch to US, back to switch and finally to the NDS. Is that right? 

All of my wires are AQ and all make a difference, for whatever reason. So in Ken's case it would seem to make sense to use decent leads on all connections. But that is simply conjecture. 

"In Ken's case, where the US takes the files from the NAS and then serves them, the files must go from NAS to switch to US, back to switch and finally to the NDS. Is that right? "

you have hit the nail on the head HH -- precisely what i was thinking and hence my question. Thanks.

enjoy

ken

Simon-in-Suffolk posted:

Of course if the US is transcoding then 'data' from the NAS will be quite different from the US. However an electrically noisy device connected on the network may be conducting through to other devices unless properly decoupled or filtered by Ethernet cables, perhaps with lower capacitance to ground. For me I prefer to get a quality engineered switch for decoupling and remove some of the variables.

Simon

not sure i understand this at all in connection with my question. my system sounds very good with the cable that came with the NAS connected between switch and it. i have now replaced that with Meicord, just out of curiosity -- messing about if you ask me. strangely when i tried the AQ cinnamon in the same place -- i eventually removed it as i felt  it was tilting the sound in a way i didnt like at all. then i heard that the NAS to switch cable is irrelevant when playing my NDS, which confused me why i heard a difference with the AQ. but now it seems its relevant.

i dont believe i have an lectrically noisy device, or it i have it, it is very selective with its effects. that is, its benign when i am using the bog std cable, but kicks in and does funny things when i use the fancy AQ -- which i gather is screened.

Ho hum... streaming is easy right? this must explain why i still have my LP12... :-)

enjoy

ken

The best solution to answering whether or not the cables are better or worse is to try them as you have found already. You heard a difference, so its clear to you that they actually do make a difference whether or not it technically makes sense. Trying to delve deeply into exactly what makes the difference is interesting but I personally was more interested in improving my musical enjoyment of my system. 

Clearly the different cables make the sound change, for better or worse. I tried a few and found the best combination for me in my system, but this seems to vary according to each installation/system. Possibly the best help here is for people to try and give their opinion on the type of musical presentation that each brand/ model of cable gives, as this could narrow the field of cables to try.

Ken wrote : 'not sure i understand this at all in connection with my question. my system sounds very good with the cable that came with the NAS connected between switch and it. i have now replaced that with Meicord, just out of curiosity -- messing about if you ask me. strangely when i tried the AQ cinnamon in the same place -- i eventually removed it as i felt it was tilting the sound in a way i didnt like at all. then i heard that the NAS to switch cable is irrelevant when playing my NDS, which confused me why i heard a difference with the AQ. but now it seems its relevant.'

Ken It was me that suggested that the NAS to switch connection wasn't as important as the US to switch and the switch to NDS because I assumed (wrongly) the the US was a hard drive version doing serving duties and the NAS was a back up device only. I then realised that your US was the SSD version and it is pulling files from your NAS and then serving those files to the NDS. In this case, all the connections are as relevant as each other and I would not suggest any particular hierarchy other that the final step from switch to NDS is POSSIBLY the key one. Which is why I still suggest (assuming you have any patience or energy left) you test the Meicord and the AQ Cinnamon on this final link (switch to NAS).

Agree with the above - my set up is exactly the same. Chord C-Stream is used for all of the links, except for a different Chord cable for the final connection (wall socket to NDX in my case).

One thing Mike B suggested to me, ages ago, when using the AQ cables, which have metal plugs and are screened, it that I should lift the Earth at the nas, so that it was only grounded at the streamer. I use a little Lindy 90 degree RJ45 adaptor. Whether it sounds different I'm not too sure, but I can see the logic. The only trouble is, for some reason the AQ wire seems to be locked into the adaptor and won't come out!

I've today put an ifi iPower ultra low noise power supply on the switch - again recommended my Mike - which has also changed the sound, making it seem fuller and more open, and quieter overall. Goodness knows how or why it can make a difference but it certainly has done. I think I like the difference, but need to give it more time.

As Ken says, it was so much easier, in some ways, just to play a record!