Audioquest Cat7 cables

Well... I don't want to give too much away but we have a prototype here at BS-Audio (BS is my brother Bert's company...)...

We have a 4 port design, with two audiophile ports (one upstream, one downstream - directional of course) incorporating my Jitt-Red technology. This is groundbreaking technology that has a significant effect on the sources of bad jitter in the system. I would tell you more, but it is a closely guarded secret we regard as our "crown jewels" that other audio companies would love to have, so we are currently lodging worldwide patent applications. Once the patents are granted, we will publish a number of deeply technical white papers on this technology and how it works on jitter and reducing RFI with some very convincing graphs that are totally meaningless. We expect these white papers to generate many hours of discussion on forums as people discuss the  meanings of every word - yes, the technology is that amazing.

The audiophile interfaces are fully reclocked (they have to be, I have to get a digital signal from the Ethernet cable) and opto isolated with specially selected isolators (they are the cheapest we could find, but we did specially select them) and fed from a separate linear power supply. This may be optionally upgraded with external plug in PSUs whilst the standard PSU takes a standard kettle lead plug in order for it to be upgradeable with third party external power cords. All LEDs are able to be switched off for minimal RFI and when delivering audio data along the audiophile port paths, the switch can be set to power off all other functions to minimise RFI (although clearly losing connectivity for the items connected through the other two ports - but for serious listening....). This is our TruQuiet technology. 

The product will be called the PerfectSwitch and our product roadmap has the PerfectSwitch2 (and 3 and 4 etc...) to be released every 9 months after the original launch probably with slogans such as: "This changes everything." and "This changes everything. Again" and "Improved Perfection." Special Edition versions - which will clearly be sonically better - will also be available at an appropriate price premium with hand selected components and tighter tolerances.

The PerfectSwitch is all housed in an elegant but understated black full width case both to improve RFI performance and for matching with other equipment. Chic designer enclosures will be available later in 2013 for those who want to be proud of their PerfectSwitches and have them in pride of place in their audio setup.

Although the results aren't fully in yet, we are finding that placing the PerfectSwitch on an isolating platform further enhances audio performance. In fact we were quite shocked at how much better the performance is on a platform - it shouldn't be, but it is - and it is plain for all to hear. The better the isolation, the better the sound with an improvement in air and space around instruments together with greater clarity and overall tonality together enabling the listener to connect with the musician more. 

Of note though we are finding that no one manufacturer has the ideal combination of isolating technologies so we are developing the PerfectLevel to be used in conjunction with the PerfectSwitch for the ultimate sound quality. The PerfectLevel really does raise the performance of the PerfectSwitch significantly, but it is unfortunately made from hand selected and sourced particleboard and 1/2" copper piping (we hand select them on shopping trips to Wickes) and is manufactured exclusively (by Bert) to our specifications and hand treated with our exclusive PerfectIso manufacturing process only available in our production facilities (my garage). Unfortunately, this is a very labour intensive and costly process (you should see how much we pay Bert!) so the PerfectLevel is unlikely to retail for much less than £800. The jump in performance it gives though has to be heard to be believed.

You also need to know that if buying the PerfectSwich, you really should be using appropriate cables. The old adage a system is only as good as the weakest link really holds here. We budget around 30% of our total system price for decent Ethernet cables (I know it is traditionally thought to be 10%, but Ethernet carries frequencies that are significantly higher than audio so we need to spend extra to make sure they work). Of note though, we have found inconsistent results with cables less than 1m long and suggest cables should be kept shorter than 2m to maximise sound quality. Whilst longer runs will work, we have found them to degrade the overall system performance and suggest that for the ultimate in sound quality, multiple PerfectSwitches should be used together as repeaters (there is a special mode called StraightAudio the intermediate PerfectSwitches switch into to maximise performance here). We also recommend that any non-audio computer equipment should be sited at least 6 metres away from the playback device so our minimum recommended system is 3 PerfectSwitches with associated cabling.

For maximum performance, we recommend using multiple PerfectLevels under the PerfectSwitches and are finding that you should use as many PerfectLevels as you have PerfectSwitches in the system. That is, if you use 3 PerfectSwitches to travel 6m, you should use 3 PerfectLevels under each PerfectSwitch (making 9 in total). 4 PerfectSwitches -> 12 PerfectLevels. Whilst the system will work with fewer (heck, it will even work with no PerfectLevels!)  you will be significantly compromising overall system performance by doing this.

We are confident that the PerfectSwitch is a truly revolutionary and groundbreaking product that will give audiophiles with well setup systems that final magic that enables them to sit back and just enjoy the amazing sounds of their favourite music. Just read what others are saying about our products, you won't be disappointed. We are confident that a BS-Audio Perfect system will be your last upgrade for a significant amount of time (our aim is to drain your bank account to make damn well sure it is )

Do you think they'll sell?

Fair enough Guy,

For me it's the listening. For you, the science test must be passed. I understand your point, but surely you can see mine?

By the way, I assume the people at APC magazine would not be so foolish as to think that the Powerline is a mains "cable for suckers"? No way would they make that schoolboy error. Naim's marketing blurb is not dissimilar to Audioquest's in that it majors on telling us how well it is made, but avoids specific scientific claims because, as Russ Andrews discovered, the Advertising Standards Authority take a dim view of that kind of thing. Despite this, the Powerline is a great product. 

My point is that the science for these types of products is debatable, at best, and yet they work (at least, some do).

Keith

As far as GIGO, I have been saying this for ages.. So in the context of the physical Ethernet level this is managed with physical Ethernet isolators. These are used insensitive medical and industrial situations. These isolators are quite pricey but I did notice a Japanese company was targeting similar devices for the audiophile... I expect a few more to appear for the domestic market.. Again much scope for snake oil.. But an engineering level.. Which is where I come from.. These do make engineering sense and can be measured and built to a specification unlike many of these so called boutique cables.

Pthe other area I have been saying for ages that affects network players is network cleanliness.. Ie the amount of broadcast frames that are on the link.. Since each such frame has to be processed by the the network player TCP/IP stack whether it is intended for it or not thereby causing processing load... I mitigate this by using a seperate VLAN for audio as I have a busy LAN  .. I do this in my professional role for clients as well for similar reasons..To me it works and measurable .. Ie less network overhead to the player.. And another reason I am more comfortable separating my DAC from my Ethernet network player... certainly if you can hear a difference between WAV and FLAC for example.

Simon

The company you are refering to is Acoustic Revive and their ethernet isolator is called the LAN Isolator RLI-1. This device is of better quality than the devices of medical standards. Test results are available on their website if someone might be interested.

And I can confirm that this device improves sound quality very noticably. It only took me 2 seconds to realise that.

Now I do realise that all my devices and connections are absolute crap since they come from our beloved audio manufacturer and some miserable companies that produce professional grade network devices.

But then are not all our devices flawed? They are all strung together compromises.

-

aleg

I have it and can confirm: it makes an audible difference.

Is it better than a $2 choke on the ethernet cable?

My recommendation: UnitiQute as a streamer into a Naim Dac/555PS. UnitiQute still remains the best VFM streamer IMHO. I'm not convinced that the NDX is that much better used as a pure streamer.

Whatever gunk is generated due to the ethernet cable should stay in the UnitiQute box and shouldn't be transmitted into the Dac.

v amusing!

Yes, it is; I tried chokes on the ethernet cable.

Unless you wish to break the laws of physics it is absolutely impossible for the ethernet cable to have an effect on the sound quality of your equipment. this is a complete and unescapable FACT verified by 100's of years of science. The ethernet cable doesn't even have anything to do with the sound reproduction - it is just a mechanism for moving the digital information completely unaltered from one place to another - I will repeat it again - ETHERNET CABLE HAS NO EFFECT OR INVOLVEMENT IN SOUND REPRODUCTION WHAT-SO-EVER.

Here's a though experiment for you snake oil believers out there:

You have two NAS server S1 and S2. You have three cables C1,C2 and C3. C1 and C2 are cheap Maplins Cat 5e, C3 is some sort of expensive snake oil cable. You have a naim streamer N1. You have a digital file FLAC1 stored on S1.

You connect N1 to S1 using the C1 and listen to FLAC1 - this should be you baseline.

You connect S1 to S2 using C2 and transfer FLAC1 to S2 - we will call the copied FLAC FLAC1.1

You connect N1 to S2 using C3 and listen to FLAC1.1

Now are you saying that FLAC1.1 on S2 will sound better than FLAC1 on S1?

FLAC1.1 will be identical in every single way to FLAC1, totally and utterly indistinguishable - it is impossible for there to be a difference unless something is catastrophically wrong with you equipment.

If you believe the snake oil cable merchants then you will have to say that FLAC1.1 will sound better. BUT FLAC1.1 has already been transferred through C2 which is a cheap cable, and yet you seem to be saying it has had no effect on your sound quality.

The same thing happens when transfering the digital information to the NAIM - it is transferred perfectly over what ever cable you use and stored in a buffer - the data stored in the buffer is a perfect digital copy of the data on the disk and it is from the buffer that it is decoded - the way the perfect copy got into the buffer is irrelevant, much like when you transferred FLAC1 to FLAC1.1 using a cheap cable.

So, if you accept the unescapable and verifiable fact that the the cheap cable can transfer the data faultlessly from one server to another, then you have to accept that fact that the cheap cable can transfer the data perfectly from the server to the streamer. 

In the days of analogue where degradation did happen on transfer then this was an issue, but with digital this issue is nonexistant.

All of this is indisputable FACT. If you do not believe it then you are denying physics, fault tolerant protocols and years of human achievement in computer science. 

It's not about the bits

More about poorly designed yet very expensive hi-fi components if RFI is getting in and affecting the sound that much.

Erm, don't understand what you are saying here - The end result of the protocol stack, for an end user, is that it does just work. If it didn't we wouldn't use it. For example if it didn't just work, I would not be able to post this message error-free on these forums.

BTW - IP Protocol Stack - http://en.wikipedia.org/wiki/TCP/IP_model

Conrad - you appear to be confusing different application error recovery methods with ethernet data transmission. Its the latter I was referring to as we are talking about ether patch leads(!). It is definitely not a case or error free for a whole host of reasons which is why the OSI 7 layer model was developed to secure digital payload.

Am intriqued about your reference to the IP Protocol Stack - we are talking Ethernet data transmission. Further if you really want to understand TCP/IP I recommend a proper technical text book than relying on a Wiki - Wikis perhaps here are leading to confusion?

So to be clear:

Ethernet patch leads are at the physical signal transmission level - there is no concept of bits or bytes and the signals can be littered with errors/distortions.

The Link layer is the first layer that converts these signals into a basic frame of binary inforamtion. This is checksummed for integrity - and encpsulated by higher protocols for framing, transport, session and presentation and finally application (uPNP/XML for streaming) - conceptually look up the OSI 7 layer model which was the foundation of the TCP/IP Stack.

The idea of 7 layer model /TCP/IP stack (which I have spent my whole career working with or when I was younger developing standards for [X.400/X.500 Series]) is to provide a more reliable and more abstracted from the physical signal transfer way of communicating information the higher you go  up the stack. This is neccessary becasue of inherent unrelabilites of data network transfer the further you go down the stack. IT is *this* reason why you can post on the forum using the internet and TCP/IP.

Finally may I respectfully say your view appears blinkered - yhou need to think outsie the box and take a more holistic engineering system view. You are stating 'FACTS' (BTW the more you shout the less people hear) that clearly are not as evidenced by people's feedback. Sound quality can be affected by many reasons that are not related the protocol mechanisms of the TCP/IP stack. Andy_S mentions one, I have another.

I spent several stressful weeks this last summer with my team having to fix low level network traffic issues that were polluting a clients network, increasing CPU load of protocol stacks and casuing all sort of side effects. The 7 layer model was however working wonderfully... !

Simon

Thank you for your response Simon.

Please could you explain to me your theory of how the performance, construction and quality of a functional ethernet cable can affect the sound quality I hear when I listen to my NDX.

I need to understand how, once the bits are in the NDX buffer, the mechanism by which they arrived in that buffer can have an affect on the decoding that occurs after that. With of course the premise that the cable is working to specification and therefore introducing no spurious digital data into the bits in the buffer. 

It's not about the bits.

Conrad, Aleg has answered.. It's not about the bits. It's about the associated analogue signals in the patch lead conductors and shield if it has one either induced and/or conducting through the leads.

Simon 

So the "everything's perfect once it reaches my streamer so the network doesn't make any difference" camp have been vanquished and we're back to  the original question: can funky ethernet cables cause a significant improvement when the science only allows for a marginal difference? My experience says yes, but I acknowledge that this is difficult to accept and rationalise.

Keith

As I understand this discussion, there is the "science" of what we want the cable TO transmit, but also the science of what we want the cable TO NOT transmit.  Some are only focusing on the first half of the sentence, if I understand.

Hi Bart,

Yes, it seems so.

Perhaps they're only commenting on a third of the science, Simon's explaining two thirds and there's another third we don't know about? That could explain why the cables work :-)

Fascinating to read that the laws of physics, particularly as pertains to ethernet cables, have been known for hundreds of years.

I also love this idea that a buffer somehow overcomes any shortcomings in transmission quality, seamlessly reconstructing any degradation of the digital information at a speed that allows replication of the transient nature of the music. One could even start to believe that computers really can endlessly copy information perfectly and wonder why manufacturers don't offer longer warranties on their hard drives, flash memory etc.

When I watch my digital television why do I sometimes get pixellation, frame freezes and high pitched squeaks? Isn't there a buffer that could sort that for me?

Or when I open an internet page  and it says 'Done but with errors on the page'. And that's a simple visual, generally static data set. 

Fascinating, truly fascinating. And asking experts being a better method than using your own ears to establish improvement in SQ. Italian earthquake 'experts' anyone?  I just can't keep up.  Shouldn't wi-fi be as good as wired if the transmission medium makes no difference? Oh my.... need a lie down. 

There's always a possibility that an unknown variable will be discovered  ( and maybe become gospel) in the future.

When 'perfect sound forever' was introduced, who knew anything about jitter?

Most engineers trashed the idea that a digital cable could make a difference.

Snake oil.

The only argument in favour I can find hidden above is the possibility of a small detectable effect on sound from RF interference

If you want to test this in your system, buy a few RF chokes for £1 each and put them on the ethernet cable end nearest the relevant unit.

Whichever way the cards fall, you still won't need the expensive ethernet cable.

@pt109 most *true* engineers obviously know that a digital cable acting as a transmission line provides  a whole host of variables that will cause many effects/differences within a system performance. Engineers know that systems work in a way that is way more involved that simplistic sales/marketing diagrams, wikis  and summaries. *True* engineers don't deny observed outcomes. Measurement and analysis is part of the discipline of a professional engineer including subjective mean opinion scores.

There was an item on the local TV news the other which included an interview with a woman with permanent nausea. She was waiting for an operation to insert a pacemaker into her stomach. Apparently, it relieves the symptoms, but no one knows why. 

it is now thought that the high frequencies which are lost in 16 bit music affect the sound, even though they are well outside the range of human hearing. This was not accepted in the '80s and not universally accepted now.

Russ Andrews has been prevented from claiming that his mains leads combat the effects of RFI by the Advertising Standards Authority, based on expert advice. The evidence presented by RA has been rejected because the expert believes it deals with the wrong form of RFI (sorry if I've got any of his wrong!).

Naim make no claims, in their marketing "puff" as to why their mains leads improve sound. They simply describe their construction, which includes a degree of mechanical isolation.  

Most audiophiles accept that mains products work, but it's very difficult to prove why or measure the effects. You just hear them.

Keith