Audioquest Cat7 cables

Keith, my understanding of the Russ Andrews cae was they couldn't back up the  assertions made in their advertising literature. Clearly if one could say anything one liked and presented as factual without evidence or backup it would be a dangerous and wild place. You can use measured subjectiveness.. E.g a cat food company had in their advertising slogan than 8 out of 10 cats preferred their cat food.. They could back it up with measured  survey data. But that company does not go into a whole load of animal food nutritional information that it can't back up.

If you look at advertising literature of the late 19th through 20th century of technical consimer products... You see there is a big reliance of the nativity of the consumer with often references to subjective words like 'amazing', 'magical', 'wonder' etc. in the late 20th century into 21st we have become more cynical and so the so called scientific 'fact' and 'scientifically proven' is often used by the vendor to counter subjectiveness by the vendor and these rightly need to be scritinised and challenged if appropriate. 

Hi Simon,

Check out the ASA ruling (Google Russ Andrews Advertising Standards). it would interesting to get your views.

The jist of it appears to be that RA submitted a technical paper to support his claims, but it was rejected by the ASA's expert. Apparently there are two types of RFI (DM and CM), but in the expert's opinion, the data only supported the effect on DM. Therefore RA's cables can have no more impact on RFI than ordinary leads.

So Either:

- The expert is wrong;

- The leads don't work, or;

- The leads do work due to an unknown cause.

I agree on marketing blurb, by the way, having worked on product development and Athens struggled to recognise the product when it's described in the brochure!

Keith

a lot of companies are making loads of profit on expensive mains leads.. a mains lead should be electrically and mechanical sound .screening may be a bonus .100s of £ for a lead NONSENSE

I don't know how many Powerlines Naim have sold, but I'm guessing they have a lot of very satisfied customers. The effect is obvious.

With all of this, what works for you (whatever the mechanism) may not work for someone else.Different setups, environments etc. To be fair to RA, he does allow a trial period which seems pretty fair. If you don't like, don't buy. 

PS - hello Andy. Hope all is well with you 

Hi James,

i have no problem with RA leads, although I wouldn't put them on Naim gear.

My point is, with these products, there is a discernable difference that is very difficult to measure or explain. Some experts try to explain it, others simply accept it and some deny it.

UKeith, thanks for the pointer and I read the ruling. It's quite interesting. To me the issue of point 1 was not that their leads reduced rfi, but by doing so they stated it would make your equipment sound better. It was this that seems to cause the issue .. That is there was no formal assessment even with a formal subjective sample of this being the case wthe ithe RA leads ... And therefore the aspect of your audio equipmentn sounding better seemed purely  subjective. 

As  far as the second point.. Yes conducted interference is usually common mode and induced or radiated is differential. But common mode is ultimately differential as well as it depends on your point of reference. Woven shields are known to impede radiated interference at fairly low radio frequencies. So I can only conclude RA somewhere said that their woven shield impeded conducted RFI which wouldn't stack up normally, typically you would need to add a  inductance (ferrite ring for example)  , but this area is very measurable and a whole science set of professionals in its own right (EMC engineers/scientists.. There is no large unknown here it is a very  well defined and highly specialised field with formal standards ) and so should have been easy to demonstrate using an official accredited test lab with no subjectivity if indeed it worked as RA stated.

If you interested the formal compliance standard on RFI and interference is EN55022, which all EU consumer devices must comply with or should comply with......

This is what is interesting with homehub Ethernet over powerline as they are in breach of EN55022 in many applications but they are currently lobbying an exclusion.. A true example of science, politics and money mixing together to cause chaos (IMO) but that's another story....

absolutely not. As a scientist and engineer I am enjoying watching the non scientists grasping at straws (and Simons rather bombastic and content less posts BTW - they contain allusions rather than any information - but don't let that fact put off those of you who are straw-graspers) justifying the purchase of rediculously priced cables. Any decent Ethernet cable will transmit perfect data in all but the harshest of environments and the amount of rfi entering a network player is minimal and will not affect the player that much - or if it does, the player is badly designed (which I admit is a possibility). 

There are NUMEROUS published works that show the brain can be led down the garden path on stuff like this (heck, I even argued long and hard with someone who claimed that a bit identical rip sounded better when the rip had been made with a book on top of the ripping CD player - totally mad). Not one of the people who hear a difference (and I don't doubt they do hear a difference) will even admit to placebo effect being a cause. Well.. I'm sorry guys and gals - you are the ones with your head in the sand. I'm just blunt enough to point it out to you and not molly coddle you with pseudo science and grandstanding...

Andy, that's a shame I thought you were sharing my view as professional engineer who works in this  field.. to vanguish the snake oil merchants by providing informed content rather than subjective assertion ..Never been called bombastic before... Fascinating..... Quite intriguing 

Ill look up your bio to see what you do.. If you are an engineer here perhaps our paths will cross professionally.. it's a small world..

[edit no bio listed apart from some hifi items.. Not much to go on... Oh well .. You are welcome to email off line to share notes and expieriences perhaps meet at an AES meeting if you are a member?]

Hello James - yes everything very well here. Hopefully, all is well with you too  Perhaps we should do some more experiments with Ethernet cables.

On the other hand....

Sorry Simon... Not out to offend, but your posts allow The Believers to hang credibility to incredibility without getting to the possible causes of this problem which basically boil down to a broken player design or placebo. As I think Simon and I proved 6 or so months ago, you can basically drive an nDAC with anything bit perfect and it sound the same - even in a relatively high end system, but that still won't stop The Believers believing massive differences between transports/cables to the nDAC. Same here with Ethernet.

Andy, I don't disagree with you  the placebo effect is significant... But I have just had to agree in a major contract recently  to cater for subjective assessment as well as technical measures to demonstrate  performance levels of how well something is working. This has introduced me to MOS (mean opinion scores) and subjective base lining... Even though it's very inconvenient as an engineer  I can't ignore it, I have to deal with it... and therefore subjective base lining to try and eliminate the placebo effect becomes another science in its own right... Because of course people tend to follow the herd and believe what they want to believe.. Which I think is one of the points you were making..

Pure electronics engineer by trade (close to 30 years ago now). Spent the last 20 or so years working for a silicon manufacturer working on software and hardware for digital audio and video systems. Currently project managing the low-level hardware and device driver software for an IP based networked media player to production. Not sure if you'd call that relevant experience or not...

Possibly a random question but -

could you sketch out those parts of a system (CD and HDD) where the signal could be characterised as 'analogue' and those parts where the signal is represented by the actual 1s and 0s - thought of as 'digital'.

I know CDs are not read digitally - I presume HDDs are also read in an analogue of the extracted digital signal. It seems transmission is 'analogue' - with accuracy checking. It might be easier to specify those parts of the system which deal with simple 1s and 0s.

I don't find Simon's posts bombastic and am glad he contributes.

However, some posts in general confuse matters by presenting lots of small discrete informational points without any conclusion or meta-analysis of the relative importance of the points. These are then taken as encouragement and/or proof by those wanting to evangelise snake oil.

You might sum up the discussion above as:

1. The data gets transmitted fine.
2. A poor streamer/dac may pick-up a small amount of sonically noteworthy RF interference from ethernet cable.
3. Interference is primarily mitigated by use of twisted pair cable (almost everyone has this)
4. If you want to, spend £3 on some RF chokes to see if it makes any difference to you.
5. Either way, you don't need an expensive ethernet cable.

Andy, it sure is.. I am more data networks and voice media now..  I suspect/hope we have quite a lot in common, but our style and approaches differ..  But I think our aims are the same .

Adam, yes slightly random as we were talking Ethernet patch leads.. 

But for CD this 'physical level' would equate IMO to the laser receiver driver and filter in the head unit. Ie the the received signals need to beshaped and tuned and the servo controllers and the laser focus adjusted accordingly.

In the HDD this physical level would be equivalent to the disc heads and servos controlling the. I am afraid I am not up on low level disc head design and electronics.

The Ethernet patch lead is interesting as it uses a differential analogue encoding that is quite distant from the digital data. A similar paradigm is with the encoding od digital signals over a satellite link or satellite TV. 

Adam yes with data networks, it is easier and appropriate to measure the effectiveness on how well you have received the payload, typically inter packet jitter (nothing to do with sample jitter before anyone asks), Round Trip Delay, and packet loss.

However this thread started to go down the side effects of ethernet patch leads, ie those side effects that are nothing to do with the digital load or primary purpose of the patch lead. Therefore you can't measure this in terms of the digital measures.. But you can measure these in terms of cross talk, noise levels, RF earth current etc. these are all analogue measures. Fortunately unless extreme these don't affect our ultimate digital/content  payload.(thanks to protocol stacks etc as we previously discussed). But they may conduct through to and affect sensitive low level analogue circuits and high precision clocks unless completely decoupled and screened such that such perturbations are below the noise floor (in frequency and aplitude) of the desired audio output signal.

Simon

I appreciate this Simon, but 10 years ago, an engineer ( who made sure I was made aware of his credentials) told me ''You are wrong''  and ''Its simply impossible for a digital cable to make any difference whatsoever''. Same with power cables.

Some so-called experts on the 'net claims a linear power supply CANNOT make any difference on a Squeezebox Touch. I must be delirious 'cos I found it made a MASSIVE difference...

When I was in class, my electronics teacher called me crazy when I told him I could detect differences using different speaker cables...

BTW, I don't mean any disrespect to you Simon.

Hi, well I guess there are bad eggs in every profession.. In my experience anyone who has to resort to hyperbole and preaching an unwavering mantra despite people's expieriences I would treat with caution or even as a quack, not least they are not going to be very persuasive unless they can listen. There are very few things that are completely categoric.

Interesting thread....

Surely the core issue here is that discounting quantum effects, everything is analogue, it's just how the engineer chooses to interpret the transmitted signals. In digital transmission different analogue levels, frequencies or state transitions are interpreted as representing digital bits. Various techniques are used to protect the integrity of this data, some of which are obviously operating in the analogue domain, others like checksums and error correction protocols operate in a purely digital sense.

Within the context of Digital Audio, the problem is perspective. If you view the system as purely digital, then it's easy to say things like patch leads etc., have absolutely no impact on sound quality. If, however, you remember that 'digital' is just a way of representing information in an otherwise analogue world, then it's easier to see how this perfect model soon breaks down. Good engineering can minimise unwanted impacts, but it's ultimately about compromise. If you engineer a system to be close to perfect in one dimension, then often you end up compromising other dimensions. The optimum solution is good enough across all relevant dimensions, so can still be vulnerable if you target the attack at one particular aspect.

Personally I don't loose a lot of sleep worrying about Ethernet patch leads - I use Maplins CAT5e as they are cheap and flexible and clip on a couple of ferrite clamps to deal with any induced interference. Happy days for me and my NDS 

With a buffer in place you can fill the buffer with music data .... you could then disconnect the patch lead and simply play music until the buffer ran dry .... during this time you should hear a dramatic improvement or perhaps not .... as you had disconnected the offending source of noise. Unfortunately, the limited buffer size may not make this possible to try.  

So lets us not forget the purveyor of these magic cables is claiming that the digital stream gets better ... when even the most ardent claim discussed in this thread makes no such ascertion. I think we all agree that the bitstream is the same under normal conditions with an on-spec lead. 

I agree that any cable can introduce RFI, EMI and electric noise ... except an optical cable. 

However, does anybody disagree that a magic cable between NAS and switch (or router as the literature says) will not affect SQ. 

Of course you should use decent quality cables, a well made switch ... but to me cables of this ilk will always be made with oil of snake. I still contest the worst part of an Ethernet cable is the RJ45 plug so I'm awaiting the development of the AirRJ45 - in the meantime these do fine. 

LOL Guido. Very good. Now as memory becomes ever cheaper the buffers may hold entire tracks enabling the audiophile to nip round the back of the system as each begins and unplug the ethernet cable before relaxing to enjoy (provided they are not using gapless playback)

The buffer on my SbT actually holds 30 to 45 seconds of music and SQ shows itself completely unimpressed with the plug pulled.

Marky Mark hit the nail on the head.

As to Adams question - everything is analogue. It only becomes digital when it is sampled by a clock. The discrete time sample is what makes a digital system work. Digital design includes a lot of analogue (try designing a memory system that runs at 800MHz without worrying about analogue) techniques, but they are very well understood. So well understood that error detection and correction are only needed in the harshest of environments or potentially lossy data sources such as CDs. Everyone's PCs run with unchecked/uncorrected memory systems for example and they work 99.999999999999999999999999℅ reliably - or close to that!.

The key point with computer based digital audio at home where the environment is relatively benign and data loss is simply not an issue due to the design of the system, the only time you need to worry about the digital signal being analogue from an audio viewpoint is the precise time the sample is converted to analogue. Everything else is right by design. 

The key thing here is that interference can affect the precise time of conversion thus introducing distortion. The clever digital designer does everything he can to get that clock to be clocked at exactly the same time every time. This is quite difficult (i.e. impossible) to do! Things like RFI and poor circuit design just make it even more difficult.. 

I know Julian (sad loss - great man I once spent an hour with just listening to him recount stories of how Naim started) talked everything being analogue in an interview once about the design and implementation of the original CDS (which I was lucky enough to own) but that was in the context of minimising clocking error (or jitter!). That is the key part in digital audio - minimising jitter... The design decisions in the CDS on treating the signal as analogue enabled a D->A converter with less jitter...

There is no magic in digital audio - despite what people might say. Good design and meticulous attention to detail really do matter, but despite what people might claim or say, the interactions and their effects on systems are reasonably well understood. For example, I fully understand how RFI affects digital audio, and I also fully understand how Ethernet leads can get this interference into the system. What I'm saying is that if this interference is getting in enough to make audible differences, the player design is fundamentally broken. If this is the case, you shouldn't be looking to a band-aid of cabling, but more at your choices of player purchase because if that does make a significant difference, you have bigger environmental problems.

That or you are imagining it and it is a placebo... (Now who is being bombastic)

PS.I had 90℅ of this written 3 hours ago, but got sidetracked. Apologies if the post looks out of place, but I'm on a tablet at the moment, so will just post what I have...