CAT 7 - Ethernet Cable - Revelation Audio Labs - USA -

[abhijitnath]

I'm not trying to be wilfully bloody-minded here, but the scientific method doesn't preclude 2. Response 1- debunking the result outright, is obviously closed minded. 3 assumes that there was a result. The reasons that could cause the result to be a false positive are very well known and have been covered extensively in every DBT every constructed. Expectation bias, the shortness of audio memory (under a second), statistically insignificant sample & number of trials, and so on. I'm happy to be a guinea pig for a well conducted trial with a completely open mind- I'm just saying that this one- asking 4 people whether they thought a cable sounded different or not, is not scientific, and hence the results are not valid (except as a fun experiment, which I'm happy to go along with). A double blind ABX test is the only way to go- and btw, there have been several of them over many, many years, whereas the reverse- a scientific test proving (even for analog cables, forget digital) differences between cables, has never been done. Maybe the accumulated wisdom of 50 years of experiments should give us pause for thought?

 

[Thad E Ginathom]

It may result in variance in the latency ( Which is the Jitter ) with too many packet lost during the transit or corrupted packet ( Which may happen with the really really BAD cable ) . But any normal cable , with the back to back connectivity , this will not be really a factor . Assuming some level of buffer is allocated in higher level application ( which can compensate for the jitter ) , I am not sure how this could show the improvement theoretically.

Jitter is caused by bit-by-bit timing errors, eg, a theoretically square wave is not square. There is no such sensing in the network connections under discussion, so jitter is not possible in this link of the chain.

It has been mentioned that it is a prerequisite that all the data should be delivered, that it should be delivered in order, and that it should be delivered in time. I don't think that the protocols under discussion guarantee the final point: we look after in-time by using a mechanism which has a capacity far greater than needed, and thus, there is plenty of time for quite a lot to go wrong and be corrected. Compare this with watching a YouTube video, which may stutter, break down, or take minutes buffering a few seconds of playback.

backtracking a bit...

Also, about the RJ45 cable, there are locking Neutrik connectors available. My Variax guitar has one- this is what it looks like. http://www.vettaville.com/images/cable_ends.jpg

Something I'd never heard of, but then I am not a guitarist ...or even a musician. It looks fascinating, but it doesn't look like it is anything related to networking. Pieces of wire get used for all sorts of purposes, and it looks like here, the wire just happens to be CAT5. And that there is an RJ45 termination inside that Neutrik shell.

Would love to hear more, but obviously this isn't the the thread for it...

 

[vijwilso]

Jitter is caused by bit-by-bit timing errors, eg, a theoretically square wave is not square. There is no such sensing in the network connections under discussion, so jitter is not possible in this link of the chain.

It has been mentioned that it is a prerequisite that all the data should be delivered, that it should be delivered in order, and that it should be delivered in time. I don't think that the protocols under discussion guarantee the final point: we look after in-time by using a mechanism which has a capacity far greater than needed, and thus, there is plenty of time for quite a lot to go wrong and be corrected. Compare this with watching a YouTube video, which may stutter, break down, or take minutes buffering a few seconds of playback.

I remember , Jplay buffers the complete track before it starts playing the tracks . In that case ,The jitter in the segment between music Server and Jplay is not a matter of concern .Should this be a equivalent of playing from the local device !!?

I worked in real time VoIP protocols and Jitter measurement is one of the key metric, even for that matter to mesasure the fwdin the switches and routers . There are cetain RFC for it RFC 2544 is one of it . But here JPlay is not real time application and Jplay buffer the complete track or part of it and then it stream it to the DAC .

 

[Thad E Ginathom]

I believe this post explains it.

SPDIF cables carry music in square waveform while CAT cables carry chunks of data with a CRC checksum that triggers an error or helps in reattempting transfer if the packets are damaged. The software player or the soundcard doesn't appreciate anything that does the work of streaming the data until that point. The errors are taken care by the ethernet card or the I/O chipset and internet protocols before even it is accessed by the player from the RAM. In case of SPDIF, the DAC has no idea if the digital waves reaching it are perfect or damaged so as to correct them, except for analyzing the clock errors. Because such a loss or damage to the actual music waves go undetected it may be making sense to use good quality SPDIF cables for interconnection. Not the case at all with data transfer. Heck, there is nothing like sampling rate in ethernet transfer which has a huge impact on sound quality. I am thinking aloud here. I may be grossly wrong.

The buffering, before or after, is not the fix, it is the actual data transmission method.

 

[vijwilso]

I believe this post explains it.

The buffering, before or after, is not the fix, it is the actual data transmission method.

I think the buffering before streaming-out the data to the DAC should help in a big way . The TCP/IP can provide a reliable mechanism for the data delivery , but it does not guarantee the In-Order-Arrival or In-Time-Arrival of the packets . So the the buffering in the Receiving application (jPlay ) has got a huge advantage

 

[Thad E Ginathom]

It guarantees in order. It has to, or the received data might turn into gibberish. It is fast enough that in time, although not guaranteed, an be taken as a given. Well, PCs are not real-time machines anyway: they don't actually guarantee anything in time.

Remember that the network protocols and mechanisms are [mostly, and, I think, for the purposes of this conversation] content-type independent. There is no "awareness" that this is audio data that has to be streamed for real-time playing.

I know people who choose whole-song play-from-RAM buffering. They say it works for them. I also know people who claim that knocking a few ms off playback latency makes the sound better --- and then they add several whole seconds of latency by whole-song buffering. It's another thing, for another thread. I don't use Windows, so cannot even try jplay, etc.

 

[vijwilso]

TCP sends a chunk of frames with the sequence number attached to the every frame . And each frame can choose a different path in the NW before it reaches the final destination .So the packet chooses the shorter path may arrive soon than the others , Also the resent packets reach outof order. but it will be re-arranged with-in the protocol itself . In that aspect , yes In-oder delivery is guarenteed to the upper layers apps .

'TCP WINDOW mechanism' buffer might not be big enough to compensate for any high latency or any issue with the transmission , while the data is consumed faster by the App .

I am not taking about the buffering by the NW protocol , i was refering to the buffering by the application which is content aware .

 

[abhijitnath]

I think the buffering before streaming-out the data to the DAC should help in a big way . The TCP/IP can provide a reliable mechanism for the data delivery , but it does not guarantee the In-Order-Arrival or In-Time-Arrival of the packets . So the the buffering in the Receiving application (jPlay ) has got a huge advantage

Exactly. We are confusing aspects of transmission protocol (S/PDIF, Ethernet) with the cable transmitting the signal (75 Ohm Coaxial, CAT7). The two are interconnected, but maybe we should debate issues arising from them seperately, i.e jitter can be caused by both (http://grouper.ieee.org/groups/1394/b/copperheads/eh990126-CAT5JitterTxMask.pdf), but the effects can be isolated.

Something I'd never heard of, but then I am not a guitarist ...or even a musician. It looks fascinating, but it doesn't look like it is anything related to networking. Pieces of wire get used for all sorts of purposes, and it looks like here, the wire just happens to be CAT5. And that there is an RJ45 termination inside that Neutrik shell.

Would love to hear more, but obviously this isn't the the thread for it...

Sorry to continue this OT digression-
The Neutrik locking connectors are used in pro networking applications too- it's just a way of fastening them more securely. I remember going to JSA (a large legal firm in Mumbai) and trying to plug into those recessed plug points that conference tables have, and finding a female Neutrik RJ45.
The Variax itself is a digital guitar simulator, i.e. it converts the guitar signal into various types (Les Paul, Strat, acoustic guitar, etc.). Each string has an A-D converter attached, which does DSP per string, sums the signal, and either does a D-A conversion, or sends it as a digital output via the RJ-45 cable.

 

[vijwilso]

1) TCP provides reliable transport for Data with the error recovery mechanisms and ensure every packet that Transmitter sends reaches the Destination ( JPlay )

2) Jplay has the buffering mechanism to compensate for any latency/Jitter in the data transfer to it ( in the Server to jPlay segment)

So the Eth cable between the Media Server and JPlay should not theorically bring in any improvement.

However since the folks have heard the difference , trying to get if there could be any explanation to it

 

[captrajesh]

Borrowing Captains Flame-retardent coat again

:lol:

For proper testing, you need to have the receiver in a constant state which is not possible for humans. Our hearing changes based on our environment and the state our brain is in, which is constantly changing. Don't know about others but my concentration generally keeps dropping with time even though music keeps playing.

Each human being is different. We all have five senses but not only all senses are not equal and also, senses of all are not equal. Moreover, one needs experience to reach a level which FMs like Bhagwan have gone. Not everyone can be a Tea / Wine tester. It takes keen sense and innate will to get there.

You have hit the nail on the head by stating what has been highlighted above. Concentration levels of each of us are different and it takes practice to improve it. You can relate with me if you had ever attended meditation classes wherein you are required to concentrate on one thing, be it a dot on the wall, a candle flame or an imaginary point between your eyebrows etc. It would be so difficult at first as the mind keeps wavering. As one keeps practising, one would be able to concentrate better and better. Same is the case here.

Another problem with humans is that our senses have limited ability e.g our eyes perceive the earth to be flat but that is not the case. I remember being taught how measurement with a tape changes with the position you look at it from, showing how deficient our senses are.

It is not the limitation of senses but limitations imposed by laws of physics. From ground level your eye perceives earth to be flat but go higher and you can clearly see the earth's curvature. Same is the case with tape measurement. The parallax error is due to the position of the sensor (eye) in relation to the thing being sensed (tape position)

@Captain - Your statement about unknown measurements is imho invalid as you are not aware of how ethernet works or for that matter what it measures.

I definitely do not. I'm not a techo and have pure sciences background academically.

There is no way to counter your argument and equally, there is no way you can prove it to be correct because nobody knows the measure of "unknown".

I don't intend to prove anything. All I'm saying is don't disprove anything relying only on whatever known parameters instead of personal experience.

 

[Thad E Ginathom]

Sorry to continue this OT digression-
The Neutrik locking connectors are used in pro networking applications too- it's just a way of fastening them more securely. I remember going to JSA (a large legal firm in Mumbai) and trying to plug into those recessed plug points that conference tables have, and finding a female Neutrik RJ45.
The Variax itself is a digital guitar simulator, i.e. it converts the guitar signal into various types (Les Paul, Strat, acoustic guitar, etc.). Each string has an A-D converter attached, which does DSP per string, sums the signal, and either does a D-A conversion, or sends it as a digital output via the RJ-45 cable.

Oh! Thanks. It's been more than a decade, and I hope I have seen the last of being underneath anybody's desk but my own . No need to keep up with the current hardware.

In all fairness to RJ45s, though, they are fine as long as they are plugged in. It is otherwise handling them that is a pain.

(offtopic: Fascinating guitar stuff, thanks for that too)

 

[Staxxx]

Not sure if this helps or causes more confusion, but most of my system is digital until it reaches the power amplifier stage where it is amplified (A/B). At its time in 2005 this was considered quite revolutionary and I had some long talks with the designer Ballmann. Specifically I was worried about the 20 odd meters of cable that he supplied between the preamp and amp with an RJ45 connection at both ends. He firmly believed that neither the length nor the connection degraded the sound. I tend to believe him as he is really considered an authority by some.
However, with S/PDIF there is absolutely no question that each cable I tried (that is to generate the signal until the pre-amp stage) presented the sound in a different way. It was not very subtle in my opinion and I could not understand why initially. Although I am quite foggy about it still, it was something to do with reflections?

 

[mowgli80]

Each human ... Same is the case here.

That was just an example of how our brain cannot always be in the same state. You are always getting affected by external factors. Generally my music listening for couple of hours. My brain would explode trying to concentrate that long, that would be for super humans. Out of curiosity, would you be able to put yourself in such a state, for a reasonable amount of time, to conduct a listening test for a another person even if you are a champion in meditation?

It is not ... being sensed (tape position)

Just to be clear, you are saying that laws of physics prevail even if what you percieve by your visual sense differs in the given examples, while laws of mathematics does not hold true over aural senses for ethernet cable because of unknown unmeasurable parameters .

I don't intend to prove anything. All I'm saying is don't disprove anything relying only on whatever known parameters instead of personal experience.

I see lot of people in this thread give a lot of credibility to Bhagwan's and his friends' hearing without having personal experience of sitting and listening with them or ever even meeting him in person. In my book, that is an unknown parameter as well and yet people are trying to disprove what I say .

PS: I am not questioning Bhagwan or his friends or their intentions or their listening power.

 

[abhijitnath]

Not sure if this helps or causes more confusion, but most of my system is digital until it reaches the power amplifier stage where it is amplified (A/B). At its time in 2005 this was considered quite revolutionary and I had some long talks with the designer Ballmann. Specifically I was worried about the 20 odd meters of cable that he supplied between the preamp and amp with an RJ45 connection at both ends. He firmly believed that neither the length nor the connection degraded the sound. I tend to believe him as he is really considered an authority by some.
However, with S/PDIF there is absolutely no question that each cable I tried (that is to generate the signal until the pre-amp stage) presented the sound in a different way. It was not very subtle in my opinion and I could not understand why initially. Although I am quite foggy about it still, it was something to do with reflections?

Although I'm not a cable believer (as might be clear from my comments on this thread), I'm willing to buy some variation between S/PDIF cables, because they need to be on spec to work correctly (75 ohms)-otherwise it's just a regular coax cable. Variations in spec may lead to variations in sound.

 

[Staxxx]

Interesting comment Abhijitnath. Sorry this is slightly OT: I settled on the Jorma cable eventually, which was not the most expensive cable, but the one I FELT was most neutral, revealing and pleasant to my ears in the final choice between it and the Stealth Varidig Sextet and Esoteric. I am not sure how these cables might vary in specifications (as I am not an engineer) but they certainly did in composition, design, sheathing etc. The Jorma was by far the thinnest and simplest looking of the lot.
This selection was done after auditioning in my system for an appropriate amount of time.
Strangely - as some other FMs have also found in previous posts - the S/PDIF cable has been the one cable that has made more of a profound effect on my sound that I could discern, which is in an appropriately revealing system. I think this may be mainly because of the time domain issues as Arj commented on earlier which is not of consequence in the CAT cable connection?

 

[abhijitnath]

To clarify something I mentioned earlier- the time domain issues are due to S/PDIF protocol, not the cable itself.
How this works is as follows- both devices that are connected with SPDIF need a common clock that they can calibrate against, to ensure that both of them are receiving, say, 96,000 samples a second. If your clocks aren't great, one will receive 96001 in the time that the other transmits 95999. This is a protocol/clock issue, not a cable issue, per se. To take another example from my parallel music making hobby, my guitar preamp (Fractal Audio Axe FX), refuses to accept any external clock when connecting through SPDIF, just its own, because most midrange sound cards have awful clocks and terrible jitter.

 

[Staxxx]

Understood. Just wondering if the construction of the cable itself can effect the transference of such protocol? or cause interference - that is even if the bits are being delivered perfectly ....

 

[mowgli80]

Interesting discussion about spdif .... Jitter is generally measured in milli to pico seconds. I have always wondered if people are able to hear timing issues to that granularity or if there is a larger implication apart from timing.

 

[Jazzhead]

Interesting comment Abhijitnath. Sorry this is slightly OT: I settled on the Jorma cable eventually, which was not the most expensive cable, but the one I FELT was most neutral, revealing and pleasant to my ears in the final choice between it and the Stealth Varidig Sextet and Esoteric. I am not sure how these cables might vary in specifications (as I am not an engineer) but they certainly did in composition, design, sheathing etc. The Jorma was by far the thinnest and simplest looking of the lot.
This selection was done after auditioning in my system for an appropriate amount of time.
Strangely - as some other FMs have also found in previous posts - the S/PDIF cable has been the one cable that has made more of a profound effect on my sound that I could discern, which is in an appropriately revealing system. I think this may be mainly because of the time domain issues as Arj commented on earlier which is not of consequence in the CAT cable connection?

Staxxx be interested in knowing , how you managed to procure the said Digi cables for in-house testing . No dealer , as far as I know or was it from your audio buddies . The Varidig Sextet is highly thought of , so surprised that the Jorma trumped it . Must be a case of system synergy , apologies for being OT. Thanks in advance .

 

[Santy]

Interesting discussion about spdif .... Jitter is generally measured in milli to pico seconds. I have always wondered if people are able to hear timing issues to that granularity or if there is a larger implication apart from timing.

Jitter may be in pico seconds causing mild aberrations in D/A converstion, but problem is, jitter is amplified.