Interesting response on switches and ethernet cables

Not sure what you are trying or your referenced article is trying to say here. You/they have confused analogue coupled interference with digital signal encoding. These are two quite different things…
A more meaningful measure would be in my opinion to measure the ground noise floor on with different data throughputs…
Your table doesn’t even mention whether the RTP is WAV or FLAC encoded.
I realise you are quoting somebody’s web article, but this sort of thing just perpetuates confusion and mis understandings amongst non technical people.

2 Likes

I don’t know whether proven, however it is a logical and plausible explanation for differences in sound arising from changes to network components, and differences in such claimed effect on sound between different streamers/DACs - and why some may exhibit no change to sound, as per the OP.

(N.B. IIRC from the many posts in the several cable and switch threads, whilst many people either didn’t recognise or believe in the possibility of psychological influences - some even claiming to be immune - there were reports of audible effects from people who assessed under conditions that suggested the risk of bias was unlikely.)

I wonder if there is a similar market for fibre optic network cabling. I wonder what sort of marketing blurb would be needed to differentiate one optical cable from another.

Is that the case? The article states:

I will measure with RightMark (newest 6.4.1) to look at the usual dynamic range, noise floor, distortion along with the Dunn J-Test signal to see if there’s any evidence of jitter anomaly in the Transporter’s RCA DAC output (rather than the XLR for the sake of convenience).

Isn’t that what we were discussing in this thread?

I just think if it were proven with measurements, there wouldn’t have to be an eternal debate about it.

One more thing about this:

Thomas wrote several times before about his results with USB and Ethernet regarding an LPS on the core, and just now provided another nice summary here in another thread:

There is also being open-minded and honest enough to change your position when something is conclusively demonstrated to you that turns your previous beliefs on their heads. Take, for example, Naim’s belief that a Unipivot couldn’t possibly work effectively. That stood until one particular R&D Engineer who had come to Naim to work on tuners, presented an idea for a Unipivot that he was convinced would work well. Allowed to make some prototypes, these went on to convince the sceptics both within Naim and beyond. The rest, I guess, is history.

A core ethos of Naim is to build and work from a knowledge base but always be open minded to new ideas and approaches. There is time allotted within the R&D week so engineers can pursue “blue-sky” projects, something that only really works with an open-minded approach.

8 Likes

No not really… that is transport clock jitter… which these days in just about all DACs is decoupled from the digital DAC clock.

We were talking about coupled noise into ground planes, Powerlines and electro magnetic fields… which can interact with sensitive DAC reconstruction electronics including the clock.

One example of this the difference between WAV and FLAC decode that some people can hear the difference between. Assuming decoder is not broken, this difference is caused by the different noise profile caused by the WAV PCM decoding and serialisation electronics with the FLAC decode and serialisation electronics.

The key thing this coupling is product and environment defined.

2 Likes

Is there any research on this so called “coupled noise” or evidence to support their interaction with sensitive DACs, and through the use of a better Ethernet cable and/or switch this is eliminated? And that the difference is clearly audible?

EM is easily measurable - so it’s it’s causing the issues you are claiming then logically the impact of the EM is also measurable.

Happy to be open minded if there is evidence.

Absolutely… my point is about denying something which by definition is not being open minded… but then subtly slipping in a change in approach. So yes we would appear to be in agreement and share the same values…

If Naim or anybody states with their latest products and technology they are now able to somehow maximise performance through using previously unused techniques or approaches that is different.

Evolving and learning affects all of us who work in ITC… but denying something and then silently going against that without a stated reason undermines trust… and you can see companies come a cropper when they loose customer trust.

1 Like

There are all the EMC regulations, you can read my undergraduate thesis if you want :grinning:, you can read the Naim and others white papers… and as Naim use TI DACs this is a good paper from TI on how to reduce noise emissions using their PHYTER chipsets. https://www.ti.com/lit/an/snla107a/snla107a.pdf?ts=1613418807195&ref_url=https%3A%2F%2Fwww.google.com%2F

Section 4 is particularly relevant…

Notice a a noise coupling jig is mentioned… often this can be a coil fed into a receiver.

Electro magnetic relationships and coupling were discovered by Faraday and later used by Marconi to develop radio.

If you are interested in learning the engineering and science of coupling and interference mitigation, an interesting path is obtaining your CEPT radio licence… you may even get question on it in your exam… I did (it was about electromagnetic intermodulation interference coupling affecting a TV.)… This is also part of my professional world now, albeit from a security perspective

I guess it was this open mindedness that got me challenging the defacto Netgear switches back in the day that were often used by home streaming setups, I guess I like to challenge assumed consensus. I discovered an induced coupled birdie causing intermodulation in my NAT03… the FM coax was parallel to an Ethernet lead connecting to the Netgear… I swapped the switch out with a Cisco 2960 that I had spare… the birdies disappeared… for me the rest is history in terms of audio and noise coupling… I undertook many experiments and even shared some of my findings with Naim back in the day (inter frame timing consistency affecting sound characterisation from first gen streamers)

3 Likes

I’m not sure what you describe is accurate when it relates to the article, but i might be misinterpreting it…

This is the test setup:

  • They use a number of different UTP cables and switches to provide a source signal to a Logitech Squeezebox Transporter streamer/DAC.
  • On the RCA output of that DAC they measure the amount of noise / jitter / distortion, and how much that deviates for each combination.

Their findings are that all the cable and switch combinations produce (near) identical output results, with only a 0.0001% deviation, which should be impossible to discern for a human being.

They are using this E-MU 0404 calibrated ADC / measurement device to capture the analog output of the DAC:

0404USB

Am i perhaps missing anything here?

1 Like

I still they are referring to transport jitter. The measurements quoted are relevant for that… I will read the article in full later to see if I have missed anything… and will be device and environment specific…

However if listening to a device, if no discernible difference is heard based on differing environments and connections, then great. If a difference is heard but you can’t see it with your measurements, you are measuring the wrong things.

I would like to look at pink noise, and then measure the power density per frequency bin … and compare … you might need to average over much time per measurement… and use quite granular bins.
I am sure there are many similar measures… but that is the one I would like to do first… and focus around very fine bins in the speech frequency range.

I readily admit, I have decoupled my transport from my DAC other than a Naim DC1 SPDIF lead… and I don’t hear the effects of networks, switches, or Ethernet leads in any meaningful way now… yes there can be ultra subtle sonic shifts, but certainly not one having better SQ over another…

The Audio Myths… food for thought thread had an interesting debate and discussion on that… if you haven’t already watch the video.

1 Like

Ultimately we should be sceptical of our subjective hearing ofcourse, especially when we enter the domain of miniscule changes such as the ones discussed here.

Human hearing is notoriously unreliable, or perhaps a better word is circumstantial. Just a tiny bit of stress, either positive or negative, can completely change what we focus on and how we perceive it. If we can’t measure it objectively, it is statistically much more likely that we are fooling ourselves into hearing something that isn’t there, or are focusing our attention on something different in the music than before.

1 Like

You are both right :slight_smile: If it is heard but not measured, the measurement is not suitable. At the same time it is difficult to be sure it is heard. Statistically relevant blind tests with tens of thousands of participants like in pharmaceutical studies are “rare” in hifi.

On the topic of unreliable hearing and the difficulty of blind testing in audio things: A German TV magazine did a blind test of violins, a few years ago: A Stradivarius, a very expensive modern concert violin, and a basic beginner’s model of good quality. The blind testers were Anne-Sophie Mutter and two other experienced people with orchestra backgrounds.

They could not distinguish the Stradivarius and there was no clear winner. Some preferred the beginner’s model. Conclusion: Either generations of violinists and conductors were deluded, or blind testing these things is more complex than naively expected

2 Likes

Reading the TI article, and let’s just focus on ethernet cable for now - the one clear recommendation is to use shielded Cat5E cables, which I think in today’s day and age everyone is doing that. But you are arguing that even that’s not enough? where is the evidence for that?

Because when reading different forums I find people sharing a real improvement with audiophile switches for all the well known dacs or servers/ dacs , be it dcs, linn, Innuos / chord Dave, Naim, and even the 130 k MSB.

Sorry Hmack, I have misunderstood your tongue in cheek post. Good response from yours :+1:

1 Like

Sorry don’t follow your argument? What are you basing your personal assertion that everyone uses shielded Cat5e these days? Is it based on your personal experience? Where is the evidence? Certainly in my personal and professional experience shielded Cat5e cables are somewhat unusual… and is indeed not part of the Cat5e specification, but is not specifically excluded. Certainly all my Cat5e runs are un shielded, and my Cat6 are.
At least you have one exhibit of evidence (web document) to understand that there are multiple sources of EMI even from an Ethernet cable. Now show me yours…

This is a very good point… measures are not limited to measurements from electrical measuring devices… subconscious bias needs other methods to measure and assess.

I feel this thread and the Audio Myths… food for thought thread have many similarities… :grinning:

1 Like