The inaudibility of Ethernet

Really !!! … Over a dozen concurrent accounts on Audiogon … Earlflynn …Cindyment … SkyPunk , etcetera , etcetera . New forum Spring 2022 … Old Tricks .
Wrong, but that’s okay. not taking the bait son.
 
Gentlemen, please STOP. The WBF is better than petty little spats like this. RoT, I expected better than this from you, and Tony? There is no need to "school" anyone here.

The fact is, there are plenty of ways to get great audio when streaming. There are deficiencies AND attributes with which way you decide to go with it.

This ends now. Let's please get back to a constructive dialog, instead of chest thumping, honestly HS type behavior. We are all grown adults here.

Tom
 
@treitz3

I'm not schooling anyone... just pointing out, in a "Measurement Based Audio Forum" that you can not "measure" the impact of ethernet on streaming data.

The "measurement" fact is undeniable unless the hardware itself is faulty.

I have, indeed, done development of audio over IP. So, I have "measured" this. I provided some of my work background to support my claim on this. It was important when we were laying out the first and second generation of multimedia over IP ( in the 90s working with ATM, optical, ethernet, firewire, USB, satellite, cellular, cable, ISDN. DSL ) to ensure that we could eliminate this. There was a LOT of money at the time flowing into the field. Cable, satellite and telcos all wanted a piece of entertainment streaming to the home.

Indeed, the biggest issue we used to have was the use of USB-1 by many commercial manufacturers up until so late in the game. We knew, and measured, the issues with the limited bandwidth of USB1 and synchronous mode. We knew, and measured, that asynchronous mode using USB 2 was the solution, but the cheap skate consumer audio manufacturers were way behind the curve. Oddly, the more High End they were, the worst they were because of their low sales volumes and the relative high cost of their non recurring R&D - they couldn't afford to keep up with the fast pace of the computer electronics industry.

By Y2K the problem was solved. At the time, if you wanted high quality streaming sound you used at least 100Mbit ethernet (from the modem) and USB2, PCI or Firewire. And a player with a big data buffer to eliminate any synchronicity issues between data I/O. As it turns out, the size of the memory buffer was the single most important key as it could eliminate the bursting of a bad datalink ( cellular was the worst at the time ).

Still is. But most datalinks today are quite robust nowadays and memory is dirt cheap.

As an aside... CD players also had a similar issue with jitter, remember? I recall that Arcam (I think) were the first to put a little bit of a RAM buffer to decouple their IO. The clocks used for reading the CD and for driving the DAC were separate. This is pretty much the same thing we do with a data link and a DAC.

BTW, as an audiophile, I provided just a hint of the complexity of the LAN in my own home. I stream data off my PLEX servers and I hear no issues with ethernet. I also mount the file systems and "pull" using VLC and Foobar and I hear no impact on the music. My current system now uses GigE drops to all wired ethernet interfaces and 802.11n for the wireless devices. I can assure that there is no impact on either on audio playback.

I do hear the difference between VLC and Foobar, but that's the quality of the decoding. I prefer Foobar. Tidal HiFi's own player is very good as well... The biggest issue that I have with all of those players that is the inability for Chromebooks to play at higher than 24/48. Whereas Android and Windows do support Perfect Bit play back. But that's the drivers in the OS, not the fault of the decoding engines.

I hear differences in the op amps in the DAC. I suggest that all High End DACs should support op-amp rolling by supplying a socket interface, not an SMD set up.

But these are issues in the audio drivers in the software, not the data link.

If someone wants to dispute this, fine with me, provide some proof. Or at least qualify your experience on this part of audio.

My simple claim is that you can not hear an impact from ethernet streaming on your audio playback. If you hear such, then the problem is elsewhere in your system, most likely in a ground/noise plane in your system.

But as always, it's up to you, if you feel happy with it, go for it. After all, I love a bit of negative 2nd order harmonic myself. Heck, even sometimes, I enjoy just a little bit of the old 3rd order harmonic too... it makes the drum kit sound faster.... seriously, and yes! you can measure that!

Have fun.
 
Last edited:
  • Like
Reactions: AndrewChen
It seems you are looking for a debate. I will tell you right now, that I personally use none of what you describe and my streaming rig can go head to head with a very well regarded transport I doubt you have ever heard in your system (along with other gear I have no desire to get into right now). As mentioned, there are many ways to achieve audio nirvana, when it comes to streaming.

Your way is NOT the only way. I ask that you stop presenting your way as the only highway. Maybe, step back a bit and learn from others on how they arrived at what they have.

That is the spirit we are looking for at the WBF. Not absolutes backed up by endless backgrounds.

Tom
 
No, I am NOT saying that my way is the only way.

I'm only pointing out that the data link does not affect the DAC. I have no issues if someone else has a different point and they want to bring it up in an intelligent, respectful manner... and provide some technical data perhaps?

Is this not the "The Measurement Based Audio Forum" ?

Note that I'm not bringing my own system into the discussion, that's for another forum and I'm not going there. You haven't lived audiophilia nervosa until you got one of five amplifiers in existence in the World and it smokes on you, twice.... nothing like being a guinea pig in the search for truly good sound (yes, the amp sounded fantastic until it smoked and took out the woofers ). Worried about the data link? Heck, I'm just worried about burning down the house!
 
Last edited:
Is this not the "The Measurement Based Audio Forum" ?
Indeed. This subforum is described as:

A forum where objectivists can ask questions and exchange ideas regarding the science of audio as well as anything pertaining to audio that involves tests, measurements, ABX testing etc. We have all indicated a desire to read about the science behind our hobby but this is one forum where the subjectivists will not be allowed to derail or flame an objectivist thread with the comment "I trust my ears". As a result to all members this is intended to be a flame free forum and the warning on the door will say "Enter At Your Own Risk"

Maybe consider to change this description if it no longer reflects how the mods interpret the nature of this subforum?
 
  • Like
Reactions: TonyE
Oh, measure away. Just keep in mind that maybe the right things aren't being measured.

Tom
 
They do a lot of testing and listening at Alpha Audio. They say that not only do jitter, switches, LAN filters, power supplies and cables make a sonic difference, but the differences can be seen in the measurements.

This article contains their advice for setting up a streaming system, with links to the technical tests:

PS The above has some interesting technical analysis and some good advice, but I don't agree with some of their expressed views. I posted to this thread so that objectivists could peruse the measurements and conclusions.
 
  • Like
Reactions: Republicoftexas69
(1) The power supplies of the switches also differ greatly. And swapping out the standard power supplies for a decent (quiet) model, lowers the noise floor significantly. Where? Where are they measuring the noise ?

(2) And switch and router settings also have an impact. Think flow control, priorities and Vlans that can isolate and prioritize traffic (by setting the QoS priority per Vlan). They offer no proof. A VLAN affecting a DAC? Priority traffic of frames and IP packets affecting a DAC? How about the the phase of the Moon? How about multicast? Unicast? WTH does QoS and VLANs have to do with the data link? Nothing, they have NOTHING to do with it.

(3) swapping out a switch lowers the noise floor and thus the jitter on the clock that drives the dac. These are reproducible measurements and results that have also proved audible. Again, just a statement, no proof. Jitter on the clock that drives the DAC. Where is this "clock"?

Here I got a comment... by DAC, do they mean a consumer box that contains a DAC or the actual sending of data via a DMA over an address/data bus to the DAC device? And with the understanding that the DAC will actually read the data from DDR directly under ITS OWN clock, meaning, it doesn't matter if there is any kind of "jitter" in the data stream that the ethernet ( or optical ) interface to the consumer box because the hardware within isolates it and reclocks the whole thing. It reads the data link, validates the data, it logically filters it and then writes it into DDR.... later a network layer will ensure the validity of the data and creates the packets and copies it yet into more DDR, later another software process, will read DDR, process it into music data and copy it to a memory location ( could be main memory or some registers in the DAC module ), then the DAC hardware will clock that music data into its own buffers.... Notice that there are several, separate hardware and software processes, all using their own clocks, within the box proper. Far, FAR removed from any jitter that might be coming from the power supply in the ethernet switch.

YOU HAVE TO MEASURE THE JITTER AT THE DAC CHIP ITSELF! Not the data stream going into the consumer box.

(4) cables also have an influence, More claims, no measurement no proof.

They claim that the jitter introduced by the power supply in an ethernet switch affects the clock driving the data into a DAC.

The make a lot of claims but offer no proof about audio.

Fundamentally, they've failed to show a schematic of their test configuration and how ( and where ) they are measuring these things.

They did not make a connection between the network switch and the DAC? How the heck is a network a switch affecting the data going into the DAC chip? If you have a device that allows external noise from getting into data going into the DAC switch, I think you are using a badly designed device... and fancy names, high prices are no garantee that the device has been properly designed.... you're just paying for amortizing the non recurring R&D, marketing and sales cost between much lower volumes... Been there, done that.

Anyhow... Now, they did show the specs about their testing their ethernet switches and noise around them... and what did they find out? THAT THEIR POWER SUPPLIES are the culprit... and their final conclusion is:

Because it ( switch noise ) also depends on where you place the switch and especially: where you plug in the power supply of the switch.

And what did I write above?

" If you hear such, then the problem is elsewhere in your system, most likely in a ground/noise plane in your system. "

QED.

Mind you, I take deep objection to the likes at Audio Science Review, since I know that we don't know why things sound good to us. As a scientist I make the claim that our psycho acoustic models are insufficient and our resulting measurements inadequate to measure what sounds good...

HOWEVER, I have spent enough time in my career with multimedia and metrology to know HOW to make a measurement that backs a claim.

These people are selling you snake oil. They are not measuring in the correct place, and fundamentally agree with me.... in most cases, their network switch noise is related to the quality of the power being provided.

BTW - speaking of DACs.... try getting one that supports op-amp rolling. THAT will make a huge difference, far more than any fancy sounding audiophile ethernet cable.
 
Last edited:
Oh, measure away. Just keep in mind that maybe the right things aren't being measured.

Tom

You know, if you read what I write, you might find out that we agree. Science is NEVER settled, by definition.

Let me quote my previous post...

"As a scientist I make the claim that our psycho acoustic models are insufficient and our resulting measurements inadequate to measure what sounds good..."

In Western Science we use empiricism to develop the science models from which we create the measurements.... rinse and repeat.

I would fully expect that a forum that is titled "Measurement based" would support a discussion... As you pointed out

"It seems you are looking for a debate"

Indeed I am. Do you have a problem with that? We debate about things at work all the time. About work that is. Such is science and R&D.

Have you ever done a Peer Review of your work? We do it all the time, that's how we learn from others, Listen to their opinions and facts. We don't insult each other... like that guy who "blocked" me and even sent me a PM to tell me he was blocking me. There's a closed mind. So be it.

Don't be like that. Always keep an open mind.

Oh well.... there's a lot of snake oil in the field of audio. Starting with the likes of Bose, the Home Theater stuff and a lot of the High End. Often they sell you stuff that indeed affects the sound, whether it's worthwhile is irrelevant, as that is subjective. And often it can be measured indeed. But often the reasons provided by the marketing dept are ludicrous...

Sometimes, like that post about that outfit that "measured" network based noise on the DAC.... they use fallacious logic, it's called "Appeal to Authority". They show a bunch of intruments as if they knew what they are doing and us, the readers, are supposed to trust them because they are authorities in the field and "know what they're doing". In reality, they have no clue, as they are measuring the wrong thing, but most people don't know enough and trust them.... yet those of us who know how to take measurements know they are just BS'ing.

Not that we can not hear differences introduced into the system, but that the causes of the why we hear differently are not the ones that are being "measured" in that marketing "report". I would expect stuff like that, their faulty logic, is EXACTLY one of the things withing scope and discussed in this forum.

However, sometimes we know WHY things sound different... take negative 2nd order harmonic and 3rd order harmonic. Yet we don't know WHY we like the former and not the latter... nor why we don't like higher order harmonics. Yet, sometimes we design with 2nd in mind because, we like it.

Take some of my amps, class A and tubes. They measure horrible according to the ASR crowd. Julian Hirsch and David Ranada would have gone apoplexic over them... yet, people love them and pay good money for them. Have you ever heard a Class A SIT amp? Gorgeous sound, believable soundstage... 45 watts into 8 ohms at 1Khz at 1% THD (*). If you went by pure, agreed upon measurements you would think that would be a terrible amplifier. Certainly Amir would think so. But when you sit down to listen to it... hmm... hey.... the thing sounds really good.

Then you got LPs. I got thousands of them. Supposedly they measure bad yet they often blow 24/48 recordings... and sometimes they even match the master tapes. Why is it that digital supposedly measures so well and yet sounds so bad?

Again, there is measuring and listening. Someday they will correlate, but you don't want to throw either of them out with the bath water. It's a learning experience.

So, yes indeed, it is a DEBATE and we learn from that. Keep an open mind.

Peace and good tunes, man.

(*) BTW, measuring THD, and THD+N, is mostly irrelevant. An outcrop of the FTC decision back in the late 70s. Ideally we want to know the frequency based spectrum of the distortion. 2nd order is not that bad, 3rd can be handled but 4th and 5th and so on are terrible.
 
Last edited:
Nope. No problem at all. You just won't be debating yours truly.

Tom
 
OK, you don't have to debate measurements.

How about negative 2nd order harmonic though? Gotta tell you, it makes the sound of the piano, voice and strings amazingly realistic. Just add a little bit and the soundstage opens up... close miked voices are THERE, in front of you, between the speakers.
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu