We Need To Talk About Ones And Zeroes

@mdalton

That thread you linked is very good indeed. It features some definitive viewpoints.

I don't need to read anything from anybody anymore. I know the truth.

There is light at the end of the tunnel.

I was trying to make my original post both polite and sensitive to various points of view, but it seems I mostly tripped over my own feet.

My goal, or wish if you will, would be that folks who haven’t yet had the opportunity to educate themselves on the fundamentals of digital audio do so, so as to know what is possible vs what isn’t, and what makes sense vs what doesn’t, instead of slapping long-held analog-related beliefs on digital where they don’t belong.

I think it is emblematic of the problem when a clearly intelligent and learned individual accepts and propagates, doubtlessly in good faith, falsehoods about digital within this community and more importantly for themselves, those beliefs may lead them to make unnecessary and / or wasteful purchases.

A lot of folks seem to confuse digital files (made of ones and zeroes) with digital signal (the analog waveform representations of said ones and zeroes on transmission lines) and undesirable (but analog) noise that might travel along over said transmission lines.

Looks like a number of folks agree that the digital audio file that "lands" in your streamer is an exact, bit-perfect copy of the file Qobuz sent you, and of the file coming out of your CD or SACD of the same, assuming resolutions match and they originate from the same master.

In other words: Despite the horrors it traversed through AWS facilities, your digital audio file has suffered absolutely zero degradation or ill effect whatsoever on its journey to your home, explaining why audiophiles largely consider streaming equal in sound quality to traditional digital sources such as the aforementioned CD/SACD, local audio files on a NAS, or DAT transports if you’re into that.

Now, how would Ethernet gear located in the last 10 or 20 feet (aka your home) of that file’s 3000-mile journey somehow manage to compromise it where AWS itself failed?

Why would a simple $18 Monoprice Ethernet switch affect a digital audio file unaffected by a trip through hundreds of super-noisy industrial-grade switches...?

What can reasonably be hoped to be achieved by swapping in a $700 "audiophile" switch, all power supplies being equal?

If that $700 switch sounds better to your ears, I respect that, and I would prefer all the stuff about confirmation bias, sunk-cost fallacy, misery-loves-company, etc. be left out of this thread. It would be really interesting, however, to understand how a device that does nothing but send network packets on their way can favorably impact sound quality.

@macg19 

I would be happy with that, personally.

@gano 

Data center equipment is selected for high throughput, low latency and other factors including manageability and cost. Low noise isn’t one of them, and reasonably so since noise has no effect on digital data transmission. As a result switching mode power supplies and cooling fans are the norm, so these are noisy environments.

respectfully, you can’t draw correct conclusions from incorrect facts.

That’s generally true, but I’m not sure what you mean in this context?

Audiophile switch sellers purchase off-the-shelf switches, ditch the noisy switching mode power supply (SMPS) that came with it, repackage the switch with a small linear power supply (LPS), and sell it at a huge markup.

The LPS will in fact yield a significant noise level reduction. In other words, the product they're selling actually performs pretty well.

The only issue then is that they're selling, say, $80 worth of goods for $500, with no other expense than maintaining a website chockfull of audiophile puffery.

In a way, we can look at these people as arbitragers, leveraging audiophiles' lack of interest in or familiarity with basic network equipment.

All the more the reason to be a well-informed community - it's good for our wallets! 😃

@kennyc 

I'm confused. You give jeffbij a +1, then you apparently trash his post.

Did you mean to give the +1 to another poster?

@jsalerno277 

Please forgive in advance my philosophical pontification

It is more than forgiven, it is appreciated!

While your response is factually incorrect as noted by others, it is thoughtful and articulate and a pleasure to read 🙏

sophisticated error detection, correction, and redundancy mechanisms built into the system. Here’s how they ensure high reliability despite the presence of noise

This, and files can further be checksummed or hashed to confirm that the received copy is identical in all aspects to the sent copy.

Meanwhile, those AI are getting pretty good :)

@kennyc 

I meant to add "Some" in front of "Audiophile switch sellers" and I never did, but you're right. Vendors like Silent Angel and those you mentioned appear to produce their own PCBs, thereby offering an original product, which is really the right / ethical way to go - especially at the $3-4,000 price point.

On a higher level, high-end switches still send network packets on their way just like Monoprice switches, and there are limits to how quiet their power supplies can be made.

Speaking of which, if you go to this page and scroll down you will find a pic of the Silent Angel Bonn NX's ($3,999) main PCB. Isn't that an onboard SMPS right there at #4?

I don’t know if any of this resonates with those following this thread

@livinon2wheels 

It certainly resonates with me. I believe the choices you’ve made show admirable self-awareness. More often than not it’s not so much what we do that matters as why we do it. Best wishes for a prompt return to normal after the tragedy you’ve experienced.

USB was developed almost 30 years ago to replace and consolidate the old serial, parallel and PS/2 ports for PC peripherals, and later became ubiquitous in small electronics chargers.

How this pedestrian interface became quasi-standard in high-end digital audio is puzzling.

USB evolved tremendously over the decades and USB4 is a powerhouse, however many of today’s high-end DACs and streamers are still stuck in the USB 2.0 era.

@richardbrand 

After seeing your post of last night I worried local winemakers were boosting their product with methanol. I'm relieved to see you're OK this morning 🙂

... anyway, UDP is used only when latency is more important than accuracy: telephony, gaming, live sports. It is not used at all for audio streaming because latency isn’t that important, but accuracy is. Accordingly, both Tidal and Qobuz stream over TCP. Not sure what Spotify does, but no one cares.

Audio streaming presents a very, very light load in Ethernet terms, and error-free transmission is a given.

"Noise" and the so-called "jitter" are immaterial to digital-to-digital transmissions, and the vast majority of quality DACs readily handle noise and jitter before they can get to where they might negatively impact sound quality.

It’s probably safe to say that digital components (ie components whose inputs and outputs are both digital), and that have no onboard processing power, such as switches, routers, digital cables, etc. do not have a bearing on sound quality (again, all power supplies being equal).

Streamers, at least theoretically, should not either (digital in, digital out) but they do, or at least they can.

Streamers are computers in a pretty case, and even the weak-kneed Raspberry Pi-class CPUs employed by Aurender and the like have plenty enough power to handle relatively sophisticated DSP, thereby giving the designer wide latitude to transform the sound at will.

It isn’t that hard for DSP to make a streamer sound like a tube amp or, why not, a broken phone booth. Wanting to harness that power is understandable; what would you do if you made and sold high-priced streamers? Presumably, you would give your customers a taste of what you know they crave, just like the chef-owner who ladles fat and sugar in every dish up to just beneath the threshold of detectability. Pulling it off is a minor art form, but you don’t catch flies with vinegar.

DACs are half-analog, so of course they can make a difference in sound quality. So can upsampling, DSD conversion, and other signal processing taking place within the digital realm, but that’s a whole other conversation.

Personally, I believe system voicing is the job of the amps / speakers pairing. If any additional voicing is desired, it should be handled at the preamp level.

If every component in your system, including your digital chain - which should be absolutely neutral - intentionally dials in some level of euphonic distortion, pretty soon things become unmanageable. It’s like newspapers where every single writer is now a columnist whose important opinion on national affairs deserves to be heard, including food writers.

@mitch2 

UDP can safely be ignored in the context of 2-ch digital streaming, together with all the AI-generated nonsense @richardbrand has been cut-and-pasting in this thread. Both Tidal and Qobuz use TCP, and TCP does ensure bit-perfect transfer of your PCM-encoded music to your streamer (I say PCM because I don’t think either service offers DSD as of now).

SACD is different in that it is encoded in DSD64 format. Now there are multiple considerations associated with this, and folks who are interested may want to read this blog post by Benjamin Zwickel - the founder and designer of respected DAC manufacturer Mojo Audio - who is considerably more qualified than I to dissert on the subject.

@richardbrand 

A person relies extensively on cut-and-pasting at the risk of appearing lazy and unintelligent.

Some highly knowledgeable folks contributed their expertise to this thread, but rather than learning from their expertise you chose to go ask google AI.

Incidentally, as you know, we all have access to google AI and we would know how to query it if we felt like it. Doing it on our behalf adds no value and merely pollutes the thread.

No one can be expected to read 2,000 words of google-generated drivel and I admit I did not - so if I missed something I apologize, but the gist of it seems to be that 100% error-free transmission of a music stream cannot be 100% guaranteed 100% of the time. To which I’m happy to concur: A listener may get caught up in traffic in an underpass and their buffer times out, a submarine may sabotage an underwater backbone cable, etc.

What should be abundantly clear at this point, however, is that music streaming via Tidal and Qobuz, the only kind that matters to audiophiles in this day and age, is error-free absent a rare unrecoverable error; if an unrecoverable error happens, it will manifest in gross failure and playback will stop; and therefore, in no case is a TCP/IP stream or file transfer ever susceptible to the kind of subtle sound quality degradation that strikes fear at the heart of audiophiles. It will either work perfectly - meaning, your streamer will get exactly what Qobuz sent - or not at all.

So it stands to reason that that TCP/IP stream will remain unaffected by components it encounters in the home just like it was unaffected by the components it encountered in the depths of AWS - until it arrives at a component capable of modifying / altering sound quality, aka the streamer.

Router, switches, Ethernet cabling and the like should have zero effect on sound quality (again, all power supplies being equal), and devices claiming to "fix problems" with your home Internet are necessarily making false claims, since there are no problems there to fix.

Happy listening! 

@lalitk 

After perusing your virtual system following @mdalton ’s compliment of same, I would say that if my own system were anywhere near the state of evolution yours is in, I would likely be more open to evaluating components I currently consider secondary.

My own system would likely look very different than yours, and I suspect we enjoy different musical and listening styles. One thing they would probably have in common though is the NADAC stack. Why AES67 isn't better known / more prevalent in the audiophile world is beyond me.

It’s a beautiful system. Congratulations!

@lalitk 

Thank you for your post. In principle, I 100% agree with you.

That’s why I was careful to write that pre-streamer digital components should make no difference in sound quality. I was also careful to insert a qualifier about "all power supplies being equal". Finally, it is conceivable that, amongst the millions of possible component combinations, a new switch might correct some kind of incompatibility or condition caused by the old switch.

In short... I don’t know. I’m aware of opinions out there holding that passive digital components may have a positive impact on sound quality, and I respect some of them.

Claims of "stunning" or "transformational" differences, however, are 99.99% certain to be hogwash (the remaining .01% being due to correcting some equipment incompatibility as mentioned above). Barring the latter, folks making such ridiculous claims likely haven’t heard any difference at all, let alone a "stunning" one, and probably couldn’t tell the difference between the cannonade in the 1812 Overture and their wife slamming the lid on the trash can.

Which brings us to, if passive digital components do make a difference in sound quality, how big can that difference possibly be expected to be? How big of a positive impact can these components make, considering that the packets they transmit at that stage represent anything and everything besides music? I believe, very very little; and whether it’s worth pursuing it is a personal choice. Personally, I am not willing to spend time auditioning network switches, because the reward is more likely than not to not justify the investment. But that’s just me.

You mention measurements. I’m aware it’s something of a dirty word for some around here, but the good news is that the bit-perfectness (?) of audio streaming is demonstrable without any measurements involved, simply by checksumming the sent and received data.

For the record, I am biased toward science and rationality (in case you didn’t notice :); however, I think measurements are only as good as what they measure, and that they do not necessarily correlate to sound quality. I do believe John Atkinson’s work adds value to Stereophile reviews, creating a balanced approach.

I think I read somewhere that John Atkinson himself said he does not review digital components because "there is nothing to measure" (correct me if I made that up!).

But I do firmly believe that our pursuit for a good sounding system benefits most when curiosity is anchored in reason and when we leave enough room for both exploration and skepticism to coexist.

@lalitk 

I could not agree more, and in fact this applies to most things in life.

I try to keep an open mind and adopt the position that very few things are 100% black or white.

(Apologies to @devinplombier for this temporary high jack)

@mdalton 

Not at all, in fact I'm curious about the super tweeters after reading @lalitk 's eloquent description in his virtual system page.

Something I would have assumed was not for me due to my hearing threshold no longer having a 2 in front of it (at least it still has 5 figures!), but cymbals can produce harmonics into the 70 KHz range. Wonder if other instruments benefit similarly from super tweeters.

On that general subject, here is an interview of Uncle Paul in which he reminisces about the early days of digital audio. Interesting

https://www.theabsolutesound.com/articles/paul-mcgowan-and-the-history-of-digital-audio/

@richardbrand - being old enough to remember the launch of CD doesn't make you that old...

Yeah, I was gonna say

@mswale +1  Still, one can benefit from a ’clean’ home network.

@sns 

By "clean network" I assume you mean an audio-only subnetwork walled off by VPN and various and sundry firewall tools.

If one has unusual network security concerns or wishes to shield their beloved streamer from the prying eyes of North Korean hackers, sure. But to expect any sound quality improvement from it will most likely result in disappointment. Clean or "dirty", TCP ensures bit-perfect delivery either way.

That said, big iron is fun! But if the yardstick is sound quality, enterprise-grade hardware will neither help nor harm, though as noted you’ll have to exile it to the basement and acquire serious network skills to operate it.

There are further steps possible via the Dejitter It Switch X if you believe eliminating  needless network traffic improves streaming sound quality.

@sns 

I do not believe it does, unless overall network capacity is being strained as @jeffbij indicated.

I've been reading a bit on Switch X, and that's definitely a good one. More on it later 😃

I found audiophile switches redundant or have very minimal impact as long as we have addressed galvanic isolation between network devices, effectively breaking ground loops and blocking electrical noise from traveling through Ethernet cables.

Too often, people only consider an item’s primary purpose and judge it by how well it accomplishes that purpose, but that’s a limited, it not somewhat blinkered approach.

Indeed, by that token no one should buy a Patek Philippe because a lowly iphone keeps perfect time and the Patek sure doesn’t, so Patek Philippe must be snake oil.

Of course, accurate timekeeping has long since ceased to be the raison d’être of fine watches. Therefore their secondary purposes of beauty and craftsmanship have superceded their primary one of timekeeping.

Similarly, if a person doesn’t like to look at items of low or average quality in their home, then a well-designed network switch milled from solid aluminum and fitted with the best components and ports available makes sense. 

If the goal for that network switch is to improve sound quality, however, disappointment is bound to ensue.

The Telegärtner M12 switch doesn't use RJ-45 connectors. Instead, it is equipped with proprietary connectors, requiring the use of proprietary patch cables.

An interesting approach for sure.

Hope this clears any confusion!

@lalitk 

It does! Thank you!

It is indeed good to know that the "M12 [switch] in [your] system contributed to a smoother and more refined sound and further improved tonal density and micro-dynamics"! Otherwise, we might have thought - mistakenly, it turns out! - that you "found audiophile switches redundant or hav[ing] very minimal impact, as long as we have addressed galvanic isolation between network devices". So, thank you for clarifying!

Happy listening!