Do Audiophile Network Switches Make a Difference?

[@Eric_the_Hiking]

I just know that if UPS or FedEx made a switch, it would drop packets.

[@RedSinter]

Ahahahahahaha ahahahahahaha Ahahahahahaha

[@alecgrolimond1678]

Good one.

[@MikeHageman]

Total packet loss. 😂

[@LynnXternal]

It's so awesome that you're making KZbin videos. Happy to see you here!

[@AudioScienceReview]

A pleasure to be here. Thanks for the welcome!

[@DBTAudio]

@@AudioScienceReviewIs there a need for a switch if you have the ability to connect directly from your router to your music streamer?

[@oztechsolutions]

Thank you for this great video. Very informative. Being a network tech guy and seeing products being sold as geared to audiophiles i just can’t help myself not to get upset especially if my friends think that it really does making some improvements and that they need this expensive gear.

[@zackschindler8334]

I worked at a GM car plant (Detroit/Hamtramck Assembly) for 30 years in the IT department. The plant was 100% dependent on the data from the servers and the network. If for any reason the line stopped running it cost $50,000 a minute (USD). Needless to say there was a lot of emphasis on us keeping data flowing. Never once was there a discussion about putting different switches in (we used Cisco) that might be better because they had magical properties. it was a noise laden environment with hundreds of welding robots running 18 hours a day and thousands of Motorola DMR radios in constant use. Oh and the plant has it's own substation too. And yet the data still flowed. Rest assured that the 1's and 0's don't care what wire they go down.

[@betochiwas]

Yeah, I have worked with a lot of Industrial Switches, Allen Bradley, Cisco, Ntron they all work quite fine undee rough conditions

[@nicktube3904]

How can you explain the original Smps of my Cisco 2960 fucks up my streamer and high end dac with nasty noise. Changed to a clean linear Farad psu and music sounded magical!

[@zackschindler8334]

@@nicktube3904 If you get a new power cable for $2k it will sound a lot betterer!

[@Ceko]

@@nicktube3904 music sounded magical because it’s magic!

[@srkbear1]

@@nicktube3904 It sounded better because you paid money for it and needed it to sound better. The more money you spend the more investment you have in the result. Amir just showed you with irrefutable evidence that the actual data arriving to your DAC is unchanged by these “upgrades”, and besides, if there were errors arriving to your DAC via a network stream the result would be dropouts, glitches and other non-musical events, not better “sound”. You’re suggesting that this switch could somehow add something analog and musical to a series of ones and zeros-making the original waveform prettier in some miraculous fashion. It’s preposterous. Use your brain in the service of logic, not in the service of aping manufacturer marketing slogans-the internet is forever and you’re memorializing silly things.

[@kaitonbuitendijk2476]

Great that you are on KZbin! Very clear explanations. Easy to follow. Please also make some beginner explanations for reading those audio graphs

[@AudioScienceReview]

I have one planned. It will be a long one though so have to figure out how to best get it done without people dropping out half way through the video. :)

[@kaitonbuitendijk2476]

@@AudioScienceReview maybe you can make a little series? One diagram per movie, then the why, what how. describe why it's important to audible effects, what is measured, how it can be effected(by what). Just an idea 🙂 do with it what you want ☺️

[@cristiantolbaru7153]

Awesome work! Congrats and keep them coming, it is pure gold Amir!

[@AudioScienceReview]

Much appreciated.

[@chrisstephenson9890]

Thank you Amir. Great theoretical and applied analysis to a subject I have been intrigued about for a while. As a software engineer I was very dubious, however I could accommodate the theory that noise could be dragged into a system via the ethernet port. I feel I'm now much better informed and justified in not spending any extra money in this area of my setup.

[@AudioScienceReview]

My pleasure Chris.

[@exarkunn69]

Truth with data to back it. Doing IT and computers for 20 years I've never understood why people buy these, now I have all the proof I need!

[@heysplangy]

Sounds compelling, but his fantastic machine can't measure how a human perceives audio. It does a SUPERB job of telling you how much noise is in a system, but not much else.

[@heysplangy]

@John Bravo nope, but I have played around enough with different power supplies on switches to know that differences in sound can be had. SINAD is not the determining factor in audio reproduction.

[@heysplangy]

@John Bravo why? because I have heard improvement with better power supplies? This is nothing new. Power supplies can make differences; it's an inarguable fact. Don't become just another of Amir's diehard objectivist cult who don't know how to think or listen for themselves. I appreciate much of his testing and attempts at debunking...I mean who's going to argue with a guy with 100k worth of machinery that does all the "listening" for you. hahaha

[@QQ-td9id]

@@heysplangy If there is no difference in sound, quantitatively, then there is NO WAY human can perceives the sound different unless placebo effect. LOL

[@heysplangy]

@@QQ-td9id hahaha. Logically that would make sense...if your premise was correct. OK. You're using all the right words that a wannabe member of Amir's cult is supposed to...but... "If there is no difference in sound..." blah blah blah. The AP can not tell you that. I know you think it can, but it can not. It's going to give you a very good signal/spectrum analysis and extraordinarily detailed signal to noise and distortion measurements. It is not going to tell what something *sounds* like. Machines can't do that(yet). Now...with some effort and some different thinking, maybe they can come close. Can the AP quantify soundstage? Depth? width? The space between? how a stereo image manifests. ...a spinorama for stereo imaging presentation. I get it...you've found the guy with the fancy machine and an impressive (but really not that impressive) resume, and he's telling you all the things you want to hear so you don't actually have to do the listening or the thinking all by yourself...since now you're pretty much an expert hahaha. Some people just need to have their reality defined for them... If it walks like a cult and talks like a cult...

[@SteveBracy]

Thanks Amirm! Been working high speed digital RF for years: IT JUST WERKS. Thanks for explaining this to everyone.

[@dq8979]

Keep em coming, this is great! Thank you!

[@AudioScienceReview]

I am glad you are finding them useful. I have more planned.

[@Lockk9]

Would love to see a test of iFi's iPowerX - will it make any difference to analogue audio that comes out of the DAC ? or snake oil ?

[@Gabriel-of-YouTube]

That's a different thing.

[@SwirlingDragonMist]

I've seen a few videos showing Ifi products improving the sound of audio output, noise that can be heard right in the video.

[@Lockk9]

@@SwirlingDragonMist Would you have a Link to a specific video where you can actually hear the difference ?

[@SwirlingDragonMist]

@@Lockk9 Here ya go kzbin.info/www/bejne/jma5Yn2FiM-ViKM

[@SwirlingDragonMist]

@@Lockk9 this one too kzbin.info/www/bejne/mIeQmIZ3mMumaac

[@Daniel-ur3sf]

I don’t think it was explicitly stated in the video, but essentially what happens is the ethernet packets arrive at the streaming device, they are buffered in a queue, then they are read out to the DAC which uses its *own clock* to reconstruct the original analog signal. Ethernet clock jitter is completely irrelevant.

[@guyboisvert66]

Exactly!

[@bwoody1090]

Knocking them down one at a time. Another great video Amir.

[@AudioScienceReview]

Thanks partner. :)

[@bobdylan6237]

A voice of reason in a cacophony of fools! So glad to see you on KZbin, Amir!

[@simontompson]

When you're next in need of ideas, a related test I'd love to see would be Ethernet cable vs Wi-Fi. Love the hard work here - thank you!

[@guyboisvert66]

Won't do anything special if you have sufficient bandwidth. The streamer use buffering, all will go well if the buffer never gets empty.

[@stopthefomo]

Welcome to the KZbin community Amir!

[@AudioScienceReview]

Thanks for the support Fomo.

[@taylorsharp5928]

But the great Hans Beekhuyzen said audiophile switches sound better than those dirty consumer switches. He has an audio analyzer for decoration in his studio, so he must know what he's talking about. 😁

[@AudioScienceReview]

Right on. Hans is very well spoken and produces great videos. But as say, uses his instrumentation for décor. He talks forever how jitter is removed this way and that way but won't power on that analyzer to realize that they have not.

[@urom]

after this "audio evangelist" starts to do this switch nonsence, i stopped watching his channel, because this discredited all the rest he talks about

[@acoustic61]

I stopped watching Beekhuyzen years before the switch review. Same with Audiostream, Darko etc. I cancelled my Stereophile subscription last year after three decades. JA was good at writing convincing reviews of so-so products. Computer Audiophile tried hard but had no technical background or writing skills. Amir is a breath of fresh air and likely is responsible for deprogramming of many recovering audiophiles.

[@decoryder]

Hans Beekhuyzen doesn't lie, he's a man of great personal and professional integrity, so when he says he's hearing something, I believe him. I don't understand why he's hearing it and I'm quite frustrated by the fact that I don't understand how he could be hearing anything at all, but I don't doubt his words.

[@MosoKaiser]

@@decoryder Expectation bias is a bitch. Also, the fear of losing on sponsorships and review units being sent one's way.

[@MachielGroeneveld]

Ethernet doesn’t carry noise into your audio or video layer. Amir is so polite for measuring all these switches proving something that is impossible for not being there.

[@jamieirwin6333]

Awesome review, a breath of fresh air! Don’t ever sellout, we need intelligent people like you Amir to stop the audio foolery.

[@Burevestnik9M730]

There is a whole movement of unprecedented morons and imbeciles led by the cult leader Chris Connaker of Audiophile Style LLC 4211 York Avenue North Minneapolis MN 55422 United States who negate the obvious: there is no f. difference between $20 switch and $640 EtherRegen. It is not the question whether it sounds "good" or "bad". It sounds IDENTICAL. Amir both listened and measured. No f. difference. EtherRegen is one more snake oil on the market intended for - sheep. Shaaave the sheep - screams Connaker and his bunch of scammers.

[@KeithHeinrich]

@@Burevestnik9M730 I agree but individuals have the freedom to waste their money as they see fit. Arguably cables and switches are cheaper than drugs in the long term and do less harm. ;)

[@richardhernandez7741]

Why on Earth does it matter to you how others listen? Oh... you care how others spend their money... How so very altruistic.

[@Pete.across.the.street]

Does nobody see the problem with his tests?

[@baredesigns1]

@@Pete.across.the.street please educate us, what it means.

[@metal571]

A wild Amir appears

[@AudioScienceReview]

Hello Metal. :) Did I hear right that you are no longer doing headphone reviews? You were doing such a great job of them. I usually seek out reviews of headphones before doing my own and invariably I would find that I agree with your assessment far more than others.

[@metal571]

@@AudioScienceReview That's true, and my final video explains it, mostly. Nowadays I try to help audio companies behind the scenes by providing subjective impressions and suggestions where possible if they are interested, and I mostly live on Twitter, talking with Dr. Olive, Resolve, oratory, etc. Appreciate the kind words. Means a lot coming from you man. Stay well

[@metal571]

@@AudioScienceReview Are you planning on doing headphone reviews on this channel too? Would be really interested to see those. The world needs as many measurement-backed reviews as possible, Tyll style of course

[@mcbpete]

Something I've always wondered about with these kinda products (be it cables, hardware devies etc.) - If it can be *proven* to have no measurable difference between 'non-audiophile' equivalents, how do the companies get away with the marketing blurb from a Consumer Protection/Trades Descriptions POV ?

[@horacekope5939]

I work, and have worked with networks for over 2 decades, and there is no way any half-decent switch can make a positive difference to an audio signal. If the switch wasn't any good, it would have been taken off the network infrastructure because it is obviously corrupting data that a typical TCP stack (whether on a L2 or L3 device) couldn't already 'correct', or more accurately, request again. A faulty switch usually speaks of hardware failure. Data traffic is far more crucial to transmit accurately than audio signal; the audio signal is carried as data. A switch is primarily created to transmit data accurately (along with routing capabilities and possibly other network management functions), bit perfect if you will. And if data is corrupt, you'll literally miss out on 'bits' of music.

[@erics.4113]

I'm a software guy and have enough networking experience to finally understand the world of BS in audio. This one hit a nerve, because thinking you may NEED to improve a network switch is hilarious to me. Talk about fixing a problem that doesn't exist.

[@AudioScienceReview]

It is amazing, isn't it? Yet every day more companies are producing more of these. It is easy game. Get a dirt cheap switch, change its power supply and maybe the clock and claim it is an audiophile switch!

[@dekanyz]

@@AudioScienceReview I think, many people want to spend money to improve their gear, even if it sounds superiour. The manufacturers just allow them to do that by creating high end cables, fuses and such devices.

[@eDXTRe]

Not to speak of 'high end CAT6 audio grade' cables🤣😂...... Pure baloney i think. Just normal switch and cat5 cat6 cable does work perfectly fine😊

[@erics.4113]

@@eDXTRe imagine the results of a panel blind testing cat6 cables

[@eDXTRe]

@@erics.4113 I guess 50/50 results may be bud more like 40/60 60/40 for furtunate gamblers. But someone who pretends to hear difference must easily be able to get 90 or 100 percent score in a double blind test, which they never want to do😅

[@joserafaelhernandezcarucci1324]

Great video Amir, thanks for this one. Any chance you will do the same for the "superior" power sources?

[@AudioScienceReview]

You are mind reader because I was thinking that was the next one that would be necessary to do! I will up its priority. :)

[@joserafaelhernandezcarucci1324]

@@AudioScienceReview fantastic, keep up the great work! :)

[@Adream-lf3mw]

@@AudioScienceReview Maybe verify or debunk the $5k Audioquest power conditioner?

[@jonathanmatthew5631]

@@AudioScienceReview please do so! I am also interested in your power supply review

[@joentell]

Welcome to KZbin Amir. Good to see you here.

[@yangarzon8994]

I Wonder if you could give us data on power line wifi connection for audio. I use that between my streamer and my nas/servir roon / at the opposit from my system without issue or noise. Bit Wonder if you could analyse it

[@MyFatherLoves]

I know this video was made quite a while but I hope you can reply. I've been searching for an all-digital surround processor that can leverage Dante AOIP. Storm Audio has one for $26k and JBL has 3 Synthesis processors that also has Dante integration, with the SDP-55 being the least expensive and actually affordable second-hand. My question is, is the audio pristine in a surround processor before it's run through its DACs? If it is, one could purchase an SDP-55, send audio through Dante, and then onto our DAC(s) of choice or Dante-In of a studio monitor, and we'd only have to worry about that performance of the outboard DAC(s).... right? Companies like HEDD audio have Dante bridges for their monitors that their very nice in-monitor DAC could resolve beautifully. If the SDP-55 truly doesn't harm the audio when sent via Dante, a whole new world of beautiful DACs and powerful DSP (such as Q-SYS) open up to us. You could sort-of DIY you're own Trinnov with far less investment and far better measuring gear.

[@GentielioGaming]

Peter from Steinway pianos was in a caliberation with Lyngdorf open baffle speakers that were crazy high in prices but for me it was the best in the show i went to. He also mentioned that 20% is the speakers that are responsible for the sound and 80% the room and acoustics. That is why he was comitted to the dsp that Lyngdorf used for their high end speakers.

[@matthiasmartin1975]

27:16 Dude, have you never heard of the EmotionEngine in the Playstation?

[@jmp622]

My TP link switch can’t handle hirez music streaming from Roon. Must be connected correctly to router on its own cable. All other devices connected to switch run fine including video. Very odd

[@Waitaminutesilly]

Anyone who understand networking would tell you that audiophile network cables or routers are completely scam. Great job Amir

[@gregsonberlin3782]

Oh Amir, you are talking so much out of my heart. It‘s like your are my tongue. Thanks for all your work.

[@mmil]

I am still waiting for an "audiophile person" to comment on the "break-in period of the switch" and a proper "warm up process" before it starts working at its "full potential". Because we all know that those 0s and 1s need to loosen up a bit over time and if you warm them up before hand you definitely get richer sound :)

[@AudioScienceReview]

As it happens, the person who sent it to me "broke it in" from what I recall by running it a while. But yeh, that takes absurdity to an entirely different level.

[@MrAlb3rtazzo]

how about ethernet filter like this AARDVARK ETHERNET ISOLATOR ?

[@ENDUROYZ250]

Great explanation, thank you for that, with your video a lot of people in Portugal are not going to be happy .

[@niallmacdonald2710]

When I saw the title of the video, I laughed. I still watched the video anyway, then laughed again. I like my hi-fi setup, and have I feel a multi-room system at about the start of what people consider audiophile, and no further. Diminishing returns after that point in my opinion. I went over to FLAC files running from a central home media server about 20 years ago, to small pcs around the house, connected to my DACS via toslinks, to prevent any ground noise from the pcs getting through to the DACS. Most of my hi-fi equipment is Arcam and Musical Fidelity, with one Yamaha integrated receiver in the office. I have never, for one second, considered network switches to have an impact on the sound, probably because I had a CCNA qualification (now lapsed) and understand how TCP/IP works I feel reasonably well. I enjoyed your video, probably in part because of confirmation bias. Over time, the connected pcs have become smaller and lower power, and the media server has had several rounds of upgrades, but everything downstream of the toslinks has stayed the same. 1990's sound still sounds good to me.

[@matthewv789]

It does depend on whether you’re using TCP or UDP. While UDP is less common, or at least less familiar since the web runs on TCP, UDP is actually very common if not standard for applications like online gaming or streaming audio or video. That’s because they are time sensitive - by the time you would receive the retransmission, it’s too late, the world has moved on and that data is irrelevant. Video would rather drop some pixels or display some noise or compression artifacts or briefly reduce the resolution rather than freeze. The action from other players in online video games continues in real time whether your network can keep up or not. Finding out some detail of what they did 10 seconds ago is 10 seconds too late, the world has already moved on. Etc. Most such applications build their own reliability protocol on top of UDP, rather than use TCP; they use something more appropriate for their use case than TCP is. On the other hand, data is data and everyone wants fast, reliable transmission of packets no matter the application. Wired Ethernet, as you say, is usually not a cause of problems in most home settings. Gigabit Ethernet has way more than enough bandwidth to easily keep up with even multichannel high-res audio, especially if there is little other network traffic. But for cases where it is too congested (which happens at a lot lower bitrate that the advertised maximum, maybe half as much or less), there is a feature called Quality of Service (QoS) which some products support, which allows you to set some traffic as having priority over others, because it needs to arrive as quickly as possible. So audiophiles might want to get gigabit Ethernet switches with QoS, and make sure to set it to prioritize their streaming audio data. But they don’t need to be specific “audiophile” versions, that’s silly. The capabilities needed are the same as for gamers, which is the most common need for this type of feature and capability. And of course, if the network isn’t keeping up with needed data rates, it’s likely to be extremely obvious (choppy or interrupted audio), not some subtle degradation of audio quality (unless you’re streaming through some service that can automatically degrade the data being sent to a lower bitrate in response to network conditions, like KZbin does; then a subtle degradation is entirely possible). Jitter? They must be kidding. Network packets come in when they come in, it could be sooner or later, by a substantial margin. It’s up to the D/A converter to assemble data that’s been temporarily stored in a buffer and use its own clock to emit them on a specific schedule. As long as there is enough data in the buffer, the timing or noise of the network switch is irrelevant.

[@guyboisvert66]

Exactly! TCP/IP and UDP/IP are used in best effort networks, like the internet, your intranet, etc. So a streamer must use buffering technique (async communication) to make sure it has enough data to feed the DAC even if the packet receiving bandwidth can vary a lot. The buffer can be as big as the RAM permit... As for UDP vs TCP, the difference is that UDP is connectionless and sender / receiver must take car of lost packets whereas TCP make it for you. For your network, just design it to get the less impact possible from the potential big data transfers between, say, your PC and your server. That's quite easy to do... For internet, yes it's good to get a router that supports QoS, check the Mikrotik line of Pro Routers that offers that and then some. It's pro stuff for ridiculously low price. If you have a more complex and busy network, then a switch with QoS could be used, but as a network engineer, again it's easy to design a network to avoid this while not having to implement QoS... And for those talking about noise, linear power supplies, jitter and all other inapplicable nonsense, please spare us your complete misunderstanding of networking... Go back study and maybe one day, you'll realize that this is complete snake oil...

[@3rdeye1983]

Can you confirm if audio transmission for Roon for example is TCP only? Also isn't audio/video usually over UDP?

[@thomaswalder4808]

UDP is normally used for applications where close real time is important and which are data in both directions. When you do a phone call over Internet you not want a buffer which delays the playback by 3 seconds as that would make bidirectional conversations difficult. Streaming has not to be real time and it transmits data only in one direction. If the playback is delayed by some seconds it does not matter. Even if UDP is used that does not necessarily mean that there is no error correction. TCP/IP handles all errors and packets arriving in wrong order on its own - so the application developer not have to worry about it. UDP does not - but the application developer can still implement its own error handling.

[@S.M.A.Batista]

Amir please keep doing videos. Huge fan. I love the honest delivery.

[@pascalcollet3897]

Brilliant thanks i obviously doubted an audiophile switch would do any good being an IT admin. You have just honeslty proved that

[@dylanneild2580]

Great video. One tiny quibble i'd make is that ethernet units are called "frames", rather than "packets". "Packets" are the encapsulated data of IP, and "segments" are the encapsulated data of TCP (or whatever other transport layer protocol is in use). It's a small distinction but an important one when talking to networking folks. Otherwise, spot on analysis backed up with clear, reproducable testing.

[@molhammhd8068]

Do we need to have an Audiophile Internet Service Provider?

[@BadGuyGoodAudioReviews]

32:26 My spirit animal. Great stuff

[@joshua43214]

Amir! Thank you so much for making vids. Love your reviews!

[@MyFatherLoves]

This is something I've wondered for a while now with my job in Pro Audio sector and Dante.

[@Michiganman08]

This is eye opening and hysterical. Thank you for work!

[@dreddguy6454]

It's nice to have an explanation of the graphs. I usually skip them on ASR and go direct to your summary. Even then I take the results with a pinch of salt. Some products that measure poorly can still sound great. But I'm glad that I now have a better understanding of the science and appreciate the way you cut through the BS manufacturers use to flog their wares.

[@guyboisvert66]

If it measures poorly, then you're just hearing a distorted / colored version of the original signal. Yes you can love it, not me... You can as well buy an equalizer if your like. An example of that are the tube amps that add harmonic distortion, some are not even linear!

[@Harald-MacGerhard]

thx a while great bunch Amir. I have been living with my TP-Link Archer wifi routers and using the hardwired ethernet ports for the critical communication, all audio equipment being hardwired ethernet. Whatever they say about wifi.... hardwired ethernet is rock solid. You saved me a lot of headache and pain and nightmares. The TP-Link strategy stays :-)

[@RedSinter]

Speaking of WiFi do you have any reviews of the best WiFi products or ones with the best ability to offset or negate potential interference and throwing distance.

[@raccerx67]

keep it up sir, this is what we need to stop the new people from wasting money and they can spend it where it actually counts, much respect and thank you!!!!!!

[@Nikkel_]

I think some or all of us audiophools should actually just donate 10% to you of what we were going to spend on snakeoil instead! Thank you Sir!

[@antoniomarsicola8608]

I'm dreaming of a similar debunking analysis on streamers such as a 1200$ SOtM sMS-200ultra compared to a 40$ Raspberry Pi4....

[@r423fplip]

Chromecast audio !!!

[@antoniomarsicola8608]

@@r423fplip great gear, i have it, but it hasn't a usb out so it's not asynchronous, it could really have some jitter

[@dilshadmashoor7566]

Can you please post a link to where you found a Raspberry Pi4 for $40 thanking you in advance

[@fieldcar]

I'm loving the videos. Keep it up!

[@AudioScienceReview]

Will do given all this kind encouragement.

[@MrKimber58]

Finally a breath of fresh air!!! Thanks for all the work Amir.

[@dingbat19]

how about you reach out to some manufacturers and see if they are willing to comment? e.g. Thunder Data, Melco, Innuos, UpTone ? Do you ever do this?

[@rogerschofield3324]

Wonderful analysis , thank you.

[@derekhughes5093]

Great and honest video Amir, well done.

[@AudioScienceReview]

Appreciate it Derek.

[@AH-wr1ir]

Amir, Please can you explain why, after installing a Bonn 8 network switch into my system, the picture quality from my ethernet connected sky mini/projector improved so significantly I had to make very large adjustments to the picture settings? I am not discounting any of your findings and the scientific method is the core to humanities advancement. I am sincerely looking for an explanation? many thanks.

[@moonlight-kh6uz]

You seem to be lagging on knowledge for at least 10 years. The concept of jitter misleads people into thinking that all you need in a digital signal is the correct bits (which is relatively trivial to transmit) with great timing (low jitter), and so all you need is a great clock. This simplistic view is highly misleading. At least three things matter - the clock, noise and bandwidth. In the image of a perfect square wave, the horizontal axis is time and the vertical axis is voltage. We will assume the clock is perfect - ie. the vertical signal lines occur at perfectly spaced intervals (the bit rate). When the signal is representing a binary 0, it is at 0v. When the signal is representing a binary 1, it is at 1v. And we will assume that the receiver of this signal decides that the transition between a 0 and a 1 has occurred when the signal rises through the 0.5v level, and that a 1 has transitioned to a 0 when the signal falls through the 0.5v level. Now imagine that there is noise added to the signal. If the frequency of the noise is below the bitrate then this perfect square wave swims on top of a longer and smoother wave. The interesting point is that the timing between the data transitions (where those vertical lines pass through 0.5v) is unchanged. So no problem, yet. If the frequency of the noise is above the bitrate then the horizontal lines get fuzzy. And if we combine the low frequency noise with the high frequency noise the effect is combined. Again, the interesting point to note is that the timing between the data transitions (where those vertical lines pass through 0.5v) is unchanged provided the noise is not extremely high. So, again, no problem. Noise, on its own (as long as the deviations caused are materially below 0.5v) is not a problem. The reason it is not a problem is those vertical lines, because noise does not change the space between them. Now imagine there is no noise. Zero noise is impossible, but something else that is impossible is the vertical line on the square wave, since it requires infinite bandwidth. The vertical lines imply the signal can achieve 0v and 1v in more or less the same instant. Whatever tools we have to transmit a signal, the demands of high bit-rate signals are way beyond what the available tools can deliver. Think about how your analog cables can mess with sound up to around 20kHz, and then think about the enormously wider frequency range required of a digital cable (and, optical cables just have a different set of problems, mainly related to reflections). The higher the bit rate the harder it gets. When we allow for constrained bandwidth, instead of transitions being instantaneous, the signal goes up a slope when transitioning from 0v to 1v, and down a slope when transitioning from 1v to 0v. If the bandwidth was the same as the bitrate then the signal would be a sine wave. To reasonably square out the signal you need to add several harmonics of the bitrate (say 7 or more) above the bitrate, and that is a lot of bandwidth - even more for higher bit rate signals. By adding harmonics, the sine wave begins to square out. Interestingly, in both of these constrained-bandwidth examples, the transitions through 0.5v are still perfectly spaced - even with the sine wave. So still no problem. But as I mentioned, a higher bitrate signal (if you think high bitrate files must always sound better) requires even more bandwidth to square out the wave, and so in a system that has a finite limit on bandwidth, a lower bitrate signal will be more accurately represented than a high bitrate signal. On top of that, if you ask anything in a music server to work faster, it will work with less precision and this is a key trade-off to be aware of when you assume higher bit rates must be better, just because the numbers are bigger. These examples only allow us to conclude that there is no problem if we can achieve zero noise or infinite bandwidth. But each of those goals is unattainable, and the problem becomes apparent when there is both noise and constrained bandwidth. So what happens if we add a low frequency noise component to a frequency-constrained digital audio signal? All of sudden, the 0.5v points are shifted right or left by the addition of the low frequency noise that lifts or drops the signal between bits. Shifting the slopes up or down shifts the 0.5v points left or right. The greater the amplitude of the noise, and the greater the bandwidth constraint, the greater is the effect on timing (jitter). Now if we add high frequency noise to a frequency-constrained signal you can see that the transition timing at precisely 0.5v is now hard to discern for any digital receiver. If the signal is vertical at the transition then noise does not affect it. But as soon as the transition is not vertical then noise changes the transition point. It is the combination of constrained bandwidth and noise that inevitably creates jitter (variation in data transition timing), regardless of how great the clock is.

[@Cygnus56]

Bravo! I feel much better now with my Netgear GS108 for my expensive NAIM gear. We audiophiles can let our addiction extend to non-essential things, if we don't do enough research. Dare I ask... do ethernet cables make a difference? I've done blind A/B testing on my system with different ones and I swear that I can notice a small difference.... ok maybe I'm crazy.

[@andyrocha]

@Audio Science Review Hi Amir great review. Have you tried or any experience with converting Ethernet to fiber then back to Ethernet before the dac? Curious if the measurements there show any improvement or if it’s one of the same, have read more about that lately. Thank you kindly

[@BL-yj2wp]

22:06 - very good and important point to make.

[@urbantone]

Yeah i was just offered a deal of EtherRegen and this opened my eyes Thanks Amir

[@gregsonberlin3782]

BTW: I found a „test“ of the Innuos PhoenixNET before watching your video. Your arguments are almost the same as my thoughts (I‘m a computer scientist). Unfortunately I think that I shouldn’t link to this test here but I wanna claim that it‘s rather criminal to fool people like they do. So keep up in educating us audiophilly!

[@ChristerJohansson]

Nice video! Would you anticipate any difference if a PoE switch is thrown into the mix?

[@AudioScienceReview]

I do not. The barrier to noise from Ethernet is so strong that it doesn't matter what is on the other side. Now, if you use POE to power your gear -- which is rare -- then its dirty nature may make a difference. I have tested one such product: www.audiosciencereview.com/forum/index.php?threads/dante-avio-review-streaming-audio-interfaces.19062/

[@Ray-dl5mp]

Amir is actually a very funny person in these videos. I really appreciate it.

[@fan2hd277]

You are throwing dynamite in the audiofool lake. I like 👍 it. Btw, your test is flawed! You didn’t test it with Audioquest Diamond RJ45 1000$ cable ;)

[@fubartotale3389]

I'm having audiophile quality air pumped into my home, (trace gasses free!) You would not believe the expansion of soundstage and improvement in imaging!

[@48mcmahon]

allitis setting packet switching priority to that type of packet that contains the music data

[@rix911]

As a 20 year networking/data specialist this video is a breath of fresh air. I especially like how you disected the mindset of the people who buy stuff like this and formulated a video to actually convince them. When asked questions like these I generally give very similar answers to people -- minus the fancy graphs and presentation. One thing you should mention is mention is the concept of CRC checks.

[@Astulock86]

Cyclical redundancy check? I have an old hard drive with a CRC error and my data is locked away, irretrievable. Any suggestions?

[@rodneyvandenoever]

The failed CRC check warns you your data is corrupt, and now prevents you from using this corrupted data, which is a good thing (dependent on how you look at it)

[@ronlysons6750]

The last five minuets of this video says it all. Don't go looking for improvements because you will most likely find some, wether you switch out a component or not. Let the improvements find you, Also, this is probably the only you tuber I believe, most of them are just salesmen. Influence's.

[@ChristianGoergen]

So the Ether regen is absolute transparent too. It doesn’t alter the signal! What a device!

[@oscarmarfori613]

Your video justified that I don’t need to get that expansive switchers I just ranned to my neighborhood electronics store and grab that cheap network switch, guess what? The streamer sounds great even just listening to my files that was sent thru the streamer or thru my internet music, thanks for sharing this review 👍

[@dogratco]

Thanks very much for this review. A very minor clarification in wording is that an Ethernet LAN switch handles frames, not packets.

[@josephwhite4043]

Frames are built with packets. If you are going to correct someone do your reading first….

[@dogratco]

@@josephwhite4043 Yes, but an Ethernet switch does not operate at the packet level. Please search for "ethernet frame" and do your own reading.

[@dogratco]

An Ethernet switch operates on frames, not packets. Please search for "ethernet frames vs packets" to do your own reading.

[@dogratco]

I first worked with Ethernet in late 1984, after the IEEE 802.3 spec was finalized. I don't need to do further reading.

[@gioponti6359]

Re unplugging the ethernet cable and eliminating a potentially existing (or non-existent) noise source while playing music: Results *might* be different between streaming from streaming service (Qobuz.., where locally stored data is very limited) and streaming from local server.

[@MrSilviut]

GENERAL ASR COMMENT. My problem with all these measurements Amir does (referring to everything) is, I’m not smart enough to know how relevant/useful they really are. It’s great that everything is so technical, but how would I know, for example, if placing a mic in right front of a speaker is the way to test a speaker? And how do I know about how the environment affects that test? I’m not really an “audiophile”, I just like music. I don’t have the technical knowledge to know if a lot of these things are relevant. The biggest things making an impact that I’ve noticed are: 1. Recording and Mastering “source file” 2. Speaker and power 3. Room treatment/acoustics, and room treatment/acoustics is arguably second. So I’m glad he shows that many pieces of equipment are irrelevant because it lines up with my experience. Nothing else I’ve bought really made as much difference as those things. Maybe the DAC could be considered important.

[@MrSilviut]

By the way I really appreciated Amir’s video explaining how to make sense of a frequency response chart and how it related to psycho acoustics.

[@awsomenesscaleb]

Please test a Shun Mook mpingo disc. I want to know if it works. 🤔

[@intothedragon]

I would love to know Amir's opinions on the Melco s100 that goes for 2 thousand Euros

[@DEALUX]

Why do you need one? Does your router not have multiple LAN ports?

[@gioponti6359]

Alpha audio has done a live test on various switches anyone can see here on youtube…

[@JustBrowsing777]

I thought jitter was always about timing and not noise?

[@Pete.across.the.street]

He was trying to measure jitter in analog too

[@Adream-lf3mw]

Amir, thank you so much for making these types of videos. We need more truthful information. I really appreciate the work you are doing. I subscribed right away and watch all the vids you put out so far. Maybe once you get around to testing all the gear you have planned you can verify or debunk the high cost of Pakedge port switches and wireless routers costing in excess of $1,000 each? Thanks again.

[@guyboisvert66]

Control4 just charge premium for something you can get really cheap... even pro stuff like Mikrotik cost much less than that!

[@michaelrovner4165]

Absolutely do ..made a huge difference going optical and using linear supply...2 inch cables

[@mariopalijan1290]

Sir, sorry for the expression, but you're kick*ss ! Love your work ! I'm ITman and an rookie audiophile. And I'll stay rookie AF... When I first heared for audiograde UTP cables, that was too much for me ...

[@brave1671]

Thanks a lot Amir! Great channel!

[@anderssvensson4554]

Thank you sir! Very interesting and educational. Hearing you explain takes your measurements to the next level of clarity.

[@johnk7451]

What about the bs usb regen?

[@meaninthemirror]

Thank you for taking time for debunking these kind of snake oil devices.

[@Pete.across.the.street]

He's selling the snake oil, measuring jitter after the dac,... Dumbest thing I've heard today.

[@Wordsalad69420]

As a software engineer who works on the networking layer, this one is hilarious to me. People say data is sent analog. Yes, but that is irrelevant, because the software will make sure the data is exact. If a packet it dropped, it simply retries. Unless your streaming protocol uses UDP (I don't think they exist), then it is not possible to have streaming issues, especially if you are streaming through your house.

[@decoryder]

Simple test: Play back some audio file (mp3, flac, whatever) over a remote fileshare on your network. Then copy that file over to your local machine, and listen to it again. Chances are that you won't hear a difference. If there's was some 'degrading' influence from whatever switch you're using, you'd be able to hear it when you're playing the file over the network. Thank you Amir for shedding some much needed light on this subject.

[@sporqist]

Don't listen to it. As humans, we are biased. Just record the audio output and compare the resulting audio on an oscilloscope or in software. Funnily, none of the "audiophile" reviewers do this.

[@decoryder]

@@sporqist Given the fact that the makers of said 'audiophile' switches all claim that their gear makes an 'audible' difference, I don't see how a simple listening test to verify said 'audible difference' requires the use of an oscilloscope and/or software.

[@neosolipsist1016]

Yes Amir is the best. I believe in the testing he does. Bottom line: My DAC buffers almost 30 seconds of minutes of data and then reclocks it on the way out! Let your components do what they were designed to do...

[@smoorenc]

Thanks for doing this. I was wondering how it could make a difference. I have a LPS on my switch just to keep the wal-wart out of the chain. Great review!!

[@guyboisvert66]

Easy to answer, you don't need an LPS on any networking gear...

[@chrisharper2658]

One can only hope that what you've shown can help at least one self proclaimed 'audiophile' understand the error of their ways. I had placed a critical comment on one 'audiophile's' review of an 'audio grade' Ethernet Switch and got slammed. I think it was that old German guy. While I understand TCP-IP, asynchronous communications and buffering, some 'audiophiles' make there own truths. Kind of like politics these days. People want to believe in something and companies are popping up left and right to meet those needs. Very nice review.

[@waqasahmed939]

It's really weird to see audio guys ask network guys about what switch they recommend for audio, and we're like "Just any decent switch will do"

[@armoth6958]

Hey, note - I’m an audiophile. Doesn’t mean I believe in nonsense claims by companies

[@sporqist]

@@waqasahmed939 -- audiophile proceeds to power his 15$ TP-Link switch with a linear power supply :)

[@goodsound4756]

I consider myself as an audiophile, I tested such a switch, but couldn’t hear *any* difference, so why should I need one? Don’t declare audiophiles generally as Fools.

[@chrisharper2658]

@@goodsound4756 Not all audiophiles are the same and to that point, it seems some audiophiles are fools and have very strong opinions too. I think the term audiophile can be very ambiguous but is mostly ego driven.

[@jr-6025]

This is already my favourite youtube channel

[@AudioScienceReview]

It is mine too! Oh wait. :) Thank you.

[@TrinitronX]

The only reason to buy some kind of audio-oriented network switch is very plainly: If you need a device with EtherCon jacks for rugged use cases (e.g. stagebox connections, etc...). Otherwise, there's no benefit to using a device that advertises some improvement in audio quality. That said, there actually _are_ some network switches that perform better than others due to their implementation of certain IEEE standards. These generally are only important if you're using Audio over IP (AoIP) technologies such as Dante, Ravenna, and/or AES67. Namely: 802.3az Energy-Efficient Ethernet (EEE) and buggy implementations of IGMP Snooping (RFC 4541) can cause PTP clock sync issues & subsequent audio drop-outs. Other than that, I've never seen a piece of network gear cause a problem so long as it allows for turning off 802.3az (EEE), and implements IGMP snooping properly (e.g. for multicast streams & PTP broadcast packets). That said, some unmanaged network switches and even an older model of MoCA adapter that I have both seem to caues PTPv2 sync issues because they implement 802.3az (EEE) in hardware, with no way to switch it off. This causes the ethernet ports to enter a low power state periodically, which can contribute to other devices with less accurate internal PTP clocks to drift and de-synchronize just enough to cause the devices to mute their Dante or AES67 streams. Note that PTPv2 can be extremely accurate, with nanosecond level precision of clock sync between these devices otherwise, so it's very sensitive to 802.3az (EEE).

[@mikede2464]

Another valuable video by a seemingly polite, mature, intelligent, well spoken, nice guy.: a rare thing these days especially on youtube. Thanks for putting these out.

[@kcfish4862]

Gotta change the channel picture to ASR

[@AudioScienceReview]

Thanks. I was blind to it being the generic one. Just fixed it.

[@SuspiciousAra]

I have to say what audiophiles are saying in this case: sir, but you didn't break in that cable and switch and whatever. you need to use them 1-2 weeks and then repeat the tests :D

[@chestermango1201]

Me new favourite KZbin channel. More please.

[@AudioScienceReview]

Thanks Chester. Was only going to do these videos occasionally but seeing the reaction, you can expect more.

[@amirjubran1845]

Just want to say that your comments at the end made so much sense and even though it is obvious I needed to hear you say it out loud to hammer it home for me. That is why more and more I count on my wife to comment on her feelings whenever I change things in my system. She has no emotion when it comes to the gear so I can always count on an honest opinion from her.