Alpha Labs - Does a Bad Network Switch RUIN your streamer?

[@jamesgiampapa]

Always appreciate the effort that goes into your videos thanks

[@BrentLeVasseur]

This is why I switched from copper network Cat6 cables to 10gbit fiber optical. At least with fiber, all the noise from each device is essentially siloed, so that it doesn’t collectively accumulate and cause jitter in the DAC. The jitter is then minimized to during the initial conversion process in the Mac or PC in the fiber cable transceiver itself, and that acts like a firewall stopping it from going further into the switch and then into the DAC. You can further reduce that noise by putting all your gear on batteries or low noise power supplies, which decouples it from the grid and further helps reduce jitter.

[@phantomplastics6582]

Amir will be having more sleepless nights after this.

[@jmtennapel]

This isn't a competition or anything. We do what we do, they do what they do. Both is fine. Just don't stir up a fire that we are not looking for, we rather put our energy on HiFi equipment and investigating why we hear what we hear.

[@nathankimmett8803]

Amir the software engineer? Who cares… he has his cult - let them go drink the red koolaid together. I’m here for the audio and music!

[@Blacksheep1042]

I can only carefully calculate Guess that if he will choose/ find it in his interest to analyse- research the influence of low quality switch on the output performance of a streamer he will sleep quite well this night and all the coming nights. In the morning, he will wake up and do what he is doing for a living… I can also carefully assume that He will focus on the end of the Chain under the common sense modern approach of - The manufacture/designer have to Execute electronic audio +electro acoustical 2023 updated engineering know how. No more, no less. Be blessed, healthy, creative, free and prosper.

[@DBravo29er]

Amir actually isn't interested in proof that these things matter. He seems to only be interested in promoting his ideology....his faithful belief....that things like categorically *can't* matter. And, because of this broad based assumption, that he is permitted to dismiss them, mock them, and gaslight them. Because he claims "sCiEnCe!", he is unrebuttable. That's his worldview. And the view of the goon squad that he cultivates on his forum; he corrects exactly zero people in his goon squad who assert the above. Therefore, he's aligned with it.

[@AudioScienceReview]

​@@nathankimmett8803 I am an electrical engineer. Grew up with electronics design, before there was such a thing as a personal computer. Once that was invented, I also got into software so have both skills.

[@jannikneumann2609]

Jaap, can you please test what those little lan-isolators from iFI or Pink Faun do in that scenario + also the question is an optical network the way to go (yet there are no streamers as far as I know with optical network input)?

[@TheAlphaAudio]

We tested those as well. They do something. But don't use them in serial config... they can reflect the noise as well.

[@grandebufalife]

The Hifi Rose rs130 has a optical network input

[@jmtennapel]

@@grandebufalife As does the iFi Neo Stream, and new Linn streamers. MSB has implemented their own version of optical connections (propriety protocol), but not for connectivity input. We have the iFi Neo Stream in the upcoming test sunday, but we will use the regular ethernet connection.

[@jmtennapel]

You can hear what an optical connection does in the livestream where we listen to 11 different switches, including an optical connection. Just browse back in the livestream archive of this channel.

[@edstromen]

Does it really affect the analog XLR output to amps and the output from amp to speakers so it really is hearable? undependet of say and stereo costing €2500 and upwards in eternity.

[@TheAlphaAudio]

Yes

[@hudosytuk]

So what switch to buy, or I can just get decent Netgear/D Link with LPSU? What about LAN galvanic isolators, would the help here?

[@TheAlphaAudio]

Yes... that would work perfectly

[@carminedesanto6746]

Ok..but when out trough a DAC ..playing music can you hear the difference between a “higher end switch “ and a regular ( industrial) switch 🤔

[@TheAlphaAudio]

Yes

[@thomaswalder4808]

@@TheAlphaAudio Why nobody has proven that in a double blind test?

[@domenicofpv6490]

Wat is nou het beste mbt het aansluiten van de mantel? Beide kanten of alleen de bronzijde aansluiten?

[@TheAlphaAudio]

@@domenicofpv6490 beide kanten... Altijd bij een netwerk.

[@rafalz5878]

That's why I use to my source (streamer) optical fiber and don't have to use "audiophile switches" :). Great job like the rest of the videos!!!

[@RA707AR]

Excelent videos as always - ultimately becoming very, very interesting. In two words : Great - thank you. As a take II, any plan to test different cat /brand cables, eventually Ifi lan silencer?

[@AllHandlesHaveBeenTaken]

Very interesting and value information. I am looking for some shielded cables. Bit how do I know if my switch is good or bad. I use basic netgear switches. I dont have scope.

[@TheAlphaAudio]

@@AllHandlesHaveBeenTaken we have some extensive tests on our website. Https://www.alpha-audio.net.

[@sviktor4]

Is your 230V power supply fully isolated? Can you make these tests with properly isolated power supply? I'll be forward please just use a 5V EI transformer with a greatz and 2 capacitors for power supply. What happens if you ground the streamer? What happens if you use the mains to power all the devices, i mean don't isolate the streamer from the system? People use their system as it is fully cabeled, tons of WiFi and mobile signals al around, 230V not clean. Can you measure the streamer with battery powered equipment in a fully assembled system?

[@Thoughtflux]

Love your videos man! Already subscribed. Question - if physical LAN connection can inject noise then why not use only wifi? I'll be grateful for your response.

[@TheAlphaAudio]

The wifi radio in the receiving device also introduces noise.

[@Thoughtflux]

@@TheAlphaAudio i plan to use a Adum3165 based USB isolator after my pi before it connects to a xmos/crystek reclocker (powered by LPS) which feeds my dac via coax. Will such an adum based isolator be sufficient to mitigate noise & jitter introduced by LAN or Wifi?

[@iBlindPanic]

Are you saying a digital signal is influenced by noise? If so you are simply wrong (onzin). Imagine software transferred via a switch from one pc to another, one changed bit would make the software unusable. That is why we have things like CRC, we simply send the data including timing info again if we detect a fail, that has been the case for cd players for a while and for computers it was always the case.

[@johnh539]

What is a switch? I have a full fibre router less than a year old and a galvanic bridge on my ethernet cable. Is a switch another thing you add to the ethernet line? If so is it only something you need if your router is not the best? I am within a few weeks of spending another £1400 on a streamer, a linear power supply and cables but I thought that was it, don.

[@TheAlphaAudio]

Google...

[@roberts3889]

Great video. For the average "audiophile", ignoring the question of a noisy switch due to a defect, how would a one determine which switch(es) are designed to be low noise?

[@nathankimmett8803]

Excellent video - thank you!

[@TheAlphaAudio]

Thanks!

[@audiotom3855]

Very interesting as always, thank you! Is there any future plan to test the influence of different routers?

[@TheAlphaAudio]

No, not really. A router does not have any influence once you have a switch connected to your streamer. I always advice that, for connecting directly to the router is not ideal in terms of noise.

[@audiotom3855]

@@TheAlphaAudio thank you. I have a switch next to my streamer and it worked wonders :) I am always interested in improving further, i think using a low noise power supply will improve things even on the router

[@jmtennapel]

@@audiotom3855 You can try this for yourself of course. I did, out of curiosity, and in my case it doesn't.

[@TheAlphaAudio]

@@audiotom3855 You can try. I never heard a big difference once I installed a decent switch with nice power supply.

[@Tacet137]

Very nice now measure the output of a dac so we can all see how huge impact it makes, right?

[@krell2026]

Thank you for yet another informative video. Your finding of measurement changes is very useful to know. I have experienced the sound differences with better switches, but the naysayers are already out and about by the looks of it. No matter what comments you get your work is valued :)

[@krell2026]

@@baronvonaux8294 each to their own, i know what i hear and wouldnt part with money if a component wasnt audibly better, to me, noone else. Alpha do the technical tests as well as actual playback sessions, far more comprehensive than others - its a lot of work so be respectful. There are many things in hifi both digital and analogue that can be changed with all sorts of options and opinions and there is the 'cant possibly make it better' and cost/benefit angles. At the end of the day you can lead a horse to water but doesnt mean it has to drink it....

[@Roosville1]

Just to clarify, 100ps of jitter is 0.0000000001second peak-to peak change, which then goes through a DAC anti jitter circuit and a PLL lock in which will pretty much intergrate this intto something an awful lot less. Nice measurements, but lack context.

[@TheAlphaAudio]

100ps jitter is very, very audible... Also, we measure more than just jitter. Phase noise LowFreq modulation, clock stability, etc is also influenced by noise from a network. I would suggest to just sit down and listen. Especially a decent power supply will be audible in a good system.

[@ClassifiedBrief]

Jitter is also additive, it will not always be removed and yes that jitter is absolutely audible.

[@thomaswalder4808]

@@TheAlphaAudio "100ps jitter is very, very audible." Is there any published double blind test that support this claim?

[@thomaswalder4808]

@@TheAlphaAudioA turntable causes also jitter as the rotation speed is also not 100% accurate. Even the signal encoded in the groove causes a varying resistance - also the vinyl itself which is never 100% flat. If 100 ps jitter would be "very audible" then listening even to a high end turnable would me an acoustic nigthmare.

[@gerihifi]

I really like your measurement, very helpfull!! In comparison with Amir's measurements, it shows it should be done from two angles! As some people said, it's not Amir not right or you have the answer, BUT it needs a combination of several measurepoints and therefore I prefer your approach!! Keep going, will definitely support your channel! Thanks a lot!

[@TheAlphaAudio]

Thank you! Appreciate it!

[@thomaswalder4808]

It is the output of the DAC that counts. So it makes a lot of sense to measure that output as Amir did.

[@thinkIndependent2024]

The Fix List!!! 1. Ferrite Beads on that network cable 2. Galvanic Isolation on the port 3 . Power the switch with a clean linear source. (The circuit are built with a variance to manage noise) A more complete test would verify any true effect confirming change from input to output

[@jmtennapel]

Or just buy a good network switch.

[@MrDannytercaefs]

Beste, niet direct een vraag over het netwerk maar wat is jouw idee over het volgende. Als je moet kiezen tussen een stream van AAC 128 kb/s of MP3 320 kb/s welke is volgens jou de beste kwaliteit? Met vriendelijke groeten, Danny.

[@TheAlphaAudio]

320 mp3

[@MrDannytercaefs]

@@TheAlphaAudio Thank You 👍🏻

[@lupoal4113]

thank you!

[@domenicofpv6490]

Sommige zweren bij 2 of zelfs meerdere switchen in serie. Jullie dit als eens getest? Wel interessant namelijk om dit te onderbouwen

[@TheAlphaAudio]

We hebben daarmee geëxperimenteerd. De isolatie kan verbeteren. Klank matig is het hit or miss...

[@domenicofpv6490]

@@TheAlphaAudio heb nu 2 netgears in serie lijkt wat meer rust te hebben, maar geen wereld verschil

[@TheAlphaAudio]

@@domenicofpv6490 precies... hit or miss... maar ik werk liever met één goede met mooie voeding.

[@domenicofpv6490]

@@TheAlphaAudio wat is een goede? Heb 2 keer de netgear gs (uit jullie test) met lineair powersupply

[@TheAlphaAudio]

@@domenicofpv6490 LHY maakt goede en betaalbare. Farad is prachtig. Pura met de Dodo ook.

[@LovelyDoetje]

All of that isn't important. The signal on the network does not care for the noise etc. This means the dac and clock shouldn't care either. If any of those parts are influenced by the noise on the network your problem is not the network but the dac (and clock). I know the expensive dacs are almost holy parts in the setup but they are goldmines for company's. In reality they are so cheaply build they can be influenced by noise on the network. With a good dac (even cheap ones) this isn't the case. A good dac has no problem with almost any level of noise on the network. As long as the digital signal can be read and the handshake still works there should be no issue. And if it fails your music pauses or totally stops playing but it will not decrease in quality, there is no loss of bass or space or whatever, it will just stop. So throw away that dac, buy a good one and be done with it.

[@jmtennapel]

Which DAC do you suggest then?

[@a0r0a7]

@@jmtennapel any😁

[@sviktor4]

@LovelyDoetje I totally agree with you, the music should stop and probably does it too, just 10 times the speed like a PC or a Bluetooth headset where you can hear the cracling sound when there is no signal. Since these systems produce sound an order of magnitude faster than your 44100 Hz PC you no longer hear crackling anymore you hear high frequency noise insted.

[@cameronrector178]

That’s why Fiber media convertors are your friend. Keep all that junk away from your dac.

[@TheAlphaAudio]

Not always...

[@cameronrector178]

Show us! Proper implementation of course.

[@TheAlphaAudio]

@@cameronrector178 we have a complete article about that. And a blind test. Check our website.

[@RAMIRO61FERNANDES]

👍💥

[@razisn]

Now everybody who can hear the difference between 9 and 10ps of jitter, please raise your hands. Thank you, now the ones that can hear the difference between -120 and -130db of noise. Thank you again. Btw, doubly shielded Cat6 ethernet cable is a few cents per meter terminated and you can get very well implemented switches with properly galvanically isolated ports (as per the Ethernet standard) starting at $30 or less.

[@TheAlphaAudio]

Ok

[@a0r0a7]

Well said👍

[@sviktor4]

Basically non of the ethernet gear is grounded, most of the power supply create a fake ground connecting the mains with 2 capacitors to the neutral it is good enough to protect the devices from human ESD, but in return 115V AC lingering between any two device what is connected with a cable. This voltage is present in ethernet cables too, you have a double isolation transformer in the chain and the high voltage AC finds its way trough, because the feke ground is connected to the ethernet transformator as the Ethernet standard suggest. The solution is very simple you just ground the negative pole of the DC powersupply.

[@IHearEverythingDude]

No it does not.

[@TheAlphaAudio]

You are funny...

[@IHearEverythingDude]

please measure what is comming out of dac outputs.@@TheAlphaAudio

[@reptidan]

Ruin, no. Negatively influence yes.

[@TheAlphaAudio]

@@IHearEverythingDude how? I would need to run it through an A/D converter.... I tried it a couple of times: it doesnt work. Also: I measured the clock that feeds the dac... how much proof do you need?

[@nathankimmett8803]

@@TheAlphaAudioyou can’t convince people to look who don’t want to open there eyes. Please don’t try to appease to the audio science crowd. There are lots of audiophiles who want to learn and see and don’t worship Amir the software engineer… Keep up the great work - thank you!