Good Explenation Hans. It's 100% in line with what we are measuring at the moment. That article will be published in a couple of weeks, I guess. I will link to this video for reference and explenation of the effects.
Hey, thanks guys! I watched your network switch contest, picked up the SW-8 and a couple Netgear GS108Ev3 switches. And OMG! What a difference these switches made in my setup. I never realized how much crap was in the stream, these switches made a profound difference. So much so my 80 yo mother noticed immediately the better sound and picture quality coming out of the TV and sound bar with bass module. The PQ was immediately noticeable, but what blew me away was the sound coming out of the Polk unit, and then the bass, if never sounded this good. I am now a believer. ;-)
Greetings Hans: Alex from UpTone Audio here. Thank you for kind acknowledgment of our research and product development. And you did a very fine job of explaining and illustrating some of the factors which affect digital transmission! It pretty much boils down to the fact that: Jitter on any input signal (not just Ethernet, this applies to ALL digital transmission!) causes amplitude noise INSIDE the chips and on the ground-plane of the circuit board. This AM (amplitude modulation) gets converted to PM (phase modulation). This jitter/phase-noise (same thing, just expressed in time or frequency domain) propagates--both with interfaces such as Ethernet, USB, S/PDIF, I2S, whatever, and in the chips on the boards. Even a DAC containing a PERFECT clock sitting right next to the DAC chip/circuits input pin is going to be affected by all the upstream jitter/phase-modulation of the ground-voltage-reference that took place in the chips and on the PCB just before it. The above is about as concise as I can be (and it is me paraphrasing the much longer and more nuanced and mind-numbing-fact-filled explanations that John Swenson has spoken to me over many conversations). It is the net of all the reasons anyone hears differences with ANY of this stuff! In addition, common-mode noise--from power supplies--travels on Ethernet cables and sails right through the transformers at every RJ45 port. That type of noise is, just like clocking, one of the many factors introducing jitter/phase-modulation of the ground-voltage-reference. Common-mode leakage/noise happens to be one aspect of what the EtherREGEN's active-differential isolation moat is blocking. And then we reclock with an expensive 10GHz-capable ultra-low-jitter flip-flop (actually we do that on both sides of the moat). As you may know, global chip shortages forced EtherREGEN production to a halt at the beginning of 2022. Very frustrating since demand was consistent and continuing-with more than 3,200 units shipped in its first two years. We have been working hard in development of EtherREGEN Gen2 for almost 1 year-utilizing more readily available parts and incorporating a number of technical advancements. These include Gigabit capability on both the ‘A’ and ‘B’ sides of our unique active-differential-isolation moat, even lower jitter clock synthesizers, and the world’s best sine-to-square wave converter (benefiting those who chose to pair with an external 10MHz reference clock). Development progress on the new model has been very positive the past two months, and we have already pre-purchased over $80,000 worth of circuit chips for the first production run of 250 units. Unfortunately last week we discovered a problem-wherein one key chip (a specialized Ethernet PHY) that we bought and were designing with turns out not to perform an advertised capability, so we have to scrap $6K of parts and await samples of a newly released chip that we know will meet our requirements. Therefore I now expect first shipment of EtherREGEN Gen2 will not take place until May/June this year. The form-factor/casework of the EtherREGEN Gen2 will be the same as the original, and we expect the price to remain under $700 (quite a bargain for the parts and techniques; I get mad seeing other firms charge similar prices for generic designs with just $50 in parts-even if they are in fancier enclosures). Again, I wish to express our gratitude and admiration for your fortitude in consistently striving to educate your audience on these complex--and often controversial--topics. The minutia affecting audible performance are real. John Swenson spent 31 years as a senior engineer at LSI Logic>Avago>Broadcom, and he specialized in designing the power networks inside the large scale chips that go into all this networking gear. By the way, John's recent week running phase-noise plots on his Jackson Labs PhaseStation (same as the $23K Symmetricom/Microsemi 53110A) has inspired him to build a board using some expensive ultra-low-noise, extremely high bandwidth differential amplifier chips he bought a while ago, to allow him to run either HCMOS clocks or Ethernet directly into the 50-Ohm input of the PhaseStation. This set up will allow him to directly measure phase-noise-off both the output of our clock synthesizer(s)--alternately based both on internal Crystek and external 10MHz AND from inline on the RJ45 connection between the EtherREGEN ‘B’ port and a streamer. The latter will be really interesting! We have never seen anyone else run low-offset phase-noise analysis on Ethernet. Jitter measurements and histograms only look at it over extremely wide frequency range-and boil it down to a single, less than useful jitter number. Yet as we have been trying to explain, it is mostly the low-frequency offset phase-noise that is affecting the whole ground-plane of the PCBs and infecting circuits. Of course he will do that Ethernet wire probe test both with internal and external clock. Good times! --Alex C. UpTone Audio LLC
This is a magnificent explainer, Hans! Well done! On the forums there are many blowhards who believe the stage before the DAC is no more than 1s and 0s. Ironically, this is quite a binary position to take. I have experienced huge improvements to my music by upgrading the digital stage through improved cables, regenerators, power supplies and OCXO clocks. The digital signal needs nurturing in the same way a traditional analogue signal does. The main advantage, however, of such digital signals is that they can be cleaned and re-clocked, whereas analogue once it is wrecked stays wrecked.
The 20 minutes of this video were one of the best investments in time (no pun intended) in helping understand the more deeper issues of jitter. Thanks Hans!
Nice work Hans. Perfectly explained and it makes logical, scientific and auditory sense. I’ve been playing with switches recently and there are clear benefits. By the time you get to the likes of an Innuos Phoenixnet switch the differences are hard to believe but clearly audible and match you description you make on high-end switches. Let the 0 and 1’s dissenters gather, but like most binary thinkers they aren’t open-minded enough to reconsider. They listen and don’t hear and they look and don’t see.
Thanks for the video with an outstanding explanation of whats going on in the digital world and whats going on when crossing into or coming from the analog world. It really helps me.
I can tell that this has quite rightly touched a nerve in the past with how dismissive people have been, your explanation is really good and helpful, if people don’t want to listen or hear it then it’s their fault. I am currently using a power line network plug adapter to my streamer; I’m sure this can’t be good? The Ethernet might have noise and maybe I’m putting noise into the mains too? Must sort something else :) Thanks for the great content :)
If you can't use regular CAT cable, at least make sure the power line network plug adaptors are plugged into the wall outlet and not in an extension cord. That does make a difference, although I don't know why.
@@TheHansBeekhuyzenChannel Thank You Hans, will try that. Also going to try plugging into a Mesh Wifi point I have just bought which does have socket on the back.
Hans: thank you for this video. It helps to explain what my ears told me was an easily discernible improvement in sound I’ve been into switches and power supplies for a few years. Finally others are starting to catch up. Ignore the people who haven’t figured it out yet. Who cares if they choose not to improve their systems. The rest of us can follow your advice and benefit from some experimental efforts and then make their decision. I’ve been so impressed with the concept I have 3 switches cascaded. With superior power supplies. You’d have to be deaf not to hear the benefits. Thanks again!😊
@@TheHansBeekhuyzenChannel Really? You recommend switch mode over linear for the digital gear? Do you have an example of one of these switch mode units you would recommend? I know Meanwell and ifi make them, is there something better? Brand? Thanks Hans!
Ik heb niet alles van je bekeken, maar deze bijdrage is buitengewoon goed en heeft extreem veel perfect uitgezochte informatie bij elkaar.. heel bedankt!!
I wish to confirm what Hans is saying in this video. I use some galvanic ethernet isolators that are designed for hospital ethernet. I use them both on the streaming device and also into my d-link switch that is providing ethernet to those streaming devices. When i use a galvanic filter on the device only i have an improvement in sound. When i add another filter to the d-link i have another improvement. The improvements are always more relaxed and natural sound l. Another improvement is adding power filtering to the device, also adding power filtering to the switch help, basically anything you can do to reduce noise Also using a good quality ethernet cable helps, I use Amazon Basic cat 7 they are heavy and appear well shielded. To anyone that doubts these things please try for yourself to verify. Also final tip, i stream with a 100mbps switch. I'm convinced streaming at lower connection speed improves sound as 100mbps should generate less noise the 1gigabit.
This was very interesting. I can't say I fully grasp it. As you allude to at the start, the ultimate measuring stick in all this is audible sound quality (i.e. differences therein). In other words, technical measurements are useful only to explain and perhaps lend credence to known audible differences. If technical measures predict audible superiority but audibly such superiority fails to materialize in a structured listening test then the measurements are effectively measuring the wrong thing. This in turn kicks the whole issue over to the science of listening tests, which is a pretty vexed issue it seems.
Audible judgement requires a lot of training, guided by people that have more experience. It also. requires a lot of equipment for comparison and years of experience. Training to do measurement takes less time but here again the measurement equipment requires large financial sacrifices. And then you need to know what to measure, which currently becomes more clear is not what most audio engineers have done so far.
@@TheHansBeekhuyzenChannel This makes sense. What I am pointing out is a certain circularity in reasoning, however: New technical explanations such as the one this video are brought forward to in a way "prove" audible differences which themselves have not even been proven scientifically to exist. I myself also believe there is an audible difference in most if not all the areas you mentioned here, but my issue is that the measurement science, for lack of a better term, can't prove that the audible differences exist, such science can only explain them (i.e. explain WHY they exist) after they have been proven to exist in the first place through listening tests. Still, there is no harm is seeking an explanation for what seems pretty clearly there in more casual listening testing.
I can only fully agree with your argument. That’s why I depend fat less on measurements than some others. Measurements can prove a device is defective or badly designed, it can’t prove a device sounds excellent.
Dankjewel Hans voor deze video! Bijzonder dat Grimm het voor elkaar gekregen heeft om minder afhankelijk te zijn voor de zaken die bij andere eindpunten of streamers wel degradatie van de geluidskwaliteit geven. Ik gebruik zelf de met glasvezel aangesloten OpticalRendu, hoewel deze standaard al heel mooi weergeeft ervaar ik nog steeds verbetering bij netwerk aanpassingen. Ook het optisch omzetten geeft weer een schakel die beïnvloed wordt door de zaken die je hebt omschreven. Succes met je volgende video en geniet van de muziek!
Dat van Grimm vond ik ook bijzonder om te horen. Ik kijk nu toch weer wat anders naar mijn SA Bonn N8+F1 en een eventuele upgrade.... Blijkbaar kun je beter een Grimm in huis halen dan je geld inzetten op een audiofiele switch (+ een aparte streamer, voeding, Roon Core, upscaler, etc.). Prima video trouwens weer Hans!
TheAlphaAudio recently did some great work about Switches. Their videos and research was brilliant and well worth a watch for anyone interested in this subject.
Hello Hans, I don’t manage to answer this question. If the usb input of my DAC is galvanicly isolated, and it is the case (Holo Cyan 2), and as you say with usb connection there is no problem with time errors from the incoming device, does it mean than nothing could ever improve the sound quality I have form my beefy PC directly connected to the Cyan? I feel like it’s not that simple but it seems so logical!
You are right. It is not that simple. There is the 5 VDC in the USB connection that can (and often will) be polluted. The galvanic separation works for DC in the signal path but not for AC signals and so on. As with every digital interface it is the way it is implemented on both sides that counts.
Hans another very detailed video on Digital Audio. I think you did a phenomenal job of explaining how jitter can cause issues with audio. However you seemed to gloss over the details of how the switch can impact the DAC and digital audio in general. Is it you suggesting that the switch can induce jitter? Or is it that noise from the switch can infiltrate the audio device be translated into the analog output of the device? Also curious if you would suggest that the problem in these cases is with the switch or the audio device? Is it not the job of the audio device to filter and isolate itself from potential noise in source devices? Thanks again for your work.
Any digital device in front of the DAC can pollute the 'digital signal' with noise on that 'square wave signal'. Also the ground plane can be raised and each device can add to the phase noise.
Hi Ben. I agree with your observation that Hans glossed over "how the switch can impact the DAC..." It wasn't until the 16:53 minute mark of the clip that he says that there is a phase oscillator in the switch itself. This hit home to me what probably should have been obvious, all the signal 'polluting' elements of an electronic piece of equipment are of course present in a switch (eg: power supply.) Hence, the square wave, analogue signal has been distorted before it's even arrived at the DAC. So, as Hans also says, any piece of electronic equipment in the transport chain will have an effect on what the DAC has to convert.
Ben Payne has a point. I enjoyed the video, but it did not really detailedly explain the problem with some of the switches. Thanks for talking about this, Hans!
Based on this, it is somewhat sad how people simply go with their emotions and accept incomplete explanations. Like most people in this comment section. I dont disagree that switches can possibly alter the audio quality due to the mentioned factors, but if, for example, the problem is that the data does not arrive in time due to a switch problem, should the player not simply have a large enough buffer to collect the data beforehand?
Thanks Hans, I am learning a lot from you and discovering great products! I am sure you already have the LHY switch on your radar, love to see a review on on that, and the Wattson Madison :)
I currently use four PCIe soundcards through my one slot via a USB splitter, my CD transport is also USB, I'm currently upgrading to a NVMe based splitter and a SATA based CD transport. I'm not currently getting any audible noise, but i'm interested in seeing if I can hear an improvement in clarity.
@@TheHansBeekhuyzenChannel yes, it's my expectation that the higher levels of signal integrity that NMVe and SATA transfer protocols have will give an audible improvement.
Wat een top video. Een verademing om een keer iemand te horen die weet waar hij het over heeft en het ook nog eens zéér goed uitlegt. Klasse! Zo meteen eens even die white paper doornemen. Dank u.
Thank you Hans, you are a pioneer among audio reviewers in attempting to explain the complicated factors relating to networks and audio quality. However this video left me with this question. Would a length of fiber (and the necessary converters) inserted AFTER the network switch and before the D/A converter serve to eliminate all the noise and jitter issues introduced earlier in the chain? I think of this as the "Small Green Computer" solution because they are well known for advocating this as a method of improving the digital signal.
It strips away the noise, but if the signal is already damaged already, it can't wipe away all the issues. I have got significant increases in SQ adding fiber into the mix. You can get a better signal to the FMC by using a LPS on the modem and on a non-wifi router. If your network is already pretty good and your speakers aren't super revealing, you probably won't notice much of a difference though fixing the pre-fiber side of things. At least that's what I have found in my system, but I have 100ft of builders grade CAT 5E in the wall between my router and my 1st fiber switch. That long run of unshielded cable might be why I didn't get much of an improvement. I have read in the forums others have got better improvements than I did in cleaning up the back end.
@@Pete.across.the.street Thanks Pete, another thought along this same line. I have a 3 unit mesh network distributing wifi in my house. This gives a strong wifi signal located right in the listening room and no physical connection to the network. Is this likely to already be yielding the same advantages of cleaning up the digital signal as inserting fiber and making a physical connection to the streamer? The mesh network is TP-Link and the streamer is an older Auralic Aries, the clamb shell model from 2015.
@@brucebarbour9583 I think the newer AURALiC have some high end wifi receiver in them that's supposedly just as good as wired connection. Idk about your model. Experiment and see what you like. I have a Bluesound Node, wifi doesn't sound better to me. I think wifi has its own "problems" when it comes to hifi streaming. Idk the particulars though.
I have tested that on the Auralic Aries G2 to find that Wifi sounds better than a consumer switch but a professional Cisco switch is slightly better. The SOtM and Uptone Audio switches sounded the best. But that might depend on how crowded the Wifi bands are locally.
@@brucebarbour9583 i also have a mesh system, with orbi. Tried also using a sattelite Wi-Fi connected, with Ethernet out from it. But I prefer the wall cable, I have 12 meters in wall, unshielded like you, why is unshielded better than shielded? I am trying the melco now, previously had the ee8 and Paul pang quad, none of them seems to do anything I can pick out in a-b testing. Is so small (if there is any) that it would be impossible to take out in blind test.
When streaming from over the internet I guess there's a chain of switches and other components involved unknown to the end user/listener. Is this a relevant factor of influence when compared to streaming from within one's own home network?
Each component in the digital signal patch can (and most times will) pollute the square waves that carry the digital signal. That can be distortion of the wave form, noise, phase noise and so on. The better those components are, the less they pollute. But they will pollute. Consumer switches and cheap streamers are worse than enterprise class switches.
@@TheHansBeekhuyzenChannel OK. Then it's a factor but probably not a problem because most internet and cable service providers will use enterprise class components for getting the signal to the consumers, and at the consumer side you can still make a difference by using such components as well. Is that correct?
the internet does not work this way most of the internet is connected by fiber and therefore electrically isolated ... only the last mile is copper or coax so variants of FTTC and for the lucky few FTTH independent of WAN technologies, even if you have FTTH, a cheap modem / router / media converter PSU brick can still add electrical noise to your home LAN and even an audiophile switch will not completely get rid of that, even with buffering and forwarding over an egress port this is why changing your home router's PSU can also improve the sound, even with an audiophile switch in between that router and your streamer but changing the PSU on servers used by Tidal and Qobuz (this would imply changing the hardware of their CDN), will have zero impact on the sound at home no matter how noisy the server in the datacenter is, there will be zero influence and all of that data will leave the datacenter bitperfect with virtually no packet loss
@@colanitower "Enterprise Class" switches can usually handle 24 or 48 outputs, can be managed remotely, have special security and firewall software or firmware built-in, and need to be more physically robust because the replacement cost for labor or the impact of customer outage time is very high. Any business that depends on the Internet will be out of business very soon, or loose a lot of customers if their Internet access is down due to a failed switch. A decent quality home network switch works just as well in the home compared to an "enterprise class switch" unless you running a very high-traffic online business from your home and/or have more than 12 ports needed on a single switch. Your home Router/WiFi box supplied by your ISP also has some firewall and security features built-in, but once you get inside your home network, any additional switches don't need that. Many audiophile switches and streaming devices only handle a maximum speed of 100 Mbps, because that is fine for audio or video streaming (including 4K video). Both my Roku TV streaming device and my iFi Zen Stream for music will only connect via Ethernet at a maximum of 100 Mbps, even when connected to a switch or router with 1 Gbps (1000 Mbps) connection speed capability. That prevents one member of the family from hogging the entire bandwidth of the home network, and also the traffic when streaming is 99%+ only in one direction (and not bi-directional out to the Internet).
Would it make sense to use an audiophile switch in my living room to connect most audio equipment, even if my Roon Server (a Mac Mini sitting in another room) is connected to the audiophile switch via another "normal" switch?
@@TheHansBeekhuyzenChannel Thank you very much - one more question: in my understanding, Roon makes certain via RAAT that Roon Ready devices 1) get bit-perfect music data and 2) get full control of the timing in their clock. To my understanding that's one of the reasons why RAAT sounds great. Do you think that even Roon Ready devices would profit from the audiophile switch?
I’ve tried the Uptone Audio EtherRegen and the Sonore Optical and could never hear a difference in sound quality. Not sure why? I use a Musica Pristina A Cappella streamer into a Terminator Plus via I2S.
Hi Hans. As I understand there is no thresshold exactly at half the voltage to distinguish between 1 and 0. For example: when Vmax is 2V, the design such that for only a signal above 1,2V is concidered 0, where all below 0,8V is concidered a digital 1. What is changing to your explanation if you take this into account?
That makes no difference. If in my example the threshold is raised by 0.1 volt, the thresholds in your example would shift to 1.3 and 0.9 volts and thus will the transition time shift.
Taking everything described as true, what does "jitter" sound like? Can you be more specific regarding "noise"? Do you mean hiss, static, pitch changes, clipping and/or no sound at all? I want to know what to listen for when determining whether there is something in the audio chain that I need to improve.
I went down that rabbit hole (how does jitter sound) last year. From my experience jitter is always there. How strongly you hear the effects of jitter in your chain depends on the DAC and music. Effects in my chain are/were: impact of drums / percussion -- smeared and not as precise with more jitter; piano sounds awful with more jitter; generally worse stereo imaging and a tendency towards listening fatigue. Some DACs seem very jitter "agnostic" as for instance the RME ADI DAC. Other (including the Holo Audio Spring and my Denafrips DACs) strongly respond to more / less jitter. I am afraid - unless you try out some options yourself in your system with your music, it is probably quite hard to understand what all this jitter talk is about. BUT when you experienced it, you will understand.
@@medonk12rs Thank you for taking the time to describe your experiences. I have Chord Qutest, Aqua La Voce, and PS Audio DACs among my 3 set-ups. Eventually, I'll sit down and compare all in one system. In the meantime, I'll...enjoy the music.
Very interesting, Hans! Thank you for sharing your knowledge. The quality of components does matter, especially in an ethernet network. Not only do the physical components matter, but also the configuration of network protocols. In my setup I have an Aruba JL680A switch, a modified power supply, and high quality shielded network cables. My Synology filer has a very fast direct network cable connection, and so has my network streamer. I do not play high-res music files from CIFS or Samba shares, but from NFS shares; these NFS packets travel much faster over the network. In Volumio it is possible to configure a NFS connection to a Synology NFS share, and this makes playback of big high-res digital files much better.
Thanks Hans! I'm seeing the Cisco WS-C2960C-8TC-S as well. Would you recommend the older one over this? I don't care about poE, and maybe it adds more noise???
You can use the cisco meraki as well. I removed the poe section in mine. I've used both switches, Cascading 2 of them with fiber gave me great improvements.
Hi Hans, I have a question about the iFi iPower dc power adapter from an older video. Will that kind of power supply eliminate noise if it plugs directly into a RPi power port, or does it have to be plugged into an alternate port on an audio hat board attached to the RPi in order to properly eliminate the noise?
First: no power supply will eliminate all noise. To get the best result from a audio grade power supply it's best o have it hooked up to the audio card directly and have the Raspberry Pi powered by another power supply (lower grade if you like).
@TheHansBeekhuyzenChannel In the early days of CD, I bought a disc and took back as it would not play. Week later bought new player. The issue was old player was a cheap single beam new was a 3 beam . Massive improvement, especially with tracking. .
Thanks Hans. At 12min of the video you explain, that jitter is not relevant on asynchronious transmissions like USB or network. I strongly agree to that! So the summary of your video is, that the noise caused by network switches (and all other connected gear) can negatively impact the sound quality of digital hifi. Isn't it that simple?
Jitter in an async data stream is irrelevant, it's a packet service. It's the phase noise after conversion to I²S and the other problems I mentioned that cause for 'timing errors' ie jitter during conversion
@@TheHansBeekhuyzenChannel is the damage being done in the streaming engine where Ethernet packets are rendered into PCM (I2s or spdif) or later in the chain? Great background info in this video by the way, but it stops short of really explaining what’s going on in a streamer.
Hans, I watched your video about switches and came away with some of the same information you have shown before. But the big question is, "is it audible?" The illustration you showed says that the jitter is happening at 22.1 KHz. Unfortunately for me, my hearing tapers off about 10KHz lower than that frequency. Therefore, I can't hear jitter and I'm not buying some fancy electronics that will supposedly get rid of it. If you can hear it, you must have superb hearing for a senior citizen and I commend you for your remarkably good hearing.
In the entire video I have not said anything that could even lead to the conclusion that jitter is happening at 22.1 kHz. Jitter is extremely audible, but to hear it you must first recognise it at jitter. Especially jitter where the digital signal is modulated with a low frequency, say 10 Hz, is very audible. You the don't hear 10 Hz, you hear the effect of the digital signal modulated at 10 Hz.
You need to speak to my mother, she could not believe what the addition of a $42 network switch did to the PQ and SQ coming out of the TV... And then the stereo... Netgear GS108Ev3, made a huge impact. Try it, it does not work, send it back. I used to think the same...
Good Explenation Hans. It's 100% in line with what we are measuring at the moment. That article will be published in a couple of weeks, I guess. I will link to this video for reference and explenation of the effects.
👍🏻
Hey, thanks guys! I watched your network switch contest, picked up the SW-8 and a couple Netgear GS108Ev3 switches. And OMG! What a difference these switches made in my setup. I never realized how much crap was in the stream, these switches made a profound difference.
So much so my 80 yo mother noticed immediately the better sound and picture quality coming out of the TV and sound bar with bass module. The PQ was immediately noticeable, but what blew me away was the sound coming out of the Polk unit, and then the bass, if never sounded this good.
I am now a believer. ;-)
Greetings Hans:
Alex from UpTone Audio here. Thank you for kind acknowledgment of our research and product development. And you did a very fine job of explaining and illustrating some of the factors which affect digital transmission!
It pretty much boils down to the fact that:
Jitter on any input signal (not just Ethernet, this applies to ALL digital transmission!) causes amplitude noise INSIDE the chips and on the ground-plane of the circuit board. This AM (amplitude modulation) gets converted to PM (phase modulation).
This jitter/phase-noise (same thing, just expressed in time or frequency domain) propagates--both with interfaces such as Ethernet, USB, S/PDIF, I2S, whatever, and in the chips on the boards. Even a DAC containing a PERFECT clock sitting right next to the DAC chip/circuits input pin is going to be affected by all the upstream jitter/phase-modulation of the ground-voltage-reference that took place in the chips and on the PCB just before it.
The above is about as concise as I can be (and it is me paraphrasing the much longer and more nuanced and mind-numbing-fact-filled explanations that John Swenson has spoken to me over many conversations). It is the net of all the reasons anyone hears differences with ANY of this stuff!
In addition, common-mode noise--from power supplies--travels on Ethernet cables and sails right through the transformers at every RJ45 port. That type of noise is, just like clocking, one of the many factors introducing jitter/phase-modulation of the ground-voltage-reference. Common-mode leakage/noise happens to be one aspect of what the EtherREGEN's active-differential isolation moat is blocking. And then we reclock with an expensive 10GHz-capable ultra-low-jitter flip-flop (actually we do that on both sides of the moat).
As you may know, global chip shortages forced EtherREGEN production to a halt at the beginning of 2022. Very frustrating since demand was consistent and continuing-with more than 3,200 units shipped in its first two years.
We have been working hard in development of EtherREGEN Gen2 for almost 1 year-utilizing more readily available parts and incorporating a number of technical advancements. These include Gigabit capability on both the ‘A’ and ‘B’ sides of our unique active-differential-isolation moat, even lower jitter clock synthesizers, and the world’s best sine-to-square wave converter (benefiting those who chose to pair with an external 10MHz reference clock).
Development progress on the new model has been very positive the past two months, and we have already pre-purchased over $80,000 worth of circuit chips for the first production run of 250 units. Unfortunately last week we discovered a problem-wherein one key chip (a specialized Ethernet PHY) that we bought and were designing with turns out not to perform an advertised capability, so we have to scrap $6K of parts and await samples of a newly released chip that we know will meet our requirements.
Therefore I now expect first shipment of EtherREGEN Gen2 will not take place until May/June this year.
The form-factor/casework of the EtherREGEN Gen2 will be the same as the original, and we expect the price to remain under $700 (quite a bargain for the parts and techniques; I get mad seeing other firms charge similar prices for generic designs with just $50 in parts-even if they are in fancier enclosures).
Again, I wish to express our gratitude and admiration for your fortitude in consistently striving to educate your audience on these complex--and often controversial--topics. The minutia affecting audible performance are real. John Swenson spent 31 years as a senior engineer at LSI Logic>Avago>Broadcom, and he specialized in designing the power networks inside the large scale chips that go into all this networking gear.
By the way, John's recent week running phase-noise plots on his Jackson Labs PhaseStation (same as the $23K Symmetricom/Microsemi 53110A) has inspired him to build a board using some expensive ultra-low-noise, extremely high bandwidth differential amplifier chips he bought a while ago, to allow him to run either HCMOS clocks or Ethernet directly into the 50-Ohm input of the PhaseStation. This set up will allow him to directly measure phase-noise-off both the output of our clock synthesizer(s)--alternately based both on internal Crystek and external 10MHz AND from inline on the RJ45 connection between the EtherREGEN ‘B’ port and a streamer. The latter will be really interesting!
We have never seen anyone else run low-offset phase-noise analysis on Ethernet. Jitter measurements and histograms only look at it over extremely wide frequency range-and boil it down to a single, less than useful jitter number. Yet as we have been trying to explain, it is mostly the low-frequency offset phase-noise that is affecting the whole ground-plane of the PCBs and infecting circuits. Of course he will do that Ethernet wire probe test both with internal and external clock.
Good times!
--Alex C.
UpTone Audio LLC
Future customer here, I have seen and heard first hand what a quality network switch can and will do, especially with a linear power supply.
This is a magnificent explainer, Hans! Well done!
On the forums there are many blowhards who believe the stage before the DAC is no more than 1s and 0s. Ironically, this is quite a binary position to take. I have experienced huge improvements to my music by upgrading the digital stage through improved cables, regenerators, power supplies and OCXO clocks. The digital signal needs nurturing in the same way a traditional analogue signal does. The main advantage, however, of such digital signals is that they can be cleaned and re-clocked, whereas analogue once it is wrecked stays wrecked.
true
Hans re-entering HIFI hobby and having a technical background you are the most concise and accurate on the UA-cam for that I thank you
My pleasure entirely
Thank you for the English subtitles. Very helpful,
👍🏼
The 20 minutes of this video were one of the best investments in time (no pun intended) in helping understand the more deeper issues of jitter. Thanks Hans!
🙏🏻
Congratlation and many thank's for the explanation!
My pleasure
Nice work Hans. Perfectly explained and it makes logical, scientific and auditory sense. I’ve been playing with switches recently and there are clear benefits. By the time you get to the likes of an Innuos Phoenixnet switch the differences are hard to believe but clearly audible and match you description you make on high-end switches. Let the 0 and 1’s dissenters gather, but like most binary thinkers they aren’t open-minded enough to reconsider. They listen and don’t hear and they look and don’t see.
👍🏼
Excellent video Hans! I learned a lot about the topic from watching this.
Great to hear!
Thanks for the video with an outstanding explanation of whats going on in the digital world and whats going on when crossing into or coming from the analog world. It really helps me.
My pleasure entirely.
I can tell that this has quite rightly touched a nerve in the past with how dismissive people have been, your explanation is really good and helpful, if people don’t want to listen or hear it then it’s their fault. I am currently using a power line network plug adapter to my streamer; I’m sure this can’t be good? The Ethernet might have noise and maybe I’m putting noise into the mains too? Must sort something else :) Thanks for the great content :)
If you can't use regular CAT cable, at least make sure the power line network plug adaptors are plugged into the wall outlet and not in an extension cord. That does make a difference, although I don't know why.
@@TheHansBeekhuyzenChannel Thank You Hans, will try that. Also going to try plugging into a Mesh Wifi point I have just bought which does have socket on the back.
👌👍 Très bon article , bravo...
Mercy messieur
Hans: thank you for this video. It helps to explain what my ears told me was an easily discernible improvement in sound
I’ve been into switches and power supplies for a few years. Finally others are starting to catch up.
Ignore the people who haven’t figured it out yet. Who cares if they choose not to improve their systems. The rest of us can follow your advice and benefit from some experimental efforts and then make their decision.
I’ve been so impressed with the concept I have 3 switches cascaded. With superior power supplies. You’d have to be deaf not to hear the benefits.
Thanks again!😊
I hope the power supplies are quality switch mode power supplies.... Linear ones work less good on switches (in fact on most digital-only equipment.)
@@TheHansBeekhuyzenChannel Really? You recommend switch mode over linear for the digital gear? Do you have an example of one of these switch mode units you would recommend? I know Meanwell and ifi make them, is there something better? Brand? Thanks Hans!
Ik heb niet alles van je bekeken, maar deze bijdrage is buitengewoon goed en heeft extreem veel perfect uitgezochte informatie bij elkaar.. heel bedankt!!
👍🏻🙏🏽
Excellent, thanks
You are welcome!
I wish to confirm what Hans is saying in this video. I use some galvanic ethernet isolators that are designed for hospital ethernet. I use them both on the streaming device and also into my d-link switch that is providing ethernet to those streaming devices. When i use a galvanic filter on the device only i have an improvement in sound. When i add another filter to the d-link i have another improvement. The improvements are always more relaxed and natural sound l. Another improvement is adding power filtering to the device, also adding power filtering to the switch help, basically anything you can do to reduce noise Also using a good quality ethernet cable helps, I use Amazon Basic cat 7 they are heavy and appear well shielded. To anyone that doubts these things please try for yourself to verify. Also final tip, i stream with a 100mbps switch. I'm convinced streaming at lower connection speed improves sound as 100mbps should generate less noise the 1gigabit.
Thanks for sharing
This was very interesting. I can't say I fully grasp it. As you allude to at the start, the ultimate measuring stick in all this is audible sound quality (i.e. differences therein). In other words, technical measurements are useful only to explain and perhaps lend credence to known audible differences. If technical measures predict audible superiority but audibly such superiority fails to materialize in a structured listening test then the measurements are effectively measuring the wrong thing. This in turn kicks the whole issue over to the science of listening tests, which is a pretty vexed issue it seems.
Audible judgement requires a lot of training, guided by people that have more experience. It also. requires a lot of equipment for comparison and years of experience. Training to do measurement takes less time but here again the measurement equipment requires large financial sacrifices. And then you need to know what to measure, which currently becomes more clear is not what most audio engineers have done so far.
@@TheHansBeekhuyzenChannel This makes sense. What I am pointing out is a certain circularity in reasoning, however: New technical explanations such as the one this video are brought forward to in a way "prove" audible differences which themselves have not even been proven scientifically to exist. I myself also believe there is an audible difference in most if not all the areas you mentioned here, but my issue is that the measurement science, for lack of a better term, can't prove that the audible differences exist, such science can only explain them (i.e. explain WHY they exist) after they have been proven to exist in the first place through listening tests. Still, there is no harm is seeking an explanation for what seems pretty clearly there in more casual listening testing.
I can only fully agree with your argument. That’s why I depend fat less on measurements than some others. Measurements can prove a device is defective or badly designed, it can’t prove a device sounds excellent.
Thank you for your videos. Great stuff and I learn a lot!
@@TheHansBeekhuyzenChannel Thank you. Well stated!
Dankjewel Hans voor deze video! Bijzonder dat Grimm het voor elkaar gekregen heeft om minder afhankelijk te zijn voor de zaken die bij andere eindpunten of streamers wel degradatie van de geluidskwaliteit geven. Ik gebruik zelf de met glasvezel aangesloten OpticalRendu, hoewel deze standaard al heel mooi weergeeft ervaar ik nog steeds verbetering bij netwerk aanpassingen. Ook het optisch omzetten geeft weer een schakel die beïnvloed wordt door de zaken die je hebt omschreven. Succes met je volgende video en geniet van de muziek!
👍🏼
Dat van Grimm vond ik ook bijzonder om te horen. Ik kijk nu toch weer wat anders naar mijn SA Bonn N8+F1 en een eventuele upgrade.... Blijkbaar kun je beter een Grimm in huis halen dan je geld inzetten op een audiofiele switch (+ een aparte streamer, voeding, Roon Core, upscaler, etc.). Prima video trouwens weer Hans!
Terrific video! Thank you for taking the time to research this topic and break it down to clear some of the mystery surrounding it.
Glad you enjoyed it!
TheAlphaAudio recently did some great work about Switches. Their videos and research was brilliant and well worth a watch for anyone interested in this subject.
I second that.
Great video! Many Thanks!!
🙏🏽
Thank you sir 🎉
Most welcome
Rare insight communicated extremely well. There are few people who can do this
Fabulous video once again. It answered a lot of questions for me. Thank you.
My pleasure!
Hello Hans, I don’t manage to answer this question. If the usb input of my DAC is galvanicly isolated, and it is the case (Holo Cyan 2), and as you say with usb connection there is no problem with time errors from the incoming device, does it mean than nothing could ever improve the sound quality I have form my beefy PC directly connected to the Cyan? I feel like it’s not that simple but it seems so logical!
You are right. It is not that simple. There is the 5 VDC in the USB connection that can (and often will) be polluted. The galvanic separation works for DC in the signal path but not for AC signals and so on. As with every digital interface it is the way it is implemented on both sides that counts.
@@TheHansBeekhuyzenChannel Thank you very much
Hans another very detailed video on Digital Audio. I think you did a phenomenal job of explaining how jitter can cause issues with audio. However you seemed to gloss over the details of how the switch can impact the DAC and digital audio in general. Is it you suggesting that the switch can induce jitter? Or is it that noise from the switch can infiltrate the audio device be translated into the analog output of the device? Also curious if you would suggest that the problem in these cases is with the switch or the audio device? Is it not the job of the audio device to filter and isolate itself from potential noise in source devices? Thanks again for your work.
Any digital device in front of the DAC can pollute the 'digital signal' with noise on that 'square wave signal'. Also the ground plane can be raised and each device can add to the phase noise.
Hi Ben. I agree with your observation that Hans glossed over "how the switch can impact the DAC..." It wasn't until the 16:53 minute mark of the clip that he says that there is a phase oscillator in the switch itself. This hit home to me what probably should have been obvious, all the signal 'polluting' elements of an electronic piece of equipment are of course present in a switch (eg: power supply.) Hence, the square wave, analogue signal has been distorted before it's even arrived at the DAC. So, as Hans also says, any piece of electronic equipment in the transport chain will have an effect on what the DAC has to convert.
Ben Payne has a point. I enjoyed the video, but it did not really detailedly explain the problem with some of the switches.
Thanks for talking about this, Hans!
Based on this, it is somewhat sad how people simply go with their emotions and accept incomplete explanations. Like most people in this comment section.
I dont disagree that switches can possibly alter the audio quality due to the mentioned factors, but if, for example, the problem is that the data does not arrive in time due to a switch problem, should the player not simply have a large enough buffer to collect the data beforehand?
Complexe mais passionnant. merci
Avec plaisir
Ignore the free range rude Hans.... They are everywhere. The rest of us appreciate you.
I know, thanks.
Thank you once again for your clear explanation Hans.
You are very welcome
Thanks Hans, I am learning a lot from you and discovering great products! I am sure you already have the LHY switch on your radar, love to see a review on on that, and the Wattson Madison :)
Not yet!
I currently use four PCIe soundcards through my one slot via a USB splitter, my CD transport is also USB, I'm currently upgrading to a NVMe based splitter and a SATA based CD transport. I'm not currently getting any audible noise, but i'm interested in seeing if I can hear an improvement in clarity.
It’s not audible noise that is the problem. It’s noise that is in the bitstream that confuses the D to A conversation.
@@TheHansBeekhuyzenChannel yes, it's my expectation that the higher levels of signal integrity that NMVe and SATA transfer protocols have will give an audible improvement.
Wat een top video. Een verademing om een keer iemand te horen die weet waar hij het over heeft en het ook nog eens zéér goed uitlegt. Klasse! Zo meteen eens even die white paper doornemen. Dank u.
👍🏼
Dit had je al een keer eerder goed uitgelegd vond ik. Maar toch weer interessant
de kracht zit in de herhaling. En deze keer vond ik het nog beter gelukt
Great video again! Thank you for that!
Glad you enjoyed it!
Thank you Hans, you are a pioneer among audio reviewers in attempting to explain the complicated factors relating to networks and audio quality. However this video left me with this question. Would a length of fiber (and the necessary converters) inserted AFTER the network switch and before the D/A converter serve to eliminate all the noise and jitter issues introduced earlier in the chain? I think of this as the "Small Green Computer" solution because they are well known for advocating this as a method of improving the digital signal.
It strips away the noise, but if the signal is already damaged already, it can't wipe away all the issues. I have got significant increases in SQ adding fiber into the mix. You can get a better signal to the FMC by using a LPS on the modem and on a non-wifi router. If your network is already pretty good and your speakers aren't super revealing, you probably won't notice much of a difference though fixing the pre-fiber side of things. At least that's what I have found in my system, but I have 100ft of builders grade CAT 5E in the wall between my router and my 1st fiber switch. That long run of unshielded cable might be why I didn't get much of an improvement. I have read in the forums others have got better improvements than I did in cleaning up the back end.
@@Pete.across.the.street Thanks Pete, another thought along this same line. I have a 3 unit mesh network distributing wifi in my house. This gives a strong wifi signal located right in the listening room and no physical connection to the network. Is this likely to already be yielding the same advantages of cleaning up the digital signal as inserting fiber and making a physical connection to the streamer? The mesh network is TP-Link and the streamer is an older Auralic Aries, the clamb shell model from 2015.
@@brucebarbour9583 I think the newer AURALiC have some high end wifi receiver in them that's supposedly just as good as wired connection. Idk about your model. Experiment and see what you like. I have a Bluesound Node, wifi doesn't sound better to me. I think wifi has its own "problems" when it comes to hifi streaming. Idk the particulars though.
I have tested that on the Auralic Aries G2 to find that Wifi sounds better than a consumer switch but a professional Cisco switch is slightly better. The SOtM and Uptone Audio switches sounded the best. But that might depend on how crowded the Wifi bands are locally.
@@brucebarbour9583 i also have a mesh system, with orbi. Tried also using a sattelite Wi-Fi connected, with Ethernet out from it. But I prefer the wall cable, I have 12 meters in wall, unshielded like you, why is unshielded better than shielded? I am trying the melco now, previously had the ee8 and Paul pang quad, none of them seems to do anything I can pick out in a-b testing. Is so small (if there is any) that it would be impossible to take out in blind test.
When streaming from over the internet I guess there's a chain of switches and other components involved unknown to the end user/listener. Is this a relevant factor of influence when compared to streaming from within one's own home network?
Each component in the digital signal patch can (and most times will) pollute the square waves that carry the digital signal. That can be distortion of the wave form, noise, phase noise and so on. The better those components are, the less they pollute. But they will pollute. Consumer switches and cheap streamers are worse than enterprise class switches.
@@TheHansBeekhuyzenChannel OK. Then it's a factor but probably not a problem because most internet and cable service providers will use enterprise class components for getting the signal to the consumers, and at the consumer side you can still make a difference by using such components as well. Is that correct?
the internet does not work this way
most of the internet is connected by fiber and therefore electrically isolated ... only the last mile is copper or coax so variants of FTTC and for the lucky few FTTH
independent of WAN technologies, even if you have FTTH, a cheap modem / router / media converter PSU brick can still add electrical noise to your home LAN and even an audiophile switch will not completely get rid of that, even with buffering and forwarding over an egress port
this is why changing your home router's PSU can also improve the sound, even with an audiophile switch in between that router and your streamer
but changing the PSU on servers used by Tidal and Qobuz (this would imply changing the hardware of their CDN), will have zero impact on the sound at home
no matter how noisy the server in the datacenter is, there will be zero influence and all of that data will leave the datacenter bitperfect with virtually no packet loss
@@colanitower "Enterprise Class" switches can usually handle 24 or 48 outputs, can be managed remotely, have special security and firewall software or firmware built-in, and need to be more physically robust because the replacement cost for labor or the impact of customer outage time is very high. Any business that depends on the Internet will be out of business very soon, or loose a lot of customers if their Internet access is down due to a failed switch. A decent quality home network switch works just as well in the home compared to an "enterprise class switch" unless you running a very high-traffic online business from your home and/or have more than 12 ports needed on a single switch.
Your home Router/WiFi box supplied by your ISP also has some firewall and security features built-in, but once you get inside your home network, any additional switches don't need that.
Many audiophile switches and streaming devices only handle a maximum speed of 100 Mbps, because that is fine for audio or video streaming (including 4K video). Both my Roku TV streaming device and my iFi Zen Stream for music will only connect via Ethernet at a maximum of 100 Mbps, even when connected to a switch or router with 1 Gbps (1000 Mbps) connection speed capability. That prevents one member of the family from hogging the entire bandwidth of the home network, and also the traffic when streaming is 99%+ only in one direction (and not bi-directional out to the Internet).
@@Bonabuster Thanks, that explains a lot
Would it make sense to use an audiophile switch in my living room to connect most audio equipment, even if my Roon Server (a Mac Mini sitting in another room) is connected to the audiophile switch via another "normal" switch?
Absolutely
@@TheHansBeekhuyzenChannel Thank you very much - one more question: in my understanding, Roon makes certain via RAAT that Roon Ready devices 1) get bit-perfect music data and 2) get full control of the timing in their clock. To my understanding that's one of the reasons why RAAT sounds great. Do you think that even Roon Ready devices would profit from the audiophile switch?
I’ve tried the Uptone Audio EtherRegen and the Sonore Optical and could never hear a difference in sound quality. Not sure why? I use a Musica Pristina A Cappella streamer into a Terminator Plus via I2S.
I don't comment on equipment I have not reviewed. Enjoy the music.
How about put the music on a hard drive and plug it into your player eg (Raspberry)? just asking
If you use a Pi, Think the noise of the drive will not matter.
Hi Hans. As I understand there is no thresshold exactly at half the voltage to distinguish between 1 and 0. For example: when Vmax is 2V, the design such that for only a signal above 1,2V is concidered 0, where all below 0,8V is concidered a digital 1. What is changing to your explanation if you take this into account?
That makes no difference. If in my example the threshold is raised by 0.1 volt, the thresholds in your example would shift to 1.3 and 0.9 volts and thus will the transition time shift.
There is A REASON why there are 2 different thresholds...
How immune or not was the streamer you used before you put the Grimm Audio in your best system - the Aries G2?
The Aries G2 was better than the SOtM sMS-200 Ultra Neo but the Grimm was clearly better than the Aries.
Taking everything described as true, what does "jitter" sound like? Can you be more specific regarding "noise"? Do you mean hiss, static, pitch changes, clipping and/or no sound at all? I want to know what to listen for when determining whether there is something in the audio chain that I need to improve.
I went down that rabbit hole (how does jitter sound) last year.
From my experience jitter is always there. How strongly you hear the effects of jitter in your chain depends on the DAC and music.
Effects in my chain are/were: impact of drums / percussion -- smeared and not as precise with more jitter; piano sounds awful with more jitter; generally worse stereo imaging and a tendency towards listening fatigue.
Some DACs seem very jitter "agnostic" as for instance the RME ADI DAC. Other (including the Holo Audio Spring and my Denafrips DACs) strongly respond to more / less jitter.
I am afraid - unless you try out some options yourself in your system with your music, it is probably quite hard to understand what all this jitter talk is about.
BUT when you experienced it, you will understand.
@@medonk12rs Thank you for taking the time to describe your experiences. I have Chord Qutest, Aqua La Voce, and PS Audio DACs among my 3 set-ups. Eventually, I'll sit down and compare all in one system. In the meantime, I'll...enjoy the music.
Very interesting, Hans!
Thank you for sharing your knowledge.
The quality of components does matter, especially in an ethernet network.
Not only do the physical components matter, but also the configuration of network protocols.
In my setup I have an Aruba JL680A switch, a modified power supply, and high quality shielded network cables.
My Synology filer has a very fast direct network cable connection, and so has my network streamer.
I do not play high-res music files from CIFS or Samba shares, but from NFS shares; these NFS packets travel much faster over the network.
In Volumio it is possible to configure a NFS connection to a Synology NFS share, and this makes playback of big high-res digital files much better.
What if I just add a “network to fiber module” to my existing switch and connect my MicroRendu to it?
That surely can make a difference if the SFP's have good clocks.
@@TheHansBeekhuyzenChannel
How can I tell if it has good clocks?
Trial and error I am afraid
Linn Audio found this ramp-up voltage issue causing jitter over a decade ago... and they too have addressed it in the DAC on their high-end offerings.
Thanks for sharing
Thanks
Thank you sir.
Thanks Hans! I'm seeing the Cisco WS-C2960C-8TC-S as well. Would you recommend the older one over this? I don't care about poE, and maybe it adds more noise???
You can use the cisco meraki as well. I removed the poe section in mine. I've used both switches, Cascading 2 of them with fiber gave me great improvements.
@@Pete.across.the.street It looks like you believe to poE section adds noise. Easy to remove?
@@d.s.cullom5461 very easy. I switched mine to linear power supply also. Search meraki 220 tuning. You can do the same things to the 2960 as well
thanks for the superb explaining!
No worries!
Hi Hans, I have a question about the iFi iPower dc power adapter from an older video. Will that kind of power supply eliminate noise if it plugs directly into a RPi power port, or does it have to be plugged into an alternate port on an audio hat board attached to the RPi in order to properly eliminate the noise?
First: no power supply will eliminate all noise. To get the best result from a audio grade power supply it's best o have it hooked up to the audio card directly and have the Raspberry Pi powered by another power supply (lower grade if you like).
Problems are also introduced with recording. Sadly nothing can be done about this on playback.
I think that the major part of the music I play is recorded properly. But on a poor stereo in poor acoustics the quality in it can't be retrieved.
@TheHansBeekhuyzenChannel In the early days of CD, I bought a disc and took back as it would not play. Week later bought new player. The issue was old player was a cheap single beam new was a 3 beam . Massive improvement, especially with tracking. .
great video - thanks!
- how about routers, do these also harm the signal?
Yes. I got a wired only business class router and I power it and my modem by a lps
Anything in front of the DAC does.
Great refresher Hans
🙏🏻
DAC PLL constitutes LPF of jitter of incoming data stream. The characteristics of this LPF are critical to audio quality.
That is a more technical explanation of what I said😁
Thanks Hans. At 12min of the video you explain, that jitter is not relevant on asynchronious transmissions like USB or network. I strongly agree to that! So the summary of your video is, that the noise caused by network switches (and all other connected gear) can negatively impact the sound quality of digital hifi. Isn't it that simple?
Jitter in an async data stream is irrelevant, it's a packet service. It's the phase noise after conversion to I²S and the other problems I mentioned that cause for 'timing errors' ie jitter during conversion
@@TheHansBeekhuyzenChannel is the damage being done in the streaming engine where Ethernet packets are rendered into PCM (I2s or spdif) or later in the chain? Great background info in this video by the way, but it stops short of really explaining what’s going on in a streamer.
Hans, I watched your video about switches and came away with some of the same information you have shown before. But the big question is, "is it audible?" The illustration you showed says that the jitter is happening at 22.1 KHz. Unfortunately for me, my hearing tapers off about 10KHz lower than that frequency. Therefore, I can't hear jitter and I'm not buying some fancy electronics that will supposedly get rid of it. If you can hear it, you must have superb hearing for a senior citizen and I commend you for your remarkably good hearing.
In the entire video I have not said anything that could even lead to the conclusion that jitter is happening at 22.1 kHz. Jitter is extremely audible, but to hear it you must first recognise it at jitter. Especially jitter where the digital signal is modulated with a low frequency, say 10 Hz, is very audible. You the don't hear 10 Hz, you hear the effect of the digital signal modulated at 10 Hz.
Great video. However on a 2Volt digital signal isn't the threshold 1Volt not 0.5Volt?
Correct.
Thank you 😊
Welcome!
Fundamental info known since a long time but always good to repeat, nothing strange. 😊
Very true! But very few know about it.
Bedankt
what jitter correction product would you suggest for a windows 7 based computer output?
Use a network bridge.
thank you
Welcome!
Cheers to another owner of the mighty MU1!
Congrats
How about class d amplification in general? Can we all agree that pulse width modulated sound waves are just dreadful?
No we can't. Listen to a Purify amp....
XTZ Edge A2-400 owner here, with the right preamp mine sounds damn good. I have A/B units too.
Hi Hans. I love your videoes, but to me these switches are pure snakeoil.
Then you haven't tried it on a decent stereo. But be happy for it saves you money.
You need to speak to my mother, she could not believe what the addition of a $42 network switch did to the PQ and SQ coming out of the TV... And then the stereo... Netgear GS108Ev3, made a huge impact. Try it, it does not work, send it back. I used to think the same...
links?
Tap 'More' below this video in UA-cam