I have seen DACs in system drop from >100db SNR to <80db SNR and traced it back to the AC\DC supply in both desktops and laptops. I have seen SNR on USB powered lab ADCs go from <60db to >80db just by changing from wall powered to battery. There is a very definite path for switching noise through EMI capacitors on the computer though the DAC into the pre-amp/amp more so with single ended audio connections and 3-wire equipment connections.
Anybody want a laugh?
189 responses Add your response
My counter dougeyjones to the USB, is that the USB is competently designed and properly isolates the USB from the output including the potential to pass on EMI from the computer through the ground connection on the output to subsequent equipment. That even happens in lab equipment, let alone audio equipment. The higher |
In the 21st century audio engineers build equipment while actively avoiding two of the most powerful tools available to the whole of science and engineering: measurement and error control. The damage to the audio industry and its reputation in the wider engineering world will remain immeasurable until we decide to take control. 💜 Bruno His amplifier designs and the THX AAA circuit topology are really the only unique things happening in the HiFi world worthy of note recently, imo of course. I also feel the need to push back aggressively on the preposterous assertion by a few people in this thread that “streamers” are somehow better than a PC based front end, with zero evidence of any kind being offered. The way I see this development is simple, I’ve been using a PC front end for 20-years, it was simply too difficult for the median age audiophile to assemble their own computer, configure the software, download the FLAC files and on top of that, there was no opportunity for you guys to get out your credit card and spend thousands on a shiny new box and get that dopamine hit. Some of you might take offense, but I’m sorry, streamers aren’t doing anything new, and they certainly don’t sound better in any demonstrable way than a middling PC setup connected to a good asynchronous DAC. I use a USB DAC with 116db SINAD specifically to ensure that my front end is outpacing the amplification it feeds in terms of noise floor. I know measuring things is frowned upon here, but I generally roll my eyes hard enough to do a mental backflip when people say things like Ethernet or WiFi is superior to async USB.. Based on what? |
Yes i understand that.... But for me to increase the positive change i must know where and how and what i have done... At the end the change is so great and big, i will take the blind test.... It will be easy to win , but in my house, with my file, otherwise how can i? To implement the test it will takes a day just to erase all the tweaks... Another day to make them in place anew... Oufff All that for what? Blindtesting a cable is easy .... Blind testing a system less easy.... My point is precidely that: blindtesting a stream of successive incremental changes has no interest at all....The end result is not equal to a change in cable.... :) It seems it is you that dont get my point without being rude.... Audio does not reduce to what you think it must be .... |
I can only assume you don't understand blind testing mahgister. The only requirement for blind testing is that you don't know what you are testing. That is it. That is all. You can take second, minutes, hours, days, months to review or make change, the only requirement is you don't know what was changed. |
Blind testing cannot even " perfectly" prove anything.... But supposing it can, it will disprove a particular change and only one... The art of audio embeddings imply a successive incremental chain of positive changes.... For that no blind testing is possible except at the end of the cumulative process of changes....I can perfectly remember how awful was the sound of my audio system before and after these 2 years of changes and experiments...The same audio system is day and night....No blind test for me :) Audiophile must focus on the basic deep problem of audio implementation and improvement not on the defense of singular superficial modification ( like a cable addition) that can be real or not and contest by any blind tester .... The only way to protect ourselves is experimenting with low cost materials.... This is sufficient to win Hi-Fi anyway if the basic amplifier, dac, speakers are relatively good and well choosen.... |
rixthetrick, 'And do you suppose that it cannot possibly be improved upon, in any way by anybody?' No one is going to argue with that. Specifications and tolerances can always be improved. However, transmitting audio signals over Wi-Fi is not considered particularly challenging nowadays, is it? It wasn't too long ago that wired connections were considered advisable for 4K etc but now most routers can manage it wirelessly. The demands required for streaming high definition audio should easily be well within the wireless capabilities of any router anywhere, shouldn't they? Therefore shouldn't we be primarily concerned whether these improvements are of any actual sonic use to us, or are they merely just another marketing ploy to add another zero or 2 (or maybe even 3) to the price demanded? That's the problem here, isn't it? The perennial problem facing all audiophiles; the fact that anyone, anywhere can sell anything regardless of whether it has any discernible effect for any price they choose. All they seemingly have to do is to imply some quasi-nebulous sonic improvement, and they have a mandate. As many of us may have already fallen for this tactic previously, what does this say about us a group? Are we really so gullible and so easily led? If so, then how should we protect ourselves against such devious attempts to hoodwink us? The one thing that might help us would be the demand for more double blind testing, but that is usually met with much hostility and resistance by almost all sides. Alan Shaw once offered a free pair of Harbeth's top of the range M40 loudspeakers to anyone who was able to come down and successfully identify a sonic difference between 2 level matched amplifiers under such conditions at his factory. Guess what happened? No one took him up on his offer. Not even after years and years of debating and arguing the point! Even more recently Gene Della Salla of Audioholics fame got into a Facebook spat with his sometimes sparring partner, none other than Michael Fremer, over the worth of power conditioners (see Audioholics Community). Needless to say Fremer's responses so far have been far from convincing, merely reactionary and defensive. In previous times Michael was not always adverse to double blind testing. Maybe times have changed and once more the scam continues as the cartel continues to protect itself. Fair enough, they all want to keep their jobs, but how do we audiophiles protect ourselves? |
Which is why I said before the amp. If the signal over noise and distortion on the analog outputs of the DAC is below human hearing then it’s transparent and all those streamers, switches and routers and whatever upstream doesn’t amount to diddley. Now after the signal hits the speaker and how it interacts with the room is a different story.Even upstream to the dac all the electrical particular grid of the house, and all that is connected to it will color the sound coming from the dac...The dac itself being a mass of electronic components, so good it is or pricey, will add noise but if rightly design will do it in a good trade-off ... Any small new electronic components, the smaller one also, will intoduce noise of their own.... The audio engineering design is mainly the art of trade-off ...Nothing more... ALL IS INTERCONNECTED in a non linear way and affect human hearing in way perceptible not only in DB measured way... Imaging for instance is not measured in DB....The instrumental perception of timbre is not reducible to DB measures.... All this is only my opinion and experience, i am a poet perhaps certainly not an engineer.... :) But i work hard to give to me the Hi-Fi experience at peanuts cost.... I succeed....I succeed with non orthodox homemade improvised systematic experiments and low cost materials, not by buying costly effective products, nor by "buying snake oil".... The title of my "virtual system page" is: audiophile experience for the poor.....And it is possible.... I did it..... :) A relatively high audio quality experience is possible for all and cost peanuts....The necessary high price to pay for audio S.Q. is partly a " myth".... |
I think that when spending beaucoup bucks on a hifi, people figure a couple hundred extra on a router with noise management as a design focus, regardless of actual merit, is not a big deal. An insurance policy at worst. Again, there are some I know with very well thought out and resolving setups, who I do not suspect of imagining things, who claim the device covered in the white paper did make a difference. So I have to give them the benefit of the doubt. To me, if there is noise on a circuit, all devices connected to the circuit MIGHT be affected. Personally, I’ve used wireless network connections to my streamers exclusively now for many years with excellent results so I am more than pleased with that approach. Always dead quiet with top notch detail, if you have the right gear to show it. |
There is no such thing that a sound without "color"Which is why I said before the amp. If the signal over noise and distortion on the analog outputs of the DAC is below human hearing then it’s transparent and all those streamers, switches and routers and whatever upstream doesn’t amount to diddley. Now after the signal hits the speaker and how it interacts with the room is a different story. |
There is no such thing : a sound without "color"....An uncolored sound means nothing in an absolute sense... All sound coming from a room, a particular house, a specific electrical grid, a particular audio system is colored in some way for some particular ears that listen to these same" colors" in their particular way and will perceive them in their own way... Tweaks by themselves are only that: tweaks useful or not.... But controlling the mechanical, electrical and acoustical embeddings amount to more than using a tweak or 2....It need to be take care of in a systematic experimental way.....It is possible, cheap, and audible by me , my dog, and even my wife (perhaps not my wife).... :) |
Don't think so. Once something is transparent it can't get more transparent. I could add things to color in different ways but not me . Streaming over ethernet isn't rocket science and as long as you have good basic components and your noise floor before the amp is -118 no need for these tweaks that can't be heard except by my dog. |
i begins to hears in an indirect way the noise floor of my audio system when i begins to controls the electrical grid of my house and the noise floor of my audio system.... Not before.... :) A hum or a hiss are only particular noise not giving direct experience of what is the noise floor of the house where is embed the noise floor of the audio system.... Harsh female vocals are explained more by defects in the audio system than by the noise floor level of the house or of the audio system by themselves.... |
@djones - well it’s a good thing that well healed audiophiles all know the fact - "measurements mean absolutely nothing", right? Well I’m into computer audio, and I’ve been following the advice of the "people" in the forums with excellent results - just saying. And as it’s known snake oil right here in Agon, then we should just let these poor unsuspecting people enjoy the music and you’ll stick to what you already have. |
I stream all my music from either services or ripped files. I have a basic Netgear modem and router. I network a NUC running roon in my office to a raspberry pi4 running Ropiee with RAAT roon's transport protocol on my network. I then connect the pi4 to my integrated amp with the USB cable that came with it, the cheap black one. I use basic canare 4s11 speaker wire. I have wall warts connected to my panamax surge protector on different banks with my integrated. This is plugged into a basic industrial grade receptacle on a regular 15 amp circuit. Pretty much everything you're advised not to do. Yet I have a noise floor so low I never hear any, the quiet parts of a song are dead black quiet, no harsh female vocals.no hum or hiss. Excellent timbre and separation of instruments. At least as good as my speakers allow. How is this possible? |
i prefer my homemade "golden plates" on my router or on the ethernet connection.... Cost: peanuts, results; success.... :) In my experience ANYTHING , ANYWHERE, connected in the electrical grid of the house matters to the resulting audible S.Q.... It is what i call the electrical embeddings of an audio system.... |
It's been built for network streaming, which reportedly has better sound quality than USB, HDMI, coax and optic fiber. The processing is done on the server computer, the data stream is sent on to the streaming device (which just does the hand off, with very little jitter) and then to the DAC all via ethernet. Fidelizer Pro is a program that is used on my personal fanless computer and it makes a significant bump in performance. Jplay FEMTO is the server software this device is really designed to work with, along with audiophile grade ethernet cables. As all three devices are going through the router, the same ideology that is used successfully in many digital systems is to get as little jitter, noise and the timing bang on. The hundreds of reviews on European audiophile forums can't all be wrong?? I personally also use JCat FEMTO clocked USB card, based off a NEC USB card that has been highly modified, which is US$500. I haven't gone to network streaming yet, as I'd need to purchase a FEMTO network card, and ethernet DAC and this exact router. If reviews have any validity will render a superior performance in my system! |
As a counter point djones51, someone proved in a blind test with a significant group of audiophiles, that audiophiles couldn't tell when 2.5% distortion was added to their own system with their typical "listening" methods. On the other hand, in a controlled AB situation, engineers could pick up the difference quickly. It may be best to replace the term golden ears with gilded ears :-) |
This is about amplifiers but it's applicable to everything in audio. It's from Bruno Putzey paper 'The F-word or why there is no such thing as to much feedback" whether you agree doesn't really matter to the quote. The avoidance of feedback, specifically global feedback, also meant that longer signal chains quickly accumulated distortion products. A relentless drive for minimalist design ensued. If everything one adds to the signal path detracts from the result, only the smallest number of components will do. |
CD318, the best is this for another product. Basically they figured out if you will fall for the marketing blurb of one, you will fall for connecting multiple in series and that you won’t question why the first one did not do the job. Cumulative process – there is an increased level of Waversa Audio Processing when multiple WNDR capable units are connected, improving sound quality On another note, and Ethernet connection is galvanically isolated. There is the potential for analog noise to be conducted over an Ethernet configuration and in a poorly designed DAC, I could foresee situations where it could cause an issue. Of course you can also transmit 100 megabits per second across a fully isolated barrier for <$10 including support circuitry. |
"Noise suppressing capacitors" as linked to in the article are not high grade capacitors. The only thing special about them is their failure mode. This is not a "special" capacitor configuration. It is almost a given that the power supply in the Cisco switch already has it, as does most switch mode power supplies with a ground connection. There are many ways to reduce and suppress this noise - installation methods, shielded cabling, filters, etc. One proven way in electronics design would be to replace lower-grade capacitors with noise suppressing capacitors. Here is an example of a capacitor design that reduces EMI/RFI and would have a measurable reductive effect on noise levels if installed into a Cisco switch: |
djones51, 'You think that switch is funny get a load of this thing. https://www.kevalinaudio.com/product-page/waversa-wrouter' Waversa WRouter $4,050.00Sale Pri WROUTER: Ultra Low Noise Audio Grade Router Although a relative unknown in the Western Hemisphere, the Waversa Systems WRouter has been on the market for some time, and it remains the only audiophile grade router. This unit was a “from the ground up” build, taking advantage of Waversa’s in-house chip design and proprietary signal processing and protocols. The WRouter achieves what it is designed to accomplish: to present streaming data without tradeoff, with black background, more focus, less tension, improved clarity, dynamics, and detail, all with a noticeable reveal of the intricacies of decay.
The WRouter is designed to handle the most demanding audio files and is capable of streaming different content to different locations simultaneously without compromise. This unit is not a high-speed router or switch. Any step in that direction proved detrimental to overall sonic presentation.
The Waversa WRouter achieves network security by “packet filtering” and, for ease of use, is not configurable by the end-user. However, WiFi Access can be custom labeled and password protected by the end-user.
The end user can set up the WRouter as a WiFi router, a switch, or a hotspot. Best overall results occur when the WRouter is separated from non-audio bandwidth by connecting the WRouter to your modem with an isolated ethernet line. ..... Clean power Switch mode power supply (SMPS) introduce high frequency noise and voltage spikes to normal sinusoidal alternating current. This noise has been demonstrated to travel retrograde and contaminate other power lines in your home and your neighborhood. Waversa products use either an internal linear power supply or lithium battery power, but switch mode power supply noise generated by other network sources can still pollute the audio signal. This issue has been considered and addressed by the WRouter’s incorporation of two isolated ethernet hubs. The normal LAN ports use linear power and the audio LAN ports are battery powered, providing complete power separation. Switching hub processors in each LAN area provide effective data distribution by separating distribution of music streaming data from distribution of data having nothing to do with music, such as behind the scenes network processing, firmware, LCD display commands, etc. Large power transformers supply main power ...Linear power supplies are more stable than SMPS power supplies, so distortion is reduced by minimizing jitter and noise. Battery-powered LAN audio ports The four audio ports are designed to operate on a lithium-ion battery. By ensuring consistent, clean electrical current, Lithium-ion batteries allow for a more precise data stream. This level of performance is difficult to achieve even with a linear power supply. In addition, the ultra-low impedance characteristics of lithium-ion batteries absorb and remove even the smallest noise signals transmitted through connected signal lines to achieve noise-free performance.
LAN ports with Isolation and Noise Filtering ...In addition, two transformers are strategically placed to eliminate high-frequency noise, one between the two hardware network switch chips that communicate between the three ‘dirty’ side LAN ports near the power supply and the four ‘clean’ side battery-powered LAN ports connected to the audiophile components. The second is located after the switch chip for audio data. The four audio LAN ethernet ports have an additional filter so the signal to the audiophile devices is filtered twice. ..... Separate Internal Parallel Switches The WRouter’s audio LAN ports and normal LAN ports are designed to be completely separated, using parallel internal switches. This separation of processors reduces latency relative to use of a single processor, reducing the chance of jitter due to overloading. This design also prevents internally processed packets from being delivered to the audio ports, thus improving the throughput of audio data.
Whole-cut duralumin chassis The WRouter’s solid, CNC cut duralumin chassis minimizes the effects of vibration. In addition, the inside and outside are partitioned to a thickness of about 5 ~ 8mm, so that noise, whether generated externally or internally, does not interfere with the audio signal.
Floating structure design resistant to vibration While the basic exterior armature is built to resist vibration, the structure itself is also designed to “float”, providing even more effective vibration suppression. This airborne structure makes it almost impossible to transfer vibrations, so there is no sound distortion due to vibration. ------------ All in all a decent stab at pointing the well-heeled yet unsuspecting audiophile towards pulling the trigger on what might appear to be a credible unit to an uncritical eye; albeit one priced at $4K. For good measure a couple of graphs are thrown in alongside snaps of some of the hardware contained therein. The usual emphasis on filtration and implications of signal purity are tossed into the mix plentifully. Leaving no stone unturned, great emphasis is also placed upon the importance of isolating against the harm caused by structural and airborne vibration control which seems to have now become the new hunting ground in snake oil audio. The bait fixed, the trap duly set ; the rest is just a case of waiting for a bite. Some might say that this kind of practice is harmless in itself. They might argue that anyone who falls for such an obvious scam dues so deservedly. Just another example of the free market at work. Fair enough if that's your view, but I'm more concerned of the harmful effect this kind of fraudulent behaviour will have upon the future of the high-end and the associated dealerships. The high-end is already a veritable 'wild west' festooned with overpriced products of unsubstantiated quality often assembled by designers of unknown repute. You only have to see the back pages of any magazine to what lurks there. For every Roy Gandy and Sean Casey there seems to be a thousand charlatans out there. Dealers, no doubt faced with increasing financial pressures, who might be prepared to join in with this charade are risking their business becoming a laughing stock. What a choice! Who'd want to be a dealer today? Only for love. |
The white paper linked to earlier is another attempt to wow with nonsense. It talks about line jitter leaking into your DAC and causing all sorts of nasties. So they use very precise clocks to control this. Only problem with this fix is ethernet line uses a 25MHZ clock, yeah megahertz . So it’s "fixing" a nonexistent problem. Not to mention the "line noise/jitter" and clocking doesn’t reach to DAC, the packets are stripped of the outer layers and any clocking of the stack by software, processed by the software in the streamer to retrieve the data, placed in a buffer then the DAC clock asynchronously pulls the data for conversion. The bottom line is with today’s software, hardware, filtering and precise clocks placed close to the DAC the last thing you need to worry about is a switch. I also find it amusing why no worry about the little $100 modem that converted the incoming analog signal to your house and built this digital signal so your devices could read the signal? That has more to with what’s going on with your streaming than any switch you place downstream. By all means decide what you want, buy what you want but try informing yourself as to what you’re buying and use some common sense. |
I read the part about EMI/RFI noise produced. I paid particular notice to this and what followed. The Federal Communications Commission (FCC) publishes specific regulations to limit the amount of EMI and RFI emitted by computing equipment. Each system meets these FCC regulationsIt warns about bad wiring practices. It is talking about installations in enterprise closets full of equipment. Lightning protection , ventilation, humidity you know thing to worry about in a high traffic business environment not someone's home system with a couple of end nodes. This particular switch is nothing but a $40 Cisco unmanaged layer 2 with a couple of meaningless "upgrades" there is nothing this thing will do to an audio signal that's audible. Evaluating these types of "tweaks" for that's all they are, by plugging it in trusting your infallible senses and saying "wow it's amazing " is a meaningless gesture. It's like eyewitnesses who manage to trust their eyes but usually get something wrong or wine tasters who can't tell red from white when the white is dyed red. People need to get over this hubris that they are infallible and their senses are not subject to normal human biases. |
Like everything audio, visual or taste just listen, view and eat and determine for yourself if you like what you hear, see or consume. That simple. The rest of this arguing and name calling is a complete waste of time and off topic. We are all big boys and girls and can make our own determinations based on our individual senses. Others can’t tell me if I will taste or hear a difference I like and enjoy. Only I can do that as my senses are mine, not theirs. We are not all exactly alike in terms of sensory functioning and in the case of Audio our systems and rooms are completely different. So simple folks and the only way this falls apart is if someone has not listened, viewed, or tasted, but feels compelled to make judgements anyway plus save me from myself. That my fellow Agoners is not funny, but rather nonsensical. Newsflash! We humans don’t know the why of most things in this world and we see our ignorance in the rear view mirror of life. Most often long after we pass away. |
I have no idea whether this modified switch will reduce audible noise on a stereo system. However, EMI/RFI noise is a major issue in building physical networks. Here is a link to a Cisco switch install guide, read the part about EMI/RFI. Note this section is included in all Cisco data center-grade switch install guides, because EMI/RFI noise produced by switches is both real and sometimes does interfere with the signal. https://www.cisco.com/c/en/us/td/docs/switches/lan/catalyst4900/4948E/installation/guide/4948E_ins/0... There are many ways to reduce and suppress this noise - installation methods, shielded cabling, filters, etc. One proven way in electronics design would be to replace lower-grade capacitors with noise suppressing capacitors. Here is an example of a capacitor design that reduces EMI/RFI and would have a measurable reductive effect on noise levels if installed into a Cisco switch: https://www.illinoiscapacitor.com/pdf/Papers/EMI_RFI_suppression_capacitors.pdf The OP could be correct that the modifications made to the referenced EBay switch will make no audible difference. But, there’s no way to know without actually testing the original and modified devices. However, the OP and some other posters are quite incorrect in asserting that switches don’t pass measurable EMI/RFI noise along with the signal - they do, as Cisco warns in their install guides. And, they are incorrect in asserting that it’s not possible to measurably reduce that noise with potential for an audible effect - Cisco gives us a few tips for how to do it, and we know that noise reducing capacitors are routinely incorporated in electronic designs for this exact purpose. Phasemonger |
While the basic exterior armature is built to resist vibration, the structure itself is also designed to “float”, providing even more effective vibration suppression. This airborne structure makes it almost impossible to transfer vibrations, so there is no sound distortion due to vibration.Hmmmmmmm...Calling Peter Belt. To his credit, many of his ideas agree or not, did not require a financial outlay. Not as witty as Donald Fagen’s notes on the making of Can’t Buy A Thrill But you get the picture. 4. Jeff Baxter had a theory that one of the secretaries at the ABC offices had several sets of interchangeable breasts which she would alternate wearing from day to day. This turned out to be incorrect. |
In a double-blind test, it would be much more difficult to determine any improvement with an upgraded network switch with a system that was not optimized. Comparing a network switch is considerably different than comparing another component - such as a power conditioner. And I’m asserting that no system, at any price, will make that $500 switch a worthwhile purchase. |
| Post removed |
dougeyjones, You aimed your sarcasm at the wrong group. You won't find a bigger bunch of voodoo snake oil lovers than right here at Audiogon. Their fascination with multicolored fuses and quantum this and nanoparticle that is a source of endless entertainment. As for millercarbons comment about resistors and capacitors, if he knew anything he'd know that they affect audio frequencies and anomalies, not those of a network switch. OK boys, now let's hear about how I never tried it or my system sucks and I'd never hear the difference or I'm not an educated listener. |
Nah, believe it or not, I’ve been reading AudioGon for over a decade and just never posted. Definitely don’t see myself, or this post as making any kind of headway. If you noticed above, Steakster not so subtly attempted to shame my peasant level Classe + JBL Synthesis 2-ch setup 😆. The vibe around here is very “Private County Club”, with a very defined pecking order based on the worst kind of implicit biases. I imagine international readers routinely shake their heads in disbelief at just how thick the bubble is around the Gon. I still enjoy it enough most days to keep coming back, so take the critique with a grain of salt. |