Sure but the "cleaned" signal again enters to DAC, cannot it again poorly extract the sampling info?
If you stream music from the internet, I can't recommend this more highly
I had been using a Roon Nucleus to stream Qobuz, with my Chord Qutest directly connected to the Nucleus. I thought I was getting pretty decent sound quality. And then I got a marketing email from Small Green Computer touting some of their optical gear. The basic idea is that normal cables and connections used to stream from the internet pick up noise of one kind or another (radio frequencies and electromagnetic something or other). But fiber optic cables and their connections/interfaces do not. I don’t know anything about anything, but it made theoretical sense to me, it wasn’t a huge amount of money ($1,400), and with a 30 day return policy I figured I could always return it if I didn’t hear any improvement. Well, I didn’t just hear a slight improvement; it was like turning on the lights in a dark room. Much greater clarity and detail, much better micro and macro dynamics, better timbre to acoustic instruments -- overall just more lifelike. Two quick examples: I’ve listened to some of Steely Dan’s top songs 100s of times over the course of my life, and this is the first time I’d ever noticed a particular and very subtle sound characteristic of Fagen’s keyboard in Babylon Sister. It’s hard to describe, but it’s like there’s a slight sound of air being exhaled by it. The other example: the specific timbre of whatever percussive instrument is used at the beginning of Copeland’s "Fanfare for the Common Man" (a recording by the Minnesota Orchestra). There’s more of a metallic sound than a drum skin sound to it that I didn’t know was there before. The metallic sound starts in the center and then projects out and to the sides, like a wave washing over you. Anyway, I’m just thrilled about having stumbled upon the whole "optical" thing and felt obligated to let others know about it. If you stream music over the internet, I highly recommend giving it a try. (The product I got was the opticalRendu, with the linear power supply option, and the Fiber Ethernet Converter Bundle option.)
156 responses Add your response
...And even with an optical converter, you have to convert the signal back at to ethernet so you can plug it into your streamer, at least with 99% of streaming devices, so the signal is liable to pick up electronic interference at this point and it doesnt take much to have negative impact on the highly sensitive electronics of the streamer. |
At the risk being perceived at being rude again...electronic interference can only affect audio quality at the endpoint where Ethernet packets being converted (decoded) back to the original bit structure of where they entered the transport medium (encoded). Any electrical interference along the transport wires has be inaudible simply because Ethernet/optical packets cannot be listening to. At the interface level there complete galvanic isolation between Ethernet/optical and original bitstream. Jitter/EMI on Ethernet/optical wires is irrelevant and has 0% effect on sq. Now, some manufactures designing "high-end" audio streamers for tons of $$$ really only have to worry about of interference from those interfaces, the converting technology, the size of packet buffers and clock signals. That’s what you pay for ($1,400) for Rendu devices. The real question to me remains: What problems are we trying to solve? If you replace all Ethernet/optical wires from the system and replace them with wifi, that would be a worthwhile discussion. Please read the few available reviews (e.g. Auralic Aries G1), which offers both technologies and try to ascertain which medium has a bigger impact on sq. The result: zero audible difference. As to main question, is $1,400 is lot of money to improve the sq of a system? I can only reply, my Allo Digione Signature/Raspberry cost about $300. This is a transport only, feeding an external DAC over SPDIF/coax. But it has complete galvanic isolation between the "noisy" wifi interface of the Berry and SPDIF transport, plus high-quality re-clocking of the bitstream before it enters into the SPDIF interface. Yes, there still is a small chance of the Wifi interface causing some residual EMI on the SPDIF circuitry. Is it measurable, will it make a difference? I don’t know. Challenge for anyone out there to put a $300 network bridge to test against a mega$$$ network streamer/DAC/Rendu with mega$$$ cable and audiograde switches in-between. I am a audiophile hobbyist just like most people here. All I am saying, please try and blend your listening experiences with the truth of technical explanations. Alternately, you are free to purchase things like the $10,000 Audioquest "directionally aligned" Ethernet cable and expect a "dramatic" improvement of the wired audio components. Snakeoil. |
This hobby cracks me up sometimes. I have to agree with ATMFrank. Please try to balance your subjective opinions with technology. Anyone that thinks a 3 foot, $1,000 AQ Diamond ethernet cable improves the sound quality of your digital over a well designed and built 3 foot, $16.95 BJC ethernet cable (which ships with test results of said cable) is spending an unneeded ~$980.00. But hey....whatever blows your hair back. |
It’s really hard to say "how much" of a difference something makes, because of subjective listening habits and inability to measure. I dare to state that the Allo DigiOne Signature is an incredible value for the money. It’s a touch better than the non-signature version which I had for the last 2 years. The reasons are mainly these: separate power feeds (clean), galvanic isolation (the previous had that too, but on the same board) and improved clocking. Allo also makes a USB bridge that conceptually competes with the Rendu and similar products. I didn’t have the privilege to test this out (money is not an object, but time is....) What do I with the Signature: I am running LMS/Squeezeplayer on it, without oversampling. Because I like to control to oversampling characteristics on the DAC itself, not in software. Although it is nice to have that software option (Foobar) for trial/error and to play around with it. I love the sound coming out the Signature. Someone should put that setup next to a BlueOS or Rendu device and compare, as unbiased as possible. I am feeding a Burson Conductor 3R (2xESS9038q2m) a SMSL SU-9 (ESS9038pro) and sometimes a Chord Mojo for A/B testing. For the ESS, my personal preferred filter setting is MP/fast (minimal phase, less pre-ringing), which comes close the "Meridian" sound of MQA (don’t get me started....). Playing with OS filter settings and being able to measure the effect on your listening experience is something that takes a long time to develop. It’s fun, doesn’t cost anything more and makes the difference between feeling "fatigue" and "addiction". Again, personal opinion only here. I use a combination of UniFi switch/cheap Ethernet cable for the Signature setup, and wifi on the other. I sometime switch. Honestly, to my ears, there is zero audible difference between Ethernet and wifi. Look, I don’t want to put anyone down for feeding the hobby. Many decades I ago spent top $$ for 20bit Audio Alchemy HDHC DAC, which had an "ear-and-mind-opening" experience. Did I regret buying it? No. But I do know that technology get’s better and cheaper all the time. And I just want to make sure I understand the technical reasons before spending more $$$ on the latest gadget. I like to make educated decisions. Happy listening. -frank |
I haven’t followed all of this thread, but is the OP stating that he swapped out ONLY Internet cabling to his device, or added a component to the device(s) too? I do software for a living, and understand the TCP and IP protocols used to move byte sequences over the Internet. TCP/IP underpins most aspects of Internet traffic: email, web, ftp (showing my age!). When the sender (e.g. Spotify) transfers a ’file’, which is really just a byte sequence, to a receiver (your music box), it adds checksums to ensure any errors in xmission are noticed by the receiver, and fresh copy sent. In essence, TCP ensures that what the receiver gets is what the sender sent.Now, if music streaming is what I think it is, byte sequences of data representing sound files, then there is no way a cable by itself can ’improve’ on the data sent from S to R. Yes, it may be faster, or be higher quality and require fewer re-transmits, but at the end of the day, an Internet cable’s job is to xfer data across its two endpoints, which is likely one small ’hop’ in the larger path from sender to receiver. The cable doesn’t ’know’ the signals its moving are sound files, its just electrical signal representing 0 and 1, each a ’bit’ and we have 8 bits to the byte. Only improved PROCESSING (better DAC etc) of the incoming byte stream can possibly improve sound quality. The D in DAC is a given, it's what TCP/IP does. It's the A in DAC that counts. And cabling alone is firmly in the D side of things. That said, I'm off to polish my two grand power cords, they do make a difference, honest ;) |
Great example of a person who is not smart when it comes to technology so they open up their wallet willy nilly and believe anything they are told. There is almost 0 data loss in wired home networks. Even when there is, those data failures are corrected. The only issues come does to EMI, whether through the Ether cable, which is possible as Ethernet transformer can pass somewhat low frequencies, or possibly through closely connected power supplies. |
douglas_schroeder Great examples of smart people without experience giving poor advice. "No way"That's a perennial problem here. Theorists, measurementalists, and armchair quarterbacks who discount - or even completely dismiss - empirical evidence, even when it is provided in abundance. |
https://www.livescience.com/21456-empirical-evidence-a-definition.html Before any pieces of empirical data are collected, scientists carefully design their research methods to ensure the accuracy, quality and integrity of the data. If there are flaws in the way that empirical data is collected, the research will not be considered valid. |
In the age of disinformation, audio marketing has lead the way for many years. I believe in civility and rational discourse, throwing verbal molotov cocktails doesn’t solve anything. Back to the topic and original question: is the amount of $1,400 spent on a well-designed product a good investment? How much "improvement" can be expected and how is it measured/justified? This can not be answered without additional context. Such as asking the question, "what is the problem that needs to be solved", which requires some technical understanding and basic scientific approach. No way around it. Audio2design is correct in stating most arguments are based on anecdotal and here-say. Even well respected community members don’t aways get it right (saying this without arrogance). Re-reading the (very favorable) blog about the Sonore Rendu in Computer Audiophile made me realize that the problem is still NOT understood. That was in 2018. But we now we have a dozen of solutions to choose from, some rather expensive ones. A 12 Meter "directional" AudioQuest Diamond Ethernet cable goes for $11,000 retail. Not bashing here, just stating that I have no respect for companies like Meridian, Chord, AudioQuest, Sonore (the Marketing dept. mostly). Although I own a few of their lesser expensive products....and you can call me hypocritical? Yes, perhaps a little. In the meantime I enjoy (subjectively) listening to my humble setup, carefully selected and budgeted. Addicted to good audio :-) Cheers, -frank |
audio2design By empirical you mean anecdotal.No, by "empirical" I mean "empirical." I understand that you tend to consider empirical data as inherently anecdotal, so it's odd that you now seek a distinction between the two. It would be best to learn the differenceI'm glad you now acknowledge the difference. |
| Post removed |
this is true, but at the same time empirical study cannot exist without a theory, these two things are tied. given a theory, you may verify it empirically if you are unable to prove its soundness theoretically. in physics, this is a common practice, and also in computer science. for example, if a constructed by you mathematical model/theory/algorithm cannot be proved (theoretically) to be good (i.e., be objective, realistic, or optimal or sub-optimal), you carry our an empirical study. In computer science, this involves computational experiments that somehow show the practical behavior of a given algorithm. pretty similar scenario is in audio stuff where your estimation is based on auditioning (these are your experiments). but here auditioning experience itself is subjective, unlike physics and computer science, for example, where the experimental results do reflect an "objective reality". in audio world an "objective reality" is subjective, it depends on our taste, mood, current environment etc. |
niodari ... empirical study cannot exist without a theory ...Actually, by definition, "empirical" means in the absence of theory. (See definition two posts previous.) Of course, I think empirical observations are most useful when coupled with theory, but the two are not inherently linked. That's something that just bugs the heck out of the measurementalists here. |
@audio2design
What part of "empirical" evidence, from a scientific standpoint, requires controlled listening tests, not ad-hoc listening tests hence anecdotal evidence is hard for the kids in the back to understand??The answer is "none." Empirical simply means "verifiable by observation or experience." Whether the observation or experience results from an orderly, documented, scientific, study, or simply a few dudes sitting around smoking a doob and listening to their favorite tunes....the difference is simply methodology - both are examples of obtaining "empirical evidence." An empirical study can be performed to verify a theory, and empirical evidence can also be obtained to verify a correlation with measurements, but you can have measurements and/or a theory without empirical evidence. Empirical evidence is obtained every time you listen to your system. I think you are all sort of making the same point. |
Empirical evidence can also be anecdotal evidence and persons postulating on how much difference a cable makes in the sound coming from their speakers is empirical and anecdotal evidence. From this evidence we build hypotheses is the difference because of the cable or some other factor? Now we can begin to form experiments to understand why this person hears differences between cables or we can just take their word for it only one of these will further our knowledge of cables and human perception and it isn't the latter. |
Copied this from a website, but ... Breaking Down the Evidence The main concern with empirical research is the collection of unbiased evidence. Researchers must carefully design the research while minimizing exposure to potential errors. In the scientific world, it is common that several scientists or researchers gather evidence simultaneously through the replication of the same study. In addition, a peer review is a primary tool in science that is used to validate the evidence provided in a study or research. https://corporatefinanceinstitute.com/resources/knowledge/other/empirical-evidence/ Stolen from Wikipedia. These methods yield only probabilities. Among scientific researchers, empirical evidence (as distinct from empirical research) refers to objective evidence that appears the same regardless of the observer. For example, a thermometer will not display different temperatures for each individual who observes it. Temperature, as measured by an accurate, well calibrated thermometer, is empirical evidence. By contrast, non-empirical evidence is subjective, depending on the observer. Following the previous example, observer A might truthfully report that a room is warm, while observer B might truthfully report that the same room is cool, though both observe the same reading on the thermometer. The use of empirical evidence negates this effect of personal (i.e., subjective) experience or time. What this means is the answer to "do you like A better than B" could be considered non-empirical evidence. Being able to tell A and B apart would be empirical evidence. However, if B was a modified version of A, then it could be empirical depending on how the question was worded. I do feel this paragraph is a bit strict in its definition. From bennpar.com Anecdotal evidence is using your personal experiences and stories to illustrate your point. Empirical evidence is measured, unbiased, and replicable.
From study.com Today's topic, anecdotal evidence, can be defined as testimony that something is true, false, related, or unrelated based on isolated examples of someone's personal experience. Anecdotal evidence is very popular in the advertising world. Every time you see a claim about a product's effectiveness based on a person's personal experience, the company is using anecdotal evidence to encourage sales. Again, like almost every single uncontrolled listening test reported on Audiogon. From yourdictionary.com Based on casual observations or indications rather than rigorous or scientific analysis. Again, like almost every reported listening test here. Oxford English Dictionary:
So yes, I am quite confident in classifying almost all the reported listening experiences on this site as anecdotal. |
@djones, Empirical evidence can also be anecdotal evidence and persons postulating on how much difference a cable makes in the sound coming from their speakers is empirical and anecdotal evidence.@audio2design gave an excellent overview, according to which these are non-empirical evidences, as an empirical evidence reflects objective reality. @audio2design has in fact justified that the most of the arguments here are based on non-empirical, i.e., anecdotic evidences. And empirical evidences are somehow related with a theory. For instance, by observing the temperature (an empirical evidence) we rely on the theory based on which thermomiters were designed. Even anecdotic evidences here rely on theory. E.g., when one judges about sonic properties of a DAC, he relies on the theory on which this dac was built (e.g. Fourier transformation used in sampling/unsampling process or whatever theoretical assumption supporting the architecture of that unit). From this evidence we build hypotheses is the difference because of the cable or some other factor? Now we can begin to form experiments to understand why this person hears differences between cables or we can just take their word for it only one of these will further our knowledge of cables and human perception and it isn't the latter. Indeed, these are subjective realities. Perhaps, thus guy smoke marijuana and every cable, including the tested one, sounds wonderful for him or he was encouraged by the seller or whatever. An empirical evidence based on theory is reliable but unlikely that may occur in these threads. |
Let me start saying I haven't read this entire thread but I have read others and I do believe people have heard an improvement with optical interfaces vs ethernet. I'm also aware SGC and Sonore make exceptional well regarded audio products. That said it is a misconception to think optical as a technology is "completely" clean for RFI purposes, now if this RFI dirtyness (residual of optical) affects audio or not that's another story. The entire RFI/EMI issue arises on ethernet and dacs, servers and such on poorly designed and improperly shielded systems. When you are modulating a wavelength (either multimode light or single mode laser) over fiber there are residual RFI which by the way are many times higher than with actual ethernet, 30 to 40 dB in magnitude, radiating in all directions and inducing into cables creating antenna effects everywhere, shielding with aluminum, copper and even cable shielding blocks at most 10 to 15 percent of this signal but it is ineffective at blocking effectively specific residual due to their magnitude. This problem is even more accentuated with BiDi devices which transmits and receives over the same fiber on different wavelengths. Finally this will get induced into ac adapters, other cables and such. Does it affects music? Apparently not since no one has complained and people I trust have tested the Sonore and SGC products and report an improvement with optical, but it does exist and it is not "completely clean" On my testing some optical equipment is better than others blocking and shielding this effect, SGC and Sonore products haven't been part of my testing as my test was on a different application than audio related directly to my work. |
just in front of my Antipodes server to eliminate EMI/RFI that may potentially be contaminating the approximately 30 feet run of Ethernet from my router to my server @mitch2 question is what this contamination might do, your tplinks will certainly contaminate based on my results, is it audible or different in SQ? TP-Link Gigabit SFP to RJ45 Fiber Media Converters cost about $20 bucks But they are terrible at shielding, think of it like most FMCs will radiate omnidirectional and through the cable in every direction. Using SFPs "usually" will radiate on controlled directions (mostly in front of the SFPs) use MM better than SMF unless you need more than 1 gb which is highly unlikely, and avoid BiDi |
It’s really hard to say "how much" of a difference something makes, because of subjective listening habits and inability to measure. Agreed Although it is nice to have that software option (Foobar Have you tried HQPlayer? Yes, there still is a small chance of the Wifi interface causing some residual EMI on the SPDIF circuitry. Is it measurable, will it make a difference? I don’t know.My point exactly, will the analog part or digital conversion stage elements suffer from this residual or will it be impervious to it? I haven't noticed any differences myself but others intriguely do, should I say they are delusional or should I give them some credit? I choose the latter, I don't like to dismiss people experiences, unless at some point I realize they are truly delusional.
I prefer these filters too |
In a home environment, with cheap cables, data losses and retransmits are near zeroThis is accurate, if I may I would just add that if the switches support flow control and enough buffers the values will be actually zero and if they don’t if your audio devices do buffering it doesn’t matter if there are errors anyways during transmission. |
Have you tried HQPlayer?Good question. I have used HQplayer for a brief trial period. The UI is a bit primitive, but it provides a great learning opportunity to study the effects of filter design on sq. Essentially, one is able to simulate internal DAC operations. But you have to learn the theory first before making good selections for the various filters. A great resource and highly recommended for anyone interested in the topic. I see a lot of people struggle with it. In my case, I only applied my limited playtime with HQPlayer to a high-quality ESS9018 based DAC and found very few (subjective) differences between the various options. Not enough for me to put up with the poor UI and make it the primary player for everyday use. As an end-point to Roon perhaps it is better suited. I also think that if one has a high-quality d/s based DAC of recent design, the benefits of using HQPlayer is greatly diminished. Luisma31, thank you for your comments. |
I haven’t noticed any differences myself but others intriguely do, should I say they are delusional or should I give them some credit? I choose the latter, I don’t like to dismiss people experiences, unless at some point I realize they are truly delusional. I struggle with this as well, because I found myself in the exact same dilemma. First of, I am not an active blogger unlike others here. And one must avoid not to fall into the pit holes, e.g. endless arguments. The OT headline of this thread caught my attention. And without becoming too confrontational, I asked "What is the problem opticalRendu is trying to solve?" Why does this device makes such a "huge" difference in perceived improvement? The "improvement" needs to be put into a different perspective and somehow quantified. My approach now: I think the measurementalists are doing a great service to the community and I read, and listen to them. But by no means are they the only authoritative decision point. Nobody can ever be fully immune to the "aha - this sounds great" effect, because after all it is wonderful hobby (for most). And therefore nobody should be put down for being excited about the improvement of a certain product, unless it is, as you said, completely delusional. Opinion vary... |
I’m a retired Network engineer. I’ve been involved with wiring, packet tracing and what not (source routed token ring yea). Where I run aground in this issue, is the data prior to the DAC. It’s all digital data packets. Electrical interference or noise makes no changes to the data packets themselves. It could cause a packet not to be received, or fragmented. I would think all/any problems with noise/interference would show up in the analogue signal path. Any one? |
Electrical interference or noise makes no changes to the data packets themselves. It could cause a packet not to be received, or fragmented.Could you please re-write this in CAPS, so that everyone can read it :-) Some people believe that expensive shielding and "directional" data cables would somehow improve audio quality. But it's a futile effort to fight a enormous industry and their believers. Only a small amount of education would remedy this. Again, I am not saying that microRendu is a flawed product, but I am saying that the value is inflated by misleading marketing and phony claims. In extreme cases, some boutique hi-end audio manufacturers will sell you a $30,000 music streamer (network -> USB or SPDIF bridge, or DAC) and make you believe it's a worthy investment. Some people on this thread here can build a better performing streamer for $300 and with expectations of similar results. |
With a very good 75 ohm spdif cable and OTG USB, this allows me to use my Android to my DAC for Tidal MQA-https://www.nobsound.ltd/usb-to-spdif-converter-xmos-xu208-coaxial-optical-toslink-support-source-ou... Sound pretty good in my system. Not bad for 40 or so bucks. |
For my first ever post here -- saw this thread last week and thought might be interesting to try the SoTM solution. After 3 days across many music types, I can hear discernable improvement in my set-up (maybe 5-10% clarity, smoothness). No measurable data just perception. My set-up: about 40ft ethernet to PS Audio DAC from computer using JRiver as my "server". Lot of electrical noise in my environment that I have spent these Covid days at home cleaning-up (separating power on front end with Adept Response & amps with isolation transformer, etc..). Went ahead and got the SGC "kit" with optical module, linear power supply and fiber ethernet converter. Not cheap (about $500) but also not the $1200+ set of boxes Since my version of PS DAC already has ethernet input, I did not any other product for USB. Kit came with no instructions so just followed the YouTube video from SGC. About 10min of connecting and the PSAudio was recognized on JRiver without any config updates. Thought I would share given some of the confusion on these responses. Not affiliated with any vendor. Just someone working from home. |
I have single mode fiber between an ER and sNH G10. I recently added a master clock to the ER. The noise level dropped a ton and the details and reverberation are amazing. However I now recognize a bit of edginess in the upper frequencies that I suspect were unmasked. My SFP's are Startech and are rated for 80 KM. I wonder if the powerful SFP's are making noise. I ordered some attenuators to try and also ordered some short range multimode SFP's and fiber from Sonore. I'm not sure if adding a Master clock is reclocking but it can have a profound effect. |
@anzaanimalclinic please report back with your findings. I use a cisco 2960 8 port fanless switch in front of an Etherregen feeding my DAC and I noticed an improvement in SQ by making this link optical where I previously had copper I put this down to improved electrical isolation. I went for second hand low range really cheap multi mode SFP modules as i figured that more powerful = more emi/fri think I pai £10 for the pair on an online auction £5 for the patch cable, the cisco was less than £40 its nice to recycle this old network gear :) I do have a short AQ vodka from the etherregen to my DAC which I picked up second hand, I can’t hear any significant difference between this and a chord C stream cable that cost about 1/6th the price of the Vodka new FWIW. |
FYI, a well-respected member of the audiophile community has taken the time and presented a technically correct assertion of why any switch or cable on the networking side of the digital audio stream *cannot* make any difference. Listen and learn: https://www.audiosciencereview.com/forum/index.php?threads/do-audiophile-network-switches-make-a-dif... |
More like a "well-known hater of the audiophile community" |
This isn’t the 70’s where all we look at it is a single frequency THD value.How true that is. I grew up with DIN 45500, the HiFi standard. THD of 0.1% and 40-16,000 Hz was phenomenal performance. We (in W-Germany) also had relatively unbiased consumer report publications. It was completely acceptable and common to disqualify certain products of questionable design. What happened to the mindset or making informed decisions? |
Being a retired telecommunications digital switch design engineer and Navy electronic technician I find all of this quite funny especially if you are streaming off of the internet. Twisted pair has its draw backs as do coax and fiber. What counts is the quality of the source in generating a true square wave, the transmission lines ability to keep that wave form intact and the ability of the receiver to decode that signal. Do you truly want to hear and see a difference. GPS satellites have stratum 1 clocks. We use these in remote situations to clock telecommunications equipment for a reason. We have IP the inferior transmission method for a reason which I won't go into. We implemented ATM in 6 switches in the network only to have to pull them out because a bigger authority decided Ethernet was the way the country would go. https://www.philips.com/a-w/research/technologies/cd/beginning.html |
| Post removed |
