Streamers vary greatly in the amount of noise generated from within, noise is the enemy of resolution/transparency. In researching and comparing streamers you'll find much difference in the lengths some makers go to lesson noise. Every single component on motherboard can contribute to a lower noise floor. Proprietary operating systems and linear power supplies also contribute greatly to lower noise floors. Streamers set the resolution/transparency floor for the entire streaming chain, extremely important component, perhaps the most important component in the entire chain.
Sound quality differences in streamers
Can there be sonic differences between moderate and high priced streamers when used for streaming only. I will not use or engage an onboard DAC or any other feature, just stream from Tidal or Amazon to DAC. If the unit is just transferring zeros and ones to a DAC can there be differences in say a $300 WiiM and a $3000 dSC streamer? Thanks
Showing 11 responses by sns
@verdantaudio I too have found Roon best with core and endpoint/renderer segregated, in my case two streamers run in bridged mode. |
The Dejitter Switch X addresses an issue most are unaware of. Having a 'clean' network is required to get the most out of one's streamer/streaming setup, this means starting at modem. I discovered the benefits of having an audio only or 'clean' network and a 'dirty' network' for the house a couple years ago. |
Audiophiles who own more expensive streamers are generally no longer comparing these to the affordable, they've moved past that. Vast majority are comparing expensive streamers, the fact a wide variety are purchased likely goes to show all expensive streamers don't sound alike, they aren't choosing their particular streamer based on some emotional attachment or bragging rights. Evaluating sound quality is what sets the audiophile apart from the casual listener, based on my experience listening to a wide variety of high end systems over many decades I've come to understand many audiophiles are very good judges of sound quality. Seems some believe they have golden ears, so if they don't hear a difference between streamers its not possible for others to hear differences. |
Its nonsensical to believe dacs act as a noise conditioner for detail obscuring noise further upward in streaming chain. The only sense of 'conditioning' a dac can possibly achieve is optimizing certain inputs, this usually usb on most dacs. Galvanic isolation is one of the ways they achieve this, dac manufacturers understand they are only one link in streaming chain, so attempt to optimize how they interface with rest of chain. Interfaces amongst all the other components within a streaming chain require the same optimization dac manufacturers provide, data is traveling across the entire chain. And interfaces are only one consideration, minimizing noise within components is fundamental/a priori to achieve lowest noise floors for max resolution/transparency. And we haven't even gotten to clocks within these components, this has everything to do with presentation. And so are we going to argue that clocks makes no difference. |
I too have a custom build, while I've built/assembled or modified both general and audio only streamers, one I'm presently using was built/assembled by one of the most senior and respected audiophilestyle forum members. This particular unit was a proven standout compared to some of the top flight off the shelf streamers. Eventually this design surpassed by the Taiko Extreme clones which last time I checked were considered to be best diy design possible. The latest Taiko, Wadax, probably some others go beyond any custom build. These custom builds generally use Enterprise level components, low noise, latency (meaning speed) and precise clocking is the goal. My particular streamer, based upon a Windows motherboard, as do most of these custom builds, requires very specialized lps as Windows motherboards use a variety of voltages for various functions. I'm presently running JCAT Optimo ATX lps, $6200 last time I looked, this took it way over top of previous attempts at lps via a variety of external lps powering modified internal switch mode power supplies. And then there are choices of operating systems, I'm running Euphony OS which disables virtually every non essential function of a normal Windows motherboard. And there are a variety of PCie cards from JCAT, Pink Faun, others for optimized network, usb, I2S interfaces, these can also be powered externally via lps of your choice. Cool thing with custom builds is you get to hear what individual components add to the mix, everything matters. Understanding what's important in custom builds will allow one to understand how streamers work and why they're so important to the streaming chain. For those interested in learning from the experts and designers of streaming equipment the Audiophilestyle forum is an invaluable resource. |
Here is long thread with technical explanations for how and why noise affects sound quality in streaming https://audiophilestyle.com/forums/topic/70440-digital-noise-discussion/
Digital noise is both self generated within components and rides along grounds to next component in chain, this why galvanic isolation effective, optical conversion or optical internet service also solves this noise traveling issue. Doesn't solve the self generated noise issue, this noise obscuring resolution/transparency which once lost can never be regained, galvanic isolation doesn't solve this problem.
As others have mentioned, clocks also critical, and this holds true for entire steaming data path, a quality clock within dac cannot repair timing issues upstream from it. Quality clocking has everything to do with presentation, generally analog sounds more natural to us because it present a continuous waveform, precision timing comes closer to presenting a similar waveform, result is a more analog like presentation with proper clocking/timing. Streamers differ greatly in clocking quality.
I misspoke in regard to windows motherboard, atx motherboard correct, these require multiple voltages. |
There is an alternative to the Switch X, and what I do. I provide a clean/audio only network segregated from my whole house network via modem>router with wifi disabled, on clean side this feeds NAS and server, on dirty or whole house side feeds another router with wifi enabled, this feeds whole house. All equipment on clean side powered via LPS. Switch X could be superior to this via isolation from the constant pinging a router gives and receives.
Quality of one's network also has audio quality repercussions, and to think that some don't believe servers, streamers, dacs don't make a difference! Just have to write that off to there are no known unknowns for some.
As for those with Roon, try a two computer or server to streamer setup. Server contains the Core, streamer contains the Endpoint. Roon specifically designed for this, keeping these functions discrete results in superior sound quality. Bottom line, streaming is a relatively fast moving technology, some people understand there are still many known unknowns, this the mother of invention. |
@mitch2 If I understand your setup correctly I don't see how any conventional routers will avoid having wifi 'contaminate' your 'clean' side network. Having the three systems interconnected to the one server is your issue. Depending on the quality of the other systems you could segregate the main system from the others by inserting another separate server for the other two systems. I don't know if Switch X will solve your issue if continuing to serve three systems from one server. A 'clean' side network cannot have any wifi interference at any point in the chain. The Switch X goes a step further and stops the constant network activity of pinging, as they mention in reviews and conversation, watch your network LED's on routers/NAS/switches, constantly flashing, sending and receiving network info/addresses, this is completely stopped with Switch X. I don't believe Switch X will solve your issue via three systems running on one server, but perhaps they would have a solution. One experiment you could try is to shut down the entire whole house and the two other audio system part of network, see if this improves sound quality.
Even with my present network solution, I'm thinking the Switch X would be an improvement via the blocking of network activity. I actually purchased a Mikrotik managed switch a couple years ago with the idea of diy modding the switch exactly how Dejitter did the Switch X, couldn't pull it off. As I recall price of switch was around $900 plus what it would have cost to mod it, internal parts upgrades and external lps, so $3500 for Switch X not that bad. |
@devinplombier That is not the Mikrotik model I had, can't recall the model but it had dual internal power supplies, I'd have to look back on ebay past purchases to be sure of model. In any case it was both the internal mods and conversion to allow for external linear power supply that complicated matters to the point I thought of the simpler method of adding a second router for 'dirty' side of network. |