That’s definitively not a problem with Ethernet or USB audio data transmission, both utilizing bidirectional communication to keep receiving end buffers happy.
So much about anecdotal references without any comprehensible and technically founded explanation.
In light of the above quote, you’re seriously parroting marketing hype about using I2S between audio components?
It’s been intended originally for the interference-prone inter chip audio data transfer, thusly needing to be kept as short as physically possible, typically a few millimeters at the maximum.
Please understand that this unidirectional protocol forces the DAC to now slave to a remote clock signal, that has been mangled by having to travel comparatively extremely long distances, passing too many interconnections along cables with dubious differential trace length and impedance matching, potentially causing more problems than it allegedly solves.
USB or ethernet protocols, on the other hand, allow the DAC to control the audio data transmission according to its on-board clock for truly optimum decoding performance.
If any undue contributions are below audibility threshold already, you will achieve no further worthwhile improvements, unless you choose to believe in marketing induced FUD and confirmation bias.
Feel free to enjoy your music via your chosen path, but don’t expect to be taken seriously without delivering any hard proof in public fora.
Torben
(- A Dane living in Hamburg, Germany - Roon Lifer)
28
USB is the best input on the vast majority of modern DACs.
It was only in the past when many DACs had limitations with USB that made it so you’d get poorer Jitter via USB than via synchronous protocols.
The main limitation of USB now is just noise rejection, but even then many products have excellent filtering and isolation. And if wanting to be super safe you can grab a JDS synapse for $50 or Neutron V1 for $100
Thanks Torben. It seems Amir used a DAC that he found to be prone to USB noise, to see if isolators worked. They do work, but that’s not an excuse for designing/using a DAC that needs them, since isolating input noise is one of the DAC’s jobs.
I2S is an ancient protocol for moving bits inside a device along carefully laid out PCB traces. It was never designed for inter-device communication, and it is very bad at it, even if one shoehorns it onto an HDMI cable (leading to multiple electric conversions). It only adds noise and jitter.
Maybe @francisco_gaston meant that the HDMI part is what was [recently] introduced? Regardless, it’s still just marketing, since it doesn’t solve anything. It’s the typical audiophile industry ‘innovation’, accompanied by bashing ‘old’ technology, so that audiophiles rush to ‘upgrade’ their gear.
So much wrong in so few lines. I am not able to put right everything.
‘analog shaped signal and its shape affect hte way clocks works with it’ - w t f?
Please get yourself educated how digital data transmission works and you’ll see why I’m simply speechless when I read something like that.
And these ’ main Hifi developers’ you mentioned are snake oil dealers who try to make profit with audiophile angst not serious Hifi devolpers or had you ever heard such claims from developers at Sony or Yamaha? No, only from socalled audiophile brands.
I don’t understand the topology where centralized service (core) is decoding a raw stream and pushes the same through the network towards roon endpoint that is maybe even more capable to do the same operations reducing the network influence instead.
So, instead using endpoints as p2p devices it goes towards old fashion topology no one uses anymore.
The same topic is if you want to stream towards multiple endpoints at the same time. The network is bloated unnecessary…
This should be adopted and properly implemented to be able stating it’s a good streaming solution.
Absolute rubbish! Roon has no sound at all! If your sound with a Roon device is bad it’s because the device you are using to render the stream is junk! Roon is essentially a cataloguing software program, it’s effect on sound quality, unless you use the DSP extensions, is exactly zero! Any sound file it handles comes out bit perfect, exactly how it was received. The ‘musical’ device in the chain is the DAC which you use.
(They say it’s not the most cost-effective solution; I say it is, if you don’t overpay for a Nucleus and use cheap endpoints like Pi machines.)
Bloating can happen if streaming ridiculously high resolutions. Stereo Red Book has a bitrate of 1.4Mbps, i.e. 0.14% of gigabit Ethernet bandwidth. If you sent 32bit samples (e.g. after DSP), that would grow to 0.28%. If your palace had 100 rooms, you’d use less than a third of bandwidth.
My emphasis. Many are not or they can’t be updated for new formats. All of this and more is explained in the documentation:
Not that it has much to do with sound quality. However, in many audiophile circles people try hard to offload computation from the streamer/DAC, so there’s that. Can’t have it all.
I do consider PS Audio within the best designers They have given excellent solution to the two problem I have comment. Others designers like Weiss take great care in their DACs to avoid those problems which are seriously studied as well by Linn engineers
The very good dacs and streamers manage correctly above mentioned problems and if you had such units , you do not need to get extra equipments.
But the I2S is just very recently introduced in the high end equipment and today’s getting more applied
By the way Measurements do not cover all aspects of sound quality
I am an engineer and although this fact I do previous affirmation. Just for your knowledge
Sound file it handles is bit perfect. Which isn’t? Are you referring into frequency or time domain issue as bit perfect doesn’t exist in one of these two
Sound quality is not well defined. Accuracy on the other hand is. We can measure differences between the expected and actual output with much greater detail than any ear can, and if those differences are not audible, the output is accurate. That’s all that matters.
Possibly, assuming that there was any real meaning intended to begin with.
It is sort of amusing though to watch people attempting to connect devices from different vendors using I2S “hdmi” cabling which tends not to work because there isn’t even standard wiring scheme…
Well, have dcs, aurender and hifi rose streaming gear that are roon ready. So their capabilities to process files is no less than intel cpu inside older nucleus. Any dsp, from my perspective, is not a philosophy for proper streaming
For me it is unnecessary bloating the network (easy traceable with kali preinstalled toolset) and the bandwidth of 1Gbps us not only available to roon passing through the router but to other network and media devices.
Having 10Gbps lan i want feel any issue but most customers with slower net would have too much unnecessary traffic.
So from architectural perspective big improvements are to be made instead adding few features no one uses…
These days it is a requirement for achieving state of the art sound quality. Although at least some streamers (not Rose, alas) are capable of doing it on device.
Is more than sufficient even for high resolution streaming to multiple devices. Everything you are proposing will not improve anything at all.
That you consider people who make DACs that are incapable of even vreproducing a sinewave explains a lot.
I’ve heard that they have hired their first engineer last year, so maybe they will produce something almost as good as a Sony boombox in a few years.
It’s 1980s technology. No piece of professional equipment uses it for anything external because it is the worst possible way to move bits between boxes.
Every bit that comes in comes out exactly the same. There are no inherent “distortions” of any kind since it’s exactly the same data. How difficult can that be to understand? If you don"t use the DSP there is no manipulation. It has ZERO effect on sound unless you opt to use the DSP where it also works absolutely transparenlyt aside from the desired effect (or not). If the sound is worse it’s due to an issue before or after the Nucleus. Simple as that! Anyone that implies differently is willingly lying or just doesn’t understand how the device and Roon works.
USB noise is a consequence of your server hardware ,Roon has NO control over that .It does however recommend a method of overcoming it using a split server and streamer as I mentioned above.
