It’s a bit marmite admittedly and I don’t have a local library or playlists from streaming services that fall in the metadata challenged category, such as Classical as the obvious example.
Given all these applications under consideration are available to trial for a useful period of time that should be sufficient to evaluate if it suits your personal needs or not and if the cost is offset by the benefits it brings.
Much like yourself, I can only relay my own findings and experiences and in that sense I’ve found Roon beneficial enough to warrant using it regularly for a number of years.
I don’t think the implication on Roon’s part is that a Roon Core won’t work and probably well over Wi-Fi it is as you say more down to the uncertainty of a Wi-Fi connection environmentally first and foremost that is the basis of Roon’s advise.
Also bare in mind that Wi-Fi and a Wired Ethernet connection work very differently from a data control and transport perspective, making an assessment on the basis of latency and throughput isn’t really exposing all the elements at play in terms of delivery of audio streams to your streamer/DAC.
No harm trying and seeing if it works for you in your environment of course, if you’re happy with what it does, all good!
In my case my Roon Core on a Mac Mini is hard wired to my LAN switch and my NDX2 connects over Wi-Fi through a Wired Access Point so it’s a mix of both.
I have attempted to read some but still have to find a fact based explanation of it.
I am the last one to convince to use good cabling, especially for analogue signals. And I can get how an ethernet cable plugged into the streamer itself can transfer RFI noise into the streamer components doing the encoding and amplification, so that also makes sense to me.
But how a signal made up of 0s and 1s would be any different if sent via copper of via wifi prior to analogue conversion still goes beyond what I understand about the technology. As long as the bandwidth is enough to let it through the result should be no different regardless of the transport. If Wifi cannot carry a PCM stream reliably then software where the difference between a 0 and 1 is crucial should not be possible with current tech.
Bit streams also carry error correction hashes so even if a bit is lost due to bad transmission the decorder should be able to reconstruct the original data. So I can accept that bad transmission would lead to gaps in the music if badly affected but I have not seen an explanation of how it could lead to the analogue signal decoded from the bitstream to being significantly changed instead of just being missing.
If there are some references about this, even mostly technical, I would love to have a look at it.
From where I am sitting I can just not see how the method of transport is not black box. There is either an intact bitstream entering the DAC or there isn’t . If there isn’t I would assume that the DAC would create choppy analogue from it, not bad mp3 quality analogue, for example. Which makes me think that if the music is not displaying judder all is fine with the point to point transmission.
By the way I am not trying to be “that guy” who just says it is all wrong. Am really looking for the physical reason this would work.
Like with audio cables, reading about the capacitance changes and such explained it hehe
A curious mind is both a blessing and a burden!
A lot of this comes down to the way that a Wi-Fi radio handles resource allocation to connected Stations/Clients. The radio in the Access Point works on a time slot basis with each carrier allocating a time slot to a client based on a first in first out basis.
The more advanced the Wi-Fi radio the more carriers and channels are available concurrently allowing for support of things like MIMO (Many In Many Out) in particular Multi User MIMO or MU-MIMO.
You then also have features such as Beam Steering and Beam Forming to focus the available RF energy towards a specific client to reduce noise and signal attenuation.
The challenge here is that the packets are transmitted and received based on the time slots available and are allocated on a first come first served basis, so if your streamer is competing with available time slots with multiple clients all utilising the same Access Point they then sit in a queue and at that point you are then relying in part on the ability of the client to mitigate data delivery delays with effective buffering. There are traffic management features such as WMM (Wireless Multi Media) you can enable to mark the traffic and to give it a chance to increase it’s priority and effectively jump the queue. In order for that to work the client has to support it and to flag the requirement in the packet header and allow the Access Point to coordinate the delivery based on the categorisation of the packet type.
Typically this isn’t enabled or provisioned and as such a Wi-Fi client is treated as best effort and will receive data based on fair use and airtime slot allocation which is shared between all connected clients which in many cases will included multiple clients like phones, tablets, laptops, cameras and so forth.
Unless you put your streamer on a dedicated Wi-Fi link it will always be contended to a degree.
A wired Ethernet connection on the other hand is primarily limited by the switching capacity of the switch or router and will be forwarded at line rate, which translates roughly to a limit of about 928 Mbit/s per 1 Gbit/s port and up to the Packets Per Second switching capacity of the switch itself.
1 Like
It is mot the digital data that are affected, but electrical (RF) noise superimposed, conveyed to the DAC, causing some form of modulation effects during the da conversion process. Ground plane modulation apparently can have an audible effect even at exceedingly low levels. Different cables or switches may filter the noise in different ways - with if course some people possibly preferring a variant of the technically incorrect signal. (Apparently one possible effect is percerption of increased “brightness”.
Eanwhile Simon in Suffolkhas given good explanations of wifi asoects
With my Hugo 2 fed by a Mac mini audirvana was the best sounding piece of software I could find
Now I know why folks are saying that why Audirvana sounds better than Roon! Make sure that your end-point is configured as fixed-volume, use exclusive mode (if available) under Roon. Voila, suddenly the SQ from Roon transforms from from dark to light.
The technical bits makes a lot of sense, thank you. A lot of this I knew but some is so darkly technical they don’t tell us developers that so we don’t get uppity 
BUT this still re-inforces, in my mind anyways, the logic that it would lead to judder in the music, not worse quality actual music wave forms. If the wifi stops sending data the music should stop and if you are not noticing that does that not mean the wifi is capable enough?
I mean it cannot actually drop resolution in the bit stream can it?
Oh I would not argue using a good cable connecting the actual streamer. That makes perfect sense. But that should be enough, right? I mean one cannot be responsible for all hardware used in the entire path from Tidal’s server to your DAC. So if I use a good cable to connect the core there could be anything in the wall or even outside the house.
You know I tried it and maybe it is mental only but I do notice a slight improvement. As the Uniti Star has electronic volume control I would expect the stream itself to remain unchanged. But hey this seems to actually DO something. Thanks.
If you have more than one Roon end point in your home network (and I have 4), then Roon does offer the following feature.
Roon is aware of the capabilities of each of my ‘streaming devices’ & DACs (Naim Muso Qb, Chromecast Audio into Marantz NA7004, Sonore microRendu into Mytek Brooklyn+ and Linn Klimax DS/1).
So, if I were to choose to play a 24bit/192mHz file from my NAS, Roon is aware of which of my systems can handle this resolution. In the event that one of my systems does not support the format, then Roon automatically transcodes to the highest possible PCM resolution that the device does support.
Similarly with DSD or MQA content, Roon is aware of which of my systems supports DSD or MQA and if it does then it automatically feeds a bit perfect file directly to the system. If one of my system does not support the format (eg my Linn Klimax DS/1) , then Roon automatically transcodes the DSD or MQA file into the highest PCM format supported by the Klimax (having first carried out the first MQA unfold in the case of an MQA source, should I choose (and I do) to configure Roon in this way).
The above process is seamless and means that I do not have to worry about format support when I play music on any of my systems. Roon simply automatically provides me with the highest possible resolution on each of my systems.
Just one of the many features of Roon that I find invaluable.
2 Likes
So here I am back to admit sheepishly that I just wired in my Roon Core with an Ethernet cable.
It seems that a weird noise I have been having on Roon before and after Tidal/Qobuz playback may have been related to this after all.
Let’s see how it goes. If the noise stays away you all were right LOL
Posted in wrong thread, perchance?
Nope, was intentional. Above there was a discussion about Ethernet vs Wifi as part of the quality discussion here. Thought it good to come back here to say I changed my mind.
I do make mention of an issue that is addressed in another thread, however.
Thanks for this!
I had a quick scan and what I see makes sense.
Only, how would it account for more noise coming from a digital signal transported via WiFi vs Ethernet cable? Even googling on that specific bit does not really yield anything.
I think the same hehe.
Thing is my question is about whether or not the Roon Core is connected via Wifi or Ethernet. My Star is connected via ethernet Cable.
Yeah I have seen the threads but there are so many replies on it and the signal to noise ratio of facts over opinions is so high that I have given up on it.
With my latest personal experiences I am suspecting that the comments are right, I can just not think of how signal noise other than dropouts or stuttering can enter into a digital stream just cause it is sent wirelessly.
Thanks for taking the time to write this summary 
That makes sense but here is the still missing piece of the puzzle for me:
You have Roon Core => Network => Streamer . If the signal from the Roon Core to the network is transfered through the air instead of through a cable then I would expect LESS of a chance for these errant electrical signals to be able to propagate all the way to the streamer, and therefore, the DAC.
In the comments above it was explained very nicely how WiFi works through. Talking about packet scheduling and such. In this I can fully accept packet loss, or delay. But in my mind that would mean there is no digital data to decode, therefore resulting in no analogue signal. Not in degraded analogue signal. This signal would then pick up any noise from the reception radio through the cable to the DAC, but the path that went through the air cannot incur electrical noise.
So I guess to summarize my question is then rather, how would Wifi “noise” manifest itself? Gaps of silence in the music? Or is it only a case that the gaps become so small that it is perceived as as lossy signal.
If you do go down the Roon route and you are handy with building PC’s etc, one thing to consider would be a Roon Nuc.
The Roon Nucleus is quite expensive (£1.5k - £2.5k) but, the good people at Roon provide the operating system from the Nucleus for free. It’s called Roon Rock.
You can buy an Intel Nuc i5 Gen 7 for around £200 from the usual auction site… add some Ram, PCIe m.2 SSD (128gb) for the OS plus an Internal SATA SSD for music and a fanless case and you can build a very good Roon Nucleus for around £400. It can sit silently under your music system.
If that’s something of interest, google Roon Rock Installation Guide and you will find a ton of stuff at Roon that tells you how to do it.
Took me about 20-30 mins to transfer the bits from the Intel Nuc case to the fanless one and install the OS etc.