Is there Life After Roon?

Sovrce Gear
21

Decided to try LMS/Squeezebox today, got iPeng going on my iPhone, and the Material Skin for my desktop to control everything. Then since both streaming server services (Roon and LMS) were running on my m1 mac mini I did some A/B testing to see if I’d hear a difference. Keep in mind I only tried this test with my ZMF Auteur headphones since I’ve moved some things around and I’m waiting on a Zynsonix speaker amp box to come before I use my replacement Susvaras off my aegirs again.

Setup is M1 Mac Mini Roon Core/LMS > Network Switch > optical isolation with 2 ethernet to optical adapters > RPI 4 with RopieeeXL > SU6 > Amber 3 > Jotunheim 2

When doing this test I never touched the volume control and I used God of War by Bear McCreary as my first test track. The first difference I noticed is that LMS/Squeeze does a better job with sound staging and placement than Roon. Also as the track gets busier I noticed that Roon tends to sound flatter than LMS, with LMS doing a better job keeping everything where it “should” be in the track. Roon on other hand, due to its flatter sound can make certain instruments come forward more, which some may or may not like. But overall it seems that LMS handles the business of this track better than Roon does as far as sound quality/resolution goes.

The next track I tested was Rolled Up by Benny Sings. This one had its fair share of differences as well, but overall LMS just sounded more “correct” to me if that makes any sense. While Roon does do a good job with placement on this song, it seems to exaggerate certain things in the mix (like the sound of the birds at the beginning of the song), while this may be nice for picking things up in the mix, the placement of things (like the birds) just sounds more correct to me on LMS. Another thing I noticed was the voices seem to have slightly less definition to them on Roon vs on LMS. Bass also has slightly better resolution on LMS, but Roon bass does sound a little more forward than LMS.

As far as the interface goes, once I got iPeng customized to my liking on my iPhone I found myself enjoying it just as much as Roon’s interface (for local music). And with the iPeng playback addon, I could use it to play music through my iPhone just like Roon ARC (as long as I’m on my home network lol). So, for my use case so far (A single server to a single endpoint and sometimes my phone) I can see myself switching entirely over to LMS/Squeeze, and saving some money in the process. The downsides to LMS though are ease of use/installation, missing lyrics (album/artist info isn’t missing though with the “Music and Artist Information” plugin), worse integration with Qobuz (while there is a plugin to use Qobuz via LMS/Squeeze, the interface for it sucks compared to Roon.), Easy EQ settings for specific headphones, and no Roon radio where it will just start playing suggested music once you get to the end of a playlist or album. But despite all of those downsides, what LMS does well may just be enough for me to switch since Local Music playback is my primary use case anyways and I don’t mess with EQ.

I know I should probably take more time to test these two back and forth, but I just wanted to leave my first impressions here. So far though I am fairly impressed with LMS/Squeeze for local network music playback and if I switch over to it completely I’ll probably get rid of Qobuz as well (and use Spotify for music discovery), then I’d be saving money on two subscriptions while still having the functionality and sound quality that I care the most about.

Hope this helps anyone on the fence/considering trying out LMS/Squeeze coming from Roon.

8 Likes
22

Hell yeah. Thank you for the write up. My experience is the most noticeable on how it handles my bass. Roon really muddies it up. I get better dynamics and detail handling with squeeze as well

4 Likes
23

Ok that is actually a cool feature.

4 Likes
24

I actually find a lot of “new” music and artists this way, let Roon radio pick music based on song choices but pulls music from Qobuz or Tidal if you’ve got that.

5 Likes
25

Yeah… I’m probably making a pass on roon after my 1yr sub ends. I already knew my streamer app sounded better than roon but after taking a break from it and having used roon again for a few weeks, there was this constant sibilance that kept pestering me during my sessions. I thought it was my power cable at first since that was the only physical thing that changed in the system, so I thought there was some synergy mismatch. But given how much I like this power cable, ain’t no way! It was dirty dirty roon :roll_eyes:

Edit: to clarify, this was local files (SSD) via streamer software vs local files streamed via roon

4 Likes
26

I’ve got to get around to doing some Roon vs. Direct streamer playback to see if I hear anything appreciable. I’ll make that one of my 2023 endeavors.

4 Likes
27

I’ve spent the last 6 months doing one month on, one month off and comparing. After this last longer session using Squeeze, I can’t go back to Roon, especially with my new speakers.

1 Like
28

What do you think is the factor and from an architecture standpoint is is the same with Roon versus Squeeze?

29

It’s interesting, your not the only one who’s reported it.

On the NS1 I can’t hear the difference between direct streamed files, and files served via Roon.
And as far as I can determine the Qobuz files should work in a similar way to Squeeze. So it might vary based on the streamer in question.
The Roon server is undoubtedly higher load than some alternatives, but that shouldn’t affect anything in the streamer.

UPnP/DLNA , Squeeze and Qobuz, basically just provide the streamer a download link, the streamer just “downloads” the bits, decompresses the file (if it needs to) and plays the bits.

I do dislike that Roon has a proprietary undocumented protocol.
RAAT doesn’t work like that, and I know it has logic to synchronize playback across multiple devices, but it’s hard to believe that would be significantly more CPU heavy, though who knows.
Neither can have any errors introduced, they are both over TCP/IP as far as I can tell.

Reminds me a bit too much to the Wav vs Flac debates, where again all it really could be was CPU load.

4 Likes
30

Out of interest, have you tried using Roon’s Squeeze support?

1 Like
31

Yes I have, so I guess that has made it 3 combinations that I was juggling with. Fortunately, Antipodes does a good job at allowing you to specify your server program and your player program independently.

Squeeze/squeeze was the best > Roon/Squeeze > Roon/Roon

1 Like
32

I’m at a total loss as to how Squeeze/Squeeze can be any different to Roon/Squeeze.

The protocol is entirely trivial, the player asks for a URL to the track in question, and the server provides a link, after that all the server does is serve the file, that should be entirely transparent to the streamer.

One of those mysteries I guess.

33

That sounds a lot like CPU load, similar to what @NickMimi found when he was testing his his Innuous. Roon server with its rendering adds to the CPU requirements seem to be higher than bespoke software.

3 Likes
34

I’m assuming the server and renderer aren’t on the same box though in @Veritas case.
The Roon Server is certainly not light weight, and I could certainly imagine that load making a difference, if both the server and renderer were on the same machine.
But with a separate streamer, especially if both are using the Squeeze protocol there ought to be no real difference.

35

You’re right there, in the Innuous that was the case, one device doing both. It would be interesting to see. I also wonder if it’s less of an issue on post Roon devices are taking the additional CPU resources into consideration. More CPU utilization, worse performance but it the case of Veritas maybe it’s a borderline issue?

I’m not going to set up a Squeeze environment but I will simply compare a playlist from Roon and use the same files but on a direct connected USB drive on my dCS bridge played with the Mosaic software to see what differences I can pick out if any.

1 Like
36

Just fwiw, my drive is connected to a separate power supply via powered hub so it won’t draw anything from the streamer. How much of a difference it made idk since I haven’t tried it without, but just to share…

37

I’ll see if I have the ability to have a network path for the music, I don’t think so though. I can still do an apples to apples comparison by playing the same file from a Qobuz playlist.

38

You might have to set up a DLNA Media server for it, the NS1 software needed that.
It’s a built in Windows feature, but is disabled by default, and you’ll probably have to add the Media path to the server after you enable it.

39

Yes, if I had to guess, it is mainly about CPU usage.

Fortunately, my UI has a CPU and Temp gauges. After staring at my gauges after 3 songs, the top picture is of Roon that fluctuated CPU usage between 5-36% and Squeeze on the bottom that only went between 0-2% lol? I don’t know how that’s possible. @Polygonhell @dB_Cooper

image
image866×1636 56.6 KB

image
image866×1720 57.6 KB

3 Likes
40

@dB_Cooper if possible, do a comparison of Roon sound quality with your Core on both sides of your mesh WiFi node.

Internet → Core → mesh WiFi → endpoint
vs
Internet → mesh Wifi → Core → endpoint

It would be very interesting if you hear any difference taking the WiFi/mesh out of the RAAT packet path from Core to Endpoint.

Should be no difference, but given RAAT is a black box, it’ll be an interesting data point to confirm with your highly resolving setup.

2 Likes