With the current build (b1470) i am experiencing pitch variations while grouping Roon Ready Zones.
Describe your network setup
Server and endpoints wired with a switch between them.
And to be clear, i am pretty used to this scenario, and having never experienced this before i figured it was worth reporting.
To flesh out the description above, i listen to music using the Roon Ready functionality of the Moon 280Dbthrough it’s Mind module. Adding the sMS-200 to the group is easy, but when i switch to listen to that transport (through a Denafrips DAC) it is obvio8s that a pitch shift is taking place. While listening for a few seconds i notice the pitch shifting back to normal, and settling. When switching between these two sources i sometime experience the shift, and sometimes it is gone.
Adding to the confusion is the fact that if i rather play to the 280D directly from server over USB, and grouping the sMS-200 Ultra Neo i DO NOT experience the pitch shift.
I just tried switching back to the 280D on it’s Roon Ready input, and grouping the sMS-200 Ultra Neo to it, and darned right, after 20-25’ish seconds i get a pitch shift (upwards in frequency) for some 20 seconds, which then returns back a notch.
I realise this ought to be related to the Clock Master Priority, and which Zone i initally selected, then added zones to.
Thank you for your post. Around the timestamp you shared, the SOtM sMS-200Ultra Neo Zone is upsampling the 44.1KHz FLAC file to 705600Hz. It’s also performing this upsampling when grouped. There’s a severe server-side resync delay.
What precise DSP are you using in this chain?
Please share a screenshot of the Device Setup dialog window for both of these Zones, as well as a screenshot of your Signal Path when playing to the SOtM sMS-200 Zone itself. Thanks!
This seemed only to happen when the Moon280D was in Roon Ready-mode. And i do run the sMS-200 underclocked, as it sounds really nice that way.
I shut down and changed the USB cables on the SOtM sMS-200 Ultra Neo and it’s txUSB Ultra yesterday also, and since then i cannot recreate?
Right now (7:07 CET +/-1 minute) i had the SOtM glitch totally while playing it as the secondary zone. Was playing it as the secondary zone, upsampling to max PCM (even multiplier) as usual.
I am running the Denafrips Ares II in NOS mode on the end of it.
And, after glitching out i ungrouped the zones and played same track to the SOtM single handedly, and it sounds absolutely beautiful! No issues while not grouped!
The SOtM just freaked out as the Main zone with the 280D Mind2 grouped! There are also some dropouts while some songs start to play from the playlist, particularly if they are in higher rez (only in grouped mode).
7:13 CET for this occurence.
Trying the same playlist (redbook->24/192 PCM) while having the SoTM grouped with the 280D on its USB input directly from server does NOT cause pitch variations or droputs/glitches. (Both Zones upsampling through MUSE to Max PCM-even multiplier)
Server idling away:
And to be honest, i don’t NEED to group the offending zones, it’s just a curiosity that i haven’t seen before. And, since my initial post i have also upgraded to Roon b1480 on all devices concerned.
One more datapoint:
I changed to the SOtM sMS-200 as the primary Zone and then added a Ropieee based Pi3B feeding the Moon 280D over coaxial, and the issue arose once again. It seems to rear it’s ugly head when the source files are high rez mainly.
This issue is present whenever i group two Roon Ready zones where i upsample to one of them. Updated to the lates build a couple of days back.
This is new, i have been running this config back and forth and never seen this behavior before.
Would it be possible for you to record a short audio/video clip that shows the pitch variation? I’m not sure how much logs would help us, I only notice one dropout instance in your timestamp, and there doesn’t seem to be much information as to why. Also, if you are able let us know the track affected alongside with the timestamp that can help us pinpoint the issue much better than just the timestamp. You can upload the clip to the below link:
Hi Noris,
i have something even better for you guys!
I have figured this out, i believe.
The offending Zone in this case has mainly been the SOtM sMS-200 Ultra Neo, which has displayed these pitch variations and drop outs.
While this may still be related to a change in the RAAT-protocol, or it’s bandwidth requirements, i have found a workaround.
For sound quality reasons i have my sMS-200 set to these settings:
The culprit seems to be the 100Base-T setting for interface speed.
It seems the traffic load, to keep more than one Zone in sync, is exhausting the capacity of the SOtM with this setting.
When changed to 1000Base-T, the issues are gone!
And here’s the deal, i have previously had issues with extreme high res material through the SOtM when running in 100Base-T mode, but they haven’t been seen for many moons…
So it seems this might be a coincidence, with the 100Base-T setting, grouping Zones with the associated traffic overhead, and high res upsampling.
And whats more strange is that, when switching back to the 100Base-T setting i can’t replicate this anymore.
The issue came came back!
I recorded a clip, showing the effect, keep eyes on the display of the Pre, i switch between the Ares II (fed by the SOtM) and the Moon 280D.
But i also set a different priority scheme in the SOtM settings, “Network Throughput”, and after that the issue does not seem to surface again. It does sound slightly “less good” but for comparisons it will do…
Since some time has passed, I wanted to check in and see if your above statements remain true? Were you able to bypass the issue by decreasing the traffic load on the SOtM?
Hi Ben,
yes, this is still true. Something has changed since a build close before b1470 which affects grouped zones, where one of these are upsampling to more than 24/192.
It point towards the underclocked SOtM which causes the pitch variations when grouped. If the SOtM is the primary zone in a group, the pitch variations affects the secondary zone, like the Moon 280D Mind2 with similar effect.
I can adress the issue by:
Not grouping zones, SOtM sounds magnificent on its own
lower the upsampling rate to the upsampled zone
upsample to a similar level to the secondary zone also
I can also set the SOtM to 1Gbps network speed
I can almost rid (not fully) the issue by setting “Network throughput” as optimization scheme in SOtM
So, while the issue remains it can be managed…
Unless you’d like to offer some insight as to why things have changed?
