Euphony OS w/Stylus player setup and issues thread

[ray-dude]

11 hours ago, austinpop said:

 

Stylus is a monolithic player that runs on the standalone machine that is attached to the DAC. As such, there is no such thing as dual PC Stylus.

 

StylusEP is a subset of the full Stylus. It is the playback piece, onto which has been added remote support via SlimProto, the Squeezebox protocol. This requires a SlimProto-supporting music server like LMS or Roon Core.

 

Now if you could persuade Euphony to take the server piece of Stylus and graft SlimProto on it, you would have a Stylus Server, which in conjunction with StylusEP would give you a distributed setup.

 

I’m happy with Roon, so feel no need to advocate for this.

 

I have been assuming that the skipping and freezing we hear going from Roon core to large buffer Squeezelite was related to Roon and how it interacts with SL (Roon has to jump through all sorts of hoops to do the multi-zone thing...I assumed this is what was messing up large buffer end points).  Do you have a sense of what they needed to do to get Roon to play nice with a large buffer end point?

 

Can't wait to hear the beast!

[ray-dude]

Interesting....in my "stuck in traffic, thinking of all the things I could experiment with but will never have time to actually do", I was contemplating hacking the Squeezelite code to present two end points to Roon: one large buffer one and one regular one (basically a control path with the small buffer end point, and a bulk transfer path with the large buffer end point).  

 

Roon Core could then stream to both, but the bulk transfer path (connected to the DAC) would sync with the control path directly, and Roon could happily stay in sync with the control path (basically a proxy to maintain sync). 

 

Plan B would have been to have the control path end point process run on the server itself, and have the two end point processes stay in sync over the network (to further minimize network traffic to the actual end point connected to the DAC).

 

Of course, the best answer would have been for Roon to actually fix the sync issues to large buffered end points, but alas, there dragons dwell.

 

Thankfully for all of us, Željko is actually doing the work to make it work!  I am very excited to hear the results and support his efforts!

 

Thanks to you and Roy for the advocacy and pushing things forward!