Developping a stereo system replacing spark + dialog

[BlackMamba]

Thanks Streamy

not "too extensive" but complete. I understand that Dialog is ok and it's then I agree it's probably better to go throught it than worth developping a specific element

However, I have to questions :

1- for how long have you been using your system ? I had bad experience in time with PLC.
2- volume control looks to be the bad point in what you say. Changing singnal (known as software volume control) is modifying the coding probably not the best way when you try to get it full digital up to the phantom. I didn't get how you manage it. Through spark ? Or is ther another way to manage phantom volume ?

Best regards

to 1: I got my first Phantoms in March 2015 earlier that Devialet was able to deliver. I got it from the french store Colette. So, I went through the entire history of updates. PLC worked for me, once I put all systems on the same power line branch and removed all surge protection hardware that kills PLC signals. However, I must say that I live in my own house and have no neighbors that use another PLC on the same power line. I run a second PLC, though, that comes with the IPTV set and the two PLC work peacefully on the same power line. So, the Phantom/Dialog system has a certain robustness. Along the updates, about 2 years ago Devialet dramatically improve reliability of buffering on the Phantoms. This was maybe the game changer. You can now see in Spark how the buffer is fast filled and it presumably already starts buffering the next track in play queue as well. Since Spark has gapless play buffering works really well. But also streaming from the Phantoms optical inputs is rather low latency and in my case 100% dropout free! Much better than all Airplay before 2.0. However, AP2 is not yet implemented and I do not know how latency of it is. AP1 had acceptable latency when using screen mirroring but else the latency was fixed to 2 seconds, which is too much for acceptable UX.

to 2: I’m not changing the signal but indeed using some code from the github that @ogs by chance in his post indicated ( https://github.com/da2001/phantom-bridge ). This code allows to change the volume of the Phantoms through a webapp like the ‘Roon Web Controller’ where I have integrated a volume slider for changing the Phantoms volume. It sends commands through an upnp port to the Dialog, same as it would come from Spark. So no change of the signal different than what the Phantoms anyway does internally. As I use Roon this Roon extension is convenient and works independent of operating system and device. But you have to install a server app, e.g. on the RPi, like I do it. I run it on the same RPi that I use as Roon RAAT endpoint. You could run it on the volumio RPi endpoint. How and if you can extend or modify volumio I have no clue. Maybe you could use Moode where you can get the source code to tweak.

Thank for this detailed answers. I am not sure of the exact "how to" for implementation but I understand your explanations. It will probably take me some time but I now know that a RPI can feed, fully digital, a system of 2 phantoms + dialog  and that the volume can be controlled almost natively (hardware). Next will be to go for real .

Last ? To go out of the RPI, is it better to use the HDMI (with a HDMI to TOSLINK box) or a HIFIBERRY DIGI+ standard version ? 

Thanks a lot for all your inputs