I have spent many hundreds of hours getting all my music, mainly classical, sorted out and playable in various ways using LMS. And then I thought, what if my Touch fails? No other system gives me the flexibility of LMS and associated plugins, such as:
- many of my own menus, the way I want them, such as Works by Composer, Albums by Conductor, and so on
- random play of sequential sets of tracks (i.e. classical works)
- track play statistics
- and so on.
So how can I replace my Touch?
I could try to put a hardware solution together, using a Raspberry Pi, Wandboard or whatever and run the brilliant squeezelite on it . But do I want to worry again about async USB vs coax, vs optical, power supplies, a whole new operating system to learn and manage and so on? No I don't.
I have 2 network connected amplifiers ('players', or 'renderers') that accept flac and wav format audio streams over http (this is not unusual with players that claim to be 'DLNA/UPnP compliant', although it may not be universal). Why don't I just send the data over my wired ethernet network direct to them? The LMS DLNA plugin is very limited in capability. What I want to achieve is the full flexibility of LMS and the web GUI control system, without an SB device.
Briefly, my solution is to:
- use a UpnP DLNA server to declare a playlist consisting of an http adddress and port on my music server (that is the only job the server needs to perform once the player is playing)
- run squeezelite on my server, behaving just like any other LMS device, and pipe the output to an http stream in a form my players understand
- on my player, select the playlist item broadcast by my DLNA server
- use the LMS web page in the normal way (on a laptop/tablet/mobile) to control what I am listening to (the player does not show any information about what is playing).
If this is of interest, your server is on on linux, and you don't care about multi-room synchronisation, see my next 2 posts. The solution isn't perfect yet, and it may even be impossible to make it so, but it does work with some limitations, and I'm hoping some other people might be able to think of ways of improving it.
Many of you will think I'm mad even trying to do this without hardware, but I'm convinced a solution along these lines should be possible, without no loss in sound quality, whilst retaining the same flexibility and ease of use.
- many of my own menus, the way I want them, such as Works by Composer, Albums by Conductor, and so on
- random play of sequential sets of tracks (i.e. classical works)
- track play statistics
- and so on.
So how can I replace my Touch?
I could try to put a hardware solution together, using a Raspberry Pi, Wandboard or whatever and run the brilliant squeezelite on it . But do I want to worry again about async USB vs coax, vs optical, power supplies, a whole new operating system to learn and manage and so on? No I don't.
I have 2 network connected amplifiers ('players', or 'renderers') that accept flac and wav format audio streams over http (this is not unusual with players that claim to be 'DLNA/UPnP compliant', although it may not be universal). Why don't I just send the data over my wired ethernet network direct to them? The LMS DLNA plugin is very limited in capability. What I want to achieve is the full flexibility of LMS and the web GUI control system, without an SB device.
Briefly, my solution is to:
- use a UpnP DLNA server to declare a playlist consisting of an http adddress and port on my music server (that is the only job the server needs to perform once the player is playing)
- run squeezelite on my server, behaving just like any other LMS device, and pipe the output to an http stream in a form my players understand
- on my player, select the playlist item broadcast by my DLNA server
- use the LMS web page in the normal way (on a laptop/tablet/mobile) to control what I am listening to (the player does not show any information about what is playing).
If this is of interest, your server is on on linux, and you don't care about multi-room synchronisation, see my next 2 posts. The solution isn't perfect yet, and it may even be impossible to make it so, but it does work with some limitations, and I'm hoping some other people might be able to think of ways of improving it.
Many of you will think I'm mad even trying to do this without hardware, but I'm convinced a solution along these lines should be possible, without no loss in sound quality, whilst retaining the same flexibility and ease of use.