I use a Raspberry Pi with a USB sound card attached to a DAC (just a hi-fi CD recorder) as a Pulseaudio sink, and an NFS server for the media. Then any old Linux laptop can play audio through the main stereo over the network.
thanks for the comment - could you give more details? i guess you're using ethernet, but that could be replaced with a usb wifi dongle? what interface are you using from the laptop? just vnc?
Hi.
My Pi is hooked up to a powerline network thing just through the RJ45 on the Pi, but I suppose you could use a wifi dongle if you want. On the laptop, you just set the pulse sink to be the Pi's pulseaudio (need avahi running on both to see the Pi in the audio preferences pane on the laptop). You can play whatever you like on the laptop itself and the audio is routed through the Pi, so there's no need to log into the Pi itself once it's set up.
For all of the slagging Pulseaudio gets, built in network transparency is very cool. The client does all of the resampling and rate conversion, so you can ramp up the conversion quality to max on your relatively powerful laptop and the Pi just shunts the bits onto the sound card.
What I'm getting from this is that Logitech has cancelled the Sqeezebox line, which is too bad, they were nice devices, though admittedly once Logitech took over, I stopped caring as much (they were a local firm in Mountain View across the highway from Google).
The streaming software was pretty good stuff at the time, and open source, hosted here:
I apologize for going meta on this, but I literally cannot make heads or tails of his writing, which I'm pretty sure is about something that I'm interested. With a name like "Andrew Cooke" I'm inclined to think English is his first language, but...wow. Is this a style thing? Can someone summarize?
logitech made hardware and software that allowed you to stream music from a linux computer to elsewhere in the house. they have stopped making the hardware and the software appears to be "decaying".
forking and fixing the software sounds like the obvious short-term solution, but it's not so easy - it's fairly opaque perl, and despite various interested people and an open bug report, nothing has happened so far. also, it's not a good long-term solution, since the hardware will fail eventually, too (the hardware was pretty dumb; a lot is done by the software/firmware).
the best alternative i have found is the audio engine d2. that works at a lower level than the existing logitech devices - it looks like a USB sound card to the computer. that means that it works with a variety of different music players (avoiding the trap i fell into with logitech's software; on the other hand, that also makes it less flexible, so the kind of setup buro9 describes above is harder).
does that help?
one piece of context that may be missing is exactly what this hardware does / did. typically in a "computer audiophile" setup music starts in a digital file on a disk, is sent somewhere, converted from digitial to analogue, amplified, and fed to a speaker. what logitech did was the "send somewhere" part (plus, optionally, conversion to analogue). so you could keep your music in one place, but listen to things elsewhere (or in multiple places - maybe your main speakers in the living room; a headphone amp in the bedroom; monitors in the office). effectively it was a "wireless digital cable" (plus software router).
it was also useful that the logitech devices could work with a purely digital signal. that let people use other (typically more expensive, better sounding) hardware to do the digital to analogue conversion.
I'm assuming that since you have the ability to fork the software that you know the API for the device. Great, then it can be rewritten in a language other than obfuscated Perl.
Beyond that, if it's as dumb as you say, how about a Kickstarter for a new hardware output endpoint that can respond to this new, rewritten software? Perhaps an Arduino and DAC can be hooked up satisfactorily in the meantime.
I mean, he doesn't seem to give too strong of a background of the situation, but I can guess that Logitech is dropping what used to be a high quality, budget audio device that linux hackers enjoyed.
I suspect it's an audience issue. He's assuming whoever's reading it knows the background. I don't, so I can empathize with your confusion
I use a playGo usb (playgo.us). It fills a similar niche as the Audioengine product mentioned in the OP's link. Not a ton of features on top of 24/96 streaming, but to it's credit, it just works when you plug it in.
Too bad, thanks for the heads up. They were great gizmos, and the original company was cool. I had (and still use) all generations of the squeezeboxes. They have been quite reliable.
I'm in the same boat as the author, and am currently just waiting to see what might emerge as a solution.
My existing solution involves a Squeezebox Transporter [1] connected to a QNAP TurboStation 669 Pro [2] which in turn runs SSOTS [3] and the PERL source of Squeezecenter [4].
It's a really nice solution, allowing me to remotely access and play my music from anywhere in the world.
In fact, I've had what I consider to be a personal version of Spotify for the best part of a decade, whilst at the same time enjoying a superlative system at home.
I do have concerns about what will happen now that Logitech has abandoned the range. At the moment everything is working fine.
A remote failed on me a while ago (my own fault) but I was able to use Net::UDAP [5] to work around it and an Android app on the Nexus 7 to act as a convenient remote.
But... what happens when the QNAP requires an update that shifts the version of PERL? What happens when the libraries for Squeezecenter are no longer supported and later versions don't work with Squeezecenter?
It's all fine saying that we could just use a hardware DAC and wireless streaming. But that doesn't give remote access, an indexed and searchable library, local remotes, etc... and suddenly you need to solve all of these things from scratch again whilst still keeping yourself out of walled gardens.
As for other solutions, I have 91,000 FLAC files. Some of which are 24bit (rips of the vinyl I own). I also use MusicIP to create dynamic mixes of music along a theme... like Apple Genius stuff but far better and more interesting to listen to.
I've always felt that the problem Squeezebox had was that the quality was too high, it worked too well. In over a decade, the only thing that has failed has been the screen on the remote, and that was my own fault for kicking it across the room (accidentally I might add).
How do you make a sustainable revenue stream when your key customers never need to buy from you again?
I'll be watching music/audio/tech forums keenly for a similarly capable system to emerge.
For the time being, I think there are enough Squeezebox users out there that will be able to fix the Perl problems.
After all, the Squeezebox server is Open Source, and you're not dependent on any online services from Logitech to make it work.
How does the logitech system work? it just flings files over a network??
Can it do multiroom? (same music many places at once) if so, how? (mainly a question re: non-insignificant data multicast and syncing with low latency, i guess?)
yes, it could do multiroom (although i never tried it - it allowed you to synch streams for multiple devices). no idea how it works at a low level - similar to airplay, i would guess.
The only reason why I don't switch to linux is that my external soundcard (Roland Quad-Capture) doesn't have linux drivers. Even just playback would suit me...
Dirt cheap and works really well.