soundcheck wrote: 
> 
> This might get improved if you run your input buffer of squeezelite like
> a full-file buffer (32MB -- what is the maximum?? or similar). (Triode
> can you confirm or explain please -- if the file is fully buffered what
> will happen to the network traffic?  And what happens to gapless
> playback?)
> 
You can increase the pre decode and post decode buffers with the -b
option.  Suggest increasing the pre-decode (stream) buffer to large
values and see what happens.  It should mean you get all the network
traffic at the start of the audio stream and then only traffic related
to running the supervision protocol during the track.  Note however that
doing this essentially creates more network traffic at the start of the
track, to get less during the track.  It will always buffer on a per
track basis and hence should maintain gapless playback.  If you are
running the server and client on the same device and localfile playback
is enabled on the server then it will read direct from disk at the time
it would stream otherwsie.

CPU load is intended to be low with PCM and as low as I could get with
flac on squeezelite.  I think its reasonably minimal.

> 
> IMO the RPi  needs a serious HW upgrade respectively redesign. IMO 35$
> are much too expensive for such a poor performance and design. 
> 
I think the PI is excellent as a cheap computer to teach kids about
computing.  It's cheap USB implementation just isnt up to what we want -
I agree other devices are better for us...  The cubox-i looks
interesting, and uses the same cpu soc family we use (Wandboard)- which
was selected as we know the usb performance will be good


------------------------------------------------------------------------
Triode's Profile: http://forums.slimdevices.com/member.php?userid=17
View this thread: http://forums.slimdevices.com/showthread.php?t=97046

_______________________________________________
unix mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/unix

Reply via email to