Re: [RDD] transmitter / silence / fallback

2015-01-29 Thread Robert Jeffares


On 30/01/15 08:23, Jim Stewart wrote:

I need to find one that will keep retrying

I have found the Barix exstreamer 100 solves the problem.

Tried VLC with the maintain connection option which works if the receive 
end fails but fails on a send end fault.


We used a browser player which was more reliable.

Barix is a bit tichy about ogg and sometimes aac but 100% good on mp3.

You need a good send setup, upload speed needs to be adequate.

Barix has built in fallback to SSD card player.

We have 2 source ip's with alternate service providers, to cover the 
occasional outage,  and generally receive service providers are up [to 
speed] most of the time.



Robert Jeffares



___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


[RDD] transmitter / silence / fallback

2015-01-29 Thread Jim Stewart

>From some previous posts:

>I think I am going to face another issue. Some players give up trying for
>good when the stream fails. I need to find one that will keep retrying or
>respawn one in a loop or something along those lines.

>Does anyone have any player suggestions for this setup?

VLC does have an automatic "reconnect" feature.  I don't think it is enabled by 
default though and the setting for it is a little hard to find.
I prefer mplayer because it simply exits when the stream is lost.  This allows 
me to log the event in the script and make some decisions about what to do 
about this, typically to simply launch mplayer again and reconnect.

Yes, many players just "sit there" with a stream is interrupted, which is hard 
to deal with.

>The biggest challenge with Silentjack is it currently has no provision
>for what to do when audio returns.

Good point.  I'm thinking you can have your script attempt some corrective 
action, then kill silent jack, then restart it and see if it trips again (as in 
see if your "fix" did in fact fix the problem).
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


[RDD] transmitter / silence / fallback

2015-01-29 Thread Jim Stewart
> ... config a small box (I hope to use a Raspberry Pi but will scale up until I
>find one with the needed power to serve in this capacity) such that when it
>is plugged into a network and amp/speakers somewhere it will connect to an
>icecast stream automatically and play it out of the speakers.

I have done this.

>It will detect silence in the stream (including the stream going down) and
>switch to playing (a playlist of) locally stored files.

But haven't done this on a Raspberry pi

>It will detect when the stream audio comes back and cut out the local audio
>and go back to stream audio.

>All headless and automatic.

Have done this with our slightly different setup.  In our setup we have a PC at 
a mountain-top transmitter site that runs Windows so it can run a commercial 
audio processing & FM stereo composite signal generation program, but we also 
have it receiving an audio stream (using mplayer in a Windows Batch file) from 
the Internet delivered from the in-town studio.  I do have Windows scripts that 
do all this (plays locally stored audio during stream failure, attempts to 
connect to a backup studio, plus logs everything for later inspection, etc), so 
at least I am familiar with the issues.  

> I am trying to use:

>jack
>silentjack (to detect the silence)
>jack_connect (to change the jack routing on setup and on command from
>silentjack)

Our main studio uses a GPI input from the silence sensor built into a Broadcast 
Tools ACS 8.2, but we have a backup Rivendell computer (in another city!) that 
simply uses silentjack to do things based on what it determines the problem 
might be.  For us it is typically that rdairplay has crashed while trying to 
pull up the next days log so we simply restart it. Other times it is because a 
live web stream broadcast got lost.

>audio players:

>mplayer (or its frontends)
>vlc
>others?

Last time I played with jack on a Raspberry pi, I found it a little bit 
unstable, but I was trying to do more with it than you are.  It seems that you 
might need to use a "real" (and stable, as in some are much better than others) 
USB sound device as the master clock for Jack unless you use netjack and lock 
it to a master on another computer.  I think some of the instability is from 
the lack of a "real time" kernel available for the pi (maybe there is one now, 
but I think there were some issues with the ARM processor that prevents this).  
In any case I think you can get it working well enough for your purposes, just 
expect to need to dial in a few seconds of latency.

>Problems I think I need to solve:

>A way to "statically name" the main stream player so that I can refer to it
>in scripts.
>So far, I am not seeing how to do that with mplayer or vlc, the names seem
>to be tied to the pid.

You can do this with mplayer, but not vlc.  I have run mplayer on the pi, but 
not VLC.
We do have VLC receiving a web stream on our studio Rivendell machine, and yes 
it always has a jack name that includes it's process ID at the end, which is 
always different.  I have a script that reads the process ID once VLC get 
running so it knows how to connect it in Jack.  It is all done in a shell 
script called by Rivendell.

>Is there a better way to go about this?
So I like mplayer for this since you can do a static jack name.  We are only 
currently using VLC because with our (somewhat now old, Debian "Squeeze" based) 
studio system, the mplayer version doesn't support the stream we are trying to 
get and VLC does.

>If it hasn't already been done, I will write this all up when I am done.

I'll be happy to share code with you, but since your needs are a little 
different than mine, it might be better for me to just help you with places 
where you might be having problems.

___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread Wayne Merricks
You'll probably have to install the liquidsoap plugins to enable mp3 
support (if you're unlucky you'll have to compile it with mp3 support).


Do an apt-cache search for liquidsoap or liquidsoap-plugin and you 
should see a whole bunch of them including jack and pulse.


On 2015-01-28 19:44, drew Roberts wrote:

On Wed, Jan 28, 2015 at 12:21 PM, Marc Steele  wrote:


Ive done similar with LiquidSoap in the past.
See http://www.dlineradio.co.uk/articles/stls-and-raspberry-pis/
[1] for details. 


Fallback is working, main stream does not play:

 2015/01/27 20:03:14 [decoder:3] Unable to decode stream of type
"mp3"!


Written originally for the Pi but should work on most Debian based
systems. Adding some on silence actions might do the trick for
you. 

Sent from my iPhone

On 28 Jan 2015, at 16:05, drew Roberts 
wrote:


I think I have seen discussions around these topics in the past. I
am hoping someone is up to date on these skills so that I dont
have to spend further hours or days trying to get this to work
smoothly.

What I want to do:

config a small box (I hope to use a Raspberry Pi but will scale up
until I find one with the needed power to serve in this capacity)
such that when it is plugged into a network and amp/speakers
somewhere it will connect to an icecast stream automatically and
play it out of the speakers.

It will detect silence in the stream (including the stream going
down) and switch to playing (a playlist of) locally stored files.

It will detect when the stream audio comes back and cut out the
local audio and go back to stream audio.

All headless and automatic.

I am trying to get this system going now on a laptop with a bit
more power than the Pi.

I am trying to use:

jack
silentjack (to detect the silence)

jack_connect (to change the jack routing on setup and on command
from silentjack)

audio players:

mplayer (or its frontends)
vlc
others?

Does this sound like a reasonable set of components for this job?
Has anyone done something similar? Are there any good docs in this
space?

Problems I think I need to solve:

A way to "statically name" the main stream player so that I can
refer to it in scripts.
I may not have to statically name the backup player but it would
probably be nice.

So far, I am not seeing how to do that with mplayer or vlc, the
names seem to be tied to the pid.

Is there a better way to go about this?

If it hasnt already been done, I will write this all up when I am
done.

all the best,

drew

--

http://nakedghosts.blogspot.com/ [2]



___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org [3]
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
[4]


___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread drew Roberts
On Wed, Jan 28, 2015 at 12:21 PM, Marc Steele  wrote:

> I've done similar with LiquidSoap in the past. See
> http://www.dlineradio.co.uk/articles/stls-and-raspberry-pis/ for details.
>

Fallback is working, main stream does not play:

 2015/01/27 20:03:14 [decoder:3] Unable to decode stream of type "mp3"!

>
> Written originally for the Pi but should work on most Debian based
> systems. Adding some on silence actions might do the trick for you.
>
> Sent from my iPhone
>
> On 28 Jan 2015, at 16:05, drew Roberts  wrote:
>
> I think I have seen discussions around these topics in the past. I am
> hoping someone is up to date on these skills so that I don't have to spend
> further hours or days trying to get this to work smoothly.
>
> What I want to do:
>
> config a small box (I hope to use a Raspberry Pi but will scale up until I
> find one with the needed power to serve in this capacity) such that when it
> is plugged into a network and amp/speakers somewhere it will connect to an
> icecast stream automatically and play it out of the speakers.
>
> It will detect silence in the stream (including the stream going down) and
> switch to playing (a playlist of) locally stored files.
>
> It will detect when the stream audio comes back and cut out the local
> audio and go back to stream audio.
>
> All headless and automatic.
>
> I am trying to get this system going now on a laptop with a bit more power
> than the Pi.
>
> I am trying to use:
>
> jack
> silentjack (to detect the silence)
> jack_connect (to change the jack routing on setup and on command from
> silentjack)
>
> audio players:
>
> mplayer (or its frontends)
> vlc
> others?
>
> Does this sound like a reasonable set of components for this job? Has
> anyone done something similar? Are there any good docs in this space?
>
> Problems I think I need to solve:
>
> A way to "statically name" the main stream player so that I can refer to
> it in scripts.
> I may not have to statically name the backup player but it would probably
> be nice.
>
> So far, I am not seeing how to do that with mplayer or vlc, the names seem
> to be tied to the pid.
>
> Is there a better way to go about this?
>
> If it hasn't already been done, I will write this all up when I am done.
>
> all the best,
>
> drew
>
> --
> http://nakedghosts.blogspot.com/
>
> ___
> Rivendell-dev mailing list
> Rivendell-dev@lists.rivendellaudio.org
> http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
>
>


-- 
http://nakedghosts.blogspot.com/
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread drew Roberts
May have had to make a few changes but in the end:

Fallback is working, main stream does not play:

 2015/01/27 20:03:14 [decoder:3] Unable to decode stream of type "mp3"!

Same basic line from both fails.

On Wed, Jan 28, 2015 at 12:25 PM, Wayne Merricks <
waynemerri...@thevoiceasia.com> wrote:

>  Hi,
>
> The Pastebin is set up to take input from a usb capture card on a full
> blown ubuntu 12.04 server edition.  On the Pi you could do the same thing
> however all I do instead is take an icecast feed into liquidsoap using the
> input.http command.
>
> I also don't need JACK etc because liquidsoap can directly access ALSA.
> So you end up with a quite simple liquidsoap config:
>
> #specify the icecast stream as an input
> stream = input.http("http://icecastip/icecastmount";
> )
>
> #specify the fail over file (could change this to be a template or
> directory of files)
> failover = mksafe(single("/var/audio/emergency/song.mp3"))
>
> #Enable the silence detect to kick in after 5 seconds
> stream = strip_blank(stream, length=5.0)
>
> #Specify what to switch to when silence is detected
> stream = fallback(track_sensitive=false, [stream, failover])
>
> # (Optional) Compress the icecast stream
> stream = compress(ratio=3.0, attack=38.0, release=85.0, threshold=-20.0,
> knee=0.5, gain=8.0, stream)
>
> # (Optional) limit the ice cast stream
> stream = limit(ratio=3.0, attack=38.0, release=85.0, threshold=-3.0,knee=
> 0.5, stream)
>
> # Output to ALSA (the pi's audio)
> output.alsa(stream)
>
> All you need to do is save that as a file (conventionally blah.liq).  Then
> run liquidsoap -c blah.liq on startup and away it goes pretty much forever.
>
> On ubuntu with things like upstart you can specify what happens if things
> crash which is handy but on the pi you'd have to do other watchdog type
> stuff to achieve the same result.
>
> *NB: *I haven't tested this config it was written mostly via copy paste
> and checking the API reference for liquidsoap (I don't have access to the
> pi at the moment)
>
> API Reference is here:
> http://savonet.sourceforge.net/doc-svn/reference.html
>
> Theres also a lot of examples on that site for doing all sorts of crazy
> stuff with liquidsoap.
>
> Wayne Merricks The Voice AsiaOn 28/01/15 17:04, drew Roberts wrote:
>
>   Hey Wayne,
>
>  if I am reading you and those scripts right on first glance, the scripts
> are checking for silence on a local rivendell playout and you have modified
> them to pull down a remote stream and that is running on a Pi.
>
>  If I have that right, what is the player in use on the Pi?
>
>  Or are you using icecast to relay the remote stream locally and doing
> some failover to a different mount locally?
>
>  all the best,
>
> drew
>
> On Wed, Jan 28, 2015 at 11:36 AM, Wayne Merricks <
> waynemerri...@thevoiceasia.com> wrote:
>
>>  I do that at work using liquidsoap on a pi.
>>
>> I adapted my Ubuntu instructions found here:
>>
>> http://pastebin.com/u/mezzfa0
>>
>>
>>
>> Wayne Merricks
>> The Voice Asia
>>
>>  On 28/01/15 16:05, drew Roberts wrote:
>>
>> I think I have seen discussions around these topics in the
>> past. I am hoping someone is up to date on these skills so that I don't
>> have to spend further hours or days trying to get this to work smoothly.
>>
>>  What I want to do:
>>
>>  config a small box (I hope to use a Raspberry Pi but will scale up until
>> I find one with the needed power to serve in this capacity) such that when
>> it is plugged into a network and amp/speakers somewhere it will connect to
>> an icecast stream automatically and play it out of the speakers.
>>
>>  It will detect silence in the stream (including the stream going down)
>> and switch to playing (a playlist of) locally stored files.
>>
>>  It will detect when the stream audio comes back and cut out the local
>> audio and go back to stream audio.
>>
>>  All headless and automatic.
>>
>>  I am trying to get this system going now on a laptop with a bit more
>> power than the Pi.
>>
>>  I am trying to use:
>>
>>  jack
>>  silentjack (to detect the silence)
>>  jack_connect (to change the jack routing on setup and on command from
>> silentjack)
>>
>>  audio players:
>>
>>  mplayer (or its frontends)
>>  vlc
>>  others?
>>
>>  Does this sound like a reasonable set of components for this job? Has
>> anyone done something similar? Are there any good docs in this space?
>>
>>  Problems I think I need to solve:
>>
>>  A way to "statically name" the main stream player so that I can refer to
>> it in scripts.
>>  I may not have to statically name the backup player but it would
>> probably be nice.
>>
>>  So far, I am not seeing how to do that with mplayer or vlc, the names
>> seem to be tied to the pid.
>>
>>  Is there a better way to go about this?
>>
>>  If it hasn't already been done, I will write this all up when I am done.
>>
>>  all the best,
>>
>> drew
>>
>> --
>>  http://nakedghosts.blogspot.com/
>>
>>
>>  

Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread Wayne Merricks

Hi,

The Pastebin is set up to take input from a usb capture card on a full 
blown ubuntu 12.04 server edition.  On the Pi you could do the same 
thing however all I do instead is take an icecast feed into liquidsoap 
using the input.http command.


I also don't need JACK etc because liquidsoap can directly access ALSA.  
So you end up with a quite simple liquidsoap config:


#specify the icecast stream as an input
stream = input.http("http://icecastip/icecastmount";)

#specify the fail over file (could change this to be a template or 
directory of files)

failover = mksafe(single("/var/audio/emergency/song.mp3"))

#Enable the silence detect to kick in after 5 seconds
stream = strip_blank(stream, length=5.0)

#Specify what to switch to when silence is detected
stream = fallback(track_sensitive=false, [stream, failover])

# (Optional) Compress the icecast stream
stream = compress(ratio=3.0, attack=38.0, release=85.0, threshold=-20.0, 
knee=0.5, gain=8.0, stream)


# (Optional) limit the ice cast stream
stream = limit(ratio=3.0, attack=38.0, release=85.0, 
threshold=-3.0,knee=0.5, stream)


# Output to ALSA (the pi's audio)
output.alsa(stream)

All you need to do is save that as a file (conventionally blah.liq).  
Then run liquidsoap -c blah.liq on startup and away it goes pretty much 
forever.


On ubuntu with things like upstart you can specify what happens if 
things crash which is handy but on the pi you'd have to do other 
watchdog type stuff to achieve the same result.


*NB: *I haven't tested this config it was written mostly via copy paste 
and checking the API reference for liquidsoap (I don't have access to 
the pi at the moment)


API Reference is here: http://savonet.sourceforge.net/doc-svn/reference.html

Theres also a lot of examples on that site for doing all sorts of crazy 
stuff with liquidsoap.


Wayne Merricks The Voice AsiaOn 28/01/15 17:04, drew Roberts wrote:

Hey Wayne,

if I am reading you and those scripts right on first glance, the 
scripts are checking for silence on a local rivendell playout and you 
have modified them to pull down a remote stream and that is running on 
a Pi.


If I have that right, what is the player in use on the Pi?

Or are you using icecast to relay the remote stream locally and doing 
some failover to a different mount locally?


all the best,

drew

On Wed, Jan 28, 2015 at 11:36 AM, Wayne Merricks 
> wrote:


I do that at work using liquidsoap on a pi.

I adapted my Ubuntu instructions found here:

http://pastebin.com/u/mezzfa0



Wayne Merricks
The Voice Asia

On 28/01/15 16:05, drew Roberts wrote:

I think I have seen discussions around these topics in the past.
I am hoping someone is up to date on these skills so that I don't
have to spend further hours or days trying to get this to work
smoothly.

What I want to do:

config a small box (I hope to use a Raspberry Pi but will scale
up until I find one with the needed power to serve in this
capacity) such that when it is plugged into a network and
amp/speakers somewhere it will connect to an icecast stream
automatically and play it out of the speakers.

It will detect silence in the stream (including the stream going
down) and switch to playing (a playlist of) locally stored files.

It will detect when the stream audio comes back and cut out the
local audio and go back to stream audio.

All headless and automatic.

I am trying to get this system going now on a laptop with a bit
more power than the Pi.

I am trying to use:

jack
silentjack (to detect the silence)
jack_connect (to change the jack routing on setup and on command
from silentjack)

audio players:

mplayer (or its frontends)
vlc
others?

Does this sound like a reasonable set of components for this job?
Has anyone done something similar? Are there any good docs in
this space?

Problems I think I need to solve:

A way to "statically name" the main stream player so that I can
refer to it in scripts.
I may not have to statically name the backup player but it would
probably be nice.

So far, I am not seeing how to do that with mplayer or vlc, the
names seem to be tied to the pid.

Is there a better way to go about this?

If it hasn't already been done, I will write this all up when I
am done.

all the best,

drew

-- 
http://nakedghosts.blogspot.com/



___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org  

http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev



___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org

http://caspian.paravelsystems.com/mailman/li

Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread Marc Steele
I've done similar with LiquidSoap in the past. See 
http://www.dlineradio.co.uk/articles/stls-and-raspberry-pis/ for details. 

Written originally for the Pi but should work on most Debian based systems. 
Adding some on silence actions might do the trick for you. 

Sent from my iPhone

> On 28 Jan 2015, at 16:05, drew Roberts  wrote:
> 
> I think I have seen discussions around these topics in the past. I am hoping 
> someone is up to date on these skills so that I don't have to spend further 
> hours or days trying to get this to work smoothly.
> 
> What I want to do:
> 
> config a small box (I hope to use a Raspberry Pi but will scale up until I 
> find one with the needed power to serve in this capacity) such that when it 
> is plugged into a network and amp/speakers somewhere it will connect to an 
> icecast stream automatically and play it out of the speakers.
> 
> It will detect silence in the stream (including the stream going down) and 
> switch to playing (a playlist of) locally stored files.
> 
> It will detect when the stream audio comes back and cut out the local audio 
> and go back to stream audio.
> 
> All headless and automatic.
> 
> I am trying to get this system going now on a laptop with a bit more power 
> than the Pi.
> 
> I am trying to use:
> 
> jack
> silentjack (to detect the silence)
> jack_connect (to change the jack routing on setup and on command from 
> silentjack)
> 
> audio players:
> 
> mplayer (or its frontends)
> vlc
> others?
> 
> Does this sound like a reasonable set of components for this job? Has anyone 
> done something similar? Are there any good docs in this space?
> 
> Problems I think I need to solve:
> 
> A way to "statically name" the main stream player so that I can refer to it 
> in scripts.
> I may not have to statically name the backup player but it would probably be 
> nice.
> 
> So far, I am not seeing how to do that with mplayer or vlc, the names seem to 
> be tied to the pid.
> 
> Is there a better way to go about this?
> 
> If it hasn't already been done, I will write this all up when I am done.
> 
> all the best,
> 
> drew
> 
> -- 
> http://nakedghosts.blogspot.com/
> ___
> Rivendell-dev mailing list
> Rivendell-dev@lists.rivendellaudio.org
> http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread drew Roberts
On Wed, Jan 28, 2015 at 12:06 PM, Frederick Gleason <
fr...@paravelsystems.com> wrote:

> On Jan 28, 2015, at 11:05 12, drew Roberts  wrote:
> >
> > Does this sound like a reasonable set of components for this job? Has
> anyone done something similar? Are there any good docs in this space?
>
> It sounds doable.  The devil, as always, is in the details!
>
>
> > A way to "statically name" the main stream player so that I can refer to
> it in scripts.
> > I may not have to statically name the backup player but it would
> probably be nice.
>
> What namespace is this “name” to appear in?  The JACK graph?  A process
> name?  Something else?
>

The JACK graph? (the name that jack_lsp will show and the name that needs
to be passed to jack_connect to make connection changes.)

>
> Cheers!
>
>
> |--|
> | Frederick F. Gleason, Jr. |  Chief Developer |
> |   |  Paravel Systems |
> |--|
> |  A room without books is like a body without a soul. |
> | -- Cicero|
> |--|
>

all the best,

drew
-- 
http://nakedghosts.blogspot.com/
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread Frederick Gleason
On Jan 28, 2015, at 11:05 12, drew Roberts  wrote:
> 
> Does this sound like a reasonable set of components for this job? Has anyone 
> done something similar? Are there any good docs in this space?

It sounds doable.  The devil, as always, is in the details!


> A way to "statically name" the main stream player so that I can refer to it 
> in scripts.
> I may not have to statically name the backup player but it would probably be 
> nice.

What namespace is this “name” to appear in?  The JACK graph?  A process name?  
Something else?

Cheers!


|--|
| Frederick F. Gleason, Jr. |  Chief Developer |
|   |  Paravel Systems |
|--|
|  A room without books is like a body without a soul. |
| -- Cicero|
|--|

___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread drew Roberts
Hey Wayne,

if I am reading you and those scripts right on first glance, the scripts
are checking for silence on a local rivendell playout and you have modified
them to pull down a remote stream and that is running on a Pi.

If I have that right, what is the player in use on the Pi?

Or are you using icecast to relay the remote stream locally and doing some
failover to a different mount locally?

all the best,

drew

On Wed, Jan 28, 2015 at 11:36 AM, Wayne Merricks <
waynemerri...@thevoiceasia.com> wrote:

>  I do that at work using liquidsoap on a pi.
>
> I adapted my Ubuntu instructions found here:
>
> http://pastebin.com/u/mezzfa0
>
>
>
> Wayne Merricks
> The Voice Asia
>
> On 28/01/15 16:05, drew Roberts wrote:
>
>I think I have seen discussions around these topics in the
> past. I am hoping someone is up to date on these skills so that I don't
> have to spend further hours or days trying to get this to work smoothly.
>
>  What I want to do:
>
>  config a small box (I hope to use a Raspberry Pi but will scale up until
> I find one with the needed power to serve in this capacity) such that when
> it is plugged into a network and amp/speakers somewhere it will connect to
> an icecast stream automatically and play it out of the speakers.
>
>  It will detect silence in the stream (including the stream going down)
> and switch to playing (a playlist of) locally stored files.
>
>  It will detect when the stream audio comes back and cut out the local
> audio and go back to stream audio.
>
>  All headless and automatic.
>
>  I am trying to get this system going now on a laptop with a bit more
> power than the Pi.
>
>  I am trying to use:
>
>  jack
>  silentjack (to detect the silence)
>  jack_connect (to change the jack routing on setup and on command from
> silentjack)
>
>  audio players:
>
>  mplayer (or its frontends)
>  vlc
>  others?
>
>  Does this sound like a reasonable set of components for this job? Has
> anyone done something similar? Are there any good docs in this space?
>
>  Problems I think I need to solve:
>
>  A way to "statically name" the main stream player so that I can refer to
> it in scripts.
>  I may not have to statically name the backup player but it would probably
> be nice.
>
>  So far, I am not seeing how to do that with mplayer or vlc, the names
> seem to be tied to the pid.
>
>  Is there a better way to go about this?
>
>  If it hasn't already been done, I will write this all up when I am done.
>
>  all the best,
>
> drew
>
> --
>  http://nakedghosts.blogspot.com/
>
>
> ___
> Rivendell-dev mailing 
> listRivendell-dev@lists.rivendellaudio.orghttp://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
>
>
>
> ___
> Rivendell-dev mailing list
> Rivendell-dev@lists.rivendellaudio.org
> http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
>
>


-- 
http://nakedghosts.blogspot.com/
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread drew Roberts
Hey Lorne,

I haven't digested it all yet and don't know how much it can be simplified,
but this person seems to be handling the run something else when audio
restored problem:

http://www.talkunafraid.co.uk/2011/03/interfacing-silentjack-and-nagios/

I think I am going to face another issue. Some players give up trying for
good when the stream fails. I need to find one that will keep retrying or
respawn one in a loop or something along those lines.

Does anyone have any player suggestions for this setup?

all the best,

drew

On Wed, Jan 28, 2015 at 11:44 AM, Lorne Tyndale 
wrote:

> Hi,
>
> The biggest challenge with Silentjack is it currently has no provision
> for what to do when audio returns.
>
> Right now it'll do the "There's no audio, I need to do something" part.
>
> But it has no "Audio is back now, so I will do something else"
> capability on its own.
>
>
>
> >
> > Hey John,
> >
> > thanks. I think silentjack is what will do that natively under linux.
> >
> >
> > I think if I can just figure out how to get players with fixed names in
> > jack I will be able to handle the rest. I ma hoping someone has already
> > done the thing or knows which plahers can be given fixed jack names or
> > knows a better way to do this in software under linux.
> >
> > all the best,
> >
> > drew
> >
> > On Wed, Jan 28, 2015 at 11:12 AM, John Anderson  wrote:
> >
> > > this might help...
> > >
> > > http://pira.cz/show.asp?art=silence
> > >
> > > On Wed, 2015-01-28 at 11:05 -0500, drew Roberts wrote:
> > > > I think I have seen discussions around these topics in the past. I am
> > > > hoping someone is up to date on these skills so that I don't have to
> > > > spend further hours or days trying to get this to work smoothly.
> > > >
> > > >
> > > > What I want to do:
> > > >
> > > >
> > > > config a small box (I hope to use a Raspberry Pi but will scale up
> > > > until I find one with the needed power to serve in this capacity)
> such
> > > > that when it is plugged into a network and amp/speakers somewhere it
> > > > will connect to an icecast stream automatically and play it out of
> the
> > > > speakers.
> > > >
> > >
> > >
> > >
> > >
> >
> >
> > --
> > http://nakedghosts.blogspot.com/
> ___
> > Rivendell-dev mailing list
> > Rivendell-dev@lists.rivendellaudio.org
> > http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
>



-- 
http://nakedghosts.blogspot.com/
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread Lorne Tyndale
Hi,

The biggest challenge with Silentjack is it currently has no provision
for what to do when audio returns.

Right now it'll do the "There's no audio, I need to do something" part.

But it has no "Audio is back now, so I will do something else"
capability on its own.


 
> 
> Hey John,
> 
> thanks. I think silentjack is what will do that natively under linux.
> 
> 
> I think if I can just figure out how to get players with fixed names in
> jack I will be able to handle the rest. I ma hoping someone has already
> done the thing or knows which plahers can be given fixed jack names or
> knows a better way to do this in software under linux.
> 
> all the best,
> 
> drew
> 
> On Wed, Jan 28, 2015 at 11:12 AM, John Anderson  wrote:
> 
> > this might help...
> >
> > http://pira.cz/show.asp?art=silence
> >
> > On Wed, 2015-01-28 at 11:05 -0500, drew Roberts wrote:
> > > I think I have seen discussions around these topics in the past. I am
> > > hoping someone is up to date on these skills so that I don't have to
> > > spend further hours or days trying to get this to work smoothly.
> > >
> > >
> > > What I want to do:
> > >
> > >
> > > config a small box (I hope to use a Raspberry Pi but will scale up
> > > until I find one with the needed power to serve in this capacity) such
> > > that when it is plugged into a network and amp/speakers somewhere it
> > > will connect to an icecast stream automatically and play it out of the
> > > speakers.
> > >
> >
> >
> >
> >
> 
> 
> -- 
> http://nakedghosts.blogspot.com/___
> Rivendell-dev mailing list
> Rivendell-dev@lists.rivendellaudio.org
> http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread drew Roberts
Hey John,

thanks. I think silentjack is what will do that natively under linux.


I think if I can just figure out how to get players with fixed names in
jack I will be able to handle the rest. I ma hoping someone has already
done the thing or knows which plahers can be given fixed jack names or
knows a better way to do this in software under linux.

all the best,

drew

On Wed, Jan 28, 2015 at 11:12 AM, John Anderson  wrote:

> this might help...
>
> http://pira.cz/show.asp?art=silence
>
> On Wed, 2015-01-28 at 11:05 -0500, drew Roberts wrote:
> > I think I have seen discussions around these topics in the past. I am
> > hoping someone is up to date on these skills so that I don't have to
> > spend further hours or days trying to get this to work smoothly.
> >
> >
> > What I want to do:
> >
> >
> > config a small box (I hope to use a Raspberry Pi but will scale up
> > until I find one with the needed power to serve in this capacity) such
> > that when it is plugged into a network and amp/speakers somewhere it
> > will connect to an icecast stream automatically and play it out of the
> > speakers.
> >
>
>
>
>


-- 
http://nakedghosts.blogspot.com/
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread Wayne Merricks

I do that at work using liquidsoap on a pi.

I adapted my Ubuntu instructions found here:

http://pastebin.com/u/mezzfa0



Wayne Merricks
The Voice Asia

On 28/01/15 16:05, drew Roberts wrote:
I think I have seen discussions around these topics in the past. I am 
hoping someone is up to date on these skills so that I don't have to 
spend further hours or days trying to get this to work smoothly.


What I want to do:

config a small box (I hope to use a Raspberry Pi but will scale up 
until I find one with the needed power to serve in this capacity) such 
that when it is plugged into a network and amp/speakers somewhere it 
will connect to an icecast stream automatically and play it out of the 
speakers.


It will detect silence in the stream (including the stream going down) 
and switch to playing (a playlist of) locally stored files.


It will detect when the stream audio comes back and cut out the local 
audio and go back to stream audio.


All headless and automatic.

I am trying to get this system going now on a laptop with a bit more 
power than the Pi.


I am trying to use:

jack
silentjack (to detect the silence)
jack_connect (to change the jack routing on setup and on command from 
silentjack)


audio players:

mplayer (or its frontends)
vlc
others?

Does this sound like a reasonable set of components for this job? Has 
anyone done something similar? Are there any good docs in this space?


Problems I think I need to solve:

A way to "statically name" the main stream player so that I can refer 
to it in scripts.
I may not have to statically name the backup player but it would 
probably be nice.


So far, I am not seeing how to do that with mplayer or vlc, the names 
seem to be tied to the pid.


Is there a better way to go about this?

If it hasn't already been done, I will write this all up when I am done.

all the best,

drew

--
http://nakedghosts.blogspot.com/


___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] transmitter / silence / fallback

2015-01-28 Thread John Anderson
this might help...

http://pira.cz/show.asp?art=silence

On Wed, 2015-01-28 at 11:05 -0500, drew Roberts wrote:
> I think I have seen discussions around these topics in the past. I am
> hoping someone is up to date on these skills so that I don't have to
> spend further hours or days trying to get this to work smoothly.
> 
> 
> What I want to do:
> 
> 
> config a small box (I hope to use a Raspberry Pi but will scale up
> until I find one with the needed power to serve in this capacity) such
> that when it is plugged into a network and amp/speakers somewhere it
> will connect to an icecast stream automatically and play it out of the
> speakers.
> 



___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


[RDD] transmitter / silence / fallback

2015-01-28 Thread drew Roberts
I think I have seen discussions around these topics in the past. I am
hoping someone is up to date on these skills so that I don't have to spend
further hours or days trying to get this to work smoothly.

What I want to do:

config a small box (I hope to use a Raspberry Pi but will scale up until I
find one with the needed power to serve in this capacity) such that when it
is plugged into a network and amp/speakers somewhere it will connect to an
icecast stream automatically and play it out of the speakers.

It will detect silence in the stream (including the stream going down) and
switch to playing (a playlist of) locally stored files.

It will detect when the stream audio comes back and cut out the local audio
and go back to stream audio.

All headless and automatic.

I am trying to get this system going now on a laptop with a bit more power
than the Pi.

I am trying to use:

jack
silentjack (to detect the silence)
jack_connect (to change the jack routing on setup and on command from
silentjack)

audio players:

mplayer (or its frontends)
vlc
others?

Does this sound like a reasonable set of components for this job? Has
anyone done something similar? Are there any good docs in this space?

Problems I think I need to solve:

A way to "statically name" the main stream player so that I can refer to it
in scripts.
I may not have to statically name the backup player but it would probably
be nice.

So far, I am not seeing how to do that with mplayer or vlc, the names seem
to be tied to the pid.

Is there a better way to go about this?

If it hasn't already been done, I will write this all up when I am done.

all the best,

drew

-- 
http://nakedghosts.blogspot.com/
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev