Speech or mp3 to speaker

[Pocster]

Hey

I'm trying to initially get speech to come out of the appropriate squeezebox client ( raspberry pi's ).
Followed all the instructions for setting this up.
When trying to send to clients no speech is heard and it just queues in the speech proxy count e.g. if I do hs.speak("hello") 3 times proxy count will be 3.
No audio is heard from the clients nor on the Homeseer pc.
It just sits there until I 'discard' them 1 by 1. Mode is 'normal'

Assume if I can get this working then playing a wav/mp3 via which ever client will be easy.....

[Pocster]

In HS control panel if I do

&hs. Speak ("hello", FALSE, "LMS:*")

Then the proxy count is 1 - nothing is heard from the LMS client

[Pocster]

Debug log for speech:

Jun-29 11:30:40

SqueezeBox Debug

Exiting speechTimer_Elapsed

Jun-29 11:30:40

SqueezeBox Debug

Exiting SpeakOutNextInQueue. New status: stopping. PendingSpeechTrackEnd: False

Jun-29 11:30:40

SqueezeBox Debug

StartSpeechTimer. State: stopping. Duration: 0:1.0 (1000)

Jun-29 11:30:40

SqueezeBox Debug

SpeakOutNextInQueue - Stopping.

Jun-29 11:30:40

SqueezeBox Debug

SpeakOutNextInQueue - Deleting TTS URLs.

Jun-29 11:30:40

SqueezeBox Debug

Entering SpeakOutNextInQueue. Current status: speak_wait. PendingSpeechTrackEnd: False

Jun-29 11:30:40

SqueezeBox Debug

Entering speechTimer_Elapsed

Jun-29 11:30:39

SqueezeBox Debug

Exiting speechTimer_Elapsed

Jun-29 11:30:39

SqueezeBox Debug

Exiting SpeakOutNextInQueue. New status: speak_wait. PendingSpeechTrackEnd: False

Jun-29 11:30:39

SqueezeBox Debug

StartSpeechTimer. State: speak_wait. Duration: 0:1.0 (1000)

Jun-29 11:30:39

SqueezeBox Debug

SpeakOutNextInQueue - Speak Wait.

Jun-29 11:30:39

SqueezeBox Debug

Entering SpeakOutNextInQueue. Current status: speaking. PendingSpeechTrackEnd: False

Jun-29 11:30:39

SqueezeBox Debug

Entering speechTimer_Elapsed

Jun-29 11:30:38

SqueezeBox Debug

Exiting speechTimer_Elapsed

Jun-29 11:30:38

SqueezeBox Debug

Exiting SpeakOutNextInQueue. New status: speaking. PendingSpeechTrackEnd: False

Jun-29 11:30:38

SqueezeBox Debug

StartSpeechTimer. State: speaking. Duration: 0:0.10 (10)

Jun-29 11:30:38

SqueezeBox Debug

Entering SpeakProxyUpdateDevices

Jun-29 11:30:38

TTS

Speak: (LMS:*):hello how are you today

[pcp]

I assume your player was detected by the HS3 plugin and HS3 devices were created. Did you set the set the speech volume HS3 device of the player to a non zero value? Does it work if your don’t specify a player argument (i.e. use *:* instead of LMS:*)?

Another thing to check is to see if the wav and mp3 file exists in the plugin html directory in HS3. You could try to play this generated file with LMS to ensure it is correctly recognized and plays fine,

[Pocster]

I assume your player was detected by the HS3 plugin and HS3 devices were created. Did you set the set the speech volume HS3 device of the player to a non zero value? Does it work if your don’t specify a player argument (i.e. use *:* instead of LMS:*)?

Another thing to check is to see if the wav and mp3 file exists in the plugin html directory in HS3. You could try to play this generated file with LMS to ensure it is correctly recognized and plays fine,

Thanks for the reply.
Yep speak volume is at maximum. Player is recognised fine - can see all of it in devices menu ( connected etc. )
Tried *.* instead of LMS - proxy count stuck at 1. No speech

Trying to play mp3 works fine through hs3 - but will only be audible through the host - nothing from any squeezebox clients.

[pcp]

What happens if you play the mp3 file from the LMS web UI (add the mp3 file to your media library or use the LMS ‘Tune in URL’ To play it. If that works, try adding TTS additional speech direction (500ms for example in the plugin Speech Proxy configuring extra as well 2 to 3 seconds before and after Speaking if not already done. Also can you collect a verbose debug log when playing it (deselect ‘Log to HomeSeer’ so that it is saved to a separate text file in the HS3 Log directory) and send it to me or post it?

[Pocster]

What happens if you play the mp3 file from the LMS web UI (add the mp3 file to your media library or use the LMS ‘Tune in URL’ To play it. If that works, try adding TTS additional speech direction (500ms for example in the plugin Speech Proxy configuring extra as well 2 to 3 seconds before and after Speaking if not already done. Also can you collect a verbose debug log when playing it (deselect ‘Log to HomeSeer’ so that it is saved to a separate text file in the HS3 Log directory) and send it to me or post it?

Thanks ! Will do all that later

[Pocster]

Hi

Sorry for the delay.
The LMS server can be a fickle thing - had to reboot/restart a few times just to get playing an mp3 working.
Added 3 seconds delay before and after speech..

The client player has a track on pause. When I do hs.speak there's a delay as expected - the paused track plays (I assume that is where the speech is meant to be ) - then stops

PM'd you the log. Appreciate your help.

[Pocster]

Right

I've had some success!
I can trigger squeezebox to pause the current mp3 e.g. a doorbell mp3 interrupts the playing music.
What doesn't happen is the music doesn't unpause and continue after the interruption.

Still can't get any speech to come out a squeezebox client though

[Pocster]

More update.

If I attempt speech it seems to stop any audio then play 3 seconds of the audio ( presumably where the speech should be ).

[Pocster]

Hey!

Had some success here!
Managed to get speech come out a squeezebox client
If music is playing and I trigger speech, the music is paused, speech occurs, music continues

If music is playing and I trigger a mp3 , music is paused, mp3 is played , but music doesn't continue. The mp3 is still on the squeezebox server not the original track.

Is that a bug?

[Pocster]

Hmmm.
Spoke too soon. No speech again!!!

[Pocster]

Still struggling with speech. Very hit or miss. Sometimes it will work - other times nothing.

[pcp]

Can you send me a verbose debug log of the squeezebox plugin when speaking and problem occurs? Also is HS3 running on Windows or Linux? The plugin is attempting to detect when LMS finished speaking the generated MP3 file to restore the original track; the log file should help to determine if that condition can be detected or not for some reason.

[Pocster]

Can you send me a verbose debug log of the squeezebox plugin when speaking and problem occurs? Also is HS3 running on Windows or Linux? The plugin is attempting to detect when LMS finished speaking the generated MP3 file to restore the original track; the log file should help to determine if that condition can be detected or not for some reason.

Will do . Once I get hs4 working !