Announcement

Collapse
No announcement yet.

Does TTS with Squeezebox Plugin not work on HS6 running HS3?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    #16
    I can't right now. If I turn on the TTS plugin and have it create something, it will interrupt the playlist that is running and keeping my son asleep. Suddenly killing the music would wake him up. Perhaps tomorrow, unless someone else beats me to it.

    Comment


      #17
      Originally posted by pcp View Post
      Can one of you post a (small) mp3 file that does not work? Thanks
      This may take some time, given what I described in the thread above. In the mean time, can you please post up an mp3 that works fine for you? Those of us who are having this same issue can then see if it works for any of us.

      Thanks,
      Raymond

      Comment


        #18
        Originally posted by rhardcore View Post
        This may take some time, given what I described in the thread above. In the mean time, can you please post up an mp3 that works fine for you? Those of us who are having this same issue can then see if it works for any of us.

        Thanks,
        Raymond
        Attached is wav and mp3 file I just generated on my system. LMS payers (sqqueezebox3, radio, boom,...) played it fine. Both LMS and HS3/Squeezebox plugin are running on a Windows 7 Pro PC when I ran this test (I have LMS on an older ReadyNAS as well and works fine as well). Give it a try. When any of you run into a problem, do you see any error in the LMS logs?
        Attached Files

        Comment


          #19
          Could any of you running into the issue check the media type (sample rate, bit rate, etc) of the files that are generated on your system when TTS is spoken? I have interst in the information for both the wav file and the mp3 file. Attached is the format for the two files from the prior post.

          The wav file is created first by HS3 using the voice that was specified in the configuration. The wav file is then converted to an mp3 file (it that is what you are using).

          Also have you tried not to use the wav file instead of the mp3 format (uncheck the option in the plugin configuration)? Ensure LMS is setup to play wav files.
          Attached Files

          Comment


            #20
            Did any of you had a chance to try the above mp3 files, and check the wav/mp3 files generated in your setup (which is voice dependent)?


            Sent from my iPad using Tapatalk

            Comment


              #21
              Your MP3 worked, WAV didn't

              The mp3 you dropped in worked. I was able to access it via the TuneIn function in LMS server. The WAV didn't. It played one time in what sounded like a tape recorder in fast forward, then didn't play again. I then switched it back to pointing at the MP3 and it worked fine again. Not sure what is different about the way TTS is creating the TTS on your system versus mine.

              Comment


                #22
                Originally posted by pcp View Post
                Did any of you had a chance to try the above mp3 files, and check the wav/mp3 files generated in your setup (which is voice dependent)?


                Sent from my iPad using Tapatalk
                See above post about testing the files you dropped in.

                I just updated to your newest version. Now, I can't even send it commands via the Immediate Script Command window. When I type:

                &hs.Speak "hello, test, test, one, two, three", False, hometroller:*

                I now get:

                <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Feb-04 13:50:02 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Error </TD> <TD class=LogEntry0 colSpan=8 align=left>Running script(2) &hs.Speak "hello, test, test, one, two, three", False, hometroller:*, init error: Syntax error </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Feb-04 13:50:02 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>System </TD> <TD class=LogEntry1 colSpan=8 align=left>Control Panel Immediate Script: &hs.Speak "hello, test, test, one, two, three", False, hometroller:* </TD></TR></TBODY></TABLE>

                Also, I DO NOT SEE MY CLIENTS in the speech client selection box in the events, nor do I see anything in your "SqueezeBox Plugin Speech" menu.

                I can not send anything to any squeezebox clients, nor will it now create any mp3 or wav files in the C:\Program Files\HomeSeer HS3\html\Squeezebox directory.

                This is getting pretty tiresome for a paid for plug in. Again, the Sonos plug in was up and working in 5 minutes. This really needs to be fixed and the documentation fixed of the typos in the script examples sections. Coding isn't horse shoes or hand grenades. There is no "close enough". It is either correct and works, or it's wrong and it doesn't. If you are creating a new syntax for us to learn, no problem; but the examples need to be correct or updated to be correct for whatever release you are on.

                I am using an official HomeSeer HS6 appliance (read: not cheap). All marketed plugins via HomeSeer's market place are required to work with the official HomeTrollers HS sells. This is the HS policy for plug in vendors. I shouldn't need to go and download anything else or rely on any OS hooks that aren't automatically part of the OS in HS6. If there is anything your PI needs to work on the HS6, then it needs to be included with your PI or have extremely clear directions of what is needed to be installed before making a purchase. I don't want a refund, I just want the plug in to work. Some clients don't want to invest in Sonos, and there are some things that Squeezebox can do that Sonos can't. Let's get this going and resolved.

                Thank you,
                Raymond

                Comment


                  #23
                  Originally posted by rhardcore View Post
                  The mp3 you dropped in worked. I was able to access it via the TuneIn function in LMS server. The WAV didn't. It played one time in what sounded like a tape recorder in fast forward, then didn't play again. I then switched it back to pointing at the MP3 and it worked fine again. Not sure what is different about the way TTS is creating the TTS on your system versus mine.

                  The difference most likely is related to the voice you are using compared to the one I am. different voices use different sampling frequencies. Have you compared the details (sampling, etc ) of mp3 files generated in your system with the values i posted in a prior post ? Which voice are you using?

                  The plugin is using the hs.speaktofile script function to generate the wav file which is then converted to an mp3 file using the lame program. The function hs.speaktofile uses a voice to generate the text. Way back i ran into an issue with the tts over t he phone where it would not work with some voices.

                  As far as the wav file, i ran into similar problems and is a lms isue. I recall i had to change some lms settings for the type of media it should suport. I don't recall details.


                  Sent from my iPad using Tapatalk

                  Comment


                    #24
                    Originally posted by rhardcore View Post
                    See above post about testing the files you dropped in.

                    I just updated to your newest version. Now, I can't even send it commands via the Immediate Script Command window. When I type:

                    &hs.Speak "hello, test, test, one, two, three", False, hometroller:*

                    I now get:

                    <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime0" style="WHITE-SPACE: nowrap" align=left>Feb-04 13:50:02 </TD> <TD class=LogPri0 align=left></TD> <TD class=LogType0 colSpan=3 align=left>Error </TD> <TD class=LogEntry0 colSpan=8 align=left>Running script(2) &hs.Speak "hello, test, test, one, two, three", False, hometroller:*, init error: Syntax error </TD></TR></TBODY></TABLE> <TABLE class=log_table_row cellSpacing=0> <TBODY> <TR> <TD class="LogDateShort LogDateTime1" style="WHITE-SPACE: nowrap" align=left>Feb-04 13:50:02 </TD> <TD class=LogPri1 align=left></TD> <TD class=LogType1 colSpan=3 align=left>System </TD> <TD class=LogEntry1 colSpan=8 align=left>Control Panel Immediate Script: &hs.Speak "hello, test, test, one, two, three", False, hometroller:* </TD></TR></TBODY></TABLE>

                    Also, I DO NOT SEE MY CLIENTS in the speech client selection box in the events, nor do I see anything in your "SqueezeBox Plugin Speech" menu.

                    I can not send anything to any squeezebox clients, nor will it now create any mp3 or wav files in the C:\Program Files\HomeSeer HS3\html\Squeezebox directory.

                    This is getting pretty tiresome for a paid for plug in. Again, the Sonos plug in was up and working in 5 minutes. This really needs to be fixed and the documentation fixed of the typos in the script examples sections. Coding isn't horse shoes or hand grenades. There is no "close enough". It is either correct and works, or it's wrong and it doesn't. If you are creating a new syntax for us to learn, no problem; but the examples need to be correct or updated to be correct for whatever release you are on.

                    I am using an official HomeSeer HS6 appliance (read: not cheap). All marketed plugins via HomeSeer's market place are required to work with the official HomeTrollers HS sells. This is the HS policy for plug in vendors. I shouldn't need to go and download anything else or rely on any OS hooks that aren't automatically part of the OS in HS6. If there is anything your PI needs to work on the HS6, then it needs to be included with your PI or have extremely clear directions of what is needed to be installed before making a purchase. I don't want a refund, I just want the plug in to work. Some clients don't want to invest in Sonos, and there are some things that Squeezebox can do that Sonos can't. Let's get this going and resolved.

                    Thank you,
                    Raymond

                    Do you players in the plugin status page? Can you control a player from devices for example stop, pause playing? The speech page will only show content (besides headers) if any text to be spoken is pending.


                    Sent from my iPad using Tapatalk

                    Comment


                      #25
                      Originally posted by pcp View Post
                      The difference most likely is related to the voice you are using compared to the one I am. different voices use different sampling frequencies. Have you compared the details (sampling, etc ) of mp3 files generated in your system with the values i posted in a prior post ? Which voice are you using?

                      The plugin is using the hs.speaktofile script function to generate the wav file which is then converted to an mp3 file using the lame program. The function hs.speaktofile uses a voice to generate the text. Way back i ran into an issue with the tts over t he phone where it would not work with some voices.

                      As far as the wav file, i ran into similar problems and is a lms isue. I recall i had to change some lms settings for the type of media it should suport. I don't recall details.
                      Thanks for responding. Interesting points. You have an option in your menus to specify the voice to use. I presume that is the voice from the system voices on the system running the HS3 an SB PI. What is the syntax for specifying this voice? Also, I can't look at the output files because the TTS function now will not run. No files are generated, so there is nothing to look at unfortunately.

                      I have two options in system speech properties:

                      IVONA 2 Salli OEM - US English female voice [22kHz]
                      and
                      Microsoft Anna - English (United States)

                      Comment


                        #26
                        Originally posted by pcp View Post
                        The speech page will only show content (besides headers) if any text to be spoken is pending.
                        I think I remember reading that now; the queue goes there.

                        I can control the players well. Selection of the LMS items when dropping the media player onto a page in HSTouch never results in anything being populated, but I can control them inside the HS3 web interface. I can also control them fine from individual buttons in HSTouch that I create and link back to the HS3 PI device properties. I hope all that makes sense.

                        Comment


                          #27
                          Is the tts speech proxy enabled in the plugin? Can you post a picture of the tts configuration in the plugin?

                          The voice string in the plugin should be the same as you select in the speaker.exe. Which one of the two is the default and used in the speaker.exe client?


                          Sent from my iPad using Tapatalk

                          Comment


                            #28
                            Originally posted by rhardcore View Post
                            Thanks for responding. Interesting points. You have an option in your menus to specify the voice to use. I presume that is the voice from the system voices on the system running the HS3 an SB PI. What is the syntax for specifying this voice? Also, I can't look at the output files because the TTS function now will not run. No files are generated, so there is nothing to look at unfortunately.

                            I have two options in system speech properties:

                            IVONA 2 Salli OEM - US English female voice [22kHz]
                            and
                            Microsoft Anna - English (United States)
                            As in prior post, the plugin uses the HS3 script command speakToFIle and in the plugin configuration "TTS Voice" you specify the voice to use. See the documentation for behavior if none is specified and how to specify one:
                            http://homeseer.com/support/homeseer...file&criteria=

                            Comment


                              #29
                              Originally posted by pcp View Post
                              As in prior post, the plugin uses the HS3 script command speakToFIle and in the plugin configuration "TTS Voice" you specify the voice to use. See the documentation for behavior if none is specified and how to specify one:
                              http://homeseer.com/support/homeseer...file&criteria=
                              Thank you. That's helpful. How do I specify the target client to stream to? Should I go by the name that appears for the client in the squeezebox controller, - Livingroom
                              Click image for larger version

Name:	Screen Shot 2016-02-07 at 11.57.41 AM.jpg
Views:	1
Size:	35.9 KB
ID:	1183203

                              the instance name from the status page, - Player_000000000000
                              Click image for larger version

Name:	Screen Shot 2016-02-07 at 11.58.56 AM.jpg
Views:	1
Size:	32.0 KB
ID:	1183204

                              or the name that exists for the device that was created? - Player 1 Name
                              Click image for larger version

Name:	Screen Shot 2016-02-07 at 12.04.04 PM.jpg
Views:	1
Size:	29.0 KB
ID:	1183205


                              I am thinking it is probably the first one. Is that right?

                              Comment


                                #30
                                What is the context of thequestion? Is it for the speaktofile script function or something else? The scrip api does not stream, it just creates a file for the text to be spoken. If you want to access the file via an url, as long as it is in the html directory, you can access it via the hs3 url. Can you describe ina few words what you attempt to achieve?


                                Sent from my iPad using Tapatalk

                                Comment

                                Working...
                                X