How do I use the Speech library on Mac,iOS and Android?
Starting with version 9.6 Livecode includes a new library that provides functionality for synthesizing speech for macOS, iOS and Android. The library allows for choosing from among the system voices, altering the speech rate and pitch of the voice and speaking an utterance.
Please note, this speech library requires Indy or Business editions of LiveCode to work.
Create a new test stack
Create a new stack and add a field and a button to it. Name the field for example "Message2speak".
Create the stack script
Before we can use the library it has to be initialized. On Android it is mandatory that you initialize it in (pre)openstack.
So the suggested approach is to call "speechInitializeLibrary" once on (pre)openstack, and then call "speechFinalizeLibrary" on closeStack.
In that way, the library will be ready for use while your app is running, and it will be finalized once your app closes.
Create a stack script and add the following script to it.
on preopenstack
speechInitializeLibrary
-- to be able to use the library we have to initalize it first
end preopenstack
on closeStack
speechFinalizeLibrary
-- when the library is no longer needed we should cleanup resource using the above
end closeStackCreate the button script
on mouseup
speechSpeakUtterance the text of fld "message2speak"
end mouseup
Standalone settings
To make use of the speech library you'll need to manually select it in Standalone Settings under Inclusions.
Make sure that you've selected "Select inclusions for the standalone application" unger General in the standalone settings
Under Inclusions select Speech Library.
Select the platform to which you want to deploy the app. Please remember that only Mac OS, iOS and Android are supported.
In this lesson we deploy to Mac OS.
Now close the Standalone Settings window and save the stack.
Testing the our Speak_Example app.
To test our app either open Development menu and select Test (1) or press the Test icon in the Livecode toolbar (2).
Livecode IDE prompts with a Save dialog. Press Save to save your stack before testing.
Livecode now shows a little status stack giving information about the build process.
After some seconds you should see the test app on your Mac OS desktop. Enter some text into the field and press the Button. You should hear the spoken text.
There are additional handlers and messages available in the library to the ones used in this lesson.
Handlers
- speechSetRate (allows to change rate of speech)
Syntax: speechSetRate pRate
where pRate can be according to the dictionary between 0.01 and 2
Example: speechSetRate 2
- speechSetPitch (allows to change pitch of speech )
Syntax: speechSetRate pPitch
where pRate can be according to the dictionary between 0.01 and 2
Example: speechSetRate 2
- speechgetVoices() (retrieves a list of all available voices. On Android this is supported from Android version 6 and above)
Syntax: speechGetVoices()
Example: put speechgetVoices() into tVoices
- speechSetVoice (allows to specifiy the voice to use. On Android this is supported from Android version 6 and above)
Syntax: speechSetVoice pVoice
where pVoice is the name of one of the available voices
Example: speechSetVoice com.apple.speech.synthesis.voice.Hysterical
Messages
speechLibraryInitialized This message is sent when the library has been initialized
Could be useful if you want to execute some code after the library was initialized.
e.g. to fill an option menu button with all the voice names.
speechUtteranceDidFinish This message is sent when an utterance is finished speaking
Michael McLean
After spending a day with my students working on this it appears this library is only included on indy/business version, not community version. Is there any way I can just buy the library extension without having to buy an indy licence? Otherwise I'll have to give up the project which would be a shame.
Heather Laine
I'm afraid this library does only work in Indy and Business editions, it is not possible to buy or use it separately. Sorry I can't help with this :(
Michael McLean
Thanks for getting back to me. That's a shame. Seems strange that the features already available in the windows version are available for other operating systems but not included in the libraries. Anyway, not your decision I'm sure. I'll try something else. Even being able to purchase specific libraries would be better than $500 q year or nothing. Well I guess it's something for the powers that be to consider.
simon
Hi, I'd like to know if there is any way to instead of playing the generated speech immediatly save it to file for later use...thanks
Elanor Buchanan
Hi Simon,
What is is that you want to achieve? You can just call speechSpeakUtterance when you want the text to be spoken.
Or do you mean you want to create a sound file?
Kind regards
Elanor
simon
Hi Elanor, the idea is to save the text to be spoken to a file to be played later instead to be played imediatly after being converted. Same question was posted on the forum (https://forums.livecode.com/viewtopic.php?f=8&t=35203) and according to the answers it is not possible at this time...
Thanks
Trevix
Lately, on Android 10 devices, sometime, not always, I receive the following error, after calling speechSetVoice tChosenVoice:
864,19,1, cannot convert value
I guess the extension does not fail silently if it doesn't like the tChosenVoice.
Is this true?
I could put a try end try, but it would be better to know what is going on...
Thanks
sam norris
Hello,
That is quite likely to be the case. Could you please file a bug report for this with a reliable recipe/specific problematic voice?
Trevix
Unfortunately the error report arrive automatically from some customer devices (Android 10) and I was never able to reproduce it.
The reported error is this one:
ORIGINAL ERROR:
863,19,1,runtime
864,19,1,cannot convert value
865,19,1,speech.lcb
866,19,1,1051
897,19,1,1
865,19,1,speech.lcb
866,19,1,455
897,19,1,1
573,19,1,speechSetVoice
253,19,1
241,18,1,mouseUp
353,0,0,button id 2049 of group id 2047 of group id 2035 of card id 1998 of stack "r_Settings" of stack "/data/app/com.trevix.it.segnapunto-9diYrj1nXb4gNoTMlDU-rg==/base.apk/mainstack".
My question is: do libraries (and widgets) are usually constructed inside LCB with a solid error handling, that pass info to LC scripts, or these reports are mostly used only by the developer of the specific library or widget?
I say this because in the past I often had "impossible" errors (like ...line 0 of..) that forced me, to be in the safe side, to rebuild the same functionality using LCS (like the segment widget, for example)
Thanks
sam norris
Hello,
We have made a bug report about this and will investigate. If you get any additional information about this, please post it at https://quality.livecode.com/show_bug.cgi?id=24526
Trevix
I just noticed that on OSX (and I guess iOS), running a speechSpeakUtterance "text B" while another phrase is being spoken ("text A"), correctly interrupts this last one ("text A") replacing it with "text B".
But this is not the case on Android. There seems to be no way to interrupt a TTS on Android when started.
This, is an inconsistency that I would like to be fixed, if possibile.
Thanks
sam norris
Hello,
I can confirm and have reported this behaviour as a bug. You can find that report at https://quality.livecode.com/show_bug.cgi?id=24583
simon schvartzman
Hi, can you please tell me how to use "speechSetVoice" in Android?
In the example it show how to do in Apple (even though I don't know where you get "com.apple.speech.synthesis.voice.Hysterical").
Anyway let assume (in Android) I want to set it to "nl_BE,nl-BE-language" as returned by "speechgetVoices"
Many thanks in advance...
Alice Norris
Hello,
When using the sample stack (having changed the speechSetVoice line to just: speechSetVoice the label of button "VoiceMenu"), the closest match I can find to the line you gave is "nl-BE-language", which when chosen causes the app to speak in what I presume is the correct voice.
The appropriate call would seem to be: speechSetVoice "nl-BE-language"