Speech engine in MobiLinc Pro (iOS)

Post your feature suggestions and questions here!

Speech engine in MobiLinc Pro (iOS)

Postby Marc_S » Thu Feb 23, 2017 8:48 pm

Took me a while to get MobiLinc Pro working with my ISY but enjoying it ever since..

Had a question on the Speakable alerts for device events, ie, in menus:

<select device>-->
Advanced Options-->
<ON/OFF event>-->
Speak Message

I use this feature and love not having to look at my device every time it alerts, but as a user we always want more of course haha

I searched these forums and didn't find the words speech speak or speakable anywhere, so maybe this is a first (or search don't work :)

In a nutshell, I'm wondering if there's a way to add different speech inflection, speed, emphasis, pauses, tones, etc commands or symbols into the text field where we type the sentence to speak. Any special characters that will work?

I searched around and think the app may be using the Ivona speech engine or Ivona Cloud, which should allow for many different speech enhancements inline (unless the program was designed to ignore them?) but I tried inserting some of the SSML tags into my sentences (from http://developer.ivona.com/en/ttsresour ... /ssml.html) and none worked (all were just read as plain text).

The SSML tags I'd really like to be able use are:

- "emphasis" (so I can distinguish between (calmly spoken) normal info alerts and (excited emphasized) urgent alerts),
- "break" - so I can add pauses in between some words to make them more intelligible or to repeat some words after a delay,
- "speed" - to vary the speed of spoken phrases based on sensor type (would be used with emphasis to adjust perceived urgency)
- "volume" - MobiLinc voices today are unheard (too soft) while playing music.. if you listen for them carefully they're there, just too soft.. not sure why Mobilinc doesn't "turn down" the volume of all other apps when playing a speech statement like many other apps, but the SSML volume tag wold allow me to bark out the alert at extra dB's so nothing can block them.

MobiLinc Developers: can you tell us what speech engine the app uses so we can go look up the inline tag options? And if SSML compliant, can you check your code to see if some option to accept inline SSML tags is disabled and enable it?? Thanks!!

That would rock!

Rather than ask for a bunch of app mods to tweak speech alerts, why not just allow/enable SSML tags that we can embed into our sentences to tweak what is said, and provide a link in the app to a list of SSML tags accepted? I'd also point out (minor but annoying!) that when you type in alert text there's no "test it" button to hear what it sounds like.. maybe a simple add? Today we have to go into the choose voice option, find the voice we're using, retype our sentences there, and presss play (lots of steps)

If anyone has discovered any tags that work today as well, please reply wth info!

And thanks! Marc
User avatar
Posts: 1
Joined: Thu Feb 23, 2017 7:38 pm

Re: Speech engine in MobiLinc Pro (iOS)

Postby AdminWes » Fri Feb 24, 2017 8:17 am

Hi Marc,

Thanks for the feedback.

A couple of comments:

- The voice engine we're using to generate the sound file is an Amazon service. The Amazon service does support SSML, however, we initially opted to not support that version until we received feedback from users. The good news is that it appears possible to add some version of this in.
- Yes, testing the test is a bit inconvenient by going into the voice list.
- The volume we can't control. This is 100% controlled by iOS since it's a native push notification played by iOS without MobiLinc input. The volume SSML tag (if Amazon supports that tag) may help, but I can't guarantee it since iOS is issuing the notifications, playback, sound mixing, etc.

I've passed your SSML request and testing button along internally.

Site Admin
Posts: 2153
Joined: Sat Feb 07, 2009 2:52 pm

Return to Feature Requests and Questions

Who is online

Users browsing this forum: No registered users and 1 guest