- It queries for your voice input (this is a Tasker’s Task, just like your Ring Alarm task in the second tab of Tasker). I’ll show you how to run this later.
- After this Task runs, it transcribes your speech into text, and MANY different Tasker profiles will kick in. All of these profiles have one same context: does the transcription have the keywords I contain? If yes, it triggers the profile, and the task gets run like any other profile.
Here’s how you set it up.
First look: This is the Task that you will run over and over again.
Choose Plugin > AutoVoice Recognize as shown:
Click Edit, then just go back out. If you wish to speak to it via a Bluetooth, select the appropriate option. For us, we leave it like it is.
You can now speak to AutoVoice, and it will transcribes what you say into text. But you’re not doing anything with that text yet, so let’s set up a Tasker profile to… I don’t know, put your phone into silent mode.
This profile will have a State > Plugin > AutoVoice Recognized context. As shown in three humongous screenshots:
Click EDIT. The next screen has three important options:
- Event Behavior: Basically, your whole voice command will be treated as an event (like plugging in or out a 3.5mm headphone jack). This means that the effect will last a total of 2 seconds at best. I’ll show you later how to get around this.
- Command Filter: write in here the words that will trigger this profile. I put in “be quiet for a moment”, which is what I’d want to say. You can put in “shut up” or whatever, even “correct horse battery staples”, it doesn’t matter. What matters is that you can say it accurately enough frequently. Do beware that whatever you say has to contain the exact phrase. If you tick “Exact Command”, you have to be even more precise and say ONLY the Command Filter phrase.
- Command ID and Last Command ID: I’ll cover this in the next blog post.
Once done, click the tick mark. You will be prompted to pair a Task to this profile (what you want it to do). Since we’re trying to make a Silent Mode profile, a Sound Off Task will be appropriate. I already have one made, you can make one yourself. Name it anything you like and have the action: Audio > Silent Mode > On.
Once done, your profile will look like this:
Now, activate it.
Go to your homescreen and make a Tasker Task Shortcut:
Choose the Voice Query ACTION/TASK.
The only thing I hate about Task shortcuts is that you need an icon for each of them. I just chose a panda for mine.
Click the panda, speak your Command Filter phrase, and your profile gets run.
There is one small problem, however. Like I mentioned earlier, your voice command is treated as a short-lived event, so as soon as it’s over (roughly 2 seconds after you finish speaking), the effect wears off. The hack around this is to set your Task (Sound Off in our example) not as the entrance Task (the one with the green arrow), but as an EXIT Task (red arrow). This means that long after the profile is done, the effect will still be applied. Like so:
Message me on Google+ if you have questions. A video should probably be easier to follow. Note that there are MANY MANY ways you can activate the Voice Query Task, not just through the panda shortcut. Swiping an icon with Nova Launcher is one good way. Custom ROMs like AOKP or RootBox, Carbon, ParanoidAndroid, etc. allows you to set up custom “Navigation Ring” targets in addition to swiping up for Google Now. It is by far my favorite method. As seen here in this awesome implementation. You choose your most familiar method. Ask if you need more.