When you make a voice search or any other voice input on Android, there's a complex process that goes on behind the scenes. Your voice is recorded, transmitted to Google's servers, analyzed and converted into a text string, then either passed on to the relevant web service (like Google Search) or sent back to your device. It's usually almost instantaneous if you have a decent Internet connection, but therein lies its one weakness: you do have to have that connection in order for it to work. The rudimentary offline system (in Android since Jelly Bean) relies on a relatively unsophisticated vocabulary and detection system that's slow and less powerful than the connected version.
João Dias, also known as joaomgcdon the Play Store, is one of those developers who are never, ever, content with the current capabilities of modern smartphones. He wants them to be more powerful, respond to more commands, allow more interactions, all from more interfaces. His AutoVoice app has been available for a while, allowing you to harness the OK Google interaction scheme to automate plenty of new actions and issue commands that Google's default algorithms don't yet understand.
Now AutoVoice is getting a lil' sister app, an AutoVoice Chrome extension for your Windows, Mac, and Linux computers. Thanks to it, you can perform the same actions on your phone, but while sitting at your computer (or from another phone too), like taking screenshots, sending messages, hanging up on calls, and more. João has made a demo video to show you the possibilities.
Most of the app updates this week were relatively quiet, with the notable exception of Play Music with its new ad-supported radio feature. That doesn't have to mean some of the updates don't have something new to offer. The Google app (formerly "Search") was bumped up to v4.8 yesterday, but it doesn't seem to have any noticeable changes right now. However, a look under the hood reveals some pretty interesting features on the horizon.
Disclaimer: Teardowns are necessarily speculative and usually based on incomplete evidence. It's possible that the guesses made here are totally and completely wrong. There is always a chance that details may change or plans may be cancelled prior to the launch of a new feature discovered in a teardown.
If you want to do absolutely everything from the "OK Google" voice prompt in Android... well, you can't. You can't, say, fire a rocket at the moon, or end world hunger, or have a dachshund puppy delivered to your doorstep. But developer João Dias is trying his darnedest to make all of that happen, and with the latest update to his AutoRemote app, it's getting closer and closer. You might even be able to make that rocket thing happen if you know someone at NASA.
Version 3.0.43 adds integration with the popular If This, Then That (IFTTT) web automation service, and by extension, all of the various web services and automated gadgets that IFTTT connects to.
Here's something that might just blow your mind. If you perform a voice search and Google misunderstands you, or you happen to garble your words, there's a quick way to take care of that. Just follow up with a second search that begins with "No, I said..." Google will then replace the incorrect word with what you said the second time.
Here's an example. Let's say you're in a hurry to get to our lovely website, and you say "OK Google, take me to Android Police." It doesn't hear you properly, and somehow you end up with Android mobile. A simple, "OK Google, no I said police" will take care of that issue just fine.
The Touchless Control feature on the Moto X and Verizon's 2013 DROIDs (Ultra, Maxx, and Mini) is already one of the coolest Android manufacturer add-ons around. Today it gets just a little cooler: the Touchless Control app was updated to add a "What's up" command. This voice command will make your phone read out the unread notifications in your status bar. This should be great for hands-free updates while driving. Or bathing.
Here's an example of a single email:
The vocal response will adjust based on your notifications - for example, if you have more than one unread email, it will simply give you a number.
Since its introduction, Google Glass has been in the unfortunate position of having relatively limited functionality. However, with a steady stream of updates and eventually the emergence of the Glassware tab in the MyGlass interface, we've known the elusive wearable was due for some more exciting things. A few weeks after announcing plans to add Play Music to Glass, Google has quietly added it to the list of apps supported on the elusive wearable. Alongside this release, the Accessories store also launched the Stereo Earbuds at the hefty price of $85.
Music can be activated by simply calling out 'Ok, Glass, Listen to' followed by the name of an artist, playlist, album, or song.
Have you noticed that sometimes your Nexus 5 reverts to the old Ice Cream Sandwich-style pop-up for voice dictation in some apps, as opposed to the less disruptive "endless dictation" on-the-keyboard style? You're not alone. Our fearless leader Artem found that his N5's microphone icon was missing from the default keyboard, and using contextual mic icons (in search boxes and other places) caused the aforesaid behavior. It was also affecting third-party keyboards like SwiftKey.
As it turns out, at least some Nexus devices and updates are being sent out with the "Google voice typing" option disabled in the Language and Input menu.
Show of hands, who uses CyanogenMod? Oh, you do? I've got some good news: your camera is about to become a little more interested in what you have to say. A few hours ago, CyanogenMod announced that the included camera app will integrate a voice-activated shutter mode. Judging by one of the pictures, a new time-delayed mode will be part of the package, as well. Take a look:
As you can tell, a few different words can be used to activate the shutter, currently Cid, Whiskey, and Cheese. We also see that the voice-activated mode must be turned on for each photo you wish to take.
When Google announced Keep last week, one of the coolest features we learned about is the ability to accept the "note to self" command that has been part of Google's Voice Actions since the Froyo days. Previously, this would send an email to your own account with the transcribed text and the original audio file. Keep allowed users to send that data to a proper note-taking app instead. Well, as it turns out, Catch wants in on that voice action, so in a recent update, it's added the ability as well.