After posting his initial AutoVoice home automation video, Doug Gregory posted another awesome video, where he uses sound effects and graphics from the Portal video game series to give his voice command system an identity that fans of the games will love. I love the idea, so I went ahead and borrowed it for my own setup. It’s not hard to do, but here are some quick details to help anyone who wants to do the same.

(If you have no clue what Portal is, you’ve missed two of the best video games ever made)

First off, you obviously need the voice files. They’re luckily very easy to get a hold of, as it seems that Valve (the company behind Portal) doesn’t  run around killing harmless fan sites, unlike most companies. As such, these files are both available and sorted online. The sound files used in Doug’s video are from the turrets in the game, and can be found here, among other places. That link has a list of all the files, with what they say, and can be downloaded directly as MP3 files. Additionally, GladOS’ voice from the game is very similar, and mixing the two does give you more options. Those files can be found here.

All you really need to do is to download the files, place them on your device, and then use them using the Music Play option in Tasker. Note that Music Play makes the task continue before the sound finishes playing, so you might need to add Wait actions if you find that the sounds interfere with other things, like voice control input.

As for the graphical interface Doug uses, I didn’t copy that, but it’s easily done using Tasker scenes for anyone who wants to use it. I also really like his idea of using AutoNotification to enable a notification that allows him to activate the voice system from there, which is done with AutoNotification.

Some examples of my own implementation of this can be seen in the video below. Some of it’s similar to Doug’s, and just taps into my existing home automation tasks. I did however do a couple of new things, like using a combination of Variable Randomize and Music Play to enable more than one possible reply to certain queries. To do this, put all the files you want to use for a specific reply in a separate folder, and name them 1, 2, 3, etc. Then, in your task, use a Variable Randomize action to randomize a variable (like %soundno) with a minimum value of 1 and a maximum value of whatever your highest numbered sound file is called. Then use the Music Play action, and input the path to one of the files. Replace the number in the file name (but not the file extension) with the variable you randomized.

I also added some easter eggs like playback of the turret opera and the two ending songs for the Portal games, based on a comment I saw on Doug’s video on Reddit. That’s done by simply using the Open File option, with default apps already configured.

So with that in mind, here’s my implementation of this: