This project is no longer actively maintained by the Google Creative Lab but remains here in a read-only Archive mode so that it can continue to assist developers that may find the examples helpful. We aren’t able to address all pull requests or bug reports but outstanding issues will remain in read-only mode for reference purposes. Also, please note that some of the dependencies may not be up to date and there hasn’t been any QA done in a while so your mileage may vary.
For more details on how Archiving affects Github repositories see this documentation.
We welcome users to fork this repository should there be more useful, community-driven efforts that can help continue what this project began.
MixLab is an experiment that makes it easier for anyone to create music, using simple voice commands. It's part of the Experiments With Google platform, launching as part of our newest category - Voice Experiments.
Just say things like "Play me a funky bass," or "Add some jazz drums”, and Dialogflow will parse your intent into an actionable response, creating interesting songs at your command.
MixLab is available on the Google Home and all Assistant platforms using Actions on Google, so you can start making music by saying "Hey Google, talk to Mix Lab".
And you can jam out with some fun visuals in our web experience, available at g.co/MixLab.
This is an experiment, not an official Google product. We will do our best to support and maintain this experiment but your mileage may vary.
As with most of our experiments, this repo has 3 parts of the production system available to you:
Dialogflow allows for easy exporting and importing of projects in json format to help facilitate both versioning and sharing, as explained here.
Included is a slightly modified version of our MixLab app, leaving specific project and webhook id's to be updated by you, the developer.
Fairly self-explanatory, the website/directory contains all the code and instructions for running your own version of MixLab locally. We've even spun up a development music lookup API available through localhost so you can use the same music and database we use in production. See below.
Dialogflow voice input is passed in, then reacts and responds based on intent action and parameters.
Finally, the webhook is where we add advanced handling to Dialogflow, using the Actions on Google NodeJS implementation.
When a request comes in from the Assistant (via Home or on device or website), Dialogflow parses intent like normal, passing us intent action and specific parameters based on that intent.
For example, the addInstrument intent will likely pass along a genre or instrument value that
can be used to create a proper call to our /lookup API to find stems to play.
This code is meant as an overview on how to handle AoG, but won't work out of the box.
This is due to not including the /stitching API, which is a processor-intense service
responsible for mixing our music stems together before returning to the Home to play.
The stitching service uses Rubberband for changing tempos and SoX for combining and looping audio files.
Create your own stitching service, and replace the link on line 36 of response.js
This should be able to injest the stem array from our dev /lookup API
Update package.json to point at your app id for staging buckets
Push your webhook to your Google Cloud App, and then use the link to that webhook in the Webhook section of Dialogflow, in the Fulfillment tab.
Then once you test it out on Dialogflow, most intents will be routed to your webhook, which you can follow along in either Firebase Functions or Google Cloud Functions logs.
http://mixlab-dev.appspot.com/ - top level is test html for /lookup POST calls, check out the Inspector to see how it works. This is open to localhost:8080 for you to use while testing.
- /song GET is test like above, POST is actual service
- /hash GET is test like above, POST is actual service
Anthony Tripaldi, LAD, Maya Man, Prit Patel, Yotam Mann, Use All Five, and friends at Google Creative Lab.
Blake Straus - Composer & Musical Director
Alek Fin - Producer/Engineer
Blake Straus - Fretted instruments, basses, keyboards
Dylan Meek - Keyboards
Roland Gajate Garcia - Drumkit and percussion
DD Horns - Brass and woodwinds
Morgan Paros - Violins
Andrew Synowiec - Fretted instruments
Corbin Jones - Basses and sousaphone
Alison Balbag - Harp
Ian Mallitz - Didgeridoo
Daniel Braunstein, Christopher Dwyer, Zac Sager, Mashadi Maximus
Recorded, Mixed, and Mastered at Pro Audio LA
Copyright 2017 Google Inc.
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.
We encourage open sourcing projects as a way of learning from each other. Please respect our and other creators’ rights, including copyright and trademark rights when present, when sharing these works and creating derivative work.
If you want more info on Google's policy, you can find that here.