Summary
We made a few decision changes to incorporate user feedback, and modified the granularity of values that we will extract. We were able to execute function calls every beat of the song. We are coming up with a clean approach on how to narrow down our function call space we can make for a certain time interval. We are working on a few small integrations with the UI as well to make it seamless. The new workflow for the project looks at playing a small snippet of a song using pyaudio by the user, have it recognized by the Shazam, and poll spotify for features, and get its MP3 version, and extract signal parameters for the entire song, playback the full song for the user, and execute commands from the lighting engine after processing the different parameters fed through. The user can interact with the lights at any time during the song. This way we have more real time updates from the user, and we dont require the song to be stored in our directory. We can have just a small snippet of the song through a microphone and have the user play with the lights as they wish.
Risks
Risks with this procedure is that the workflow is dense. Since we added a few components we need to test all our pain points, and make sure that we are able to execute each and every component and get to the next component with a low latency. We had some issues with Python versions, and installation of packages that made it difficult to test across different laptops. Since we are incorporating real time feedback, we need to make sure we process all user commands, and get to our original execution queue with less delays.
Changes
Each of us have taken up additional things, so we have alot of things on our plate. Parth and Abhishek will be working on function selection, and fine tuning the light selection, and the aesthetics. Rachana will be working on integrating the audio modules with Spotify, and the signal processing engines, and she will be working on processing user feedback real time. We want to test with a few user focus groups to see how appealing it is as well.
