I don't have a ton to add outside of the pictures - it seems like everyone picked it up pretty quickly this semester. It feels like it took longer when I originally took the course, anyway.
Starting this week, I laid out my endgame plans for the project: - The Arduino will use three sensors: colored light, pulse, and accelerometer. - Pulse (or if none is detected, accelerometer) will both adjust the tempo and offset the key (up to 1 octave / 12 semitones in either direction) based on speed or heart rate. - As for actual compositions, there are four or five "layers" to go through, the main three of which were cannibalized from an old composition I wrote in Greg's class back in Fall 2016. The three musical tracks are a fairly barebones variant (uniform note times, volatile instrument changes), a jazzy version with frequent smaller notes splitting up the bigger ones at 90bpm as well as having lengthy instrument "solos", and a drum and bass variant at 192bpm that rapidly cycles through different instruments in a demoscene sort of fashion. - In the ambient state, the Arduino alternates between whistling / theremin tunes and chirpy sounds. - Af
Started working on ADSR envelopes: boop While most of the code (firing / silencing notes, making checks, etc.) is only fired as needed (the entire thing doesn't use loop() in any capacity), the ADSR envelopes are being fired as part of a do-while loop when the note is playing. Due to their nature of changing volume in semi-realtime, achieving properly smooth envelope effects requires a fairly frequent rate of adjustment.
Not a whole lot to show visually this week, although I followed up with Thomas directly to show off some of the new sounds, now that everything's been transposed and stored in memory. There are roughly 30 unique instrument presets and well over 60 unique composition segments, many of which are between 5-10 seconds long apiece. These compositions can be seen in the blog post for the final week. The code is 100% complete from the musical side of things - everything works at a minimum expected standard with a level of consistency, and likewise all of the "canned" aspects like the music and instruments are written out. It's mainly a matter of fully integrating the sensors at this point.
Comments
Post a Comment