12.16.17 | Week 16. Blog Debrief
(final week)
video of the final thing in action
video shown before hands-on example to explain some of the synthesis stuff
The housing obviously isn't as polished as it could be, but it can stand up on its own and it can withstand some pretty rough handling without the speakers detaching in any capacity (I can't say the same for the power and ground of the light sensor), which were the main goals of the housing. A more elaborate or even "fixed" sort of housing were initially part of the plan, but since two of the sensors fell through it was ultimately capable of sustaining itself using direct wiring and a single battery.
A heartbeat sensor was going to be included as well, and I had the code (mostly) up and running, but polishing it the night before all three of the wires connected to it snapped - it came pre-soldered. I didn't have an opportunity between the night before and the presentation to try soldering it back together.
The other sensor was the accelerometer, the main limitation being that both my custom music library / libraries and the accelerometer library took up too much space to fit on a single Arduino Uno. Even using PROGMEM for the bulk of the variable calls, dynamic and program memory ended up at something like 94 and 70%, respectively. Part of that could be resolved by utilizing dynamic array sizes for the compositions (as not every one is 64 notes long), but that wasn't something I could realistically accomplish before the presentation.
The colored light sensor and speaker were working properly, however. Initially the plan was to have the Arduino start out in an "ambient" state alternating between random whistling / theremin loops and chirping sounds and to have someone activate it via pulse sensor to start the real sequencer, although since the light sensor was all I had integrated it uses the sequencer state in the event of a dominant color and the ambient state in the event of a "tie" or no input whatsoever.
The colored light sensor worked very effectively, only really having "hiccups" (read: switching to ambient noises between actual segments) if it was moved around quickly. I used some cardstock (read: I cut squares out of notebook covers) to quickly alternate between color detections, which made it fairly easy to debug.
Overall, I'm pretty happy with how it turned out. Obviously, there's more that can be done with it from here, but I'm just glad to have had another chance.
video of the final thing in action
video shown before hands-on example to explain some of the synthesis stuff
The housing obviously isn't as polished as it could be, but it can stand up on its own and it can withstand some pretty rough handling without the speakers detaching in any capacity (I can't say the same for the power and ground of the light sensor), which were the main goals of the housing. A more elaborate or even "fixed" sort of housing were initially part of the plan, but since two of the sensors fell through it was ultimately capable of sustaining itself using direct wiring and a single battery.
A heartbeat sensor was going to be included as well, and I had the code (mostly) up and running, but polishing it the night before all three of the wires connected to it snapped - it came pre-soldered. I didn't have an opportunity between the night before and the presentation to try soldering it back together.
The other sensor was the accelerometer, the main limitation being that both my custom music library / libraries and the accelerometer library took up too much space to fit on a single Arduino Uno. Even using PROGMEM for the bulk of the variable calls, dynamic and program memory ended up at something like 94 and 70%, respectively. Part of that could be resolved by utilizing dynamic array sizes for the compositions (as not every one is 64 notes long), but that wasn't something I could realistically accomplish before the presentation.
The colored light sensor and speaker were working properly, however. Initially the plan was to have the Arduino start out in an "ambient" state alternating between random whistling / theremin loops and chirping sounds and to have someone activate it via pulse sensor to start the real sequencer, although since the light sensor was all I had integrated it uses the sequencer state in the event of a dominant color and the ambient state in the event of a "tie" or no input whatsoever.
The colored light sensor worked very effectively, only really having "hiccups" (read: switching to ambient noises between actual segments) if it was moved around quickly. I used some cardstock (read: I cut squares out of notebook covers) to quickly alternate between color detections, which made it fairly easy to debug.
Overall, I'm pretty happy with how it turned out. Obviously, there's more that can be done with it from here, but I'm just glad to have had another chance.
Comments
Post a Comment