Month: March 2023

What did you personally accomplish this week on the project?

This week,  I finished up a bit of platform integration code, and after working with Katherine to discuss software integration I decided that it would be best to go back and rewrite large portions of my own code. This is both for ease of integration, and ensure that the code is easy to maintain, modify, and work on for someone else. I’ve gotten through parts of it, mainly the Python end of things. This involves some code that manages sending messages to Arduino and representing notes in the code. The note representation hasn’t changed much, but it has been refined. For testing purposes, it was mostly just a number, a length, and a checksum. Now, it’s a list that contains a length, a list of notes that are represented as their MIDI values, and finally a checksum. The current implementation has better support for chords, because the previous simply relied on the Arduino interpreting them fast enough to queue them into the same time step. The clocks of the computer and Arduino won’t be synced, so there’s no need to add a start value (rather, it would be difficult to, and it is currently unnecessary due to the speed of the current pipeline). For more info on the representation, notes will be sorted into measures, which are grouped into phrases. There will be one phrase that constantly updates which contains eight measures. The measures have no set length (in Python), but do have a maximum length of timeSignatureNumerator*timeSignatureDenominator*8. This means that in 4/4, you can have four sets of eight 32nd notes, which is generally the smallest subdivision anyone will have to deal with on a daily basis. Measures are ordered, as are phrases, but note lists are not. However, ensuring that there are no duplicate notes is important.

If I decide to make a version that does not enforce strict timing, I can just add a flag to the note list that determines whether or not they will be played immediately or sent to the phrase+measure list set.

Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?

I had initially planned on finishing up integration and testing around this time, so it’s debatable as to whether or not I’m on track, but I plan on finishing as much as I possibly can before the interim demo.

What deliverables do you hope to complete in the next week?

I plan on expanding upon what I already have and making it work with Katherine’s code, as well as letting the Arduino sequencer interact with the solenoids.

What did you personally accomplish this week on the project? Give files or
photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week (12+ hours).

I finished up the interfacing code for having our computer talk to the Arduino. It sends data in chunks, where certain symbols (like ! or @) delineate portions of data. This way, we can ensure that the data we send is accurate by marking different blocks of data and including an ending checksum. It also allows for easy chord representation and reading (parse through bytes until you see a block footer, add every note in the block to a “play next” list). This code works on a small scale, but I’ll need to make sure it can work with large amounts of data as well.

Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?

As it stands, I’m currently on track.

What deliverables do you hope to complete in the next week?

I want to tighten up my algorithms and ensure they produce at least somewhat nice sounding results. This involves tweaking Bayesian Updating and ensuring that the note queue functions properly at all times. I’m also going to add phrase support for the chord estimation, so that we can generate cycles instead of walking along one never-ending chord line. This also helps with deciding on weights for certain chord probabilities.

What are the most significant risks that could jeopardize the success of the
project? How are these risks being managed? What contingency plans are ready?

HW: Before this week, my biggest concern was not having enough power but I realized that the power supply I use have +30/-30 and 2 A each — and the maximum power rating is 120W. So, I realized that I can just add 30V/2A and -30V/2A to create 120W power source for the circuit. But now, my only concern is that how I am going to build the mounting system and how we are going to integrate the generative mode. If the generative mode taking too long to integrate by the interim demo, we may have to resort to only doing simple-note-playing mode by the interim demo.

Music SW: The concern this week was data transfer. There are ways to send data to an Arduino over serial, but this data generally isn’t marked as anything specific. So, we needed a way to ensure we can read in data and send it down the proper paths (note queue, actuator system, etc.). Generative mode needs a bit of work, but the hope is that it should be done in time for the interim demo (no point in just hoping, though. I’m also working on finishing it and setting up a data pipeline!).

SW (CV): My biggest concern at this point is detecting without environmental interference causing an issue – the plan is to try to switch it to symbol detection. We could have a specific symbol drawn out on a small piece of paper that is then stuck on a hand, for example, that the OpenCV uses to track where they want to signal. This would solve the issue of someone wearing the signaling color or someone in the background wearing it. The contingency plan would to go forward with color signaling, since that is working, and just isolate the environment with a black backdrop.

Were any changes made to the existing design of the system (requirements,
block diagram, system spec, etc)? Why was this change necessary, what costs
does the change incur, and how will these costs be mitigated going forward?

HW: Not yet!

Music SW: Ditto.

SW CV: No.

Provide an updated schedule if changes have occurred.

HW: None!

Music SW: Same!

Software: None, except working to switch to symbol detection this week in addition to the feedback.

This is also the place to put some photos of your progress or to brag about a
component you got working.

HW: I was able to power three solenoids simultaneously!

Music SW: Nothing to see, but I set up some pretty nice data structures for data transfer, at least in my opinion.

SW: Looks cool on video, but image captures don’t look much different. Both normal note generation and generative mode is working and you can switch between them which is cool!!

What did you personally accomplish this week on the project? Give files or
photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week (12+ hours):

Before Spring Break, I implemented a somewhat simplified version of Bayesian Updating. This will allow us to predict chords as the user plays. One thing to note is that the probabilities of every chord that doesn’t contain the played note will slightly decrease, but the values in the array are normalized between 0 and 1, so this isn’t apparently. This also means that the selected chord will always have a probability of “1.0” with this algorithm. Although it does simplify some aspects (like visually picking out the right chord), it can be slightly misleading, as we aren’t really “sure” that the most probable chord really is the chord being played over.

Some output from the algorithm can be seen below (array indices are [C, D, E, F, G, A, B]):

“ Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?

Thankfully, I am on track and can immediately begin working on integrating all of the musical systems together. This includes computer communication with the Arduino.

“ What deliverables do you hope to complete in the next week?

I plan to integrate the Bayesian Updating and general signal transmission code with the Arduino’s actuator code.