Carnegie Mellon ECE Capstone, Spring 2023 Project by Anita Ma, Kelly Woicik, Anna Gerchanovsky
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). As reiterated in Anna’s report, we have …
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). This week was the interim demo. It …
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).
So I started off the week integrating the PDA with the feedback generation mechanism. I assumed the format in which the input from the PDA would be fed into my feedback generation mechanism– a singular note. However, you know what they say about assuming…
I assumed wrong. Instead of a single note, I would be obtaining a list of values representing frequencies both the user and the song. So, I had to backtrack, and started ideating how I can deal with a list of inputs instead of a singular frequency. But this backtracking was a sign to me to clarify the input output connections with my teammates. I reconvened with my teammates, who were doing the much more laborious task of connecting the front and back end. Kelly and Anna were running into some bugs with headers and the wav files sent over the network. I didn’t completely understand the problem because I am not too familiar with front end systems and thus am not working closely in this progress.
We had some discussions on the module connections– while our design report laid this info out, this level of detail was not covered because we didn’t know about these issues that we would run into. For example, how are we going to sync the backend and frontend together to know where to index into our list of expected note values? We ideated and drew up the hypothetical flow diagram of information.
Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?
I am currently blocked– I will have to wait for my teammates to link the front and back end together for me to make substantial progress in linking the PDA and feedback system together. We are slowly getting more behind, but if we grind next week I think we can make it back on track. I will get frequent updates from my teammates so I can start as soon as I get a concrete answer on the type of input that I will be given.
What deliverables do you hope to complete in the next week?
I am to finish the interface in the next week.
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). This week, I continued working on the …
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? The most significant risk as of right now is linking the backend PDA with the front end web application. This …
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).
This week, I took a deep dive into integrating pitch detection with our web app and also figuring out what signal processing needs to be done.
First and foremost, I got the pitch detection algorithm to work alongside our website! Our website is now able to display a pitch in Hz while the user is singing.
The problem with the RIFF header was resolved by passing in the following to a new wav file: number of channels, frame rate, sample width, and frames that had been written by the Media Recorder API.
After this was successful, I started to look into doing some preprocessing on the audio and what exactly we want to filter. For now, we get the last valid pitch (i.e not 0 Hz and less than 1000 Hz) and pass that pitch to the web interface.
One pitfall I did not expect was that I had been recording in GarageBand using their input monitoring and so by extension all signals were preprocessed. However, after removing the input monitoring and testing the same recorded audio on our website in real time vs on our pitch tracking algorithm after the fact, I came to the realization that we will also need a low pass filter to block pitches < 100 Hz.
As integration took most of my time this week, testing has been limited to singing into our interface, changing notes, and seeing if the pitch on screen changes depending on if I go higher or lower. Additionally, there was a small error that would cause aubio to crash that was removed.
Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?
According to the Gantt chart, I’m integrating pitch detection to the website. Therefore, I’m right on schedule.
What deliverables do you hope to complete in the next week?
Next week I hope to have more robust tests on the pitch tracker for the website and a plan to deal with how much preprocessing needs to be done on our input signal.