Week 10: Debugging, Analyzing and Starting to wrap it up!

Hello everyone, welcome to week 10 of my senior project. This week, I finished
my first draft of my senior project (and Library of Congress) presentation and
half of my paper. Besides, I figured out the issued I faced about tracking
together with my on-site mentor Mr. Alyea and other programmers who are
working on Weaver. I have started some baby-steps on audio analysis and
prepare to wrap up my whole internship.

Writing the presentation

Following the advice from the Senior Project Committee and the presentation
guidelines, I finished writing my presentation in the following structure:

– My experience in the internship and major achievements.
– Result showcase
– Description of the internship
– My research questions
– Past research (literature review)
– My methodology on my research questions
– Results
– Analysis
– Citation

While making the presentation, a main difficulty that I faced is illustrating
the algorithms. I solved it by drawing flowcharts on Lucid.app, which provides
flowchart tools for free.

Solving the issues on tracking

Remember the ghosting light issue that I mentioned in the first few weeks of
my internship? I encountered it again when trying to produce better audio.
When trying to get better tracks out of the scanned image, I applied color
filters with different parameters of color weights. However, when the filter
has the red color of weight 1 and no blue and green weights, something weird
happens: there seems to have two track paths where there should only be one. I figured out how to solve the problem by trying to apply filters with different color weights. Finally, I found out that the best weight is (Red: 1.0, Blue: -0.2, Green: -0.2).

Other issues of tracking are solved by simply update my version of weaver, as
known as applying “git pull” on the software repository

## Starting audio analysis

After cleaning up the files and renaming the audios files, I loaded all the
audios I have in the software Sonic Visualizer:

I will provide conclusions of analysis next week. Here are the problems I try
to answer:

– Which way is the best in reducing noises?
– How good is my smoothing method?
– Is there any other obvious problems in the spectrum and how can I address
them?

Next week

ShowcaseI will finish writing my paper and revising my presentation next week. The
progress now is adding some more relevant literature review content, and I
will show them next week. Besides, on next Friday, I will present my work to
the Library of Congress, and I will summarize my experience. Stay tuned!