[β οΈ JOURNEY] Building MP3 AI Manager for Music Hoarders (Day 31)

Day 31
Hours worked/Total: 2/98.5
Day Recap
Tried low-pass filtering directly on Wave Source, but frequency cutoff doesn’t make a huge difference—very subtle between 400Hz and 4000Hz. Most of the signal is in the low frequencies, making it hard to get a meaningful visualization.
What I Did
- Implemented a Low-Pass Filter:
- Expected clearer separation of frequencies, but differences were barely noticeable.
- High-pass works as expected, letting only high frequencies through.
- Low-pass is too dominant—most audio seems to be in the low range anyway.
- Mid-Pass Experiments:
- Tried visualizing mid-pass frequency bands—still struggling to extract something meaningful.
- Need to find a way to better highlight mids without losing clarity.
Remarks
Filtering works in theory, but in practice, it’s too subtle.
Either my filter implementation is too soft, or most music just sits in the low-end.
At this point, I might need to normalize frequency distribution so mids and highs aren’t drowned out.
Next Goals
Try visualizing low, mid, and high frequencies separately with more contrast.