Senior Member
Joined: 02/09/2020
Location: United StatesPosts: 222
| Posted: 03:08pm 02 Nov 2025 |
Some thoughts from listening to the YouTube video:
* The sequence is over 40 minutes long. I was able to convert it to MP3 using one of the on-line "YouTube to MP3" converters. The resulting mid-quality file is 36 MBytes, far too large for MMBasic to play unmodified. Message me and I'll send you the MP3 file.
* If you use GoldWave or another audio editor, you may be able to break up the 40+ minute recording into individual songs, then save them individually to the SD card. Your MMBasic program could pull each in sequence, or randomize the sequence if you'd rather.
* There are a number of paid and free tools to convert an MP3 to MIDI; "Basic Pitch" from Spotify hung up and never completed. I didn't download "AnthemScore" - it's free trial only will do 30 seconds of audio, and I don't have a use for this otherwise.
* Listening to the first two songs, I noticed that the span is over an octave. If the main range is from C to C, there are times when a lower B flat is played.
* Playing the MP3 file in the background while writing this, I hear that most are in the key of C, but some aren't, with an even broader range of notes.
* There are also times when multiple notes are played at the same time, for harmony. It sounds like the originators may have had spring-wound music box mechanisms to record. The music box "pins and springs" mechanism may provide you with the mental model you need.
* Harkening back to the dawn of my career in the 1970s, you may be able to set up "tone detectors" using LM567 phase-locked loops, one per Xylophone key. Your set of ~20 of them would be tuned to each of the notes, and when detected, strike the Xylophone key through one solenoid per key. This would not accommodate the dynamics (softer vs. louder) that the music sometimes employs.
Do you envision writing a MIDI interpreter for the Pico?
Edited 2025-11-03 01:16 by NPHighview