Wouldn't you want to automate all the preparatory work for music analysis, Watson?
I absolutely adore Jeremy Brett’s portrayal of Sherlock Holmes. Whenever I delve into complex compositions, not only by Ryuichi Sakamoto but also by Mussorgsky, Mozart, John Williams, or Hosono, I find myself wandering to 221B Baker Street in my mind. There, Sherlock, voiced by Japan's great actor Shigeru Tsuyuguchi, resides. I engage my brain as if conversing with him, and my inner Sherlock always manages to solve the mystery that troubles me. The most brilliant part of my mind takes on his persona, while my confused self assumes the role of Dr. Watson. We engage in a back-and-forth dialogue until we eventually uncover the truth. That's the feeling it evokes.
By the way, recently, another genius advisor within me is on the verge of becoming personified. He's the protagonist of a manga series that was serialized for an astounding 40 years in a Japanese weekly manga magazine. He’s a police officer stationed at a small police box in front of a park in a town a bit away from Tokyo. Despite his role as a police officer, he's almost a gangster, extremely clever, skilled, and well-versed in the underbelly of society. If he applied his abilities to his duties, he would undoubtedly rise through the ranks, but he always uses them for his childish hobbies or interests (usually to make money), ending up in big trouble in the end. He's also exceptionally proficient in IT.
Analyzing music is very enjoyable for me, especially during tough times, but the preparation can be quite demanding. So, I dropped by the police box in my mind to consult with him. He immediately said, “Delegate such tasks to machines.”
Inspired by this, I tried seeing if I could read sheet music PDFs into a music sequencer. I knew such apps existed, but just recently, I found out that the free music sequencer I always use, “Musescore,” has that function. When I tried it, it read and cleaned up the score perfectly.
Then I asked ChatGPT if I could input an image of sheet music and have the solfège syllables superimposed on each note according to movable-do. The answer was yes, and it could be done using Python. When I tried it, the process stopped midway with a message saying, “You have reached the limit of free usage for ChatGPT Plus, please register to proceed,” which was a bit disappointing.
I've always been hesitant about upgrades in everything in my life. Detaching something that’s already familiar to my body with my own hands is cumbersome and even causes psychological conflict. Perhaps you can empathize when I say it’s akin to the elderly’s reluctance to upgrade Windows. However, as Ryuichi says, “Art is long, life is short.” To analyze music more deeply and extensively, it seems it’s time for me to take on a modest challenge.
Do you know any other technological tricks for analyzing or explaining music as I mentioned above? If you don’t mind, please bestow some wisdom upon this 21st-century primitive.
この記事が気に入ったらサポートをしてみませんか?