Audio Analysis and Symbolic Music Note Generation

Sarvesh Satish

doi:10.36948/ijfmr.2026.v08i05.89740

Audio Analysis and Symbolic Music Note Generation

Author(s)	Mr. Sarvesh Satish
Country	India
Abstract	This report documents an analysis of a short trumpet recording, working through the full pipeline fromraw waveform to a playable MIDI file and, eventually, engraved sheet music. I used the Fast FourierTransform (FFT) to look at the recording's overall spectral content and a Short-Time Fourier Transform (STFT) spectrogram to see how that content changed over time. The pYIN algorithm provided a frame-by-frame estimate of the fundamental frequency (f0), which, combined with peak-picking on the staticFFT spectrum, was used to identify musical notes. Mel-frequency cepstral coefficients (MFCCs), spectral centroid, spectral bandwidth, and a chromagram were also extracted to characterize the recording'stimbre and harmonic content.The f0 contour was grouped into fifteen discrete note events with defined start times, end times, and durations, which were then written out as a MIDI file using mido and imported into GarageBand toproduce sheet music. The pipeline worked end-to-end, but it was not clean: the f0 estimate trackedclose to a single low pitch for much of the clip even as the note table shows a wider melodic range, shortnotes near the 0.1-second mark were borderline between real events and tracking noise, and the last~1.5 seconds of the recording — mostly breath and decay rather than a clear pitch — still registered as asustained F4. These issues, and what might fix them, are discussed in Sections 4 and 5.
Keywords	audio signal processing, pitch detection, pYIN, Short-Time Fourier Transform, music information retrieval, au- tomatic music transcription, MIDI
Field	Computer Applications
Published In	Volume 8, Issue 5, September-October 2026
Published On	2026-10-14
DOI	https://doi.org/10.36948/ijfmr.2026.v08i05.89740

View / Download PDF File

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9687-828-838	editor@ijfmr.com

International Journal For Multidisciplinary Research

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Audio Analysis and Symbolic Music Note Generation

Share this