Automatic Music Transcription Quiz

Answers and explanations appear as you take the quiz.

1. What does Automatic Music Transcription primarily output?
   • A symbolic representation like MIDI or sheet music
   • A louder, remastered version of the audio
   • A text description of the song's mood
   • A compressed audio file
2. Why is polyphonic music especially hard to transcribe?
   • The audio files are too large
   • Overlapping notes share and blur harmonics in the spectrum
   • Microphones cannot record more than one note
   • Polyphonic music is always too quiet
3. What was notable about Google's Onsets and Frames model?
   • It generated new songs from scratch
   • It only worked on spoken word
   • It separated onset detection from frame-level pitch detection
   • It removed background noise from recordings
4. Why is the Constant-Q Transform useful for music?
   • It deletes silent sections automatically
   • It converts stereo to mono
   • It speeds up the audio
   • Its frequency bins are spaced logarithmically, matching musical pitch spacing
5. What does an 'onset' refer to in transcription?
   • The precise moment a note begins
   • The end of a song
   • The volume of the recording
   • The tempo of the piece
6. Which model is known for transcribing multiple instruments at once?
   • JPEG
   • MT3
   • BERT-base
   • WaveNet
7. What information does a symbolic MIDI note event typically encode?
   • Only the waveform’s microphone brand
   • A compressed audio bitrate and codec license
   • Pitch and timing such as onset and duration
   • The listener’s preferred musical genre
8. What makes MT3 a multitrack transcription system?
   • It converts every instrument to a single speech transcript
   • It only estimates the global BPM
   • It identifies songs from short audio fingerprints
   • It predicts note events across multiple instrument parts