Using computer vision to read the existing tab from the page would be more efficient and precise than trying to use audio processing to isolate individual notes from a single instrument in a potentially busy audio track. The existing tabs also nail down which string and fret is used for each note, which would be a difficult task for audio processing because there are multiple combinations of string and fret for any given note. For example, an open G can also be played on the 5th fret of the D string, or the 10th fret of the A string, or the 15th fret of the E string.
I understand but would assume it could be solved by chord/positional groupings or similar. Manual/human tab transcription need to infer positional assumptions anyway; some artists will play the same songs different ways too ala Bob Weir
the existing tab from the page would be more efficient and precise
But what about songs that aren't made into a YT video?
Comments
Using computer vision to read the existing tab from the page would be more efficient and precise than trying to use audio processing to isolate individual notes from a single instrument in a potentially busy audio track. The existing tabs also nail down which string and fret is used for each note, which would be a difficult task for audio processing because there are multiple combinations of string and fret for any given note. For example, an open G can also be played on the 5th fret of the D string, or the 10th fret of the A string, or the 15th fret of the E string.
I understand but would assume it could be solved by chord/positional groupings or similar. Manual/human tab transcription need to infer positional assumptions anyway; some artists will play the same songs different ways too ala Bob Weir
But what about songs that aren't made into a YT video?