Quote:
Originally Posted by
ardis
➡️
So can this detect complex jazz chords? Can it also transpose?
The transposition is extremely good (as one would expect from zplane) but not only can it not "detect complex jazz chords", it's not very good at the basics.
deCoda allows one to name chords manually as follows:
maj, min, dim, aug, sus2, sus4, maj7, min7, 7, dim7, hdim7, maj6, min6, 9, maj9, min9
Even in the simplest tests it can't usually detect them itself.
I created a 16 bar wave file with each bar being four crotchets of a chord in root position. I used a piano sound and the results were as follows when the velocities were fixed at 72:
Chord deCapo's interpretation
----- -----------------------
C maj C
C min Eb
C dim A dim
C aug Ab aug
C sus2 Bb
C sus4 Bb
C maj7 C
C min7 Eb
C 7 C
C dim7 A dim
C hdim7 A dim
C maj6 A min
C min6 A dim
C 9 C
C maj9 C
C min9 Eb
With velocities that varied slightly, the results quickly became more inaccurate and with real-world material, it can be very wide of the mark.
deCoda has some promise but the harmonic analysis isn't satisfactory at the moment.