Tuning Center Note Calibration

[baconpaul]

As reported by @JackyLigon on slack

@baconpaul Hi. Just conducted a small SCL measurement test, and found that with both rational tunings (just-intonation specified in ratios), as well as tunings specified in cents, that they are generally tuned with their base pitch at 261.6 Hz, however, in every case where the tuning had greater or less than 12 notes to the octave, that there was an offset of so many degrees.

For example with 10-equal, while the base pitch is dead-on accurate, the starting MIDI Note is -2 lower, and starts on MIDI Note 58.
Jacky Ligon 4:15 PM
While loading a rational 6-tone SCL file based on harmonics 6-12 has an offset of -6 MIDI Notes, and the tuning starts on MIDI Note 54. Again, the sounding C @ 261.6 Hz actually starts on MIDI Note 54 on the controller:
|
HD2 06-12 - Harmonic division of 2: Harmonics 06-12
|
0: 1/1 0.000000 unison, perfect prime
1: 7/6 266.870906 septimal minor third
2: 4/3 498.044999 perfect fourth
3: 3/2 701.955001 perfect fifth
4: 5/3 884.358713 major sixth, BP sixth
5: 11/6 1049.362941 21/4-tone, undecimal neutral seventh
6: 2/1 1200.000000 octave
When the time arrives, I can share a small archive of SCL/KBM files so that we can work from a common starting point when testing. The good news, is that your concert C base pitch is accurately tuned, although offset in terms of the 1/1 mapping of the starting MIDI Note.
For 15 tone equal temperament (specified in cents) the pitch of C @ 261.6 Hz, falls on MIDI Note 63.
For 17 tone equal temperament (specified in cents) the pitch of C @ 261.6 Hz, falls on MIDI Note 65 on the controller.

My plan of attack here is to tackle this in the unit test framework first so I'm ready for when I tackle KBM.

[baconpaul]

Just an update before I put my laptop away for the year

1. baconpaul/pitch-detect-unit-tests has unit tests to check tuning
2. turns out the problem @JackyLigon reports is due to the fact that I tuning center at note 48 / frequency 130.8 not at note 60 / frequency 261.6. Simply changing the constants in SurgeStorage would adjust this I think but that requires testing and will break folks who have tuning saved with != 12 note scales.

So more work to do here. Also #1041 would solve a lot of the problem