How to assign notes to each phoneme? #3

Hi,

I tested a batch of samples using the STARS project and noticed that in the output output.json, the length of ph_durs and note_durs are different.

I'm trying to build a GTSinger-style dataset for SVS model training.
Could you please explain how you assign musical notes to each phoneme?

Is it possible to share the code or the general implementation idea for this mapping?

Thanks a lot!