-
Notifications
You must be signed in to change notification settings - Fork 26
Open
Description
Hi there,
First, thank you for putting together this dataset!
I have a question regarding how the pitch range label is created. According to the paper: "We classify samples into 7 different groups of pitch ranges based on the mean pitch of the notes.The sorting criteria is as follows: very low: C-2 - B0, low: C1 - B1, mid low: C2 - B2, mid: C3 - B3, mid high: C4 - B4, high: C5 - B5, very high: C6 - G8."
When I inspected the dataset, I found a mismatch between the description and the real data. For instance, in commu00001.mid, the mean pitch is 69.03 which should fall into the mid high range. But it's labeled as mid in the metadata.
Could you please clarify this? Many thanks!
Best,
Nicolas
Metadata
Metadata
Assignees
Labels
No labels