00:04:120 -> 00:08:120 you dont represent any of the vocal as LN but now suddenly you map them at 00:08:120 (8120|3,8370|1,8620|0,9370|0,9620|3). The sudden jump in LN is quite inconsistent and abrupt. I think a better Idea so that both sections arent inconsistent is to map the more pronounced vocals as LN such as 00:04:620 (4620|1) - 00:05:120 (5120|1) - 00:05:620 (5620|0) - 00:06:620 (6620|0) - etc. This would also be good because in hard diff there are a ton more LN compared to normal diff
I changed the rice at 00:04:620 - 00:05:120 - 00:05:620 - 00:06:620 - 00:07:370 - to 1/2 LN
I decided to changed LN at 00:08:120 (8120|2) - 00:08:370 (8370|1) - to rice
I'm reopening this since it's still messy
Following more pronounced vocal is nice but we will miss some vocal like 00:07:120 and 00:09:120 which are forced to be single due to the criteria
so my suggestion would be to ignore vocal for this section and just follow the instrumental here and maybe create like a consistent pattern
My suggestion would be to actually keep this pattern of 00:04:120 (4120|1,4120|0,4370|3,4495|2,4620|1) - and repeat it