00:01:050 -> 00:21:049 - this and normal diff feels very monotonous without any significant differences and interesting cues. i feel like going for vocal-only focus would do a benefit. example would be deleting non-vocal - 00:03:383 (3383|3) - and changing some into LN - 00:07:049 (7049|3,7383|1,8049|3) -
02:11:716 - applicable to all similar section
00:07:049 - till 00:09:049 - ,
the vocal LN usage is a bit questionable here.
the sudden break at 00:07:716 (7716|2) - doesn't feel natural since the long vocal sound is still there.
the same long vocal sound is also available at 00:08:383 (8383|3) -
there are two ways to fix this.
or
I personally would suggest to go with option number 1
Technically, there's nothing wrong with this but if we compare it with normal diff, the gap is quite wide.
I would personally suggest to split this 1 4/1 LN into 4 1/1 LN.
if you agree, could also do a bit LN split here 00:38:383 (38383|0) - , 00:41:049 (41049|2) - , 01:14:383 (74383|2,75049|0) - , 02:07:716 (127716|1,128383|0) - , 02:30:383 (150383|2) - , 02:49:049 (169049|3) - , 02:51:716 (171716|0) - , etc
how about to remove 00:21:549 (21549|0) - ?
to avoid using more than 5 consecutive 1/2 notes.