oops can't remember why I did that. added them expect this 05:15:698 - it is already very dense and the sound is actually not that obvious to heard. As for 05:17:948 - it is also not that obvious to heard imo and I want to emphasize the violin sound 05:18:055 (318055|0,318055|4,318162|1,318162|5,318269|6,318269|2)
I agree with that the sound at 1/4 line is more like echo, but mapping it as normal note seems more confusing since they are still the violin sounds, so I prefer to keep these notes.
hmm at 140bpm I think this should not be a issue. Anyway changed the first one, the second is a little bit weird for me and I will keep current pattern.
I agree that the guitar sound and drums are together here, but I'll keep the pattern since this song is basically full of drums, it is ok to change the taste when it suits.
The idea for these four repeated parts was explained here https://osu.ppy.sh/beatmapsets/1221767/discussion/2541393/timeline#/1808073. For 01:04:802 - the previous pattern is filled with 1/2 short LNs therefore I need these two LNs (01:04:969 (64969|4,64969|0)). For 01:15:469 - I switched to focus other sound and the pattern is then filled with 1/4 LNs, therefore I just map these two LNs (01:15:469 (75469|6,75469|0)) and make a different SV to match it.
hmmm actually it was previously suggested that I should buff this part https://osu.ppy.sh/beatmapsets/1221767/discussion/2541393/timeline#/1846870, and I think the reply there can also answer your question.
I believe the previous part 01:47:802 (107802|3) - is mapped like this, and this part should be harder than that, so the density has increased by a little
The change is to match the fade out here. If I continue the previous pattern, the density at the next section will be suddenly droped by a lot and without any buffering.
The reason is simply no space since first I have to leave enough room for LN, and second this repeated section should be slightly easier than the final repeated section. So I just mapped them as 1/2 jacks as possible
The density here for both two diffs are basically the same, but insane has used lots of stair-like streams so it is a lot easier. Besides that, here for these two repeated parts (01:26:469 - 01:37:136 and 01:37:136 - 01:47:802) I want to make some small but exist diff transitions, so for the first part I mapped less notes, this was intentional.
01:25:469 (85469|0) - it's weaker but still clearly enough for me. 01:25:969 (85969|4,85969|6,85969|1,86136|3,86136|6,86136|0) - yea I agree, these two are definitely louder. As for LN I think only if I change the streams to stairs (as what you do) that would be fine, otherwise it would be messy. So that's some tradeoffs, I personally don't like stairs here, so not gonna use a LN.