i figured i'd give this some attention because its nice that someone else is out there exploring the neat behaviours of NaN.
for ages, i tried to brute force a method to place hit objects so far off screen that i could create "dormant" sliderbodies with hitcircles which weren't meant to be hit by the player. it was inspired by how poor Flashbacklog was in forcing misses constantly, but I was incredibly interested in using it as a design choice to create things that were normally impossible. imagine XNOR's tetrahedron or Transform's butterfly, but without needing to be hit as a stream. i loved aspire mapping and i was obsessed with the implications of this.
and amazingly, the issue i struggled the most with was getting the sliderheads far enough off screen to make the approach circles off screen as well. not even the forced misses or anything. i had solutions for that. i just could not get the damn hitcircles far enough off screen. i knew that hit objects could go down to y:512 even though the playfield ended at y:384, but the approach circles could still be seen. of course, i could change the circle size to 7 or higher, but even as an experimental mapper, that was crazy to me. i could set the coordinates of the object itself to NaN, but that put the object so far off screen that the slider couldn't reach the playfield. i actually put up a batsignal on these forums asking for help. nothing came of it, and i kinda gave up completely.
eventually, i did come across this method myself. but, at the time, i thought it was so completely deranged that i'd never be able to use it in a proper map. in retrospect... well, i was still somewhat right, but i should never have discarded it completely. hey, being deranged is the point, right?
the fact is, this method kinda just works. it lags, but really only in the editor, and only in super weird circumstances. all of the things i was having trouble with (every single object causing stacking forever, having to align the animations with respect to stacking, severe lag in longer maps) can just be worked around, and they don't even matter as much as i thought.
so, kudos! you pretty much nailed an objectively perfect method for a problem i spent a year on. i got stuck using the y:512 method and creating a black combo color to make the approach circles invisible. your method is way cooler as it can use actual colors and isnt even dependent on colors being turned off. i am not even slightly frustrated, as i recognize that i would never have done this myself. i credit you entirely for it, as you are the one who stumbled on this and thought it was neat enough to share.
now for the hard part, which is getting another Aspire tournament to happen.
