#rSL NG 3lap

1 messages · Page 2 of 1

plucky night
#

so the qm lock are the wallclips in the cave. we can be a couple ms ahead after shroom before that's a problem. i think it's worth being 100% sure if we can reach them or not

plain zodiac
#

Thanks !

plucky night
plain zodiac
plain zodiac
#

btw this "crack" turn is as fast as the ghost one

#

i didn't managed to get ejay's slip start

#

the big airtime at the cave exit save .5f compared to the one of the ghost

plucky night
#

nice; there's definitely no way that ejay's slip cannot be replicated though, we need to figure that out before continuing (it's likely just really finicky)

#

i also cannot guarantee that anything after turn 1 is optimal, i did not finish testing

bitter bridge
#

How‘s the progress going?

random trench
#

by looking at the message above yours, you can see that the last progress was made on the 25th of January

#

or in other words, progress is not going

plucky night
#

i plan to return to this very soon -- ryf and other alego collabs have been my priority for now

steel spade
#

That's cool

#

I would like to work on this again too

tawny summit
#

bump

#

@plucky night which part of the track are you working on

plucky night
#

as of yesterday i'm nearly finished optimizing up to the cave wallclip. i saved a bit more over alego, but i'm getting pacelocked at the wallclip. need 2 ms (i think) to beat the pacelock which i don't think i can find

#

i want to try a couple more things at the shroom then i'll work on the cave

tawny summit
#

nice gl

#

and you reckon AGC syncing rSL is easy?

#

(i think someone said that)

plucky night
#

i'm worried about one penguin on l3, other than that it's easy yes

tawny summit
#

i guess the ability to do +/-1s helps

plucky night
#

if someone wants to help, the only thing i haven't tried yet is delaying the shroom

tawny summit
#

how much time did the changes from 0 to 0:04 save

#

0.01f?

plucky night
#

i started changing inputs from the bounce after first turn. i couldn't improve the first turn more

#

if you meant the different line on the straight, something like 0.2 ms

tawny summit
#

Good work

plucky night
#

ty

plucky night
plain zodiac
plucky night
#

great job man, i was trying earlier today but got nowhere, i never got that low airtime at 2nd clip

#

do you know how this airtime works? i was struggling with it. i'm assuming with the faster pace you're forced to go wide to get it

image.png
plain zodiac
plain zodiac
plucky night
#

i tried a lot of things and could never avoid the airtime after the 4th cave turn.
the best thing i found was hopping late, slow down to avoid airtime, and the late mt release keeps part of the timesave... but i'm clipping the wall 1 f early now, so i doubt that the wallclip is possible without slowing down

📎 rSL_72.350s_909276ee.rkg
#

really frustrating that all this effort lead to basically zero timesave but i guess at least we ruled it out

plucky night
#

the penguin in question is this btw (we will be about .1 ahead)

image.png
steel spade
#

If only we had this done today and make everyone guess whether baby daisy is actually faster or not as an april fools lol

plucky night
#

too many people spilled the beans publicly already + the track has just not been fun to be honest

#

i may drop it again

plucky night
#

it's possible to realign with a quickhop, but it gives a much different exit line; i'm not sure which is better, will have to test both

tawny summit
#

and just to note, one other possible improvement i suggested included getting 1f air instead of 2f at the slipdrift frame

#

since it should result in a wider mt charge (means we can take the actual line slightly tighter)

plucky night
#

i never got anything but a 2 f bounce for the record so it may be impossible

#

also i just realized this should be about 4 f ahead of my first baby daisy run by l1, so 2:00.4 final time, that's cool

plucky night
tawny summit
#

Yes

#

Think about an extreme: imagine you get 10f airtime but you only cancel the wheelie on the 10th frame: it'll give a tighter line

plucky night
#

i couldn't improve the slipdrift part

plucky night
#

after the crack jump i get a 21 f bounce, where the ghost gets 20 f. that should save a bit of time. otherwise i think changing the wheelie turning before the drift can lead to timesave

plain zodiac
plucky night
#

yeah, it may also depend on the first turn, since the approach is different on l2/l3 and i'm not 100% sure that copying the l1 line for the spinhop is optimal

#

the beginning improvements we made can be used for l2, but not l3, bc there will be a penguin in the way at cave entry (we may need to take a wider line into cave in fact)

plucky night
#

i'm ~1 ms ahead compared to l1 first turn. should i keep trying for more improvements, or do i trust that the cave is fully pacelocked and proceed with agcing l2?

random trench
#

i reckon keep going in case it's not fully pacelocked for some reason

#

you mayregret not checking that in the future

plucky night
#

does pycore have some funny script that lets me warp the player in the forward direction by an arbitrary percentage of frame

#

so i can test pacelocks like this easily

random trench
#

nope but that probably wouldnt be very hard (you could combine it with the wallclip bruteforce script too)

#

though, couldnt you just briefly activate and deactivate agc with delay values <1 to test if the ghost desynced or not

plucky night
#

what how

#

wait deactivate?

#

oh shit

#

that's like the first time i have ever been outsmarted in my life

#

i'm going to bed

random trench
tawny summit
#

inb4 corvid

plucky night
#

i resynced l2 and start of l3. the penguin i mentioned a couple times does get in the way slightly. best i could do to avoid it loses 3 ms to l2 (ghost) which is enough to get pacelocked at the wallclip.
so we need to save a little bit of time somewhere between the last crack jump of l2 and the penguin dodge

📎 rSL_115.117s_ee76978b.rkg
#

some of the l1 improvements i found but didn't carry should be useful here, however not the alternative line after shroom (goes too far left)

plucky night
tawny summit
#

Is the last turn done in the correct way at least

plucky night
#

yes

#

but compared to my previous bd run, it's several ms slower

#

have to go out now so don't have time to grind it further atm

tawny summit
#

trying more things out now

tawny summit
#

got 497

#

more timesave possible but i gotta get some sleep

simple portal
simple portal
tawny summit
#

Great

plucky night
#

nice job

tawny summit
#

cause we did a lot of testing last time and after back and forth improvements, found that a delayed slip was slower

#

but now it seems to be way faster, weird

plucky night
#

to be honest; i have absolutely no idea. this specific turn on l3 could be the turn i understand the least out of everything i've worked on so far

tawny summit
#

how lenient is the crack hop

#

let's say someone improves the cave exit by 2 milliseconds, what are the chances that it's carryable

plucky night
#

forward position is not a factor at all

tawny summit
#

how?

#

if you're too far forward compared to the ghost then you'll get airtime 1f earlier

#

or to compensate do you just be more to the right

plucky night
#

you can make the crack jump from a variety of approaches. the most important detail is landing a certain distance away from the edge of the road, and adjusting pitch rotation + sideways momentum (or the bike bonks against the wall. it's not a matter of getting too little height surprisingly)

#

i was always able to carry small improvements through crack jumps because there are a lot of ways to adjust those variables

tawny summit
#

Okay

#

I think we should have another 3 people minimum look at lap 3 from the last bonk clip in the cave until the ending

#

Don't want to rush this

plucky night
tawny summit
#

I thought more right yaw makes the miniturbo charge path better

plucky night
#

it's a worse line. we want as much left rotation as the shape of the road lets us get away with

#

how much exactly, i have no clue

plucky night
plucky night
#

would like someone else to try this out. i don't think i will touch rsl for a while

plain zodiac
tawny summit
#

rsl finally finished

#

december 2023 --> april 2025

plucky night
#

491 is prob possible

#

oh wait i missed alego's message lmao

#

i can try it again soon. it doesn't really make sense that it wouldn't be possible to carry that but maybe

random trench
plucky night
#

it started in december 2024 and paused until late march 2025 let's be honest

random trench
#

yeah this was really part 2

tawny summit
#

I was going off of the earliest date in the rSL NG group chat

#

which Corvid isn't in anymore

#

so it started December 2023 at the latest

random trench
#

oh i was never in the group chat so i assumed it started in 2022

#

interesting

tawny summit
#

I know flurry was working on it by himself before even that

plucky night
plucky night
#

i would like to upload this myself, but i would need someone to teach me how to record properly. all my framedumps look like dogshit

tawny summit
#

iirc i tried to send you the steps for recording but it still didn't look good or something

#

but i can't remember if that was because you didn't switch dolphin versions?

#

i can help you again

plucky night
#

i tried to use both latest lua and pycore with 1 million bitrate but it looked really pixelated with both

tawny summit
#

DMs