Account for RB4 whammy accumulation rate

[cpurules]

Hello! First of all want to say this is great, have been using it a lot to verify manual paths and to find sneaky little things that we wouldn't have found otherwise, both for official RB midis and custom generated ones.

One thing that I wondered if it was possible to account for is RB4s whammy "feel good" rate - i.e. if you are whammying a sustain, you actually accumulate all of the OD possible before the sustain ends. The best example of this I have is Less Talk More Rokk, since it starts at a slow bpm and with a ~16bt OD sustain. In this video the sustain starts at approx. 0:24, the whammy stops accumulating at approx. 0:34, and the sustain ends just before 0:37

https://www.youtube.com/watch?v=VoAc7VvI81k

Math-wise, we have a 15.746bt sustain ( see chart ), and the whammy stops accumulating at approx. 12.5bts, yielding a ~1.25x rate of whammy gain compared to sustain length (capped at the max whammy possible from the sustain).

I would love an option in the optimizer to account for this increased whammy rate if it's possible. I know it is not exact, and since we don't have access to the game's internals it will never be exact, but even a slider setting akin to the squeeze % / early whammy % for this would be soooooo helpful for the handful of songs where you need to finesse a whammy to abuse this mechanic (I believe the last G act of RB's Crazy On You is one of them)

[GenericMadScientist]

Wow, that's wild.

I don't have RB4 and would need a lot more information before adding any feature. The following questions come to mind:

1. What happens if you don't whammy for the first beat or so?
2. What happens if you activate mid-sustain?
3. What happens for long sustains at a constant BPM, like the one at the end of YYZ's guitar solo?

[cpurules]

Happy to provide any info I can!

1. If you don't whammy for the first beat or so, you just delay the whammy accumulation. You are still hard capped at whatever the sustain can give you. In the example for LTMR, I could delay whamming for up to 3 seconds and still obtain all of the OD that the sustain could provide, because of this mechanic. (As an aside, this is usually what you want to do when this comes up optimizing - delaying the whammy could delay you hitting a full bar of OD to the end of a sustain you cover, rather than you starting to deplete OD during the sustain)
2. To my knowledge, activating doesn't change whammy OD accumulation other than you are also draining OD at the same time
3. Good question! I think it would be functionally the same as Less Talk More Rokk posted above, but I went ahead and tested.

• LTMR's opening OD sustain is ~15.75 beats long; YYZ G solo OD sustain is 23 beats long
• As mentioned in OP LTMR sustain stops awarding whammy approx 3 beats before the end of the sustain
• For YYZ G... unfortunately the sustain is long enough that unless you're actively draining OD or purposefully miss whammy, you're going to fill up your bar (7.35 beats of whammy for a quarter bar x 3 = 22.05 beats to fill up).
• So for testing, I activate with a half bar shortly before that long OD note, and you can see that the OD starts depleting rather than increasing after approx. 18 beats of whammy. So 18 beats account for the full 23bt length of the sustain, which is roughly a 1.27x whammy accumulation rate. Here is a video of that: https://www.youtube.com/watch?v=GpaMG8CCcIE

Let me know if there's any other testing etc I can do for ya!

[GenericMadScientist]

Interesting. 1.25x feels like a reasonable value to assume, which is equivalent to 20% of the sustain being dead weight (ignoring EW). For completeness a shorter sustain that is in the ballpark getting a half bar would round out the testing, Snow (Hey Oh) being the first example that comes to mind. It's also worth checking bass. Not that I expect either of these tests to deviate from the ~20% dead weight finding.

Adding support for this to CHOpt would not be simple. Certainly within the realm of possibility and I can see how you'd do it, but it'd be a big headache. Right now my priorities are finishing up initial GH3 support, stress testing HOPO/tap detection and then drawing them (since which notes are HOPOs affects GH3 pathing) and maybe RB drums. If you think it'd be valuable to have a cheap version that just always gets OD from the first 80%, I can add that. But barring that this'll have to go on the back burner.