NorwayHi! I noticed we haven't really linked this anywhere, so I'll put it here for any new runners who are looking to chat about the game. We're a helpful bunch and the community is great, so feel free to join!
This category was made both to provide a suitable category for new runners that might have trouble with the range of glitches in All Stages, as well as to offer a good alternative to the traditional categories for experienced runners. The following glitches are banned from the category:
- Backflip-grabbing the key
- Double-tumble glitch
- Basejumping
- Single-tumble invulnerability
- Walljumps (includes elevator jump in 0-3 and waterfall jump in 8-9)
- Clipping through ceilings, floors and walls (includes phaser clip, Crusher clip, elevator clip, Shovalot clip, frog clip...)
- IFGs (Infinite Fall Glitch, used in 5-6 in All Stages)
- Going out of bounds
- Extended handspring jumps off trampolines
- Letting DK Jr. out of his cage in 8-7
This list might not be complete and will be edited when needed. Please check back to review the list of glitches before submitting a run in this category.
I kind of mentioned it the other day when we were talking about it, but a somewhat in-depth explanation would go something like this:
There are two parts to the walljump: the angle/distance in which Mario approaches the wall, and the jump itself. A good example here is 5-3, where you want to walljump with only a tile of space to backflip from. There's a certain spot (possibly a pixel?) where you can get it, which leads me to believe that you can get the walljump from certain distances only. You'll know if Mario jumped from the right angle/distance if he "turns to the right" as he hugs the wall you're trying to jump from. If Mario's sprite doesn't change so he faces right, the angle/distance was wrong.
Once Mario get the correct jump into the wall, you have a short window to press the jump button. It's not measured completely but I believe having read somewhere that the window is 3 frames long. This can be seen in Bizhawk if you inspect the memory address for whether or not Mario is in a jumping state; the value will shortly switch from 1 to 0. When it comes to different stages, the stage physics probably change how easy it is to get the walljump. In a stage with wind physics applied it's really easy to get it, but when you do it off an icy flooring such as in 7-2 it's oftentimes more tricky. I don't believe we have looked into this very thoroughly but I'll make a point to look into it more when I'm finished moving.
Doing this trick RTA is pretty hard, and as I've also mentioned before, you just develop a feel for it with lots of practice. We have different methods of getting it; SparkCity mashes the jump button as fast as he can from the moment he executes the backflip, I just press the button once or twice after the initial jump, and I'm sure the others have their own way of dealing with it. My best advice is to spend some time practicing walljumps only so you get used to them, because it really is a trick that everyone does a little bit differently.
I don't know how much help this was to you, but hopefully it's a somewhat thorough explanation of how the mechanics work. :D
Hey,
After checking out LudwigVonKoopa's run (https://www.speedrun.com/run/zq60jl8y) I noticed an admin notice saying that GBC emulation + initial SGB frame is "probably the best for DK94". I decided to check it out since I got excited, but I don't actually think that is the best way to emulate the game if you are playing on BGB, the reason being that it seems like the game is still being emulated using SGB.
The simple way to check it is by opening the meny and choosing Other > Cart info. If you're using GBC emulation, it will read "emulating as GBC in DMG mode" or something to that effect. However, GBC + initial SGB frame reads "emulating as SGB". Not only that, but the SGB timer lag is also there. When you finish a level and Mario exits the screen, there will be a short sound, a small delay, and then the timer starts ticking down. That small delay is actually a bit longer on SGB timing, and is very noticable when emulating using GBC + initial SGB frame.
Because of this I have some doubts about this method of emulation being GB accurate. I was wondering if there has actually been any research done on this to confirm that it does indeed use GB timing, or if we are just assuming that this is the case?
