Donkey Kong 64 any% Tool-assisted Speedrun Explained

sequence. If the lamp falls when the gauntlet throwing animation timer reaches 0, the first animation of the lamp falling on your head will be skipped immediately. So by shooting targets at the right time, part of their animation is skipped, saving more time than spent waiting for the right moment. The game has a failsafe if the player somehow ends up outside the ring. Warps the Kong instantly back to the middle. In this case, it helps to save a little time by placing the banana next to K. Rool without having to travel a long distance with the barrel. Otherwise, this phase is pretty straightforward.

Tiny's round takes place almost entirely on rails. That's why RingRush spends most of its time gaming. Here, he demonstrates how the orientation angle does not change in the first frame of a new walking animation when walking backwards. The only thing that can be optimized is to enter the barrel of the Mini Monkey, travel to the shoe and deliver the final blow to each toe. Each finger takes one more hit than the last, but there is always a fixed amount of time between attacks, so only the last hit on each finger matters. As we look at this phase, let's learn some fun facts.

If you go far enough out of bounds to charge K. Rool, the fight stops, including the timer, until he charges again. The out-of-bounds safety mechanism is used a lot in this phase to reach the barrel of the Mini Mono without having to deal with the increased delay of K. Rool's shockwaves. Tiny automatically pulls out his gun even though he doesn't actually have it. If he kept it, he wouldn't be able to take it out again. That would gently crash the game, since there would be no way to damage the toes without a gun or oranges, and there would be no way to get out of the fight.

Main Menu mode prevents pausing and taking damage, and the round timer is paused inside the shoe. . It is possible to get to the press box where the kremlings are, but doing so causes an additional delay, so it was not shown. Additionally, although Rambi Arena offers Mini Monkey, it is possible to skip purchasing the move. When any Cranky barrel appears, it can actually be used by default and only deactivates a few frames later. So if you time it right, both Rocketbarrel Boost and Mini Monkey can be used in this fight without purchasing them. Here, RingRush shows how generous the toe-tapping window is by hitting it on the last possible frame.

Some of the toe attacks are based on the toe Tiny is currently facing. By standing completely to the left of the shoe, he can be in front without toes, in which case no attack occurs. Unfortunately, this doesn't save time. As you can see, the crowd is a simple textured surface and not actual models or even sprites. Additionally, the game treats K. Rool as the boss of Hideout Helm. If you quit at any time during the battle, you will be placed at the entrance to Hideout Helm, just as you would with the boss of any other level.

Using first-person view helps reduce lag during many parts of the fight, but especially within K. Rool's shoe. However, the delay in Tiny's phase is not emulated well. On Nintendo 64, reducing lag would be crucial, but here it doesn't really matter. So RingRush opted to spend half the time demonstrating extreme lag reduction methods and the other half fooling around without wasting time on lag. If you listen carefully here, you will hear the sound of entering in first person. There's a small window where you have control, long enough to go into first person, allowing for phased play immediately after the cutscene.

Here, Chunky uses the out-of-bounds safety mechanism to his advantage to hit the center switch faster. And from there, the rest of the fight is simple. With this, Donkey Kong 64 is defeated in 26 minutes and 46 seconds from when the game is first turned on. Some of you will know that the real time world record is 24:51 by SigNa. There are three reasons why this TAS has a longer runtime. First, the real-time recording was done on the Wii U Virtual Console, which has virtually no lag, saving about a minute and a half compared to the Nintendo 64. Sometimes the lag is emulated inaccurately to DK64, so this emulator will likely run.

It only loses about a minute compared to the Wii U version, but it's impossible to measure this exactly. Secondly, the timing methods are a little different. Real-time runs begin by pressing A in the menu to start a new file and end at the first frame where black bars are visible at the beginning of the cutscene where K. Rool is knocked out. TAS timing begins when you turn on the console and ends with the final input needed to finish the game. By real-time standards, this TAS would be 26:13 instead of 26:46. Last but not least, there is an important difference in the rules.

The number of fairies collected is permanently saved in the cartridge's static random access memory. This means that after collecting fairies once on any DK64 cartridge, the mystery menu functions will be irreversibly unlocked forever. As a result, the DK64 community decided that using these features was fair play in real-time in any %

speedrun

s. The TAS community prefers different rules. They recommend that runs start from a completely clean state, including static random access memory. So the TAS has to collect the banana camera, 6 fairies and enter Rambi's box in Jungle Japes to unlock Rambi Arena, while the real-time race has access to it from the beginning.

Again, this is very difficult to measure exactly, but the estimated time difference is about 4 and a half minutes. Ultimately, these three factors add up to about 6 minutes, so this run is comparable to a 20-minute run on Wii U based on real-time rules and timing. And that concludes this explanation of

tool

-

assisted

speedrun

ning. Thanks to RingRush and theballaam96 for their help and thanks for watching!