DKC Atlas Forum

What's everyone's best times for this level? (just racing through to the finish, not 100%)

SDA has a record up boasting a time of 54.70 seconds by Cyberwrath, but numerous people have beaten this in the past; and personally speaking, I'd love to see a video get posted up there with a better time if somebody is able to.

NTSC version works best; I can't get better than 58 seconds using the PAL version, but when I use an NTSC emulator I can get better quite easily despite the rough control scheme I use.

This guy has achieved a time of 53.20 (and 53.14, although that was on an emulator) so I'm wondering if anyone can get better than this; and if so, would they be interested in fiddling around with a few recording methods to get a run submitted to the SDA in decent quality? Either way, see what you can achieve and perhaps record a video or two of your run attempts (on an emulator if console isnt possible) so that you can see where you're losing time.

Another thing I'd like to see: if anyone would be interested in TASing this. I suppose this isn't really the place to ask, but it'd be cool to see what time is possible and what runners should aim to achieve.

I've had this one about a year ago:

http://www.youtube.com/watch?v=slj2vq1zXAw

So my best time is 0:56:75. Not the best, but it's only about two to three seconds away from the best. I made a few little mistakes (for example, when I jumped out of the water once, I clipped part of the wall).

Also, I used an emulator for this, but I didn't slow the game down, and I didn't use save states. I only used an emulator because it was easier for me to record the video that way.

One last thing: I have Mario Kart Wii, and the guy who got the best time, mjf314, shows up in the best times for almost every track.

Lol; I didn't actually realise who he was before; but I do remember the name Michael Freid from the SDA if I recall correctly.

I've actually been speaking to Michael Fried (mjf314) on AIM for nearly three years now. In late 2005/early 2006, we played Mario Kart DS (and were within the top 100 players around the world) together for months. I met him through Donkey Kong Country 2 GBA; I had Google searched some words similar to "Diddy's Dash records" and found some of his times.

My best time is a little under a minute. I can't recall exact, unfortunately, and my DKC3 cartridge is missing.

One thing people should be aware of when they try to beat the time, is that it's faster with only one character. As you have to wait for both of them to get into the barrel before you are blasted away.

I tried to do this, TAS style, awhile ago. Didn't fully optimize it but got a time of 51:97. Another thing that's ass is that you can get a slower in level time, but is a faster real time.
This is easy to check in the emulator going frame by frame, like this:

51:34, 51:35, 51:46, 51:47, 51:40, 51:41, 51:42, 51:43. Really lame to be honest... If I first get the 51:40 time and then improved it by one frame I would have got 51:47.

Wow, you're right Tompa - the milliseconds jump back and forth... It's something you don't see with the naked eye, but frame advance reveals the erratic counting. I wonder what the heck's going on with that. Makes it a bit pointless going for the lowest possible race time.

This timer is a bit strange for another reason. It changes every frame. Typically changes in SNES games occur each 2nd frame, right?

Graphical changes occur mainly every 2nd frame, but there are things that are updated every single frame. I think such an example (though not in DKC3) would be the Factory Foreground of DKC. It's updated every frame, but you can only tell if you do a frame-by frame inspection. And it varies also depending on how fast you're moving.

The reason for most things being updated every other frame is likely to give the SNES time to process graphics or something. I haven't figured out exactly why it does this yet.

As for that weird timer, I wonder why they didn't just use the game's timer (or one at the same rate)...? I mean, it's 60 frames to a second, that's so easy to calculate it's ridiculous! This makes me want to hack the timer to get it to display in real game time, instead of pseudo-random numbers.

Also, I would like to figure out what it is they did to get the timer to show those random numbers and why...

My best was 1.12 or something on the GBA version. Pretty good.

Rodent wrote:My best was 1.12 or something on the GBA version. Pretty good.

...not really. Although the GBA's gameplay is much slower than that of the SNES, my best time from back when I bought the game was something like 1.02 or 1.03. You've got quite a bit of improvement yet.

That's really odd that the game would have such bizarre methods of keeping time for the SNES, though. I wonder if anyone still at Rare would still know the answer to that weirdness.

There are a lot of weirdness to be found in games when playing in frame advance, it's awesome.
But they were on to something when they did that timer.

In the GBA version you can get the time very low as the timer still goes on when you take the level warp. Haven't played the GBA version though...

^ yeah, I realise that; I just meant running the level without the warp.

Kowbrainz wrote:

Rodent wrote:My best was 1.12 or something on the GBA version. Pretty good.

...not really. Although the GBA's gameplay is much slower than that of the SNES, my best time from back when I bought the game was something like 1.02 or 1.03. You've got quite a bit of improvement yet.

Oh, that was a joke.