This just gets crazier and crazier.
Note that for all practical purposes, as long as you can pull off the last jump without a ring trail and with time stopped, Cannon's Core 1 and 3 are the same. For one you go down at the end, and for the other you go up, but it's not like time is ticking or anything.
That said, the previous unified 1-3 record (1:15.94) was set just under a year ago, and I decided I wanted to try another run at the level, this time using the new trick I unveiled in the A-button challenge video for CC1, with the goal of getting closer to (or possibly under) 1:15.
After a short bit of experimentation, I found out that unfortunately, the new route wasn't going to be good for saving any time. But since I already had a personal record through three parts, I wasn't just going to throw that much of the run away. Instead, while I still had the old route available as a backup plan, I thought I'd spend a bit more time just looking around to see what was possible, and...the rest is history, they say.
While it's still true that the break out at the top is, on its own, no faster than breaking up through the acid room, that fact that you start higher up leaves you in better position for another timesaver. The "normal" path for Knuckles' CC starts: dive down to stop time, get back up to go past a laser wall, hit a switch to kill another set of lasers underwater, then dive down again, go past where the laser wall used to be, enter a room...That room has a time stop switch on its diagonal wall; somewhat relevant here is the fact that it's the longest-lasting switch in all of Knuckles' part, and extremely relevant is...the fact that you can activate this switch from OUT OF BOUNDS. Just get close to where the switch would be, on the other side, and you can stop time without getting sucked back in the wall! This lets you keep the time stop going all the way to the security door switch, cutting several seconds at once.
For those of you who want to try this at home, I will tell you that Knuckles' part alone took in excess of 4 hours to get right during this recording session, and the repetitive read-write-erase stress ended up killing two DVD-RWs from my rotation.
And for the first time in all my experience with Cannon's Core, I'm now truly wondering just how low a TAS of the level would be able to go. I can imagine as low as 1:02 might be possible, but I may not even be ready to try for as much even if the tools were available to me.
Somehow I don't feel that it counts as a genuine *SPEEDRUN* if you aren't clearing the level as fast as possible, you actually spent a lot of time hitting the timestop switches instead. Ignoring loading times and the short real-time cutscenes, this is a 5 minutes run of the whole stage, when probably you could beat it in under 3 minutes if genuinely sped-through. IMO it'd be more interesting to see a genuine speedrun of this stage, without *hours* of hitting switches along the way! Go for it man
AlexxShadenk777 1 year ago
Seeing as time switches are the whole point of the stage...
Also, I did attempt a realtime speedrun (on mission 4 so that even segmenting isn't an issue) and got 4:06.52 if you want to look for that one. Under 3:00 for that would be nonsense.
SadisticMystic 1 year ago
Hey SM! They've done a TAS of this level (m1) but it doesn't even come close to this, they weren't aware of all the possible tricks. You probably spent a lot of years finding shortcuts and improving this level, didn't you?
Also: if you know that Cannon's core m1 is identical to m3, Why don't you claim 1:09:95 as your m1 record in TSC?
afolita 1 year ago
For the same reason that you can't take your best Sand Ocean mission time and replicate it four times.
Unless the mission goals are identical, such that any set of input that completes the more stringent one also completes the other, and not just "informally" identical so that all the key steps from a competition perspective are the same, the resulting stats can't be. (This lets you submit a mission 4 to its 1 chart, but not interchange the 1 and the 3 here since they have different endpoints.)
SadisticMystic 1 year ago