Posts for Kirkq

Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
If I'm understanding when nighttime occurs correctly, I think chapter 1 is using one extra step to get into Izmit because from walking around town you can only go into town on "even" step counters. There is one way I can think of to break out of being stuck on even steps, and that would be to spend one extra step prior to going into an entrance that is two tiles wide. I think this would work for the Izmit situation using the cave entrance, and I don't think it works for the Frenor situation since there's no 2-tile entrance. I'm mostly just mentioning this if the situation comes up again later.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
I was poking around at some stuff for RTA runs. It looks like going to the inn in Burland on the way out of the castle in chapter 1 would save about 3 seconds. It sets the day/night step counter (62ED) to 40 instead of 0. (Saving 640 frames of walking)
adelikat wrote:
It looks completely random that any given step will fail. However, it looks like there are rules. For instance you can't have 2 failed steps in a row.
It looks like it's intentionally coded to not count a day-night step randomly at a probability of 1/32 per step for whatever reason, assuming $0012 is well distributed. (BCC instruction at CDF1) I think it can fail twice in a row, it's just a 0.1% chance that would happen. I failed two discontinuous steps, I didn't try for failing two continuous steps.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
Regarding the bottom floor. Walking diagonally is 3,3 or 4,4. Walking straight is 5,0 or 0,5. Walking up-left followed by down-left is slightly slower than just walking left. If the crook wasn't in the way you could just walk straight left. Also the bottom floor crook throws a flash grenade which loses a few frames. For finding memory addresses, please refer to this page, it's really easy to learn how to find character position: http://tasvideos.org/MemorySearch.html local CharX = 0x00A1 local CharY = 0x00A3 local CameraOffsetX = 0x0044 --from character local CameraOffsetY = 0x0045 local Floor = 0x0065 local CameraX = 0x006F local CameraY = 0x0070 local LifeDiff = 0x00A6 local Crooks = 0x00AB local Life = 0x00AC local Lock = 0x00AD local TimerFrames = 0x00AE local Seconds = 0x00AF local Minutes = 0x00B0 local FeetUnit = 0x0797 local Feet = 0x0798 --8 feet bad, 0 feet good
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
Technical criticism: - Walking diagonally alternates between X,Y=4,4 and X,Y=3,3 This route walks in 3,3 more often than required, which probably loses around 20 frames over the course of the run. - I'm pretty sure you can manipulate the guy with the key to make it faster and not have the radio overlap. This probably saves around 25-30 frames. Just get him to walk left slightly after the first shot hits I think. Maybe I'm wrong. - The green crook on the bottom floor seems like he's wasting more time than required. It seems like the crook pattern on the roof isn't great. The zip to the foot power was new to me. I personally prefer "fastest ending achievement" over "fastest input ending". This movie is slightly slower to "finish the game" in exchange for "dropping the controller sooner". It's a debatable topic though. I don't have a strong opinion on publishing in the current state. It's far better than the previous run. It has minor technical shortcomings. If you want feedback in the future, feel free to reach out. For historical/timeline purposes: Flobeamer1922 originally posted this glitch Aug 20, 2017. I optimized the foot power route more in July 2018. My notes for unoptimized TAS routes (Spent maybe an hour on each) were the following: C: 4 + 240/256, 11 stairs, Don't Grab Foot Power - 6752 frames. D : 4 + 240/256, 11 stairs, Grab Foot Power - 6733 frames. Route D appears fastest non tool-assisted. Route C may or may not come out slightly ahead when rigorously optimizing movement in a tool-assisted setting. (I think the foot power probably makes this Route D TAS faster now, but it might still be worth investigating the route C.)
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
0xwas has obsoleted my Magical Pop'n WIP with his WIP in Bizhawk, so that can be removed from the continuances list.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
There have been numerous glitches found recently, particularly one that allows you to get like 50 levels (TAS mashing) while dying at the same time as you cast, and battle arena clips and various other methods that can get out of bounds in pretty much any area with the right seam. If you plan to take this on, please get in touch with us: http://www.speedrun.com/quest64
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
Replying to this with some new information. I was helping baldjared look into this a bit for his real time runs. I'm not planning on TASing this. The optimal pattern for the TAS should have the real guy in the middle door. You must logically solve the truth table, so asking 3 questions, middle, left, middle, solves the table fastest. It might skip like one quick text box to have the guy on the left talk about a specific thing, not sure. When Anconda comes down the left or the right he increments 2 RNGs. When he comes down above you he increments 1. You can wait a round in the Anconda fight to set up a better key, but that's likely suboptimal for the TAS. Breaking an additional pot for the key or to increment the RNG costs upwards of 4 seconds, so setting up the RNG better at the beginning will always be better. There's 256 possibilities in the RNG. They get permuted at Nick, and then added to for the rest of the calls. Optimal Nick ~25%, Optimal Anconda ~25%, Optimal Key ~33%, Optimal Boxes ~25%? = 0.005, which times 256 gives about 1 RNG possibility. It might be worth skipping a box to get a better final room though. My guess is the final randomized tower room matters for a few seconds of gain. In real time play, since the RNG increments once per frame on the starting screen, and you can start the game with any of 4 face buttons or start, I recommend mashing the 4 face buttons, which can usually net the first frame or two. Not the best table I ever created. R = Real B = Beard Lie T = Tomato Lie D = Drums Lie # = Frame number you start the game on: TOP LEFT MID RIGHT 1 R B T D 2 D R B T 3 B T R D 4 B R D T
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
Should've posted this a while ago: Since speed oscillates every two frames between a high and a low value, it looks like it is possible to force the high value two frames in a row by pausing. This buys some extra pixels on the CC Puzzle Early - No Shock pad attempt, but I wasn't able to make it work with brief testing. It's probably slightly more theoretically possible now. Also regarding Comicalflop's earlier points: - The 10th jiggy has an extra animation so it take a lot longer. - Hyper's explanation of speed is probably correct. Speed definitely does not care whether you are skidding or jumping. Thus the sprite contortion might be the best explanation.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
Requesting continuance on my N64 Bomberman Hero TAS that I've been doing on and off (mostly off obviously) since like 2009 or something. It's like 80% complete.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
Rattleman: I highly doubt it's lag differences. My guess is it's a consequence of how the character rises/falls being suboptimal in this video. I'd like to see this category published at some point. Unsure how it compares to Rattleman's/Dennisbalow's previous runs overall. I didn't watch this whole run.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
Statistically more people bought NTSC versions of games than PAL. (Something like 90% of NES games are NTSC, I cited this a long time ago.) Even if the game has the exact same content plus addtional, in my opinion there is still great value in having the original present, as it represents the fastest time in the version that the most people played by a large margin. People will see this run and think "This is not quite the game I played, where is the TAS of the one I played?" I don't think they should have to dig through obsolete movies to find it.
electricslide wrote:
Also, I really dislike replacing the NES originals with the WII remakes. Consideration should be given to the history of the NES. We wouldn't replace the NES super Mario with the SNES all-stars version.
Similarly if NES Super Mario Bros had an extra level on PAL, I don't think we should want to obsolete the NTSC run.
GoddessMaria15 wrote:
Even with the additional stage, I feel that it should be side by side with the NES DK due to the originally packaged game being its own game on the native system and this one is more faithful to the original Arcade version while being packaged as an NES ROM stored into a Wii Virtual Console WAD.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
ALAKTORN wrote:
From what I’ve seen the Retreat glitch is not fully understood and has more potential than how it’s currently used…
This is not really true. Pretty much every normal door in the game has been checked for where it leads to with a "normal" retreat glitch. "Save" retreat glitching involves doing a similar thing and saving to translate your coordinates to the first floor of the dungeon to go off map. Once off map, after underflowing your coordinates by walking off the top or left there are various glitched doors that lead to various rooms in the dungeon. I made a brute forcing script to record numerous instances of these glitched doors (literally thousands), where they are, and where they lead. Each dungeon typically has like 5-10 options. All of my notes were available to werster (he also helped run the script in various dungeons) a couple years ago, so I can only assume they've for the most part been implemented into the real time run where applicable. In short the glitch may not be well understood, but the effects have been much more rigorously tested than your post speculates.
ALAKTORN wrote:
Edit: at 1:59:48 the movement looked bad… you went too far to the left and hit the wall while going up.
That's probably like 4 frames bad. If that's the worst thing you found I'm inclined to believe the movement is pretty okay overall. - No retreat glitch is an okay category in my opinion. It's not spectacular, but it's a pretty sensible way to play through the game. It's probably really roughly around 20-30 minutes longer than a normal run, but it does not skip key dungeons. There are a few aspects of the RNG that are more random than others. I know that who gets attacked is much more random than which attack the enemy chooses. Real time runners are pretty familiar with the intricacies. It's also complicated to manipulate the normal RNG without starting from reset. The movement in this game is pretty straightforward. 1 frame can be saved around most corners due to precise TAS level movement optimization, but other than that, just hitting directions, reloading, and frame advancing should get you within 1-2 frames of optimal. As far as the optimality of this run, I would look at the battles, leveling, djinn, and equipment more than anything. I glanced at the Saturos fight. Crystal powder wasn't manipulated from a random encounter in mercury lighthouse. (I can't guarantee it would be faster overall, but it could possibly save a turn.) The player can't manipulate eruption entirely without resetting. Some things could be done in previous battles to try and set up a minimal number of eruptions, but it would not save that much time. He ends with 5/6 of each djinn type, which I really don't feel is optimal, but it's very hard to calculate the time gains/losses. Bosses are either weak to fire or water, and most major bosses are weak to water, but a summon from 4 earth and 4 jupiter isn't bad at all. Djinni also make your characters stronger when not in summon mode. My best guess is probably 4/4/4/4 is optimal, but that's mostly speculation. Adding more would save time in battles and lose time acquiring them. I'm not sure if it's faster or not. I can't really say for certain how good this run is, but for what it is, it appears to be an average level execution (or maybe slightly below average, I'd have to look closer). Things seem to be done for sensible reasons, the characters are probably a little stronger than they minimally need to be. I did not watch the whole run, I just glanced at notable parts. EDIT: I'm wondering if sacred feathers would've been a good investment. They double distance (EDIT: maybe my memory is bad?) to next encounter for a while. Most encounters are run from pretty quickly though, so I'm not sure. They take some time to buy and some time to use as well.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
Was that MMM jump on NTSC console?
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
I might help look into this a bit later on. Feel free to contact me if you have questions. Thoroughly looking at some of the any% TAS routes I posted a page ago may give you some good ideas with 100%. No one who was working this game ever figured out how to reliably get Conga to hit that orange pad. Unsolved mystery. I would give the talon trot slide to Clanker's Cavern puzzle a 50/50 on being possible. You would need to leave at the perfect pixel/height, and maintain a perfectly straight line while the camera spins. Maybe some concept testing and/or math would help verify if it's possible. It's hard to hack your position reliably from my experiences. I never actually verified the pause buffering thing fully, it was somewhat of an afterthought. If you look into banjo's velocity, you'll see that it oscillates between two values every other frame. I think I may have seen weird effects from pausing allowing me to get two higher speed gameplay frames in a row. I have no idea if that's accurate, it may have just been an impression or speculation. Definitely mess around with it though.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
I haven't been keeping up with the precedent on these credits glitch runs, but it does seem arbitrary to choose "all levels appearing beaten" as criteria for beating the game. Like Finalfighter posted, if he could plausibly theoretically complete arbitrary tasks like making levels look complete with this method, would it really change anything? I generally think calling a game end routine which runs mostly normally and runs to the completion state is sufficient to say the game has been beaten. The question to me is something like "Are you convinced you just watched the ending of the game even though the levels don't look beat?" For me, Yes.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
As a case for perspective: The Final Fantasy 1 console runner Feasel has been dealing with this. In between runs he must let RAM decay for about 10 minutes before booting the game back up or the RNG does not clear correctly. I think he leaves the console powered on but not plugged in to do this, but I am not sure. In this sense it may be beneficial for him to find a different game that can leave ideal values in the RNG spots (probably 2 addresses) in RAM to set up extremely good RNG. He would then power off that game, pop in Final Fantasy 1, and start with whatever RNG he wanted (I think). He has been using the "cleared" state up until now. There may be desire to have the "pull the NES out of the box and play it" state, and there could also be desire to have the "This NES was just played 10 minutes ago with a different game with strategically manipulated RAM" state. Personally I don't really see the reason to exclude the latter case. (If I misstated anything, please correct me. This is my current best understanding of the topic.) As a follow-up question, could someone explain how 0000 0000 0000 0000 would break a game? I would speculate this is what you get when you first plug the console in. Does that get modified at all by providing a power source?
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
The 7th Saga TAS Nitrodon and I did used an A* or similar algorithm to traverse between towns on the grid while avoiding enemies. Nitrodon gets the credit for the original script, but I did some modifications with the randomness and brute forcing. To my understanding it first loaded the map from ROM and distinguished between walkable and non walkable tiles. Following this it determined the minimum number of steps required to get from point A to point B via a shortest path algorithm. As the algorithm ran, it rechecked to see that taking a given step made the path one step shorter. As enemies literally get in the way due to the game's encounter system, any time we would run into an encounter, we would reload a past state and take a different path. Originally the step direction was deterministic, but as we found the need to manipulate enemies on the screen to alter RNG incrementation, we switched to a randomly chosen step sequence with some randomly inserted polarization towards the extremes. We needed to start a few boss fights on extremely specific RNG values, as boss fights are entirely deterministic. Using a shortest path algorithm in our TAS saved a ton of work, and using it to influence RNG incrementation produced some otherwise nearly impossible results. In one case I ran the algorithm like 2000 times and saved the lowest number of RNG incrementations while still maintaining shortest path. Using a shortest path algorithm to solve a game entirely is pretty much impossible as most others have discussed in this thread, but I would like to highlight that in a specific case usually of map traversal, especially on a grid, it can be extremely useful.
Post subject: Calculations in MHS based on emulator memory (real time)
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
I show my method of accomplishing showing calculations in MHS based on Mupen memory in real time. I have not seen thorough documentation of this method anywhere, though I imagine at least a few people have done it. I learned this largely from the help files and the MHS forums. Does anyone have any better suggestions for doing this in emulators that do not have lua integrated? I know CheatEngine has similar possibilities with lua, but is there thorough documentation of how to do that? Does anyone have a CheatEngine-Dolphin script that essentially does this same thing? A separate question: Has anyone managed to incorporate float big endian data type into MHS searches to allow for practical use with Dolphin? -- MHS Tools -> Script Editor: Code 1: HotKeyBreak.lss This piece of code lets you press a hotkey to execute a function to change a variable to end the loop in the second code.
int Scriptbreak = 1;
//In MHS go to Tools: Hotkeys, define key to press to start the script, parm1 = 24 (not 0x24), parm2, parm3 don't matter.
VOID On_HK_24( DWORD dw1, DWORD dw2 ) 
{
    PrintF( "Scriptbreak was %d", Scriptbreak);
    Scriptbreak = 0;
    PrintF( "Scriptbreak is %d", Scriptbreak);    
}
Code 2: BKVelocityInterp.lss In Banjo Kazooie this uses X and Y speed vectors to calculate overall speed and update in real time. This was fairly complicated to implement due to the need to have a second thread for the program to not crash on a potentially endless while loop. I did a lot of reading of the MHS help file and the forums to come to this method. While the code is not particularly elegant or formal, I tried to comment thoroughly to explain what problems can arise.
//MHS Help File defines these functions.
VOID On_HK_23( DWORD dw1, DWORD dw2 ) {
	//In MHS go to Tools: Hotkeys, define key to press to start the script, parm1 = 23 (not 0x23), parm2, parm3 don't matter.
    Clear(); //Clear Display console
    DWORD dwParm = Random( 90 ); // Create a random parameter. (Left in for demonstration)
    
    if ( StrICmp( GetCurProcessName(), "mupen64-rerecording.exe" ) != 0 )
    //this may need changed for different versions of mupen
    {
		PrintF( "You must be attached to a process named mupen64-rerecording.exe !" );
		return;
	}
    
    // Create the second thread with this parameter.  We write “SecondThread” ourselves.
    HANDLE hThread = CreateThread( "SecondThread", dwParm );
    // It may have failed!
    if ( !hThread ) {
        PrintF( "Creating the second thread failed." );
        return;
    } 
    // It succeeded!
    PrintF( "Second thread created with parameter %u.", dwParm );
    CloseHandle( hThread );
}
 
///////////////////////////////////////////////////////////////////////////////////////////////////////// 
 
//We need this second thread because main will freeze if stuck in a while loop.
DWORD SecondThread( DWORD dwParm ) 
{	
  
   extern FLOAT XPos = { "mupen64-rerecording.exe", 0x867C98};  //Must use base addresses found by right clicking on values.  Process name must match.  
   extern FLOAT YPos = { "mupen64-rerecording.exe", 0x867C9C};  //Height
   extern FLOAT ZPos = { "mupen64-rerecording.exe", 0x867CA0};
   extern FLOAT XVel = { "mupen64-rerecording.exe", 0x867CB8};
   extern FLOAT ZVel = { "mupen64-rerecording.exe", 0x867CC0};
   extern FLOAT YVel = { "mupen64-rerecording.exe", 0x867CEC};  //Height

   float HorizontalSpeed = 0;
   float ThreeDSpeed = 0;

    PrintF( "From Second Thread: %u.", dwParm ); //(Left in for demonstration)
    Scriptbreak = 1;  //Break on hotkey global variable declared in HotkeyBreak Script.
    PrintF( "Scriptbreak = %d", Scriptbreak);

    while(Scriptbreak!=0)
    {
		HorizontalSpeed = Sqrt(XVel*XVel + ZVel*ZVel);
		ThreeDSpeed = Sqrt(XVel*XVel + YVel*YVel + ZVel*ZVel);
    
		PrintF( "2DSpeed: %f \n", HorizontalSpeed);
		PrintF( "3DSpeed: %f \n", ThreeDSpeed);	
        PrintF( "Velocity: X: % -015.4f Y: % -015.4f Z: % -015.4f \n", XVel, YVel, ZVel);	
        PrintF( "Position: X: % -015.4f Y: % -015.4f Z: % -015.4f \n", XPos, YPos, ZPos);

		Sleep(100);
		Clear();
		if ( StrICmp( GetCurProcessName(), "mupen64-rerecording.exe" ) != 0 ) 
		{
			PrintF( "Mupen Closed!" );
			return 0;
		}
	}
}
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
Here's a test run on beginner: http://www.youtube.com/watch?v=GELTQMyevn0 FM2: http://dehacked.2y.net/microstorage.php/info/151742259/DH.fm2 I'm still using the same lua script in my previous post, no improvements yet. Final run will be played on advanced mode. If anyone is interested in discussing stuff feel free to contact me. The rest of this post is a copy of the youtube video description: Timing from power on to final shot fired. Most of the time to be gained is by better managing the elevators at the end. The elevator timings are quite random, and are affected largely by your position on the map or the act of you calling the elevators. Shooting doors/windows or picking up objects also seems to change elevator timings sometimes? I need to figure out how the elevators work better before I work on a final TAS. Also it might be possible to kill more crooks before going to the final floor. This run will allow me to better estimate the time that could be expended to kill a crook before going to the final floor. Of note: Before you drain your foot power the diagonal movement alternates between X,Y of 3,3 and 4,4, so fastest movement would always move on the frames that are 4,4. After you drain your foot power you always move 4,4 diagonally, but you only move half as often. A TAS to submit to tasvideos would play on Advanced mode (enemies take ~2x as many hits to kill) The Theo cutscene plays at the end because I accidentally grabbed a radio before the fight. It clears after a period of time and does not technically contribute to the final time. This will not be present in the final run. (My personal best speedrun time on emulator with no lua/turbo/savestates is 4:55 by SDA timing.)
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
TTC: 8 Jiggies + Witch Switch, 100 notes, 5 or 6 Mumbo Tokens 1: Get 4 notes at beginning. Bottom right, top right, top left, bottom left. (It is likely faster with no talon trot.) 2: Jump over the edge and swim to the blue jinjo, swim to land 3: Climb tree to right of Blubber's ship, get 4 notes (Maybe talon trot? It's close.) 4: Jump to "Cliff Plateau" (Can be made barely without talon trot.) 5: Fill Leaky the Bucket and activate bottles during the cutscene 6: Get the shock pad mumbo token nearby 7: Get notes on sandcastle, get notes in sandcastle (You can get ejected by the entrance texture a bit to save a few frames.) 8: Go back around to the crates by bottles, jump up to crates and get all 12 notes on the "Cliff Plateau" area going clockwise. 9: Jump to the nearby tree with notes. Get the 4 notes. Jump into the hole in the side of the boat 10: Get notes and gold treasure. Exit the boat. 11: Swim to the left of the boat. Go up the left net of the boat getting the notes. Talk to bottles. 12: Jump down to the mumbo token on the pole, ground pound the hole in the boat. Get notes, gold treasure, and mumbo token. 13: Go out. Manipulate blubber position if possible, give blubber treasure, get pushed into the jiggy maybe? (I'm unsure how to do this cutscene fastest.) 14: Go up right side of boat nets getting the notes, take flight pad (Do not get the green jinjo yet.) 15: Get the jiggy and the single note in some order. 16: Fly up out the front or back of the overhang to the jiggy at the very top of the lighthouse. (Do not grab the yellow jinjo yet.) 17: Land on the jiggy and get the top notes, ground pound witch switch and *take 1 damage*. 18: Start a talon trot while falling off the ledge and get the 3 notes on the ramp slightly below. 19: Get the yellow jinjo on the tree. 20: Jump to the green jinjo on top of the boat, *taking 2 damage* intentionally on the top of the post. 21: Flap to cancel the damage animation, and land on the flight pad. 22: Take flight, get the 3 notes in the air above the shock pads 23: Fly into the cliff alcove jiggy. 24: Get the pink jinjo, then get the 6 notes on the high cliff. (It's probably faster to land than to turn around and keep flight, but it can be tested.) 25: Go get the notes in the nearby box. It will maybe be fastest to talon trot jump to the tree with the red feathers to avoid fall damage. 26: Go out to the pier ramp, getting those notes 27: Jump over to the stairs area, getting all these notes. 28: Get the two mumbo tokens in the chest. 29: Continue up the stairs getting all notes. At the top, talon trot drop into the water to grab the jiggy. 30: Drop down one floor and grab the three notes. 31: Go out to the flight pad like 6 seconds away, fly back to grab the orange jinjo 32: Drop and grab the alcove jiggy, fly back up to grab the jinjo jiggy. 33: Fly over to nipper if it is faster than landing and talon trotting. 34: Grab the mumbo token behind nipper if it costs less than 4 seconds. 35: Talon Trot or fly to nipper, do the fast fight strategy. *Taking 1 damage.* Attack nipper before cutscene to skip cutscene if faster? 36: Get notes inside nipper, grab the jiggy and die to a crab. -------------------------------------------------------------------------- CC Route: 7 Jiggies + Witch Switch, 98 notes, 3 or 4 Mumbo Tokens Taking 4 damage needs routed still. 1: Jump on top of start. 2: Get the 4 notes to the right, then talon trot jump to the flat platform with the fire logo on the side. 3: Get the 4 notes on the left. (You don't have to drop talon trot to get up on the pipe.) (EDIT: Consider post on next page.) 4: Go back towards the beehive and jump to the pipe to Clanker. (We do not get jinjo jiggy.) Get all the notes in this pipe. 5a: Go through long pipe of notes on the right, swim down to the keyhole, get all notes and go through keyhole 3 times. (This one should be slightly faster if Gloop's air bubbles cooperate.) 5b: Swim down to the keyhole, get all notes and go through the keyhole 3 times. Swim up and go through the long pipe of notes ending by mutant crabs. 6: Go to mutant crabs, get all notes and jiggy. 7: Jump up Clanker's tail or his side depending on the floating cycle. Get the notes on the back half of Clanker and the jiggy. 8: Go all the way up Clanker's tail, get the jiggy and drop down to the pipe under you. 9: Collect the notes from ending at the side on Clanker's personal left by the grate with the honeycomb inside. 10: Go along the slope on the wall and jump to the grate. (you probably have to beak) 11: Talon trot, get the notes here on this pipe ledge -------- 12a-18a: Assuming Clanker's bolt cycle (10 seconds) doesn't require you to wait more than about 3-5 seconds. 12a: Jump down, beak next to the ground (you probably need to kill the enemy in the wall), shock pad forward up to the pipe. 13a: Get the notes on the climb pipe, get the mumbo token near the top. 14a: Jump over towards Clanker's blowhole bolt. 15a: If we have to wait a couple seconds, it may be fast to poop out eggs at Clanker's tooth to break it for later. (or even shooting it out normally may be useful) 16a: Go up Clanker's blowhole bolt. (Getting the notes on the front half of his back won't save any time.) 17a: Get the notes and jiggy up the blowhole. 18a: Drop down and break Clanker's tooth (jiggy side) if it hasn't been done yet. Enter Clanker's tooth (jiggy side) -------- 12b-18b: Assuming route A just misses the bolt cycle and has to wait closer to 9 seconds, this is an alternate path that could maybe end up faster. 12b: Jump towards Clanker's blowhole, if the blowhole isn't leaving yet, poop eggs at Clanker's tooth to break it for later if there is time and the floating cycle is right. 13b: Go up Clanker's blowhole bolt. (Getting the notes on the front half of his back won't save any time.) 14b: Get the notes and the jiggy up the blowhole. 15b: Go back near where the blowhole took you to, 16b: Talon trot slide rejump to the climb pipe with the notes. 17b: Get the notes and the mumbo token. 18b: Drop down and break Clanker's tooth (jiggy side) if it hasn't been done yet. Enter Clanker's tooth (jiggy side) -------- 19: Get the Clanker's tooth jiggy and all notes in this room. 20: Get mumbo token if applicable (HASN'T BEEN DECIDED YET. Kinda depends on if we got the TTC Nipper Mumbo Token and other decisions.) 21: Swim down through the tunnel to clanker's belly 22: Exit out the left side with the 5 notes. (It may be fastest to swim up to the crate, talon trot, jump to the second ring, and talon trot dive) 23: Jump onto Clanker's back, get the remaining notes, go inside the blowhole. (The cycle lines up perfectly here.) 24: Hit the witch switch. Get all the notes and the jiggy. Exit out behind the jiggy. 25: Take flight, fly to the right and get the two notes over here while maintaining flight. 26: Fly to the room across from where you took flight. 27: Do not learn gold feathers. 28: Get all notes, grab the jiggy, and die to a spinning blade. ----------------------------------------------------- FP Route: (3 jiggies, 100 notes, 5 or 6 mumbo tokens) 1: Grab the first 4 notes. Avoid the Boggy cutscene. It may be faster to grab the rest of the notes after the Boggy sled. 2: Go to the left to the Christmas Tree. Get the mumbo token at the base, go inside, get all the notes, then exit. 3: Jump to get the notes around the Christmas tree ending nearer to bottles. 4: Go to bottles and learn beak bomb 5: Go around the present stack and grab the mumbo token, then the 4 notes on top (no jinjos again), then take flight. 6: Break the snowman's buttons in two shots similar to how Sami does it. 7: Fly up high, clip through the bottom of the snowman's pipe and cancel the jiggy animation while retaining flight. 8: Fly up to the hat, land and grab all notes. 9: Ground Pound at the last note, drop off and land on a button or the side of the snowman and flap down below the snowman. 10: Get the jiggy and all notes and mumbo tokens at the snowman's feet ending by the blue present with the snowman on top and 4 notes around it. It may be possible to damage cancel this jiggy with a snowball. 11: Go get the aforementioned 4 notes (if you can take damage here from a snowball to get the mumbo token inside the sir slush in under 4 seconds, do it - you may have to break a beehive later) 12: Go up by Wozza the walrus's cave and get those notes. 13: Talon trot slide rejump your way to about halfway up the hill by the base of the scarf. 14: Get all notes on scarf, ride boggy's sled down. 15: Get the rest of the notes around boggy. 16: At the bottom of the hill go to the right and get the notes by the race start 17: Get all notes in the village, taking damage, then go to mumbo's hut. 18: Get all notes in mumbo's hut and then die. ------------------------------------------------------------- MMM Route: (7 jiggies, 90 notes, 11 mumbo tokens) *Note that we are down one health from FP witch switch jiggy. 1: Get 4 notes at beginning 2: Go off to the left just a bit to manipulate the bat down. Jump to the wooden deck, poop an egg to break the door, and roll towards the left pipe to kill the bat. 3: Do not grab the honeycomb. Climb the left pole and get note on top. 4: Enter the cellar and get jinjo, notes, mumbo token, and jiggy in some order. 5: Exit the cellar, climb the pole and get the note, jump down and get the 3 notes by the gate. 6: Get the mumbo token around the corner. Enter Tumblar's shack and get the 4 notes, then exit. 7: Get on top of Tumblar's shack (probably requires shock pad), and get the notes and mumbo token. 8: Jump onto the hedges and *take a damage*, jump to the well area, get all the notes then enter the well. 9: Get all notes, mumbo token, and jiggy in the well, then exit. 10: Go to the fountain and get the notes, the jinjo, and the mumbo token under the whiplash. *Take Damage* (Ground pounding to get under the whiplash has a low knockback, but there may be a faster way.) --Get two health back somewhere after this. 11: Go out towards the house and go to the right and up the pole (not the one the pumpkin goes down.) 12a: Shock pad to the upper part of the house and get the one note on the left. 12b: Break the nearby window, back flip and ground pound up and get the one note on the left. 13: Break the window and talon trot while entering. Get the notes and the yellow jinjo, then exit. Get *one health* in here. 14: Get the right note, then go across and get the other note, then enter the window 15: Get all notes in this room counterclockwise. (don't forget bottom right note) Kill the second wall enemy before getting the 5th note to set up the honeycomb. Talon trot initiate while crossing the loading zone. 16: Get the other note at the corner, get the green jinjo, then go in the chimney. 17: Get the mumbo token, all the notes, and the jiggy inside. Then exit out the door. 18: Get the health from the bat you killed earlier. 19: Climb the left pole, then talon trot rejump to the nearest note in the hedge maze. (This is probably slightly faster than just going there normally.) 20: Get the first 4 notes in the hedge maze. --Get two health back before this. 21: Attract a ghost and have him *damage boost* you up on top of the hedge. 22: Attract a ghost and get the orange jinjo, then *damage cancel* the jiggy. 23: *Damage boost* back onto the wall. 24: Jump to the notes near the beehive. 25: *Damage boost* up onto the back wall and jump towards the pot in the back corner. 26: Fill all pots with eggs clockwise and grab the mumbo token. (Maybe kill a tombstone to set up the jiggy bounce from the last pot. Do not grab jiggy until pumpkin.) 27: Jump on the roof, go directly up top skipping some notes. Get the notes up top and the jiggy. 28: Jump down to the far part of the roof, get all notes ending nearest mumbo's skull. 29: Jump to mumbo's, transform into pumpkin 30: Get pots jiggy and go to the roof of the chimney house via the ramp. (grabbing the mumbo token.) 31: Go directly down the storm drain to grab the jiggy, grab the notes, and exit the level. CCW Route (0 jiggies, 78 notes, 5 mumbo tokens) *Note that we reset after pumpkin water switch, so we are at 5 health *We need to take 4 damage before the end, it is not fully routed. *This might not read 100% correctly due to recent changes. 1: Go hit the switch to open spring. Get the two notes to the left of spring. Enter spring. 2: Get the mumbo token in the plant by the start. 3: Backflip up twice, get the 3 notes to the left, climb the tree. 4: Get the mumbo token out on the first branch. 5: Continue climbing and hit the switch to open summer. 6: Either ground pound down the tree or jump down and bounce off an enemy at the bottom. Leave spring. 7: Get the two notes to the right of spring and enter Summer. 8: Backflip up and get the two notes on the leaves. 9: Go left and climb the tree again. (There is an alternate path up the leaves, but it seems slower.) 10: Get the 3 notes by the Zubba Hive, continue climbing. 11: Get the 4 notes by the house on the branch, continue climbing. 12: Get the 5 notes across from nabnuts house from left to right. 13: Continue up the tree slightly, get the mumbo token. 14: Work your way around the tree to the switch to open fall below in the drained river. Probably falling one tier at a time from somewhere near here would be optimal: http://www.mediafire.com/?d7g7beeba1hbe76 15: Hit the switch to open fall 16: Go to the beaver dam, ground pound or beak barge it open. Get the 2 notes and the mumbo token. 17: Exit summer, enter fall. 18: Get the mumbo token in the flytrap by the beginning. You should be able to maintain talon trot. 19: Go up the left stack of leaves. Get 9? notes counterclockwise (to the right.) (Stop before the note in front of the bird?) 20: Drop down and get the 3 notes in the flytrap. 21: Go towards the outer edge of the level and get the 5 notes from right to left. 22: Go back towards the center of the level and climb the leaf ramp right by the water. 23: Get the note in front of the bird. It might be possible to jump, get the note, get damaged, and get knocked back to the side you want. Otherwise just jump across and jump back. (Possible 1-2 damage) 24: Get another 5 notes counterclockwise. 25: Start climbing the tree, get the 1 note, continue up the tree to the Zubba Hive. 26: Get the 4 notes in the Zubba Hive. (Unsure if clockwise or counterclockwise is faster.) 27: Continue climbing the tree. Get the mumbo token on the leaf you pass right by. 28: Enter nabnuts house from the right. Get the 3 notes. Exit nabnuts house from the inside right (which puts you on the left outside.) 29: Climb up the tree and get the 8 notes on the nest in the clockwise direction. 30: Fall down towards the ramp far to the right of mumbo's. (Maybe it's faster to land on the branch under you so as to not cause fall damage animation.) 31: Go out of bounds back to the beginning. 32: Enter spring. 33: Backflip up the leaves, go around the stage counterclockwise collecting the rest of the notes.. 34: 3 notes on the ledge near the start, 3 notes on the higher ledge, 4 notes in the dirt by the plant, 3 notes on the low ledge 35: Enter mumbo's hut, change into bee, die to torch and exit the level.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
I found the random trajectory value in MHS. I'm going to describe it rather than posting a value, because everyone has different mupen builds. We will treat it as an unsigned long. The 4 individual byte components might mean different things. Upon crossing a loading zone it always refreshes to 0 (unsigned long). Each hut ground pounded in Mumbo's Mountain sets it to a specific value regardless of what has happened previously. When you break the hut closest to ground pound bottles in Mumbo's Mountain it becomes 1096810496. Counter-Clockwise from that hut, the other 5 huts are 1134985216, 1124139008, 1125318656, 1129906176, 1133412352. Killing enemies (termites, beehives, other stuff) sets it to somewhat random looking values. This is probably due to knockback angles, random explosions, etc. It also seems to have something to do with banjo's approach angle. There are a lot of factors. This means that we should be able to kill a termite in Mumbo's Mountain to stack the Conga jiggies. I will edit my MM route post soon to reflect this. At worst this means we can kill the enemy, load a savestate, hack in the trajectory values and see where the jiggy goes for faster testing. At best maybe someone can figure out how the game actually interprets this value. :P This may only be useful for the Conga jiggies and the Boggy sled jiggy. We will be sure to look for anything else that uses this value.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
There was some recent confusion in the SDA thread: https://forum.speeddemosarchive.com/post/banjokazooie_all_categories_102.html Blank63 found some PAL clips and was able to do the same clips on NTSC Mupen. However, these clips have only been found to work on Mupen and NOT on Project 64 OR console. So now we know that mupen collision isn't 100% accurate for this game. Future new clipping tricks will be confirmed in Project 64. (Most everything Cronikeys tested has been confirmed in Project 64.) 1: We're not doing the TAS on PAL 2: Clips not known to be possible on console will not be used in the TAS. We're still relatively safe to assume that Project 64 is a pretty reliable representation of the game at this point.
Post subject: Re: Mumbo's Mountain
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
AnotherUser wrote:
Also, in TTC, is it faster to collect the blubber jiggy during the blubber cutscene, or through flying as done in the previous TAS?
Unsure. We'll test it when we get there. No other levels have been fully routed yet. There are still some notes and mumbo tokens that may get cut. It is determined that no matter what is cut, MM is still all notes, jiggies, and mumbo tokens. Regarding CC: the 2 notes in the alcoves may be cut. We're aware of the jumps from the top. You can jump further than the jump you linked with a talon trot slide rejump as shown in the CC Puzzle jump attempt a few pages back. My current idea revolved around entering the tooth, grabbing the jiggy and the 8 notes, and then re-exiting the tooth, then later entering and leaving the left gill to get the 5 notes. The 2 notes off on the right can be grabbed while flying I'm pretty sure, but they need timed. I wanted to die by the gold feathers jiggy, since dying by mutant crabs is I think impossible. Much more needs routed before we start TTC. What to get and not get has largely been calculated by "amount of deviation from the nearest object we need to get", as opposed to actually routing the level.
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
Experienced Forum User, Published Author, Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
Okay, the route is going to CC as Banjo and BGS as Bee now. This was a pretty big oversight. Whoops. Why is this faster? - The proposed CC as Bee takes about 120 seconds longer than the proposed BGS as Bee. - The proposed CC as Banjo takes about 45 seconds longer than the proposed BGS as Banjo - Opening BGS takes about 30 more seconds overall than "opening CC without CC puzzle jump." - Opening BGS costs 2 extra jiggies, which equates to about 60 seconds lost. 120 + 30 + 60 - 45 = 165 seconds saved. There are some small considerations to be subtracted from this (such as less notes available in BGS as bee), which got me down to 124 seconds saved by going to CC as banjo. -------------------------------------------------------------------- This leads us to a new list of jiggies. All available jiggies going to these worlds are listed. The ones we will not get are struck out. - Available: 57 - Needed: 41 - Deleted: 16 EDIT: BGS WITCH SWITCH JIGGY NEEDS ROUTED IN, YOU CAN CLIP IN. GL:
  • First Jiggy
  • GV Witch Switch
MM:
  • Conga Eggs
  • Conga Oranges
  • Chimpy
  • Juju
  • Termite Hill
  • By Talon Trot
  • Hill 1
  • Mumbo Skull Eye
  • Huts
  • Jinjos
  • Witch Switch
TTC:
  • Blubber Gold
  • Inside Crab
  • Sandcastle
  • Lighthouse
  • X Marks
  • Fly to Chest
  • Shock Pad Alcove
  • Fly to Narrow Path Alcove
  • Pool Dive
  • Jinjos
  • Witch Switch
CC:
  • Keyhole/On Clanker
  • Mutant Crabs
  • Gold Tooth
  • Wonderwing
  • In Blowhole/Spinning Blades
  • 8 rings
  • Up blowhole bolt
  • Clanker's Tail
  • Long Pipe Swim
  • Jinjos
  • Witch Switch
FP:
  • Sled on Boggy
  • Presents
  • Snowman Pipe
  • Snowmen Beak Bomb
  • Christmas Tree
  • Walrus Race
  • Boggy Race
  • Wozza
  • Snowman Buttons
  • Jinjos
  • Witch Switch
MMM:
  • Inside Well
  • Tumblar
  • Pots/Eggs
  • On top of church
  • Chimney/Green Ghost
  • Cellar
  • Inside Loggo
  • Rain Gutter
  • Organ
  • Jinjos
  • Witch Switch
The shortest jiggy that got removed is Loggo, timed at about 29.5 seconds. The longest jiggies to remain include MM witch switch, GV witch switch, and FP witch switch, all estimated to be below 27 seconds all things considered. -------------------------------------------------------- A note is valued at 5-6 seconds. A mumbo token is valued at 3-4 seconds. Now that jiggies have been finalized, notes and mumbo tokens can be measured more accurately as deviations from the shortest path. It has been determined that everything in MM will be collected except the two honeycombs. No single note in MM takes more than 5 seconds, and no single Mumbo Token takes more than 3 seconds. (A lot of the 5 seconds notes occur in RBB by having to cross loading zones.) The next goal is to finalize notes and mumbo tokens, and then following that individual level routes will be theorized. MM has been fully routed for some time, as the stuff to be collected hasn't been changed. MM Route!