I've started to examine LordTom's last published any% TAS and already tested a few things.
I believe the GameResource page is pretty outdated and I didn't found any mentions about movement speed trick/glitch.
There's a weird movement system in this game that allows you to change direction on specific frames without losing speed with decreasing the sub x speeds (exponential/linear). This is good since,
1. You can get a different luck (at least I got magical lag frame reduction while all enemy spawns were the same)
2. You can exit doors faster (but your pos can be worse if you press at wrong time)
This isn't abused in any of the Metroid TASes, am I right?
PhD in TASing 🎓 speedrun enthusiast ❤🚷🔥 white hat hacker ▓ black box tester ░ censorships and rules...
Currently the published and my WIP uses the P1 "T", P2 "A U" combo to show the password and "initiate" it (start it).
To get the "kraid died, have missile" password, I need to wait ~1 (kraid dies) + 24 (in-game delay timer) seconds.
So... would it be eligible to...
1 - Save the password from the aforementioned place (frame ~6986)
2 - Remove these frames (~25*60) and reset the game (using reset) (frame ~5488)
3 - Input password (few frames)
Obviously this saves time, however it would look so fake... while Kraid object really dies once his health drops to 0, the flag "Kraid's dead" activates after the 24 second delay (music thingy).
I'm not sure is it "pathetic" to reset the game once Kraid object is dead (but not flag) and reset the game to input a password that will actually give me the Kraid flag activated which would also happen if I would just wait 24 seconds.
PhD in TASing 🎓 speedrun enthusiast ❤🚷🔥 white hat hacker ▓ black box tester ░ censorships and rules...
To me it seems morally equivalent to playing the TAS until you get a password for just before the final boss, noting what it is, starting the TAS from the beginning and immediately entering that password.
Kills kraid, music starts, 24 seconds wait, uses in-game reset, password shown and started. (kraid killed, 75 missiles, etc)
Using in-game reset earlier (before music starts) doesn't gives you the "kraid died, 75 missiles" thing.
So killing Kraid, "reset" button and writing in the password would be faster, but technically it could be considered as cheating.
PhD in TASing 🎓 speedrun enthusiast ❤🚷🔥 white hat hacker ▓ black box tester ░ censorships and rules...
The password here is basically a stand-in for a save. We've had similar arguments before; generally people have said that once you've "unlocked" a password (i.e. seen it in-game from a clean start), you're free to reset the game and use that password. What you shouldn't do is use passwords that you haven't seen yet.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
This is effectively skipping a cutscene via editing the save file to just after the cutscene. Skipping cutscenes via resetting is only allowed for battery saves if the game autosaves at the start of the cutscene, and loads at the end; if it doesn't autosave until the end of the cutscene, then resetting to skip the cutscene doesn't work. Password saves should work the same; the game has a point at which it saves, and if it's at the end of the cutscene, bad luck.
You are right (just like ais523). I won't abuse it in my TAS.
Btw this is what happens after you kill Kraid (distributed through 2 frames):
#1 Kraid died flag
#1 24 second timer added
#2 75 missiles added
However I wonder if your (Derakon's) "idea" is also valid if the password would be shown in the memory. In Metroid's case, it's only calculated when the game returns to "MainRoutine" (when the password shown).
And what about further exploiting the "TAS universe" analogy somewhere from the "initial state ram" thread?. Obviously I know that after kraid dies, I will get a password using in-game reset. Let's say I made the TAS, the next time I rerecord it (same input), I will know what will be the password (since I experienced in that universe). So I could (hardware) reset the game and insert that password I will get. It will be obviously a "not right" password (checksum, in-game time differences compared to what I should get with calling the MainRoutine at that time), but this also could be fixed since password generation is disassembled and known.
About my TASing progress: more than 50% done (start ~ after ice beam got), currently revisiting the whole TAS. Currently 231 frames faster.
PhD in TASing 🎓 speedrun enthusiast ❤🚷🔥 white hat hacker ▓ black box tester ░ censorships and rules...
Joined: 5/25/2007
Posts: 399
Location: New England
I'm impressed to hear you're 4 seconds ahead as of the ice beam, I was aware of an maybe 1-2 seconds improvement for the whole run based on making the low % and 100% so wherever you end up it'll be bigger than I expected.
I have to say I'd likely consider any password-entry strat to be invalid for this title, though I'd have to see the strat to form a final opinion.
I may be confused/misunderstanding, but I don't think the delay after killing Kraid is even close to 24 seconds. Watching on YT, it seems there's about 5s from when K takes the last hit until the reset can occur, then about another 7 to get moving again afterwards (which wouldn't be decreased using early reset-->enter password).
Finally, does OoB just mean "out of bounds??"
Anyways, great work & looking forward to the final product!
I'm impressed to hear you're 4 seconds ahead as of the ice beam, I was aware of an maybe 1-2 seconds improvement for the whole run based on making the low % and 100% so wherever you end up it'll be bigger than I expected.
I don't noted which techniques made a lot of improvements, but fundamental ones are X subpos and Y pos optimizing and a lot of tricky lag removement tricks.
Lord Tom wrote:
I may be confused/misunderstanding, but I don't think the delay after killing Kraid is even close to 24 seconds. Watching on YT, it seems there's about 5s from when K takes the last hit until the reset can occur, then about another 7 to get moving again afterwards (which wouldn't be decreased using early reset-->enter password).
Sorry, I totally forgot that. The delay timer's unit is only 10 frames, so 24 delay (for example at kraid) is 240 frames (249 at worst delay) = 4 second and not 24 as I mentioned. It looks so much longer when I frame advance =)
Lord Tom wrote:
Finally, does OoB just mean "out of bounds??"
Yes, whenever Samus is not on the screen.
Lord Tom wrote:
Anyways, great work & looking forward to the final product!
Thank you! Well, I'm still improving up until to Kraid, I'm nearly tested all possible paths, so currently I reset the game 159 frames faster, I need 3 frames to improve it with 9 frames (delay timer).
Also since it looks like I will have to do even more tests for the rest of the TAS even after 110k rerecords, I think I will eventually release my WIPs, since I already "fuck up" the TAS with random sound effects and music. Would like to know what others think about that, so wait until I tear apart Kraid.
PhD in TASing 🎓 speedrun enthusiast ❤🚷🔥 white hat hacker ▓ black box tester ░ censorships and rules...
Hello, I was just curious if this glitch was known or if it could possibly be explained to me why this happened: http://www.twitch.tv/gari/c/5095471
Basically around the 45 second mark I go into a door that normally leads to the big climb, but for whatever reason the game took me to a completely different part of the map with a broken door, leading me to falling out of bounds. I've seen similar glitches on youtube, but those all involve wallclimbing out of bounds into the secret worlds, which my video did not. Thanks to anyone who answers this and I'm hoping for the very tiny offschance this is something new and potentially useful to a tas.
FYI, console verified 100% run. Requires the console RAM to be cleared with the clearing cart (writes FCEUX 00/FF pattern). Will put videos up when I get my visualization boards in and I write the firmware to utilize them.
Hello, I was just curious if this glitch was known or if it could possibly be explained to me why this happened: http://www.twitch.tv/gari/c/5095471
This doesn't look like a wrong-warp to me. It looks like your map coordinates are correct, but the game loads the wrong room number ($00 instead of $04). You can replicate what happened in the video using this Game Genie code: AAAXPYGA
Your guess is as good as mine how the game got the wrong room number. Memory corruption would be my guess, but I have no idea how the corruption would have been triggered.
Memory corruption would be my guess, but I have no idea how the corruption would have been triggered.
While I know you can trigger a memory corruption which affects the game many (2000?) frames later by not pressing R for 1 frame, the video footage probably just happened because he continued the game and used the left-over data.
Pretty sure you couldn't do this consistently from power-on but using continue you probably able to recreate it in real time (depending on what addresses you can manipulate).
PhD in TASing 🎓 speedrun enthusiast ❤🚷🔥 white hat hacker ▓ black box tester ░ censorships and rules...
Can memory corruption be used to save time for this game?
The potential corruption will probably only lead to $00 room, so as long as you can't manipulate other values before hand (continuing the game from another gameplay), you can't.
PhD in TASing 🎓 speedrun enthusiast ❤🚷🔥 white hat hacker ▓ black box tester ░ censorships and rules...
Hello there...I was wondering if anyone has advice on what to change in a Metroid ROM to achieve the same Kraid position on Powerpak as on console. On Powerpak, whether from poweron or reset, Kraid is further away from door. Thanks.
speedrunner of SMB2, Metroid, Pitfall II, Blinky Goes Up
Hello there...I was wondering if anyone has advice on what to change in a Metroid ROM to achieve the same Kraid position on Powerpak as on console. On Powerpak, whether from poweron or reset, Kraid is further away from door. Thanks.