1 2
10 11
Posted: 5/21/2016 7:38 PM
Quote
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
TASer of 2010 NES TASer of 2010 DS TASer of 2010 Arcade TAS of 2010 NES TASer of 2008 NES TAS of 2008 Funny TAS of 2008 TASer of 2007 NES TASer of 2007 Funny TAS of 2007
TheAxeMan wrote:
That's actually pretty good progress. Would you have done anything differently in the first three chapters if you knew Alena and Ragnar were staying down in favor of Taloon? Fewer levelup stats?
Honestly yeah, lower HP on Alena for sure, she was hard to kill off :) And certainly I spent extra frames manipulating better stats for them, that is now unnecessary.
It's hard to look this good. My TAS projects
Posted: 6/3/2016 12:38 PM
Quote
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
TASer of 2010 NES TASer of 2010 DS TASer of 2010 Arcade TAS of 2010 NES TASer of 2008 NES TAS of 2008 Funny TAS of 2008 TASer of 2007 NES TASer of 2007 Funny TAS of 2007
So I've made progress, I just finished the Bakor fight. I have an idea and a decision to make which is the main reason I'm posting. I'm puttiing it first to make sure people see it, then I'll put my status update: Now that I have the final key, I have an option. I was looking for better weapons for Taloon to turn him into more of a beast. And I noticed he can equip the Sword of Decimation. It is cursed but the downside is no defense, but I need no defense in a TAS. Compared to the Sword of Malice it has an additional +75 damage, which translates roughly to +75 to my critical hits (and +150 after building power). And there is one conveniently in the town of Konenbur behind a final key door. It isn't that far out of the way, but it also only saves time in boss fights, which are going very quickly anyway. If it pays off, I won't really know it without basically TASing the rest of the game twice, which I lack the sanity to do, this TAS is killing me! I want it to be faster though, because I think there's a lot of entertainment value here. Using a cursed weapon as a legitimate technique is hilarious. Also, Taloon would be killing the most powerful bosses in just one attack round while the Hero becomes completely useless for fighting. He's just there to hold the pretty weapons and armor and get all the credit, while Taloon saves the world :) I'm not sure if I should go this route or not though, it could very well be slower over all :( Ok, here my status update: The decision to just do a Hero, Taloon duo is paying off big time. Taloon is a beast. I gave Ragnar's Sword of Malice to Taloon. After building power, his critical hits are 200+ damage! I defeated Balack in one set of Merchant Reinforcements. Also another innovation is that I'm manipulating taloon to cover the enemies mouth to stop spell attacks which he can do in addition to his regular attack. This replaces the slower method of wasting taloon's action to trip the enemy. Bakor with 1000hp went down in just two rounds. I"m saving a lot of time in the fights and a lot more by not having to level up Alena and Ragnar. I also saved some time by getting the Zenithian Shield on my way to Gardenbur since it was near by.
It's hard to look this good. My TAS projects
Posted: 6/8/2016 9:52 AM
Quote
TheAxeMan
Ambassador, Experienced player (709)
Joined: 7/17/2004
Posts: 985
Location: The FLOATING CASTLE
Tough decision. Dragon Warrior 2 also used a cursed weapon for more power though there the decision is easier with no crits on the final boss. You could probably figure out how much time it takes to get the weapon and equip it fairly easily. Then you can estimate how many hits you'll need to save to make it worth it. Would you use the defence lowering effect of the sword of Decimation? Infernus Shadow, who needs to be beaten anyway, drops a Sword of Miracles that is almost as strong. The downside would be a short healing message after every attack though. Watching your old run, you beat him first in the four barrier area, probably to get the sword from him. You would still have sword of Malice for Esturk. It wouldn't take long to get the sword in Konenbur but it is completely out of the way. I guess it is a pretty tough call. Good luck!
Posted: 7/1/2016 11:23 AM
Quote
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
TASer of 2010 NES TASer of 2010 DS TASer of 2010 Arcade TAS of 2010 NES TASer of 2008 NES TAS of 2008 Funny TAS of 2008 TASer of 2007 NES TASer of 2007 Funny TAS of 2007
So just an update. The Sword of Decimation was in fact a good idea. And I did finish the TAS! ...Then instead of submitting, I restarted. It just really bugged me that there were some sloppy things in the early chapters I could improve and that I could improve more things based on knowing that Alena and Ragnar aren't going to be used in chapter 5. As of right now I'm already back up to the Chapter 5 balzack fight. I saved over 2900 frames as of the beginning of chapter 5 compared to the wip I've posted here. In chapter 1 I made sure to manipulate less HP for Ragnar so I could kill him off in 1 hit in chapter 5. Also, all the level ups in that chapter are after the boss fight, so I made sure not to waste any frames trying to maximize strength or agility. Any delay frames I used were to avoid stat ups, and to lower HP. The big time saver is chapter 2. In the previous versions I get into a metal babble fight to make the tournament faster. The reason that was a good idea is that those levels I would get anyway in chapter 5, at most it costs me 1 level up by the end of the TAS. Knowing that I won't use Alena, this time I avoid the metal babble fight. It saves about 30 seconds, plus it makes the TAS more interesting since Alena goes into it even more under-leveled. Also, in general the delay frames are far less this time, by the end of my previous TAS I had really refined my bot scripts and luck manipulation techniques, so I was able to apply those techniques this time around. It is grueling going through the TAS yet again, but I'm glad I did, it looks a lot more polished this time around! This game deserves the highest quality TAS even if it is an insane amount of work :)
It's hard to look this good. My TAS projects
Posted: 7/11/2016 12:27 PM
Quote
Justin_Z
Joined: 12/29/2014
Posts: 16
Really excited to see your end product, as I have no doubt from reading your posts that you've put in a monumental effort!
Posted: 7/12/2016 2:19 AM
Quote
juef
He/Him
Player (154)
Joined: 1/29/2007
Posts: 208
Location: Québec, Canada
That is going to be quite the improvement, in terms of speed as well as in terms of quality! I admire your dedication and can't wait to see the result.
Posted: 2/9/2019 5:03 PM
Post subject: Yes, I am a masochist
Quote
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
TASer of 2010 NES TASer of 2010 DS TASer of 2010 Arcade TAS of 2010 NES TASer of 2008 NES TAS of 2008 Funny TAS of 2008 TASer of 2007 NES TASer of 2007 Funny TAS of 2007
Here is a WIP through chapter 2 of an improvement to my previous movie that I've been working on. At the end of chapter 2 I'm 2050 frames ahead. However, some of that will be lost to the fact that the day/night cycle is off at the beginning of chapter 3 (the game doesn't reset the day/night cycle between chapter 2 and 3!) The main motivation for this movie is that I have a much nicer computer than before and could bot increasingly unlikely events with shorter delays. Previously I took the approach of attempting to have generic scripts for things like manipulating a critical hit. This time I made an easier framework for which I could write custom scripts for every situaiton I encounter and the ability to chain unlikely events together. From this I've been able to remove nearly all the delay frames even from insanely unlikely events (such as rnd 1 of the first chapter 2 boss fight). In general, I'm doing more manipulations, and more chains. And the likely hood of success is orders of magnitude more unlikely than the previous movie's maniuplations. In addition I have a few mistakes in the previous movie, some of which I'm even embarrassed by: 1) The day/night cycle is manipulatable! Believe it or not, a step doesn't necessarily increase the cycle. Not catching this in previous movies is sloppy on my part. I had the wrong address for day/night and never noticed lol. 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. From this I was able to manipulate night in chapter 1 about 60 frames faster. And a much bigger boost from not needing to use the inn in Endor on chapter 2 by making sure it is still day when I get there (however, there may be little to no gain from this, depending on how I deal with the bad cycle in chapter 3). 2) Screen transition lag In the previous movie, I didn't notice that lag is random in screen transitions. Yes, random, as in the rng. Is there anything that isn't random in this game? No. I figured this out somewhere in chapter 5 in the previous movie. Not a lot of time is saved from this but it does add up over the course of such a long movie. 3) Better (worse?) stat manipulation. Ragnar for instance has about a 1/1000 chance of not getting vitality on early levels (up to lv 5). I was able to manipulate 3 levels to be only strength from this, saving a good deal of time. 2 of those manipulations were back to back, which given that the rng has almost no ability to be manipulated through level ups is incredible. I was able to use some of the boss fight (which had its own difficult manipulations) to manipulate essentially a 1 in a million scenario! In addition I was able to minimize Alena's stats too in some 1/1000 situations. 4) Better skeleton fight in chapter 2 This is a mistake I figured out too late in the making of the previous movie. I did realize that the lowest HP for a Skeleton was 42. However, It failed to occur to me that if I manipulated 1 more str (1/1000 chance) then Alena's max critical would be 21 and therefore make the fight only 2 rounds long.
It's hard to look this good. My TAS projects
Posted: 2/14/2019 2:17 AM
Quote
pirohiko
Skilled player (1785)
Joined: 5/7/2008
Posts: 187
Location: Japan
NES TAS of 2020 Glitchy TAS of 2020 Speedy TAS of 2017 NES TAS of 2009
It's an awesome RNG manip! By the way, SubNESHawk which can subframe input has been developed recently. I examined it and found that there were two input responses per frame. As a result, the cursor can be moved continuously once every frame. I tried that and made "Arbitrary Code Execution" of Dragon Quest IV. http://tasvideos.org/forum/viewtopic.php?p=481230#481230
Posted: 2/14/2019 10:50 PM
Quote
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
TASer of 2010 NES TASer of 2010 DS TASer of 2010 Arcade TAS of 2010 NES TASer of 2008 NES TAS of 2008 Funny TAS of 2008 TASer of 2007 NES TASer of 2007 Funny TAS of 2007
Yeah, it's unforutnate. Likely having subframe input would improve the TAS. However, there's no way I could do the botting that I'm doing using NesHawk. It would take many years longer than using quickNes
It's hard to look this good. My TAS projects
Posted: 2/15/2019 1:49 AM
Quote
Alyosha
He/Him
Editor, Emulator Coder, Expert player (3821)
Joined: 11/30/2014
Posts: 2829
Location: US
GBA TASer of 2023 Homebrew TAS of 2023 Gameboy TASer of 2022 NES TASer of 2015
Hmm, this made me think, NESHawk is currently very slow, but it is also in a form where one can run the entire underlying core one cycle at a time. You also have access to the entire state at all times. So you could basically extract the entire core as is and directly translate it to C for example and be able to increment the core one cycle with one function call. You could then wrap any machinery you like around that and have no extra BizHawk overhead. You could probably also simplify things a bit for a specific application for additional performance benefits . I’m not sure if this would be faster then QuickNes, but you could do some serious stuff with it. And the results would still be valid on BizHawk. Just a thought . I might consider trying it in the future.
Posted: 6/25/2020 1:07 AM
Quote
Kirkq
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.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 6/30/2020 10:35 PM
Quote
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
TASer of 2010 NES TASer of 2010 DS TASer of 2010 Arcade TAS of 2010 NES TASer of 2008 NES TAS of 2008 Funny TAS of 2008 TASer of 2007 NES TASer of 2007 Funny TAS of 2007
Kirkq wrote:
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)
Wow! And dang, that’s a lot of TAS to redo :( I’m hopelessly stuck in chapter 3 anyway at the armor selling. I’ve bottled millions of attempts with recreating the luck in the old TAS. However all my testing suggests it’s possible.
It's hard to look this good. My TAS projects
Posted: 2/12/2023 12:09 AM
Quote
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
TASer of 2010 NES TASer of 2010 DS TASer of 2010 Arcade TAS of 2010 NES TASer of 2008 NES TAS of 2008 Funny TAS of 2008 TASer of 2007 NES TASer of 2007 Funny TAS of 2007
So I managed to finish a redo of Chapter 1: User movie #638117571365360120 It is 227 frames faster, about 180 from staying at the inn. I also got an even more absurd level up. About 10 million bot attempts went into the level ups at the end of this chapter. Other frames were scavanged from screen transitions. More and more I am finding that to truly optimize this game, I have to do more and more precise bots that bot larger chunks of the game. At this point, almost nothing is done by hand anymore. The good news is the most unlikely events of the whole TAS are done. Bad news is that there is stil an awful lot to do, and the Chapter 2 boss fight is also absurdly unlikely, will probably be a few million bot attempts as well. Finishing this TAS will be a long journey, yeesh
It's hard to look this good. My TAS projects
Posted: 3/14/2023 9:47 PM
Post subject: Chapter 3 done
Quote
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
TASer of 2010 NES TASer of 2010 DS TASer of 2010 Arcade TAS of 2010 NES TASer of 2008 NES TAS of 2008 Funny TAS of 2008 TASer of 2007 NES TASer of 2007 Funny TAS of 2007
User movie #638144270152713770 Chapter 3 done! This was the major blocker that stopped my progress years ago, and the last of the super incredibly unlikely manips. I get the 3 max armor offers without declining a single one, which saves some seconds, but the odds of this are about 1 in a million or more. I'm 2475 frames ahead of the published movie with about 600 because I end 1 level ahead but will still end up on the same final level. The strat here is to get a 2nd metal slime so that I have more strength for the drop encounters and not need to do critical hits. The experience works out that it will not cost another level (but 3rd metal slime would have), so this will be realized as saving by the end of the TAS I'm officially optimistic about this TAS and super motivated to keep going! The worst is behind me at this point
It's hard to look this good. My TAS projects
Posted: 3/19/2023 6:41 PM
Quote
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
TASer of 2010 NES TASer of 2010 DS TASer of 2010 Arcade TAS of 2010 NES TASer of 2008 NES TAS of 2008 Funny TAS of 2008 TASer of 2007 NES TASer of 2007 Funny TAS of 2007
User movie #638148478993374149 Another update, got through Chapter 4, and am up to the first encounter of Chapter 5 (with the 2 Liclicks). Improvements in Chapter 4: - Getting the gunpowder jar then Orin, this way the death warp that needs to kill off 2 characters is with the more powerful enemies, which saves an action - Better luck manipulation in the level ups, I get no vitality/hp in every level - Transition lag manipulation, particularly in the ending scene of sailing to Endor, a surprising amount of frames were found just in a scripted cut scene
It's hard to look this good. My TAS projects
Posted: 3/10/2024 7:22 AM
(Edited: 3/10/2024 7:22 AM)
Quote
Raistlin212
Joined: 3/3/2024
Posts: 1
Location: USA
I'm trying to do some research on the initiative system for this game, but can't I seem to grasp the fundamental rule. Is it that you roll 25-100% of your Agility with an equal chance for any result (i.e. it's not curved)?
Posted: 3/10/2024 2:52 PM
Quote
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
TASer of 2010 NES TASer of 2010 DS TASer of 2010 Arcade TAS of 2010 NES TASer of 2008 NES TAS of 2008 Funny TAS of 2008 TASer of 2007 NES TASer of 2007 Funny TAS of 2007
Unfortunately I don't know the exact math. In the TAS, anything that can happen can be manipulated to happen so I was concerned with if it could happen. I was also concerned with approximately how likely it was so I could have an idea of how many delay frames I might expect to get. Odds were a secondary concern. If you want to try to reverse engineer the math, you can find all my scripts and understanding of the RAM here Specifically these ram addresses They show that there are 8 battle slots that are rolled at the beginning of the round. Once set, they stay that way for the entirety of the round. Agility of the characters/enemies absolutely is a factor. It's possible the enemies have other factors that force some odds. The way I tested what was possible was by getting to a specific point with a savestate, stopping my movie, setting up and running this I'd be happy to work with you on how to use this script.
It's hard to look this good. My TAS projects
Posted: 3/31/2024 1:46 AM
Quote
Kirkq
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.
Twitch Stream, Twitter, Youtube, TASVideos Profile
1 2
10 11