Joined: 11/14/2005
Posts: 1058
Location: United States
Damn, I was way wrong. I found the memory address. It is 7E0158, there are 255 (unless there is some pattern) variables, not 9. To test this, load up this save state. Press (and hold) the up button when the memory address reaches a certain number (any number will do). Load up the same state again and wait for the memory address to cycle through 1 time. Then press the up button when it hits the same number that you press up for on the first time. You will notice that Chrono's sprite will reach the exact same subpixel as before. Pressing up at different frames will result in different sprite positions for Crono, but as long as you press up at the same time on the same number in this memory address the position of Crono's sprite will always be the same.
edit:
It makes perfect sense now. Every time you move Crono the memory address resets to 1. When you stop movement, it starts to cycle again.
Joined: 11/14/2005
Posts: 1058
Location: United States
pretty much, yes. If you looks at the input on my movie file, you will see that I completely stop movement for 1 frame before walking towards Crono's mom. I also delay several frames before I move during the candy sequence skip.
That's pretty cool, that kind of remind me what happened on console long ago, I was able to get Crono to get up before bell ring again. Maybe that script skipping apply to that too?
I could be wrong or don't remember correctly, I just remembered that something "different" happened at Crono and Marle collision.
Joined: 11/14/2005
Posts: 1058
Location: United States
I was also thinking of that too. I hope it doesn't crash the game. The game will freeze if Chrono goes downwards after the telepod event, so there is a chance that this may screw other things up later on.
EDIT:
ahhh, very lame news. I just played through the entire cathedral/yakra thing. When I get back to the present, the candy scene is automatically forced when I walk downwards with marle. Oh well, it will still save time, but not really that much.
ahhh, very lame news. I just played through the entire cathedral/yakra thing. When I get back to the present, the candy scene is automatically forced when I walk downwards with marle. Oh well, it will still save time, but not really that much.
Hmmm, how would it save even a slight ammount of time at all then? You would just delay the scene for later, those it would be same fast in the end, no?
(Actually since you have to slowdown a bit to pass that scene the 1st time, it might be even slower in the end?)
Joined: 11/14/2005
Posts: 1058
Location: United States
Nitrodon wrote:
The scene moves you to the window of the candy shop, so it saves time to activate it while moving south.
Thats exactly why I thought it would be slightly faster.
Either way, I am gonna keep it the way it is. It is one of only a few things that might surprise the viewer in the whole movie.
Can you also use the event skipping trick to avoid triggering Marle's escape event when you exit the present-day castle the first time? The exit from the castle is quite close there.
See also this idea: http://tasvideos.org/forum/p/125623#125623
Also, remember to play with camera angles in the race with Johnny. You neglected to do that in the previous movie.
Joined: 11/14/2005
Posts: 1058
Location: United States
Bisqwit wrote:
Can you also use the event skipping trick to avoid triggering Marle's escape event when you exit the present-day castle the first time? The exit from the castle is quite close there.
I was thinking of that too. There seems to be a good chance that this scene will be skipable. I will do thorough testing once I reach that point in the run.
Bisqwit wrote:
I only visit a few item shops throughout the entire run, maybe 3. I will keep this I mind when I visit one though.
Bisqwit wrote:
Also, remember to play with camera angles in the race with Johnny. You neglected to do that in the previous movie.
I didn't know that it was possible to change the camera angles. The link you provided doesn't have any seeds, so I can't download it. I will surely mess around with the camera angles in the new run, after I learn how :P
The link you provided doesn't have any seeds, so I can't download it. I will surely mess around with the camera angles in the new run, after I learn how :P
If you're still trying to download it, my autoseeder should pick it soon and start seeding it. It seeds those files which have demand. Because it can only seed so many files at once, it has to prioritize those which have demand over those which have no demand.
Joined: 11/14/2005
Posts: 1058
Location: United States
Just watched the lavos battle and some of the other boss battles. The strategies are ground breaking. He consistently changes to battle speed gauge setting in order to achieve better enemy attacks, which is something I completely overlooked in my run. I will probably spend the next few weeks slowly studying what he does in his run. From what I have seen so far, there is not much room for improvement in his strategies. This is a truly genius speed run.
Great work inichi!
Thanks for making this run, and thanks for mentioning my name in the credits.
Just watched the lavos battle and some of the other boss battles. The strategies are ground breaking. He consistently changes to battle speed gauge setting in order to achieve better enemy attacks, which is something I completely overlooked in my run. I will probably spend the next few weeks slowly studying what he does in his run. From what I have seen so far, there is not much room for improvement in his strategies. This is a truly genius speed run.
Great work inichi!
Thanks for making this run, and thanks for mentioning my name in the credits.
When I started Trigger run first, I just tried to keep up with your run for each section. By doing this, I learned how to avoid fixed encounts, why you used that strategy in any boss battle and finally was able to make somewhat new strategies in some part. So nothing makes me happier than to hear you say that.
Hero, you don't have to worry about someone's thoughtless words. Because everyone knows how epoch-making things you did on Trigger run. It's true that I shortend your record by more than 20 minutes. But, as I said before, without borrowing your great strategy in most boss battles until Magus, I could never have achived such a record. So I should be the one thanking you. After all, I'm only standing on the shoulder of giants.
Anyway, now that my run is uploaded on SDA, I'd like to be help of your run to optimize it as well as I can. For the time being, I put here some improvment I found after I submitted the run.
Segment 26 Exchange a kind of enemies in 5th battle with one in 6th battle.
In Chrono Trigger, when you learn a new technique, the more you get TP, the less required TP to learn next technique is. The enemies I fight in Tylano Lair is two kinds. Some has 18 TP, and the other has 22 TP. Unfortunately, in 5th battle, I fighted one which has less TP, then Ayla learned Tail Spin. That is why Ayla's TP lackd 3 for learning Dino Tail in my run, and I had no choice but to fight 3 Mutants in Lab 16. It would be prevented from happening if I fighted enemies which has more TP in 5th battle.(10 seconds)
Segment 28 Use Robo instead of Marle in the Tylano battle.
Strategy: Use Twister 2 times, and Azala falls. After Tylano's defence weakens, use Twister 4 times, then Tylano breath out the fire. Wait until Tylano's defence weakens again, then use Twister 3 times. Perhaps, a few phisical attack will be needed. Less number of attacks would save a time. (10-15 seconds?)
Segment 33 Replace some of Crono's confuse with some Ayla's Dino Tail in the Lavos Core battle.
In fact, the insane luck and accurate key responce is needed to reach the Lavos Core battle and the center bit changes the ages in the ideal order. I was so afraid to spoil the best chance that I chose killing the right bit before it revives the center bit as a safe move. But, There are some times between it relives and uses Dreamless. so it can be replaced some of Crono's confuse with some Ayla's Dino Tail. Much less number of attacks will surely save a lot of time. Besides, Crono's second Confuse to the center bit is an unnecessary attack by my mistake. I think it is segment 33 that most improvement could be found in my run. (20-30 seconds?)
On a final note, I think you definitely will break my in-game time record though the English version has less text speed than the Japanese one. Considering some improvement being found in Saturn's run even within 10 minutes again and again, It is really natural. If you have any questions about my run, feel free to ask me.
Perhaps, you can skip getting Swallow.
I have thought that bigger damage of Confuse is necessity to beat Lavos Core until recently. So, I went all the way to get Swallow spending more than 1 minute and 30 seconds in my run. However, as I wrote above, I found it is not very necessary. Of course, skipping Swallow will make the some form of Shell Lavos and Retinite battle a longer. So I'm not sure if it would work well, but it would be worth considering at least.
I guess a real 100% TAS would be rather tedious to do. Because you would have to collect one of every item, weapon and armor, learn EVERYONE'S Techniques and complete all quests and side quests.
I don't call the current SDA 100% a true 100% Speed Run for exactly that reason.
If I am understanding correctly, are you doing a TAS like the newest Speed Run, or are you going to revive Crono after Lavos desintegrates him?
Joined: 11/14/2005
Posts: 1058
Location: United States
Okay, so I was working on the new CT run today when a horrible realization occurred. The the run on the SDA cannot be beaten with the (U) rom. It is 100% impossible. I am roughly 10 minutes into the the new run, right after the part when lucca tells crono about marle being the descendant of queen leene. Up to that point I was making great efforts to optimize the run as much as possible. When I make the comparison to inichi's run, I was over 20 seconds BEHIND!!! 20 seconds behind an un-assisted run, within just 10 minutes of gameplay.
So I decided to make a little test to see just how much the text speed really affects the length of the run. I made a short demo on both the (J) and (U) version of the run to get a more accurate comparison between the two. What took 52 seconds to play on the (J) version took 56 seconds to play on the (U) version, mind you not all of this time was even dialog. roughly 4-5 seconds are lost within every minute of gameplay using the (U) version. According to my calculations, my currently published run would have been 20 minutes faster using the (J) version. Possibly more when you consider how much dialog occurred in the end portion of the game. This is staggering.
Should I even finish the run now? It will still be significantly slower than inichi's run. Ignorant people will always compare my run to his despite the version difference.
