Secret of Evermore (SNES)

2014-02-04 11:43:03 am

I love YaBB 1G - SP1!

MetaSigma: good call. i flaked out, and was thinking too narrowly when ruling out Speed abuse.

i've had the patch for the stat boost + dying bug largely coded since July, and still haven't released it. delaying it so much has apparently rotted my mind. much of the wait is due to the Readme (i.e. i don't feel like writing it), and now i'll be sure to describe this aspect of the bug somewhere in it.

i want to say you'll cast Pixie Dust first, since you want its non-boost in the 1st status slot, as that's what'll be subtracted from all 4 status slots' boosts. but i could be wrong, as this last brainfart shows.

TheAngryPanda

2014-02-04 12:05:57 pm

Speed > Safety

Okay, so the question remains, is this actually worth the time/money spent setting it up? Is there a solid location to perform this in during Act 2 (Act 1 isn't going to work, there's no way monetarily to afford Pixie Dust at that point)? Are there any conflicting effects when used before the atlas glitch is performed?

It would absolutely have to be done before invulnerability, which implies that we'd have to do it before Atlas glitch. This also means that perhaps we wouldn't want Invulnerability and would instead opt to have high hit/evade alongside an underflowed attack/defense stat. Arguably this is the same speed, but obviously has it's major drawbacks since we HAVE to pick up defend in 100% and lose the 20s buffer it provides. In fact, we would have to figure out what amount of speed is considered "safe" for the Invulnerability glitch in order to even consider this. Then there's the detail of the Pixie Dust purchase, which can be bypassed via purchasing all the wings later on during the market run. Perhaps all this plus the optimizations required to achieve this will make the idea too slow or complex for actual use at the optimal time? Another issue with this would be that in order to make use of Speed at the optimal time, we'd be delaying the market timer by a full 8x cast more than likely, more if it needs to be leveled. This alone makes me see the trick as not being worth it in any sense.

TBH, I shouldn't of even mentioned it because I didn't account for market timer being delayed, silly mistake. It's not worth it at all if it can't be done in Act 1 rapidly.

Okay, so the question remains, is this actually worth the time/money spent setting it up? Is there a solid location to perform this in during Act 2 (Act 1 isn't going to work, there's no way monetarily to afford Pixie Dust at that point)? Are there any conflicting effects when used before the atlas glitch is performed?

It would absolutely have to be done before invulnerability, which implies that we'd have to do it before Atlas glitch. This also means that perhaps we wouldn't want Invulnerability and would instead opt to have high hit/evade alongside an underflowed attack/defense stat. Arguably this is the same speed, but obviously has it's major drawbacks since we HAVE to pick up defend in 100% and lose the 20s buffer it provides. In fact, we would have to figure out what amount of speed is considered "safe" for the Invulnerability glitch in order to even consider this. Then there's the detail of the Pixie Dust purchase, which can be bypassed via purchasing all the wings later on during the market run. Perhaps all this plus the optimizations required to achieve this will make the idea too slow or complex for actual use at the optimal time? Another issue with this would be that in order to make use of Speed at the optimal time, we'd be delaying the market timer by a full 8x cast more than likely, more if it needs to be leveled. This alone makes me see the trick as not being worth it in any sense.

TBH, I shouldn't of even mentioned it because I didn't account for market timer being delayed, silly mistake. It's not worth it at all if it can't be done in Act 1 rapidly.

TheAngryPanda

Edit history:

TheAngryPanda: 2014-02-17 09:26:59 pm

TheAngryPanda: 2014-02-17 11:12:07 am

2014-02-16 06:34:11 pm

Speed > Safety

So while working through the ruins on the Any% TAS, I came across a new-ish spawn manipulation. Unfortunately, I'm unsure if it's consistent or not. That being said, here you go:

Enter ruins main area and immediately run Up+Left, continue along this wall all the way until the dog has to be sniffed at the next entrance. Once you get to the button across the invisible bridge, the enemy shouldn't be there if done properly.

Sounds pretty vague doesn't it? Well that's because I have no real way of testing most results outside of the TAS, where this is consistently not spawning the enemy. Perhaps it's timing? Maybe the rats have something to do with it? What if it's positioning of the dog? There's lots of potential reasons at the moment and I'm more interested in actually finishing the TAS at the moment. Someone that does attempts of the game might want to sit down and look into this a bit more for the time being, otherwise it will have to wait until I'm done with the TAS to get my full attention.

Something else to consider is perhaps a camera manipulation to despawn the enemy, I didn't try this.

edit:

More stuff in Act 2:

- The despawn in ruins is likely tied to the rats "following" you up to the door from the entrance while the rats to the right of the upper left pillar move towards you from the right as you progress to the invisible bridge. This would give us 5 sprites visible, 2 sprites non-visible (rats following the dog) on screen, which is enough to despawn the dancing fool that's by the upper left button to unlock the first door. Now... the BAD part of this is that there's currently no literal way to handle this situation. I'm only able to test (properly) things that can occur in my game's specific set, it's very possible that this might end up being a positional despawn by taking a specific route as those do exist in certain things.

- Early Mad Monk isn't going to make it into the TAS because the time it takes me to acquire all the needed materials is around 15 frames slower than the time it takes me to arrive and kill the mad monk under ideal conditions anyways. I THINK this is attributed mostly to the fact that I don't need to level alchemy anymore due to me getting 90 CB's in pyramids. If it was NOT like this, Early Mad Monk would save a fairly large amount of time. Another thought is that maybe we could pyramids first, however, that means we don't have the spear for pyramids, so for every perfect sprint, we lose 1 frame of distance. This obviously adds up rapidly and ends up being around 44 frames longer than the alternative, which does ruins without early mad monk first. I do want to point out that Early Mad Monk should and probably WILL be a thing in a 100% TAS because you'd acquire spear earlier. We also don't particularly HAVE to do ruins first either, which means we could stock the Beads up instead in pyramids since the spear isn't a factor due to having the one from Act 1. This actually makes me very interested to see what other places can be altered because of previously non-accessible things. Should be fun.

- So far not having invulnerability has cost me exactly 5 frames. How do I know this? I have another TAS copy of my current version that is running invulnerability so I could compare them properly. It's quite a distance behind the other one at the moment, but that's to be expected until late-Act 3 when they should line up after the non-invulnerability version kills some things on Gomi's tower that are otherwise bypassed.

More to come I'm sure... stay tuned.

edit2:

So the path to the Mad Monk is weird, I save time in movement, but lose time in movement due to AI. Seems this map is just doomed to be essentially the same aside from other optimizations found within it. Unfortunately, this leaves me knocking the mad monk around like a beach ball for a while as the dog, which I suppose has it's own rewards... lol. The amount of frames between this map now and it's previous version 1.6 completion is within 120 frames, this is pretty amazing granted how much was discovered since then. I'm still upset we can't use Early Mad Monk, it would have made a massive difference. I'm actually considering using it anyways just because why not. Then I remind myself we have a 100% TAS coming at some point and it will have a definitive impact on that route.

So while working through the ruins on the Any% TAS, I came across a new-ish spawn manipulation. Unfortunately, I'm unsure if it's consistent or not. That being said, here you go:

Enter ruins main area and immediately run Up+Left, continue along this wall all the way until the dog has to be sniffed at the next entrance. Once you get to the button across the invisible bridge, the enemy shouldn't be there if done properly.

Sounds pretty vague doesn't it? Well that's because I have no real way of testing most results outside of the TAS, where this is consistently not spawning the enemy. Perhaps it's timing? Maybe the rats have something to do with it? What if it's positioning of the dog? There's lots of potential reasons at the moment and I'm more interested in actually finishing the TAS at the moment. Someone that does attempts of the game might want to sit down and look into this a bit more for the time being, otherwise it will have to wait until I'm done with the TAS to get my full attention.

Something else to consider is perhaps a camera manipulation to despawn the enemy, I didn't try this.

edit:

More stuff in Act 2:

- The despawn in ruins is likely tied to the rats "following" you up to the door from the entrance while the rats to the right of the upper left pillar move towards you from the right as you progress to the invisible bridge. This would give us 5 sprites visible, 2 sprites non-visible (rats following the dog) on screen, which is enough to despawn the dancing fool that's by the upper left button to unlock the first door. Now... the BAD part of this is that there's currently no literal way to handle this situation. I'm only able to test (properly) things that can occur in my game's specific set, it's very possible that this might end up being a positional despawn by taking a specific route as those do exist in certain things.

- Early Mad Monk isn't going to make it into the TAS because the time it takes me to acquire all the needed materials is around 15 frames slower than the time it takes me to arrive and kill the mad monk under ideal conditions anyways. I THINK this is attributed mostly to the fact that I don't need to level alchemy anymore due to me getting 90 CB's in pyramids. If it was NOT like this, Early Mad Monk would save a fairly large amount of time. Another thought is that maybe we could pyramids first, however, that means we don't have the spear for pyramids, so for every perfect sprint, we lose 1 frame of distance. This obviously adds up rapidly and ends up being around 44 frames longer than the alternative, which does ruins without early mad monk first. I do want to point out that Early Mad Monk should and probably WILL be a thing in a 100% TAS because you'd acquire spear earlier. We also don't particularly HAVE to do ruins first either, which means we could stock the Beads up instead in pyramids since the spear isn't a factor due to having the one from Act 1. This actually makes me very interested to see what other places can be altered because of previously non-accessible things. Should be fun.

- So far not having invulnerability has cost me exactly 5 frames. How do I know this? I have another TAS copy of my current version that is running invulnerability so I could compare them properly. It's quite a distance behind the other one at the moment, but that's to be expected until late-Act 3 when they should line up after the non-invulnerability version kills some things on Gomi's tower that are otherwise bypassed.

More to come I'm sure... stay tuned.

edit2:

So the path to the Mad Monk is weird, I save time in movement, but lose time in movement due to AI. Seems this map is just doomed to be essentially the same aside from other optimizations found within it. Unfortunately, this leaves me knocking the mad monk around like a beach ball for a while as the dog, which I suppose has it's own rewards... lol. The amount of frames between this map now and it's previous version 1.6 completion is within 120 frames, this is pretty amazing granted how much was discovered since then. I'm still upset we can't use Early Mad Monk, it would have made a massive difference. I'm actually considering using it anyways just because why not. Then I remind myself we have a 100% TAS coming at some point and it will have a definitive impact on that route.

MetaSigma

2014-02-17 12:58:34 pm

Evermore Extrodinare ᕦ(° Д°)ᕤ

1:18:58...I finally broke the sub 1:19 and recorded it too Smiley

GreenAmbler

2014-02-19 03:22:03 pm

What's the obstacle for early Mad Monk kill? I realize you have no alchemy to use but can't you just attack through the wall like I did in my highlight? IIRC the room takes about a minute to do the normal RTA way, whereas I've finished the room in about 20 seconds once. Unfortunately it will not really be beneficial for the 100% TAS because you need to get Fireball still.

TheAngryPanda

Edit history:

TheAngryPanda: 2014-02-19 04:55:40 pm

TheAngryPanda: 2014-02-19 04:09:00 pm

2014-02-19 03:55:13 pm

Speed > Safety

The only place to acquire the amount of beads required for that trick to become viable in the Ruins (which absolutely have to be done before pyramids, otherwise it's slower due to no spear) is located on the lower level of pyramids. Since I can't do those first, in any% at least, I can't really utilize the Early Mad Monk thing under this route's conditions.

That being said, I have a secondary version of the TAS where it level Sting up once and has BARELY enough to kill the Monk, I'm testing if this is going to end up being faster, but have some serious doubts due to the amount of menu delay introduced, and at that point, I'd be leveling Sting up instead of just 90 CB'ing it all the way.

Another little issue that's arisen is that on Aqua, the optimal pattern for CB casts using the route I think is the fastest uses 8x Flare + 3x Storm, hopefully twice, but we'll see, I imagine a 3rd menu operation sequence will be required because numbers aren't adding up in theory/practice, but that's okay. so two of those and probably 1-2 Storms immediately after should do it, which is annoying. However, I did figure out that by using storm, it seems to "cancel" Aqua's storm/corrosion/whatever casts as well, which is weird and useless.

As for the comment on Fireball, I actually think it might end up still being worth it for spear's extra speed per sequence of frames, it's just a matter of calculating once I'm at that point. I did a lot of testing with the bridge skip to back over, I can basically do it completely blind now based off coordinates via ram watching. 100% is likely to be an infinitely more interesting TAS it seems, there's just so much more to it, really wish I would have done that first instead of Any%.

As for the timing, that's kind of thrown off in a TAS due to other factors. It would absolutely be faster, but leveling Sting wouldn't be. The trick itself is amazingly quick, even with poor RNG, but other factors absolutely crush it in the TAS with it's current routing. I'm looking back through everything though to see if I can't find some way to squeeze a bit more out of Act 1 or Market Timer in order to acquire enough things to pull it off quicker than simply going around, nothing's really stood out though. A thought I did have was maybe manipulating some drops post-bugmuck for 10 extra crystal and simply buying some more clay so I could save CB's from Magmar, but I'd need at the very least 6 and only have access to maybe 2 at that point without detouring immensely. There is something to be said about having spare HB's to minimize the amount of CB's needed, but I'm not sure what the perfect balance is. It's something I'm definitely working on as I think this is the only possible way to include the trick into the current route without completely losing all the time. As a somewhat side note to this, I honestly don't think I could do the 14 minute timer again, it's literally the most demotivating thing I've ever experienced in a game. If someone would like to have the .bkm file and work on this improvement themselves, I'm fine with that, but I personally wouldn't be doing it again considering I've done it around 4-5 times now.

edit:

Some clarification on my thoughts about the Act 1 more HB thing:

-You need 6 CB's minimum, or 1008 damage (which is 6 CB storm max damage) combined from all sources
-If you use nothing on Magmar, you lose at minimum 200 damage and at maximum 405 (average 302)
-To make up this damage loss, it's automatically an entire extra 8x HB cast + #CB's required to obtain the same damage.
-You only have 1 extra 8x cast post-original sequence (minus CB menu operations) before he jumps back in under the TAS route, and it's a VERY tight timing (like 12 frames? probably less after manipulations)
-You'd have to manipulate a double maggot drop for sure to obtain the right amount of crystals, and possibly even include to IFG on the inlet near BBM entrance for another 3.
-On top of this, you'll absolutely have to pick up the CB in volcano entrance.

There's more to it obviously, but that alone should give anyone interested an idea of just how much adjustment will be needed, it's honestly hard to say if it's less or more than the amount of frames saved from Early Mad Monk, which is why I don't have numbers to throw out for it. I am slowly working on it... very, very, slowly.

edit: See highlight below, should be fun to play around with! Cheesy

[s]The only place to acquire the amount of beads required for that trick to become viable in the Ruins (which absolutely have to be done before pyramids, otherwise it's slower due to no spear) is located on the lower level of pyramids. Since I can't do those first, in any% at least, I can't really utilize the Early Mad Monk thing under this route's conditions.

That being said, I have a secondary version of the TAS where it level Sting up once and has BARELY enough to kill the Monk, I'm testing if this is going to end up being faster, but have some serious doubts due to the amount of menu delay introduced, and at that point, I'd be leveling Sting up instead of just 90 CB'ing it all the way. [/s]

Another little issue that's arisen is that on Aqua, the optimal pattern for CB casts using the route I think is the fastest uses 8x Flare + 3x Storm, hopefully twice, but we'll see, I imagine a 3rd menu operation sequence will be required because numbers aren't adding up in theory/practice, but that's okay. so two of those and probably 1-2 Storms immediately after should do it, which is annoying. However, I did figure out that by using storm, it seems to "cancel" Aqua's storm/corrosion/whatever casts as well, which is weird and useless.

As for the comment on Fireball, I actually think it might end up still being worth it for spear's extra speed per sequence of frames, it's just a matter of calculating once I'm at that point. I did a lot of testing with the bridge skip to back over, I can basically do it completely blind now based off coordinates via ram watching. 100% is likely to be an infinitely more interesting TAS it seems, there's just so much more to it, really wish I would have done that first instead of Any%.

[s]As for the timing, that's kind of thrown off in a TAS due to other factors. It would absolutely be faster, but leveling Sting wouldn't be. The trick itself is amazingly quick, even with poor RNG, but other factors absolutely crush it in the TAS with it's current routing. I'm looking back through everything though to see if I can't find some way to squeeze a bit more out of Act 1 or Market Timer in order to acquire enough things to pull it off quicker than simply going around, nothing's really stood out though. A thought I did have was maybe manipulating some drops post-bugmuck for 10 extra crystal and simply buying some more clay so I could save CB's from Magmar, but I'd need at the very least 6 and only have access to maybe 2 at that point without detouring immensely. There is something to be said about having spare HB's to minimize the amount of CB's needed, but I'm not sure what the perfect balance is. It's something I'm definitely working on as I think this is the only possible way to include the trick into the current route without completely losing all the time. As a somewhat side note to this, I honestly don't think I could do the 14 minute timer again, it's literally the most demotivating thing I've ever experienced in a game. If someone would like to have the .bkm file and work on this improvement themselves, I'm fine with that, but I personally wouldn't be doing it again considering I've done it around 4-5 times now.

edit:

Some clarification on my thoughts about the Act 1 more HB thing:

-You need 6 CB's minimum, or 1008 damage (which is 6 CB storm max damage) combined from all sources
-If you use nothing on Magmar, you lose at minimum 200 damage and at maximum 405 (average 302)
-To make up this damage loss, it's automatically an entire extra 8x HB cast + #CB's required to obtain the same damage.
-You only have 1 extra 8x cast post-original sequence (minus CB menu operations) before he jumps back in under the TAS route, and it's a VERY tight timing (like 12 frames? probably less after manipulations)
-You'd have to manipulate a double maggot drop for sure to obtain the right amount of crystals, and possibly even include to IFG on the inlet near BBM entrance for another 3.
-On top of this, you'll absolutely have to pick up the CB in volcano entrance.

There's more to it obviously, but that alone should give anyone interested an idea of just how much adjustment will be needed, it's honestly hard to say if it's less or more than the amount of frames saved from Early Mad Monk, which is why I don't have numbers to throw out for it. I am slowly working on it... very, very, slowly.[/s]

edit: See highlight below, should be fun to play around with! :D

GreenAmbler

2014-02-19 04:37:31 pm

No sorry, I meant you don't need alchemy or call beads. You can take advantage of the game's lousy shallow wall attacking mechanic, since the hidden path is right next to the wall your attack is "stunted" but still has a hitbox, which extends to the area the Mad Monk can walk to. I assumed you saw this already, but here's the highlight I made:

http://www.twitch.tv/greenambler/c/3343257
Could be a lot faster, I just got a 16 second room. You will need to manipulate the spawn so it's close otherwise he doesn't walk towards you (I think, I plan on using the walk thru walls cheat to observe his movements more)

TheAngryPanda

2014-02-19 04:54:57 pm

Speed > Safety

No, I haven't seen that highlight before, that's very interesting. Going to go play around with this a bit. Do you know when his spawn is set? Entering area, rough radius, etc...?

This will definitely save time, awesome. Back to Ruins for me. Cheesy

GreenAmbler

2014-02-19 05:00:52 pm

I think when entering the room it is set, and once you get close enough to him he becomes active and starts walking around in a normal monk pattern (awkwardly aggressive). So if it spawns to the right or above the center of his room, he won't ever walk towards you, which is what kills this for RTA. Roughly a 50% chance of ruining a run since you pretty much have to commit, but the other 50% you will save around 0-45 seconds over current route depending on very much randomness of course.

TheAngryPanda

2014-02-19 05:13:48 pm

Speed > Safety

Yeah, I actually have already shoved it into the Any% route after seeing that highlight for when we get someone willing to actually push the game as low as possible. So much RNG though, doubt anyone's gonna be up for that in a looooong time. lol.

MetaSigma

Edit history:

MetaSigma: 2014-02-19 05:59:22 pm

2014-02-19 05:54:08 pm

Evermore Extrodinare ᕦ(° Д°)ᕤ

Yeah I remember seeing that highlight a while ago and thinking damn if that was possible to pull off consistently and get a great spawn pattern omg ruins would just get crushed. Yeah I'd like to know if there is a way to manip his spawn position and if so I will def do it in a run. So it's 0 to 45s saved? Or is it more potentially?

You know while I myself don't want to do this it might be something to play a slightly safe Act 1, just to run some RTAs up to Act 2 ruins and try this out. If you save an average of 30s and get a semi-consistent run out of it. Times would drop drastically. Then just plug that into the hardcore route and suddenly you could potentially be in the 1:15:xx range maybe? I mean if 1:17:40 (I think) is what is possible with no BBM and Defend that means that sub 1:17 (What like...1:16:45?) is possible on the hardcore route. Throw that in and you could get what...a 1:15:5x?

TheAngryPanda

2014-02-19 06:17:26 pm

Speed > Safety

you save an average of 30s, 50% of the time. The other half that is complete time loss because the AI isn't being triggered, which you won't know until X amount of time has passed anyways, which is also variable. Arguably this could be seen as a run killer. There's also no way of defining the ideal time to "walk away" from the trick because you have no way of knowing where the monk is at without the camera being scrolled. So in order to effectively use this, we'd probably need to find a way to force the camera to move around.

As for the time you could potentially get, it's hard to call really. I wouldn't put any guesses down without doing significant testing within real runs as opposed to individual testing. Personally I don't feel as if we'll see a 1:15 without the cave warps ever. If I had to guess, I'd say the game would likely max out with current knowledge around 1:16:20-30.

The real thing people interested in this should be looking for is a setup that generates the RNG for the monk's spawn to be near the edge. If it's determined upon entering the room, this isn't possible. But there are VERY few enemies that are generated upon entering a room. If this isn't the case with the monk, we can figure out his radius via some nonsense on emulators. Once this is accomplished, you can further refine this by determining what positions are set for spawning the monk in at which points on the radius, the same way we manipulate certain spawns in bugmuck to run past them with the dog at the entrance. IF THIS is possible, THEN we can generate a setup that will work 100% of the time at least in terms of SPAWNING the monk in the right place. The AI movement itself is very, very, random in real-time. Manipulating movement post-spawn of enemies is basically frame-by-frame operations.

As for the TAS, I've already confirmed it saves time, i'll keep how much a secret for now, but I'm quite happy with it. Cheesy

you save an average of 30s, 50% of the time. The other half that is complete time loss because the AI isn't being triggered, which you won't know until X amount of time has passed anyways, which is also variable. Arguably this could be seen as a run killer. There's also no way of defining the ideal time to "walk away" from the trick because you have no way of knowing where the monk is at without the camera being scrolled. So in order to effectively use this, we'd probably need to find a way to force the camera to move around.

As for the time you could potentially get, it's hard to call really.  I wouldn't put any guesses down without doing significant testing within real runs as opposed to individual testing. Personally I don't feel as if we'll see a 1:15 without the cave warps ever. If I had to guess, I'd say the game would likely max out with current knowledge around 1:16:20-30.

The real thing people interested in this should be looking for is a setup that generates the RNG for the monk's spawn to be near the edge. If it's determined upon entering the room, this isn't possible. But there are VERY few enemies that are generated upon entering a room. If this isn't the case with the monk, we can figure out his radius via some nonsense on emulators. Once this is accomplished, you can further refine this by determining what positions are set for spawning the monk in at which points on the radius, the same way we manipulate certain spawns in bugmuck to run past them with the dog at the entrance. IF THIS is possible, THEN we can generate a setup that will work 100% of the time at least in terms of SPAWNING the monk in the right place. The AI movement itself is very, very, random in real-time. Manipulating movement post-spawn of enemies is basically frame-by-frame operations.

As for the TAS, I've already confirmed it saves time, i'll keep how much a secret for now, but I'm quite happy with it. :D

MetaSigma

2014-02-19 06:31:34 pm

Evermore Extrodinare ᕦ(° Д°)ᕤ

Hearing you say you're quite happy with it means that it's really good. =P

TheAngryPanda

Edit history:

TheAngryPanda: 2014-02-19 09:20:39 pm

TheAngryPanda: 2014-02-19 09:19:58 pm

TheAngryPanda: 2014-02-19 08:32:11 pm

2014-02-19 07:06:24 pm

Speed > Safety

Anyone feel like digging through the RAM to figure out how we can tell exactly where a specific enemy is at? (I'm pretty sure it's the only mad monk on this particular map ID, #45) It's also the one with 1000 hp as opposed to the normal ones. I have the stats ROM address from assassin17's FAQ, but I'm not really sure where to go from here or if that's even remotely close to where I should start, lol.

~~It appears that it's not generated upon entering the map, at least not from what I've noticed, but that could just be a false positive from early testing on my end.~~
It's definitely spawned upon entering.

Gonna try and list some spawn locations from testing in excel, will post details if anything interesting pops up...

-How I'm Testing-

-Savestate outside entrance
-Adjust entering area by one frame each time
-Initial test frame will be zero, going to go up to like... 10-20, not sure
-Hopefully we can find a pattern somehow that points to an indication of things working properly
-I'm poking the addresses to teleport to where the monk should be located near and guesstimating the ACTUAL spawn location on the map. Keyword here is guess, I don't know the actual RAM address for enemy position, trying to figure that out now via RAM Search.
-Once this is done, I'm going to use TASstudio to store an input sequence and copy it into each test segment to see what frame the monk arrives on in each setting, if the monk arrives at all near the wall.

If anyone wants to help test this, please do, even RTA runners can just enter, see if they get a spawn, leave, try again. Write the data down if you do please, finding a solid %chance this is possible would be fantastic to know. I'm getting anywhere between 16 and 22s times for kills on the monk in the TAS, but I think you can get quicker ones, possibly in the 8-10s range if really lucky, I have no way of confirming though until I obtain the positional addresses. At least I have something to do tonight.

Anyone feel like digging through the RAM to figure out how we can tell exactly where a specific enemy is at? (I'm pretty sure it's the only mad monk on this particular map ID, #45) It's also the one with 1000 hp as opposed to the normal ones. I have the stats ROM address from assassin17's FAQ, but I'm not really sure where to go from here or if that's even remotely close to where I should start, lol.

[s]It appears that it's not generated upon entering the map, at least not from what I've noticed, but that could just be a false positive from early testing on my end. [/s]
It's definitely spawned upon entering.

Gonna try and list some spawn locations from testing in excel, will post details if anything interesting pops up...

-How I'm Testing-

-Savestate outside entrance
-Adjust entering area by one frame each time
-Initial test frame will be zero, going to go up to like... 10-20, not sure
-Hopefully we can find a pattern somehow that points to an indication of things working properly
-I'm poking the addresses to teleport to where the monk should be located near and guesstimating the ACTUAL spawn location on the map. Keyword here is guess, I don't know the actual RAM address for enemy position, trying to figure that out now via RAM Search.
-Once this is done, I'm going to use TASstudio to store an input sequence and copy it into each test segment to see what frame the monk arrives on in each setting, if the monk arrives at all near the wall.

If anyone wants to help test this, please do, even RTA runners can just enter, see if they get a spawn, leave, try again. Write the data down if you do please, finding a solid %chance this is possible would be fantastic to know. I'm getting anywhere between 16 and 22s times for kills on the monk in the TAS, but I think you can get quicker ones, possibly in the 8-10s range if really lucky, I have no way of confirming though until I obtain the positional addresses. At least I have something to do tonight.

AsFarAsIGet

Edit history:

AsFarAsIGet: 2014-02-19 09:38:35 pm

2014-02-19 09:37:35 pm

www.youtube.com/user/HumanMYTH

Hmm ok here are the adresses for the mad monk X/Y:

7E3E8D for the X-Value (want him to be on 340)
7E3E8F for the Y-Value (want him to be on 1163)

When those values are met, the mad monk is in the secret passage and u are able to hit him through the wall. Don't forget to watch the adresses as 2byte and remember, the adresses only contain the values when you are at the mad monk area.

Hf

TheAngryPanda

Edit history:

TheAngryPanda: 2014-02-19 09:58:57 pm

2014-02-19 09:40:16 pm

Speed > Safety

If you don't mind AsFar, could you PM me how you found those? Preferably as specifically as possible. I'd really like to learn how to do that so quickly. Sad

edit: The AI is pretty damn stupid, I'll say that much already. It REALLY likes hugging pillars and getting stuck on them.

AsFarAsIGet

2014-02-19 10:00:18 pm

www.youtube.com/user/HumanMYTH

Haha yeah, I tried to manipulate him somehow to get closer but he seems to have it's own will. Cheesy

TheAngryPanda

Edit history:

TheAngryPanda: 2014-02-21 08:21:55 am

TheAngryPanda: 2014-02-20 11:37:28 am

TheAngryPanda: 2014-02-20 06:42:52 am

TheAngryPanda: 2014-02-20 06:34:39 am

TheAngryPanda: 2014-02-19 11:19:49 pm

TheAngryPanda: 2014-02-19 11:05:41 pm

TheAngryPanda: 2014-02-19 10:27:42 pm

TheAngryPanda: 2014-02-19 10:19:51 pm

2014-02-19 10:01:33 pm

Speed > Safety

He seems to only move when you're literally "in" the wall, or on the pixel next to it. I guess RTA runners will have to sit there holding right and down in the little corner, or along the side like GA does in his video. Not doing this, he seems to stand still after moving a little bit.
*Proven wrong during a few test attempts.

Resulting locations after frames of adjustment after 600 (10s), 1200 (20s), 1800 (30s), 2400 (40s) frames pass:

-Note: These are static points of data, I do not move at all during the frame advancing. Think of it as a control test. Movement complicates things and can be factored in later. If you're trying this in real-time, run along the wall between the pillar and the bottom-right corner. This will spawn an enemy below you, so be aware of that.

Frames Altered from start point | [V.Pos, H.Pos]

00 | [932 , 480] - [968 , 456] - [1010,394] - [1035,402]
01 | [1136,391] - [1155,327] - [1147,359] - [1130.455]

These are more or less just to see what happens when movement is NOT being considered into the equation. Data will get filled in as I progress...

edit-1: Using a better starting point. Specifically, the starting point is when you gain control of the character to move throughout the area after entering. V.Pos = 1272, H.Pos = 152 for the TAS. This has the monk spawning as close to the wall as possible, but he's an adventurous little jerk.

edit-2: Any input of R will affect the RNG of the AI movement. Note: Not doing anything IS an AI option. An indicator of how radically this can change things is the 00 and 01 comparison, that is a one-frame release of the R button. 00 is holding R down indefinitely. The boy and dog both stay in the exact same positions as well.

edit-3: I'm actually going to stop tracking the numbers and just leave the above as an example. If this is actually how the game decides enemy AI movement, there's nothing you can do about it in real-time other than fool around a bit and hope. THAT being said, IF this RNG isn't generated until you load the map, you COULD hold R before entering the room and then hold a menu button immediately as the area load (and i mean immediately...). This will let you open the menu on the first frame possible, release R, press Y and HOLD X IMMEDIATELY AFTER. This will advance the game by exactly one frame. Once the menu is back up, continue holding R until you reach the wall. You should have the mad monk appear around the 16-22s mark depending on stuff you have no control over.

This is under the assumption the map's RNG for movement isn't pre-determined in any form by anything before entering the area. There's also no sure-fire way to say it will always work, but it's the best we got right now aside from just getting lucky.

As for the TAS, this is going to be an absolute nightmare.

Basically I need to manipulate every frame, then do a redundancy check on the frame before and after it, for every frame, every single time I advance the manipulation. It's basically flowers all over again. Sad

edit-4:

Pressing R, or releasing it, for X frames definitely affects the RNG of the enemy. This doesn't help RTA currently, but it's something keep in mind for creating setups if this ever progresses to that level.

edit-5:

The best I've gotten is what GA's best is, which is around 16s in the area. Unfortunately the Call Bead quick kill isn't so quick... In fact it's quite slow because of the menu cancel operations. So double melee would be the next best option, except then you can't leave quickly. I'm getting savings of around 35.4s with a 16s visit, this is likely attributed to the old route actually being optimized very well. I plan to work on it some more throughout the next few days before moving on, but so far I haven't found any sort of consistency regarding movement inputs that could lend themselves to a setup for RTA.

edit-6:

Alright, so I've spent the last 2 days, or rather, since I learned of this trying to get a perfect manipulation in the TAS. I've gotten a 16s (960 frames) manipulation numerous times now, but not a single one has been lower than that. I did notice a very specific pattern though during the initial frame by frame manipulations which shows that when the Mad Monk travels left from the upper-left/center spawns, he tends to immediately walk down+left towards the player right after. This is the ONLY pattern I've noticed, and it's consistent throughout every single 16s manipulation I've gotten. It's funnily enough, why they aren't lower. Perhaps this is something like the boy's AI when we try to force him over the pit in the beginning of Act 2? Some "X" needs to happen for "Y" to occur, else "Y" cannot occur. The data from my testing is showing this statement to be true, but I have a very difficult time believing it. Another theory is that it's simply a "bug" with the AI from that specific point which doesn't allow the Mad Monk to continue down a different path without going straight left first. This could also be simply because he's then pushed towards the player's aggro circle and thus chases us down, slowly... The issue with suggesting this, however, is that in testing I've noticed him getting to the wall and then just simply walking away quite often, as if the player were not a factor at all.

All this being said, I'm happy with my current results and am going to progress in the TAS. If someone determines a quicker method of execution/manipulation, I will absolutely start over from this point again.

[s]He seems to only move when you're literally "in" the wall, or on the pixel next to it. I guess RTA runners will have to sit there holding right and down in the little corner, or along the side like GA does in his video. Not doing this, he seems to stand still after moving a little bit. [/s]
*Proven wrong during a few test attempts.

Resulting locations after frames of adjustment after 600 (10s), 1200 (20s), 1800 (30s), 2400 (40s) frames pass:

-Note: These are static points of data, I do not move at all during the frame advancing. Think of it as a control test. Movement complicates things and can be factored in later. If you're trying this in real-time, run along the wall between the pillar and the bottom-right corner. This will spawn an enemy below you, so be aware of that.

Frames Altered from start point  |  [V.Pos, H.Pos]

00  |    [932 , 480]  -  [968 , 456]  -  [1010,394]  -  [1035,402]
01  |    [1136,391]  -  [1155,327]  -  [1147,359]  -  [1130.455]

These are more or less just to see what happens when movement is NOT being considered into the equation. Data will get filled in as I progress...

edit-1: Using a better starting point. Specifically, the starting point is when you gain control of the character to move throughout the area after entering. V.Pos = 1272, H.Pos = 152 for the TAS. This has the monk spawning as close to the wall as possible, but he's an adventurous little jerk.

edit-2: Any input of R will affect the RNG of the AI movement. Note: Not doing anything IS an AI option. An indicator of how radically this can change things is the 00 and 01 comparison, that is a one-frame release of the R button. 00 is holding R down indefinitely. The boy and dog both stay in the exact same positions as well.

edit-3: I'm actually going to stop tracking the numbers and just leave the above as an example. If this is actually how the game decides enemy AI movement, there's nothing you can do about it in real-time other than fool around a bit and hope. THAT being said, IF this RNG isn't generated until you load the map, you COULD hold R before entering the room and then hold a menu button immediately as the area load (and i mean immediately...). This will let you open the menu on the first frame possible, release R, press Y and HOLD X IMMEDIATELY AFTER. This will advance the game by exactly one frame. Once the menu is back up, continue holding R until you reach the wall. You should have the mad monk appear around the 16-22s mark depending on stuff you have no control over.

This is under the assumption the map's RNG for movement isn't pre-determined in any form by anything before entering the area. There's also no sure-fire way to say it will always work, but it's the best we got right now aside from just getting lucky.

As for the TAS, this is going to be an absolute nightmare.

Basically I need to manipulate every frame, then do a redundancy check on the frame before and after it, for every frame, every single time I advance the manipulation. It's basically flowers all over again. :(

edit-4:

Pressing R, or releasing it, for X frames definitely affects the RNG of the enemy. This doesn't help RTA currently, but it's something keep in mind for creating setups if this ever progresses to that level.

edit-5:

The best I've gotten is what GA's best is, which is around 16s in the area. Unfortunately the Call Bead quick kill isn't so quick... In fact it's quite slow because of the menu cancel operations. So double melee would be the next best option, except then you can't leave quickly. I'm getting savings of around 35.4s with a 16s visit, this is likely attributed to the old route actually being optimized very well. I plan to work on it some more throughout the next few days before moving on, but so far I haven't found any sort of consistency regarding movement inputs that could lend themselves to a setup for RTA.

edit-6:

Alright, so I've spent the last 2 days, or rather, since I learned of this trying to get a perfect manipulation in the TAS. I've gotten a 16s (960 frames) manipulation numerous times now, but not a single one has been lower than that. I did notice a very specific pattern though during the initial frame by frame manipulations which shows that when the Mad Monk travels left from the upper-left/center spawns, he tends to immediately walk down+left towards the player right after. This is the ONLY pattern I've noticed, and it's consistent throughout every single 16s manipulation I've gotten. It's funnily enough, why they aren't lower. Perhaps this is something like the boy's AI when we try to force him over the pit in the beginning of Act 2? Some "X" needs to happen for "Y" to occur, else "Y" cannot occur. The data from my testing is showing this statement to be true, but I have a very difficult time believing it. Another theory is that it's simply a "bug" with the AI from that specific point which doesn't allow the Mad Monk to continue down a different path without going straight left first. This could also be simply because he's then pushed towards the player's aggro circle and thus chases us down, slowly... The issue with suggesting this, however, is that in testing I've noticed him getting to the wall and then just simply walking away quite often, as if the player were not a factor at all.

All this being said, I'm happy with my current results and am going to progress in the TAS. If someone determines a quicker method of execution/manipulation, I will absolutely start over from this point again.

TheAngryPanda

Edit history:

TheAngryPanda: 2014-02-21 11:12:05 am

TheAngryPanda: 2014-02-21 11:03:07 am

TheAngryPanda: 2014-02-21 10:52:45 am

TheAngryPanda: 2014-02-21 10:47:13 am

TheAngryPanda: 2014-02-21 10:38:23 am

TheAngryPanda: 2014-02-21 10:37:39 am

TheAngryPanda: 2014-02-21 10:03:15 am

2014-02-21 09:57:29 am

Speed > Safety

So I just ran into a pretty massive problem...

After using this, the map for Megataur refuses to load, it hardlocks like when doing the verminator skip on console. Has anyone tested this PAST just that room? Please tell me this is just something unrelated to the trick... I'm gonna be really sad if another amazing trick has to be cut to some bullshit.

edit:

Specifically:

-Run from Mad Monk area door
-Trigger event to throw spear
-Run across door (it gets some lag here every 3-4 frames)
-Enter Megataur area
-Game proceeds to run, but never actually exits transition, mapID IS loaded as well as positional data, but all control is disabled.

I'm going to try some things and see what can resolve the issue, hopefully it can be resolved easily. Perhaps we HAVE to trigger the button to lower the gate? If so, this trick just got blown out of any RTA route pretty much instantly.

edit2:

*Going to try lots of things, will let people know what works*

Indicators/Tested Aspects of the Lock:

- Going straight from the Mad Monk area door to the event results in a hard lock due to having too many events going on at once. It's similar to the Verminator Skip hardlock, but not quite the exact same.
- There is an audible glitchy/scratchy/tear that is an indicator that you're going to hardlock upon screen exit. So far anytime I've gotten this audible garbage, the game locks up on transition.
- This is DEFINITELY what's happening, just like Verminator's Skip where you go back through Dark Forest with Wings. The game doesn't load the music, so it gets confused and says to hell with you.
- Waited over 2200 frames to see if it was just a timed thing, it absolutely is not. The sound is definitely causing the issue here.

Solutions to this specific hardlock (and possibly Verm Skip lock?)

1) Re-entering the Mad Monk map, then going back into the main area.

[Yes, this is currently the ONLY solution I've found because in the process of doing it you trigger both musical pieces in maps where it doesn't matter]

A small note about this glitch that's happening, it seems that in RTA you won't be pulling off a few frame perfect character transitions, but will instead be letting the weapon obtained music play out normally. This music is what is overriding the area's music, and in turn, causing the glitch later. Now, that being said, if any RTA runners would like to test this out, I'm sure people would appreciate it that plan to use this glitch. This also gives us at least some specific direction to work in towards figuring out verm skip for console, which hopefully we can get working, but it has a rather interesting issue with it's musical BS.

RTA runners will have to re-enter the Mad Monk area as well, if everything is done optimally.

So I just ran into a pretty massive problem...

After using this, the map for Megataur refuses to load, it hardlocks like when doing the verminator skip on console. Has anyone tested this PAST just that room? Please tell me this is just something unrelated to the trick... I'm gonna be really sad if another amazing trick has to be cut to some bullshit.

edit:

Specifically:

-Run from Mad Monk area door
-Trigger event to throw spear
-Run across door (it gets some lag here every 3-4 frames)
-Enter Megataur area
-Game proceeds to run, but never actually exits transition, mapID IS loaded as well as positional data, but all control is disabled.

I'm going to try some things and see what can resolve the issue, hopefully it can be resolved easily. Perhaps we HAVE to trigger the button to lower the gate? If so, this trick just got blown out of any RTA route pretty much instantly.

edit2:

*Going to try lots of things, will let people know what works*

Indicators/Tested Aspects of the Lock:

- Going straight from the Mad Monk area door to the event results in a hard lock due to having too many events going on at once. It's similar to the Verminator Skip hardlock, but not quite the exact same.
- There is an audible glitchy/scratchy/tear that is an indicator that you're going to hardlock upon screen exit. So far anytime I've gotten this audible garbage, the game locks up on transition.
       - This is DEFINITELY what's happening, just like Verminator's Skip where you go back through Dark Forest with Wings. The game doesn't load the music, so it gets confused and says to hell with you.
       - Waited over 2200 frames to see if it was just a timed thing, it absolutely is not. The sound is definitely causing the issue here.

Solutions to this specific hardlock (and possibly Verm Skip lock?)

1) Re-entering the Mad Monk map, then going back into the main area.

[Yes, this is currently the ONLY solution I've found because in the process of doing it you trigger both musical pieces in maps where it doesn't matter]

A small note about this glitch that's happening, it seems that in RTA you won't be pulling off a few frame perfect character transitions, but will instead be letting the weapon obtained music play out normally. This music is what is overriding the area's music, and in turn, causing the glitch later. Now, that being said, if any RTA runners would like to test this out, I'm sure people would appreciate it that plan to use this glitch. This also gives us at least some specific direction to work in towards figuring out verm skip for console, which hopefully we can get working, but it has a rather interesting issue with it's musical BS.

RTA runners will have to re-enter the Mad Monk area as well, if everything is done optimally.

AsFarAsIGet

2014-02-21 12:43:35 pm

www.youtube.com/user/HumanMYTH

Oh yeah you are definitely right panda. I know why it happens, its because u dont wait till the boy holds up the spear and the victory music comes... So you maybe test what is faster. Waiting for the hold up spear victory before you leave the room or reentering.

Greetings

TheAngryPanda

2014-02-21 01:06:36 pm

Speed > Safety

in the TAS it's faster to re-enter since you can do everything frame perfect, however, in real-time I think it'll end up being just waiting it out because the risk of bugging the main area is just too high.

So let's talk Verm Skip then, and how we might be able to find what the actual "cause" of the problem with it is. That would be a nice little time saver.

AsFarAsIGet

Edit history:

AsFarAsIGet: 2014-02-21 01:11:14 pm

2014-02-21 01:10:53 pm

www.youtube.com/user/HumanMYTH

I think the verm skip is not possible, since it wants to load the cutscene we skipped earlier. The only way of making this skip possible is to throw off the banquette skip which just isn't an option. Reentering the prison isn't an option either.. Sad

TheAngryPanda

2014-02-21 01:22:05 pm

Speed > Safety

I need to re-watch/re-test some stuff for it before actually making the call about the prison cutscene being the cause of the crash. That's a bit too vague of a conclusion really, I wonder if it was only possible because of SNES9x, for whatever reason... Maybe someone could check it out on Bizhawk/Higan? Also, I wonder if something as basic as opening the menu or re-entering one of the Dark Forest areas would reset the chain of things the game is questioning when it locks up. At the very least it's worth looking into again to double check the findings and re-confirm our previous conclusions don't you think?

That being said, I do think this (the Early Mad Monk trick) is definitely worth going for and also think it should absolutely be considered a run killer should the runner not get it. Why? Because at most you save around 40s, at the very least you break even, that's a LOT of "free" time. However, if you continue after attempting this, you waste an equal amount of time that you would have saved.

I'd like to see what people come up with in terms of menu buffering inputs, which would enable precise control over how many times R is pressed/released when entering the area. If we could find anything that shows any sort of consistency, that's a lot of free time, even with buffering involved. I have my doubts though, as usual.

I need to re-watch/re-test some stuff for it before actually making the call about the prison cutscene being the cause of the crash. That's a bit too vague of a conclusion really, I wonder if it was only possible because of SNES9x, for whatever reason... Maybe someone could check it out on Bizhawk/Higan? Also, I wonder if something as basic as opening the menu or re-entering one of the Dark Forest areas would reset the chain of things the game is questioning when it locks up. At the very least it's worth looking into again to double check the findings and re-confirm our previous conclusions don't you think?

That being said, I do think this (the Early Mad Monk trick) is definitely worth going for and also think it should absolutely be considered a run killer should the runner not get it. Why? Because at most you save around 40s, at the very least you break even, that's a LOT of "free" time. However, if you continue after attempting this, you waste an equal amount of time that you would have saved.

I'd like to see what people come up with in terms of menu buffering inputs, which would enable precise control over how many times R is pressed/released when entering the area. If we could find anything that shows any sort of consistency, that's a lot of free time, even with buffering involved. I have my doubts though, as usual.

GreenAmbler

Edit history:

GreenAmbler: 2014-02-22 09:57:53 am

GreenAmbler: 2014-02-22 09:57:06 am

GreenAmbler: 2014-02-22 09:55:43 am

GreenAmbler: 2014-02-22 09:54:56 am

GreenAmbler: 2014-02-22 09:54:09 am

2014-02-22 09:30:30 am

Did some more testing. It seems pretty impossible to manipulate the Mad Monk for RTA based on how he's spawned. Also, I'm getting only like 20-25% chance he will ever walk over to you. He definitely stops moving completely sometimes after awhile, even just barely off screen. Not sure how often this happens though. He will not become active until you are within a set amount distance of both X and Y coordinates, that is, it's not a regular circular radius around him that you must be in to trigger him but rather a square/rectangular radius. From what I can tell, the useful spawn you need seems to be where he's to the left, I think it doesn't matter whether he's higher or lower (except for shorter walking distance). So if he has 9 spawn points in a 3x3 grid, there's a 1/3 chance he will be spawned in a good spot. And then of course it's completely RNG dependent whether he gives you that ~0-40 second gain, with still a chance he will just stop moving completely at some point. Also, since there's the hardlock from skipping the fanfare sounds, you have to gain another 5 seconds having to re enter and leave mad monk room (which is definitely faster than letting both sounds play out).

So if I had to guess, around 10-20% chance you save any time (with <5% chance of saving more than 30 seconds), and around 75% chance of run kill. So if you have a not so great run going, maybe try to stick it out and hope for a lucky Mad Monk, but I wouldn't risk a good run unless you really don't mind resetting.

I'll look into the Verm Skip glitch some more now there's another thing to reference, but not getting my hopes up. The issue seems to be entering the jail in Ebon Keep for the first from the castle door makes the game freak out because it wants to carry out the actions of entering from sewers but can't do so properly. I'll edit this with my findings in a bit

Edit: Seems like the Verm Skip glitch is definitely tied in with audio. Normally upon first entering, it plays the Dungeon music and has the gate shutting sound, but entering from the castle keeps the castle music, and the gate sound never plays (or I think that's what the first audio issue you hear after entering is). As you kill stuff in the dungeon, the sound glitches kinda seem like they get progressively worse. The raptors especially seem to make horrible sounds, especially if you let them run around attacking. Unfortunately, my emulator will not crash, so I cannot test whether the hardlock is still present after trying different things. IIRC it usually freezes after entering the sewers? Perhaps entering the cell to Ebon Keep town and re-entering the jail would resolve the audio issue and allow you to progress normally.

Did some more testing. It seems pretty impossible to manipulate the Mad Monk for RTA based on how he's spawned. Also, I'm getting only like 20-25% chance he will ever walk over to you. He definitely stops moving completely sometimes after awhile, even just barely off screen. Not sure how often this happens though. He will not become active until you are within a set amount distance of both X and Y coordinates, that is, it's not a regular circular radius around him that you must be in to trigger him but rather a square/rectangular radius. From what I can tell, the useful spawn you need seems to be where he's to the left, I think it doesn't matter whether he's higher or lower (except for shorter walking distance). So if he has 9 spawn points in a 3x3 grid, there's a 1/3 chance he will be spawned in a good spot. And then of course it's completely RNG dependent whether he gives you that ~0-40 second gain, with still a chance he will just stop moving completely at some point. Also, since there's the hardlock from skipping the fanfare sounds, you have to gain another 5 seconds having to re enter and leave mad monk room (which is definitely faster than letting both sounds play out).

So if I had to guess, around 10-20% chance you save any time (with <5% chance of saving more than 30 seconds), and around 75% chance of run kill. So if you have a not so great run going, maybe try to stick it out and hope for a lucky Mad Monk, but I wouldn't risk a good run unless you really don't mind resetting.

I'll look into the Verm Skip glitch some more now there's another thing to reference, but not getting my hopes up. The issue seems to be entering the jail in Ebon Keep for the first from the castle door makes the game freak out because it wants to carry out the actions of entering from sewers but can't do so properly. I'll edit this with my findings in a bit

Edit: Seems like the Verm Skip glitch is definitely tied in with audio. Normally upon first entering, it plays the Dungeon music and has the gate shutting sound, but entering from the castle keeps the castle music, and the gate sound never plays (or I think that's what the first audio issue you hear after entering is). As you kill stuff in the dungeon, the sound glitches kinda seem like they get progressively worse. The raptors especially seem to make horrible sounds, especially if you let them run around attacking. Unfortunately, my emulator will not crash, so I cannot test whether the hardlock is still present after trying different things. IIRC it usually freezes after entering the sewers? Perhaps entering the cell to Ebon Keep town and re-entering the jail would resolve the audio issue and allow you to progress normally.

TheAngryPanda

Edit history:

TheAngryPanda: 2014-02-22 03:20:42 pm

2014-02-22 09:35:33 am

Speed > Safety

Yeah, the Verm Skip is definitely a long shot, but it's also worth researching again since it would undoubtedly change all routes due to item management and ingredient planning.

edit:

hmmm.... Can you even enter the town when coming from the forest initially? I thought there was a cutscene that prevented you from leaving via the door. I can confirm that if the audio plays normally and doesn't have that tearing/scratching/skipping sound you don't get a lock on the ruins issue. It's probably the same for Verm Skip, but I think Asfarasiget tested this and couldn't find a way back out of the prison once you entered, which we'd have to do (and leave again) to reset the music, which in theory, should work. I don't know if the loading zone behind the locked door is actually there or not either w/o the cutscene attached to it being played out.

And yeah, the emulator won't crash, that's why it's "banned" from use in attempts right now, because console players literally don't even get a chance to make it work, while emulators can power through. I'm starting to wonder if some kind of spell could override the glitch since it's seemingly sound based. Maybe Heat Wave or something? Highly doubt it, but if it hasn't been tested yet, it couldn't hurt.

Also GA, after everything was said and done, the Melee Early Mad Monk saved 35.4s in the TAS, so that's a pretty epic find. Thanks for pointing it out. Cheesy

Yeah, the Verm Skip is definitely a long shot, but it's also worth researching again since it would undoubtedly change all routes due to item management and ingredient planning.

edit:

hmmm.... Can you even enter the town when coming from the forest initially? I thought there was a cutscene that prevented you from leaving via the door. I can confirm that if the audio plays normally and doesn't have that tearing/scratching/skipping sound you don't get a lock on the ruins issue. It's probably the same for Verm Skip, but I think Asfarasiget tested this and couldn't find a way back out of the prison once you entered, which we'd have to do (and leave again) to reset the music, which in theory, should work. I don't know if the loading zone behind the locked door is actually there or not either w/o the cutscene attached to it being played out.

And yeah, the emulator won't crash, that's why it's "banned" from use in attempts right now, because console players literally don't even get a chance to make it work, while emulators can power through. I'm starting to wonder if some kind of spell could override the glitch since it's seemingly sound based. Maybe Heat Wave or something? Highly doubt it, but if it hasn't been tested yet, it couldn't hurt.

Also GA, after everything was said and done, the Melee Early Mad Monk saved 35.4s in the TAS, so that's a pretty epic find. Thanks for pointing it out. :D