Posts for Raelcun


Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Some context about the game, the tricks you're using, and why you're doing certain things would be helpful for feedback purposes. How much time does the first flying dog thingsave? Because there's a decently long animation and you're not on it for very long, for instance.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
ALAKTORN wrote:
Yes I know I’ve been insulting and baiting for responses etc. but still what I’m saying is right and right now I’m trying to educate Raelcun, because if he understood my point he could actually improve as a TASer.
You're being pedantic for the sake of being pedantic. The fact that you need to hold the input down doesn't matter in the slightest in the actual scale of things. I'm new to TASing, and I will be the first to admit that I can improve. I am still learning and playing around with it, but that's why I'm working on one game at the moment. That being said, this thread has made it clear that none of those improvements will come from you.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
ALAKTORN wrote:
You don’t understand. I explained how the move works, I never said my method was faster. Your method is perfectly fine but it was the explanation of the move that I’ve been given by you and others in this thread that was wrong. I was just explaining how the move works, because clearly nobody here knew. I haven’t bothered with trying to improve the TAS yet, I’ll look into that later on.
This is wrong, lol. “Suicide” is initiated from a dash.
The only things you've done in this thread are be unnecessarily combative about basically everything, gotten aggressive at people for not explaining things to you properly, including being insulting to the TASer and other people trying to explain it to you, and then said I was wrong about my input which actually results in the same thing. If you're going to improve the TAS do it instead of shitting up this thread. Actions speak louder than words.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
ALAKTORN wrote:
I’ve actually gone and got the game to prove that I was right all along. What idiocy are you seeing? The one I’m seeing is not coming from me. Raelcun I didn’t mean to offend you by the way, but seriously you’re wrong.
I did read your post. If you've played with the game to "figure it out" then you very clearly didn't figure it out properly. Your method results in the same thing as what I described, go look at the frame counter. Both versions wind up with the enemy taking damage on frame 52. Congratulations you found an alternate input that makes no difference at all. Care to stop shitting up this thread now? How about next time you try actually TASing it, counting the frames and trying all of the other methods instead of just yours.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
The fact that you just suggested we be able to skip the blank input means that you have actually no idea what's going on. This has been explained to you in depth, if you still do not believe us, then yes, getting the game and playing with it is the only way to answer your questions. You don't NEED to dash and do the back stance in order to do a spinning harakiri because doing the input I described will do it for you. I provided you a video where you can very clearly see I only did inputs on 4 frames. Watch it again if you're confused.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Finally caught back up to where I was. Not going to bother with a WIP video because it's basically the same minus a few very small things. Changes: -I pick up the burn out in the chest on the first screen. (+3-4 seconds. Didn't time it. ) -I sell all of my healing potions except one at the first vendor trip for extra gold -I buy an extra burn out from the vendor and 1 more charm potion (~3/4th of a second on the entire shopping trip) -I couldn't manipulate the double guard fight to be perfect like the last one without wasting more time manipulating it. As a result, I take one extra hit from the guards. (+2 seconds) -I buy 2 extra spark nets from the vendor in the prison. (+4 frames) -I take an extra attack during the Fruegel fight. (~3 seconds) -- I couldn't manipulate a perfect fight on Fruegel again, no matter how I tried. The perfect starting order came up very rarely within the time frame and the RNG advances very slowly prior to the fight. I wound up with 3 similar patterns with slightly different times. I took the one that I thought was more entertaining. I should have the Prairie and Cave segment leading up to Bale done soon. The Prairie is very straight forward and the boss in the Cave is much easier to manipulate. The reason why I did the inventory changes is because having 2 more spark nets saves about a minute on the Cave boss, and carrying 3 burn outs to disc 2 will save even more than that on Lenus. Testing of some later routes shows that I'll have the inventory space to do that with no problems. It's worth noting that the time saved from enemies attacking you vs having them skip a turn with a charm potion active is very deceptive. It looks like it wastes a lot more time than it does because even when an enemy skips their turn, the game will stop and take about 100 frames to actually make them skip their turn. This is as annoying as it sounds, but means taking an extra hit, as long as it's a fast animation doesn't waste a large amount of time.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
That smirk is godlike, I'll work on it. Edit: Looks pretty good to me.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Bisqwit wrote:
Drats. In Gimp, it's possible to choose a 3x3 brush, or a 5x5 brush, but not a 4x4 one. Radius of 1.6 gets interpreted as a 3x3 brush, and a radius of 1.7 as 5x5.
Make a 4x4 pixel square then copy it, and it will show up in the custom brushes so you can have a 4x4 brush. Incidentally getting a really good Kappa with a 4x4 brush is quite hard. The face is awful, but I like the hair. We need a really good Kappa if we're going to use it, though.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Ragnarokia wrote:
So it is slower than the previous run, so that's a no vote. I just watched the previous one with the youtube link and it looks totally fine for an upload it isn't graphically fucked up beyond viewing or anything. So unless it can be beaten quicker or with interesting gimmick goals (100% etc.) I see no point at all with submitting another inferior timed run just for different emulations.
TASVideoAgent wrote:
Overall about 4 seconds are saved in gameplay, but the movie is still 12 seconds longer due to more accurate emulation. ... <p>Nitsuja's run is quite old, and uses an emulator that itself had some bugs (apparently even messing up the level 1 music.) This seems to not play a role in any of the tricks used however, as I was able to recreate pretty much the entire run. It did however leave out a lot of lag. About 16 seconds of lag and loading time changes were added as a consequence.
If he did indeed save 4 seconds of gameplay then there is no reason to write it off simply because it's "slower." More accurate emulation is actually a plus side even if it leads to longer load times. I actually asked about this issue recently because of PSXHawk and being new around here and was told that if the movie was longer due to loading time differences but faster in gameplay then it would be acceptable.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
If you try to do this in 3 frames it doesn't work. This requires tapping forward, then holding forward and pressing 1+4 If you do Forward, frame advance, Forward+1+4 the game interprets this as tap forward twice and not holding it. As such, it is required for you to do the following inputs Frame 1: Forward Frame 2: Blank Frame 3: Forward Frame 4: Forward+1+4 If you try to do this in 3 frames Yoshi will punch. Link to video Just so you can see a 3 frame attemp resulting in the incorrect input.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Because he didn't bother rotating the video.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
King of the iron "fist" tournament. Not sure why all the hate, really watching Yoshi be Yoshi is still pretty amusing.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Radiant wrote:
Fredrik wrote:
Ksutra is the 12th most popular WAD for speedrunning, including the four official IWADs (http://doomedsda.us/top20.html)
Yes, we already know ksutra is famous. The question is, is it high quality? How's the level design? Does it do anything the base game doesn't? How does it stand out, other than from having a naked chick in the last level?
He answered that by saying that it's harder than Plutonia, with a large amount of gameplay variation including a slaughter map. I'm not really certain how a question like "How high quality is it" is supposed to be answered objectively anyway. If the only reason it was so popular was that level, would there be so many demos of the other levels? Would people care enough about it to put it in the top 20 out of thousands of WADs? Are those answers we can answer objectively in the first place? Beyond having a bunch of people play it, I don't really see any way to try and answer that properly, and a solution has bias issues in of itself no matter how it's handled.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
That's the type of answer that is very helpful. As I said, it's hard to judge, given how the run skips so much of the maps and I'm not very into Doom myself, just have secondhand knowledge.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Big disclaimer here: I'm not huge into Doom, most of my knowledge of it comes from hanging out with people who are. It's also difficult to really tell how good of a WAD this is for speedrunning, because some of the maps are plain straight forward, and others are skipped immediately with archvhile jumps, death slides and rocket jumps. That being said, as for quality of play: -There's quite a bit of time, upon a simple watch through without paying much attention where Doom Guy appears to be running straight forward instead of strafe-running. -Overall, I didn't get the sense of wow factor I normally get, even from seeing RTA Doom runs just from watching the movement. It felt slow in places until the last 2-3 minutes. If everything was done as cleanly, movement wise, as the last 2-3 minutes, then I'd probably be reacting differently. And for the WAD: -A lot of the maps seem to be very straight forward including those that have parts where you are actually just waiting on time based triggers. -The difficulty of the monsters and density of them doesn't seem very high for a WAD. There's only a few instances where you're making what seem to be rather hard dodges around monsters, with lots of rather empty hallway movements. I could always just ask my Doom friends about how notable of a WAD they think this is, though.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Mastania wrote:
Do you intend to keep Shana/Miranda as the main magic thrower for the course of the game? I think that would be the right call,
No, we should only be around level 30 by the end of the game, so her magic stat won't pass Meru's at that point, even if we had her in the party the entire time. Shana/Miranda's MAT stat starts low but gets better gains than everyone else. This means by the middle of the game (level wise) she's still behind even Rose who starts high and gains slowly. Plus on top of that, Meru is faster and gets more turns.
Meru's party-wide healing probably won't be as important in a TAS.
Dragoon form is very slow in general and is going to be avoided whenever possible. Before the final boss the only dragoon form you should see is on The Grand Jewel fight as you can use it manipulate his AI into not healing for 1350 when he hits red arrow.
I suppose dual-magic or dual-physical users might even be worth it in some cases, provided they don't steal too much exp from the end party.
Dual magic until we get Haschel most likely, though, I'll have to test out keeping Rose in the party even then. In general, magic items do more damage per the amount of frames used than physical attacks and in a TAS where I won't need healing items, I can carry a lot more of them than I did in the RTA route. Haschel might be the way to go just because of the final boss and he does a large amount of damage when low health with his best weapon.
I'd been wondering if it wouldn't be worthwhile to fight the Polter Armor at some point for the Soul Eater + some exp, but according to this Loner Knights have a chance to drop it without the detour. In that case I'd be very surprised if the boost it gives Dart doesn't save time, even if only because of how long that final battle is.
That's something I have to test in the TAS, in real time it's faster to stick with the Fairy Sword so he can go Divine Dragoon more often in the final fight. In a TAS, I might stick a Soul Eater) manipulated from psyche bombing a loner knight (you won't need to mash it to kill it) on him and not dragoon at all. The final boss is a really big question mark right now for me, and I'm sort of routing ahead of where I'm TASing to figure out what the plan is, because then I need to re-route the game backwards from that fight because of how long it is. Hopefully with the inventory space saved from not having healing items, I can just take a ton of attack items into the final dungeon and utterly destroy him. One big thing is to see if I can manipulate only single target attacks from Melbu and have him miss the character wearing the Magical Greaves repeatedly since it gives evasion. If I can do that, then I can just carry a lot of attack items, no healing and just go ham.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Well, it looks like it's time to start over. I was aware that something I chose to do early on might result in me needing to do this, but since I was mostly just playing with the game, testing its bounds, and trying to figure out how muhc I could break certain things, I was fine with TASing out the start. I'm going to keep playing through on this file, just to test more things like the trigger I want to break in Bale. But we're only about 30 minutes in and I tested a lot of what I wanted to test.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Part 2 until the end of Hellena Link to video The single guard fight is easy, Lavitz and Dart need to both go first. Found out that in the triple guard fights that it's not possible for me to keep them from getting a turn, so a single turn is all I aim for. Basically Lavitz is too slow, and even when he acts first, by the time you're on the third guard, he will act after the last remaining enemy. If charm potions lasted 4 turns this wouldn't be a problem, but three turns it is. I tried all sorts of in battle manipulation to stop this from happening, but nothing worked. His speed is simply too slow. Trying to manipulate both of the triple guard fights was a pain since there's only text between the two fights and I can't manipulate RNG between the two fights. Instead, I managed, via going into the menu to get a favorable RNG pattern by making the guard spawn next to me and detect me, which then changed the RNG to a good pattern. Movement in the partial spiral staircase leading up to the walkways was a huge pain as the walls are in very weird places. More hilariously fast menuing. I found out that RNG advances from being idle in certain menus but not others, but it advances very slowly and unpredictably. In some cases, it changed in 3 frames, others took another 20 to change. Eventually, by using one potion (which saves time anyway) I was able to get the RNG pattern I wanted for Fruegel by wasting 36 frames in the menu. Battle RNG is weird. You can affect what enemies do, and even turn order inside of combat, but the RNG for every character's first turn is set in advance. So in order to make sure I got all 3 first turns and Fruegel acted immediately after, I had to manipulate that RNG in advance, while everything else could be done in the battle. Battle RNG is affected by a lot of strange things. Extraneous inputs don't do anything, but every time an attack item does a different amount of damage, it changes things very slightly. Completing an addition properly vs missing it also affects it, but depending on which attack of the chain you miss on different things happen. For this reason, we use Volcano. Volcano if you hit 2 times and miss the third does 5 damage and the first two guards have 12 health. This means I can save some time by not finishing Volcano and do exactly 12 damage to that guard with a Harpoon from Lavitz while also setting up favorable RNG with the attack item. I wasted some time mashing the first attack item to 214% to affect the RNG favorably so that only one enemy would get a turn and it was Fruegel's fastest animation by far. As mentioned above, though, each step up in damage affects the RNG, so by mashing to here, we get basically the perfect fight which saves far more time than used mashing the item. Dart throws an attack item at the second elite warden and doesn't kill it because this manipulates the AI by leaving it at 4 health. The enemy party has a single use item called Power Up. Power up doubles all outgoing damage and halves all incoming damage. If Fruegel goes past a certain HP threshhold he will immediately use it. However, if one guard is dead, and the other is low health, he will use it on his turn instead, leaving us free to kill Fruegel very quickly. The damage down rounds down, so instead of doing 7 damage, Lavitz does 3 and leaves the guard at one health, sadly requiring us to finish him with a single hit from Dart. Next portion will go through the cave until the entrance of Bale. I want to try and break a trigger in Bale to save a large amount of time. edit: Also found out that the guard at the bottom of the ladder doesn't detect you by jumping down. It's only triggered by movement, so we jump down and doing the menu before he turns around because it looks cooler. Leaving the text box open in the bridge scene doesn't actually save or waste any frames, just thought it was more amusing to leave it open.
Post subject: Probably won't be sub 8 hours...
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
dwangoAC wrote:
Can you please upload your next movie file WIP to userstorage? I think I may like to follow along. Keep up the great work!
Completely forgot, I went ahead and uploaded that bk2 to storage. Slanted walls always made me curious because of situations like 13m50s. So I actually managed to test it out once I started TASing it. That particular wall doesn't work if you hit it too far to the left because you'll get trapped into a corner, but you can see see how holding straight down shoots Dart right very quickly. As for bugs and glitches I've still never found anything useful, but that's mostly just dealing with things in realtime. I've come close to breaking some things in realtime only to find frustrating invisible walls barring my way. (Like the time I managed to force the character's hitbox through a guard's who was blocking the way until a story trigger was met only to hit a wall behind him.) So I'll be playing around with some of that stuff in the TAS to see if I can make it work. I do know that there is one huge glitch that can be abused greatly for a "new game+" run, but that would take a large amount of time to set up a file for it since it involves save corruption with saving over a file at the end of the game and resetting the console so you start the game with extra items and armor. The big thing about that glitch is you keep all the key items in your inventory, and a lot of the time story progression is tied to key items so you can skip large sections of the game like that. I'm finding this game remarkably difficult to toy with the RNG. It seems to not advance in combat due to many factors which will make getting enemies to do exactly what I want them to do, more difficult, but easier once I get further into the game, hopefully. Using additions affects them as opposed to guarding, or tossing an item. I haven't found any sequence of inputs that affects it yet beyond that, during battle, but missing an addition versus completing it also affects it, so I might wind up attacking enemies instead of guarding and missing additions on different hits of the combo to mess with RNG later on in the run. I'm probably going to have to actually attempt to find the RNG addresses since I'm finding messing with the RNG so difficult, but right now I'm focusing on other things.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Because I'm talented at making good life decisions and a 11 hour TAS seems like one of those. Especially given I've actually run this in realtime and already have a headstart on a route. Link to video Still working on routing out how a TAS route would differ from RTA. Can obviously take little to no healing items, freeing up inventory slots for things that can save time like carrying burnouts to The Last Kraken. Might even be able to skip a shopping trip here or there to save some extra time. This video has a bunch of mistakes (we're talking missed frames) and a few more optimizations to make via being stubborn with manipulating fights, but I'm not being super thorough since I want to test things and plan out how the route will be different. Around 13m10s is a very silly menu. (There's no delay in shop menus, so I can buy 13 charm potions in under half a second.) Also found out 266% is the highest you can get unless you throw the item on the perfect frame which allows 268% to be possible. If you mash above 220% it takes the same number of frames to mash to 268% but if you throw a lower damage item, it saves frames to stop. Charm potions are amazing because they stop all encounters, and even in the RTA route, I was close to having no random encounters in the entire run. With extra item slots from less healing items, this might be possible now. Charm potions also make lesser enemies ignore you for 3 turns in combat, which is used to save quite a bit of time in the guard fights in Hellena. Missing double strikes with Dart and having Lavitz finish with a harpoon saves ~72 frames over finishing double strikes. Using the charm potion on the first guard fight with Dart fighting alone saves almost 2 seconds versus perfect encounter RNG. This game was designed to be played with a dpad, there is no analog movement, only 8 directional movement. Because of this, the physics of running through rooms is very odd. If you hit a sharp slant on a wall but keep running into it, many times it will forcibly push you sideways at a high speed to maintain your current velocity in the direction you're travelling. This means you can save frames in almost every room by running into slanted surfaces. As far as I can tell, the only way to manipulate turn order in combat is by wasting frames before the combat, but doing it immediately before combat doesn't seem to affect it. Results are inconsistent so far. This will make manipulating boss fights like Fruegel a royal pain. edit: Entering any sub menu also advances RNG. Missing double strike by hitting it early or late is no difference in frames. Current plan for Fruegel: -All 3 characters to take a turn killing both of his guards followed by Fruegel's turn coming immediately after. --This will cause him to summon the next two guards immediately -Shana then should get a double turn, throwing an item at each guard. Killing one and leaving the other at 1 hp -Surviving guard takes a turn and uses the power up so Fruegel can't -Dart kill guard with 1 hp by missing a double slash. -Kill Fruegel as fast as possible, hopefully by him only using turns to throw rocks as it's the fastest animation. -If possible, get Lavitz killed so Shana levels up.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
Well, it works now. It makes no sense on why it works, but it works. First, for moving the movie file: I made a movie of the same name in the new directory and then opened it in the Bizhawk that was working. While having the directory open through Bizhawk, I copied it into the new folder and selected move and replace. Second, for the performance issues. I tested to make sure it wasn't a cpu problem by running multiple instances of the working Bizhawk at the same time. For some reason, doing this made running Bizhawk as admin, and from the new location work properly and run at 60 fps instead of 40. I have no idea why that solution worked, but it did.
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
And there's the simple answer I was overlooking. Thanks. New question, since I still haven't been able to figure it out yet. Is there a way to save a movie to a new location instead of to the original? Because while this helps for new files, the current movie is still over in the other folder, with Windows claiming it to not exist. (It didn't even copy over when I moved the folder.) Edit: Running into the same slowdown issue again when it's running from C:\ or from the desktop even when not running it as admin.
Post subject: Bizhawk behaving strangely
Experienced Forum User, Published Author, Player (57)
Joined: 8/22/2015
Posts: 73
I glanced through the forum and did a search, but couldn't find anything like this particular issue. The simple explanation is as follows: 1) BizHawk creates files in a way that windows doesn't think they exist unless I run it as admin. (Even screenshots) 2) Running BizHawk as admin makes the game I'm playing with run under 40 fps, but makes the files visible to Windows. I've checked every setting in BizHawk that I can find between the two, since running as admin uses different config files, to make sure all the options are the same, yet the unexplained slowdown from running BizHawk as admin persists. I'm not very good at explaining this issue in detail, so have some screenshots. Running Bizhawk normally is on the left, admin is on the right. http://i.picpar.com/BR3b.png http://i.picpar.com/CR3b.png http://i.picpar.com/DR3b.png http://i.picpar.com/ER3b.png Extra details: -Opening up the command prompt as admin still cannot to find the files in question. It shows an empty directory. (Same goes for trying to find the files in safe mode. I tried it just because why not.) -I tried changing the permissions on said files, but the only way to even open the properties dialogue is through BizHawk. When I actually attempt to change the permissions, Windows throws a "Windows is unable to find this file" error at me. -Changing the priority and affinity for BizHawk seems to have no effect on the performance issue. -I tried taking ownership of the hidden "Movies" folder that exists if I delete the Movies folder from the admin version of Bizhawk. Even taking ownership, removing the system permissions and setting everything to be fully visible and editable by the admin account, along with the option to make these changes to all subfolders and files makes no difference. I get the idea I'm missing something simple, but I cannot search out what that simple answer might be. Obligatory Computer details: Windows 7 64 bit i7 920 2.67 GHz 12 Gigs of RAM Geforce 550Ti graphics card Memory usage from BizHawk never seems to go above a negligible value, and watching the performance tab never shows the CPU going above 50% so I have no idea why BizHawk is running poorly as admin.