Can you pinpoint an example to compare between two videos where the audio sounds worse?
for example you can watch aruk's run from about 0:25 until 0:45 and then watch the same part in my run, if you hear no difference in the audio there, then i can't help you.
How do you have unmatched gameplay? I see you say that multiple times, but I don't see it watching the runs.
i have no idea what you are looking at but it must be something else but my submitted run here... if you can point me out even 1 part in the run that is worse in my video, then i can take this request seriously.
im not claiming it is perfect, i only said it's the currently cleanest performance yet.
and im quite confident that there will be nothing better for decades if ever...
but feel free to talk me down.
I do too, and it's why I think in certain areas, arukAdo makes better quality decisions in the presentation of his run. For example, in the third level with the spikes coming up from the floor, he doesn't try to get to far away from the floor, so he has several near misses which look really entertaining. In this run, I just see you try to stay far away from them, and don't have any moments where I think "oh no he's about to be killed, *nail biting*, whew..., that was a close one".
you guys are fucking twisted, on one hand you only care for speed, but out of a sudden, you care for entertainment tradeoffs?! how does that fit together?
i obviously have no playful tradeoffs in my run because there is no place where it wouldn't lose time, hence why it's a tradeoff...
if entertaining gameplay was the objective i would have made a run that was twice as long. or maybe i would rather just make a TAS of 102% instead of any%, that's way more entertaining by default. (btw that is still up for grabs if anyone is interested)
but in a serious any% there is no room for playing around, that's why you don't see it in my run obviously.
thanks for your understanding...
i see someone is really tollerant here -_-
i tried my best to stay reasonal and this is how you picture it, thanks a lot.
only because someone's way of talking doesn't please your ears, does not mean the person has bad things in mind. i was and still am open for every question and i always try to answer to my best knowledge.
hmm.. this is a very hard debate on both ends, i admit :/
but maybe my story about this can give you an idea of my motivation:
you might or might not know that from the very beginning, i always only submitted runs that were played on the original version of CVA.
so the question is, how did i even decide for this version in the first place?
why not the KC version that was already used for the showcase?
and also, why did i never switch to KC even when i got rejected twice for it?
my reasoning is simple: quality!
i obviously watched all existing TAS runs of this game i could find back when i picked this game up to TAS it and i always felt like the KC version was kind of a fanmade hack or something, it never felt real even when it had flashy colors, i just couldn't take it seriously since it felt so cheap.
and after my first run got rejected for the reason of not being played on KC version, i went to investigate on it further and then i found out about the lag differences and that KC was somewhat cheating about the playtime.
it made me sad that only this tiny emulation difference seemed enough to justify this ugly version over the real deal, so i tried to find out why i even dislike it so much and if maybe i could just get over it and make another run on the KC version next time...
but upon testing and watching runs, i only got more and more turned away from KC because the audio sounded so much worse than the original and the visuals looked more dead without the hit detection part and a lot of the color choices never made sense to me.
the only true point i could not deny was the fact that it has colors and the GB version was just black and white or greyscale which is indeed a huge point for the viewability and quality of entertainment.
but even with that color argument, i was still not willing to prefer KC over GB, since i would rather watch a black and white run than having this tortured audio and lifeless objects with wrong coloring in my next project...
so i once more went with the original version and to outweight the color argument i had a lot of gamebreaking new glitches found and used in my second submission, so i thought when my gameplay and my performance would just be good enough to outshine the existing showcase, there is no way anyone would care for color when they can get a complete overhaul of the gameplay instead...
well, seems i was wrong again :(
my second run got rejected too and soon after, when everyone saw my crazy new tricks, arukAdo rushed a dirty ass TAS using the new stuff and that shitty thing got accepted just because of KC version, even when it was way worse than my rejected run... maybe there was some reasoning and voting involved, but it doesn't matter. to me it seemed like people give a fuck about quality and good presentation, they just accept the fastest emulation time, no matter the circumstances :(
i was very beat down and because of this realization i also saw no point in even trying to put any more work into this direction since there was seemingly no appreciation for actual quality whatsoever, the only person who seemed to try to ease my depression was ThunderAxe31 and he also created this LUA-script for me to test stuff in the game. he also guided me through a lot of settings and other stuff so i could start a new TAS on a GBC emulation.
and that kinda sparked my hopes again, i could use this crazy LUA tool for precise movement and data on everything going on in the game, so that i could make an even more polished run than my other attempts without the tool.
and on top of all that, i now had colors too in my run since the GBC adds colors to GB games when you play it on that system, so i now had the ultimate version of this run, unmatched gameplay, good audio, good visuals with hit detection, colors, everything! the only single thing left was the runtime itself, but they would certainly not care for such an insignificant detail, would they...?
what im trying to say is that i always believed in quality.
and until now i only got pushed aside for having that ceed.
this whole submission thing here is not about my run or my pride or whatever, i already know i'm unmatched in this game in every aspect, so i need no confirmation on that.
for me this submission is more of a last attempt to convince you guys here that in the long run quality holds more value than other aspects of a product.
you are now at a point where you only have 1 single argument of a few seconds of faster runtime on one side, against basically everything else on the other.
in the end there is still nothing i can do, you will have to decide for yourself and for the representation of this website, so im fine with whatever happens.
i did what i can, the rest is up to you i guess.
hope that helps to understand my position.
and again sorry if i accidentally insulted anyone with anything, i didn't mean to.
It's sad that you see it that way and it's even more sad hearing that you don't care. Here everyone just want to make you understand what there are reasons. In the staff of this site I've met the most nice people in my life, so I really think that there is a huge misunderstanding if you feel that you've being mistreated.
i didn't mean that the people here are bad or whatever, i just came here for a simple reason to provide the worlds finest run, so the interaction aspect is not my main deal here.
and the part where i said this is the last chance i give for the submission, is also not meant to be threatening or dramatic, i moreso meant that i will not make any more runs after this, so if i submit it or not has no influence on my desicion to make no more runs in the future. (i wouldn't even have remade this run here if it wasn't for this new tool and the still existing pain of insult that i had to shove down from last time here...). that's why it sounded so dramatic, but it's not aaaas much meant to.
ps: if hard mode is a thing, then that's probably the case here.
this is basically what i tried to bring across all the time, i just didn't find the right words or thought about it in this context yet, but it's exactly that, hard mode.
it's so damn tiring to correct every small and insignificant detail, only to prevent it to attract the others in this forum to add their assumptions to it, which only makes things more and more complicated to revert back into a reasonable direction for this whole discussion.
i have to step in once again here!
Interestingly enough the IGT actually clocks down faster on the KC version.
Yes you can argue glitchyness or what have you but that's thrown out due to how different the two timers are.
Meaning there's not really a fair comparison to give this run to the KC version.
ingame timers running at different speeds in each version is absolute bullshit!
no matter the version, the ingame timer always ticks down every 32 ticks on the global timer! this means that relative to the games speed, the ingame timer always runs exactly at the same speed.
*edit*
wait a second... maybe i have misread your post.
did you mean the ingame timer actually using different global timer values to tick down? because that would be very interesting!
*edit end*
technically you could now argue that a higher ingame timer value would display a slower runtime, but unfortunately this is also not neccessarily true, especially in precise performances like a TAS run.
there is a hypothetical amount of minimum gameplay time that is needed to play through the game in the most optimal way (obviously not measurable) and there is a set amount of lag that always happens in relation to said gameplay time. those two parts of the runtime are 1:1 equal in every version of the physical games: KC EU, KC JP, GB EU, GB US, GB JP.
they could be comparable, but there is no way to isolate those parts from the rest of the run.
that rest consists of additional optional gameplay which itself causes additional lag in response (the amount might depend on the game version this time).
but there are also cases where you purposly want additional lag in your gameplay to do certain tricks or just finish a complicated room quicker in the end.
here also comes that linearity issue into play again:
the more lag your variable gameplay is connected with, the more options you have to work with it to play through a section optimally.
let's imagine you have 50/50 of gameplay and lag in the original version, that means that in theory you could reduce all lag and only use 100% lagless gameplay for your stategy or do all lag on purpose to achieve whatever things you need to benefit your route or every other combination of the mix, like 20/80 or 70/30 etc...
but if the game only gives you access to 50/0 of gameplay/lag then you have only one option to make it through an area. no variants of lag and gameplay, only bare linear movement without any possibilities to get closer to the never reachable optimal way of doing it.
(this is of course an extreme example and both versions have some wiggle room to play with, but obviously the original has way more options there)
ah damn, i can't explain this properly, it's weird af and anything but intuitive.
so if you didn't understand that last paragraph, then never mind, just ignore it.
Also to quickly touch the eyeball example preventing the spawn is faster by 3 frames at least on KC.
I'm not going to look into the GB version since it's a completely inferior version by creating more lag and having slower movements all round.
(Also the eye falls faster in the KC version meaning that it will hit you no matter what on the way down so you have to make sure it doesn't spawn on top of you first.)
and this is also very likely not true.
i don't know what you tested there with your 3 frames, but getting through the room itself is already hard enough, even without the eyeball problem, so i doubt you optimized every step down while also trying combinations of pause delays with different lag patterns... there is no way your test was accurate, at least not when it was done this fast.
and regarding the eyeballs initial spawn... :D
i already knew that aruk's run fucked up that part, because the fact that he even went for that strat is just a proof of how unoptimal the run is.
the eyeball spawns on the second multiple of 16 on the global timer after the first frame you load up the room. that means if you enter it at 0 you have 32 frames until the eyeball spawns, which is plenty of time to walk away from it.
if you enter at 15, you only got 17 frames of escaping time, which is not enough.
so, if you didn't plan out your previous rooms to end you up on a number that lets you walk far enough until the eyeball spawns in, your only option left is to pause delay it at least once so you get enough time to escape from under it.
i assume since the run was already fucked up at this point and he was already forced to used a pause delay to escape the eyeball,i think aruk just decided to go the unoptimal route and continue the pause delay since the first one now favoured this strategy for his position already being further than the other strat would be.
i hate explaining this madness to people who have basically no idea what im even talking... it feels so draining and ineffective to me, but yeah... there you have my big two cents.
honestly, i don't care too much if you gang up on me or not, so even if there was bias here, that's not really my point in the first place.
my biggest and probably only problem here is that either im soo much off from understanding this websites purpose, or the majority of people here are somehow too blind to see an obvious better showcase handed to them for free including bonus knowledge and whatever else you could ask for...
so maybe it will actuall help me if you can summarize again what tasvideos.org is even about? i always assumed it was a place that gathers the best TAS of each game and puts them out for the public for entertainment?
what am i missing here?
why is it so damn hard to communicate?!
i know that word juggling is not my best disciplin, but is it really that bad...?
come on, i even tried to warn everyone that my frustration meter is already overflowing before this discussion started.
and also im just a human too, so my emotions are somewhat relevant to the words i use and maybe the train of thoughts that i have while reading or writing something.
just don't be too picky about my word choices and forgive my probably agressive tone, thanks.
but back to the acutal topic again, let me try to answer the posts above.
Someone can correct me if I'm wrong, but I think the issue here is that this game appears to have identical gameplay compared to the other version, except with significantly more lag. If there's no distinct difference(s) besides lag, it doesn't matter how much lag reduction you did as the resulting time is slower than the published TAS. As TiKevin83 explained:
how about just click on the x2 speed button on my youtube video and you will have an 8 minute run? because your oh so much better KC version basically does just that.
don't you wanna showcase gameplay on this site here?
anyone who watched the comparison video can clearly see that most parts are just way worse in aruk's run compared to my new (and even my old run, that's why im so triggered every time i watch this insult!)
do you not care for gameplay at all? is it really only the bare emulation time?
if so, then i appologise for misunderstanding this entire websites purpose and i will leave with all my drama and my wrong version choice...
How is it a cleaner run by not despawning the eyeball? Would despawning it result in a slower time?
it's not that easy...
maybe at first glance it looks like a good option, but delaying the eyeball's spawn is not of too much benefit here.
in this screen the first eyeball will try to spawn every 16 ticks on the global timer and to prevent it from doing so you have to pause on every 16th frame obviously.
now since the screen is 256 units long and your character takes about 2-4 frames to move 1 unit, you will have to pause about 13 times to make it across the room and each pause action not only takes 2 frames to perform but also does something to the game's updating tasks that most times feel like they get delayed too, meaning you still have to wait them out at a later time.
soo... 13x2 frames excluding the other delay factor it the minimum best case scenario for preventing the eyeball for the entire room.
however if you just whip it and let it vanish on it's own, you only have to waste about 12-ish frames for the whip attack and some overlapping frames that display the burning and disappearing of the eyeball (i would estimate about 20 in total) and then it takes idk... a multiple of 16 until the next eyeball wants to come in, but that time is waaay enough to get past the room even nonTAS.
you see killing the first enemy in every screen by conventional means is very likely always the best option with current knowledge.
but this is actually one of the hardest rooms in the game, so i can't even blaim aruk too much for doing it the cheap and easy way, but that doesn't change the fact that it is way slower than my gameplay.
oh and btw pausing obviously prolongs your playtime since it freezes the gameplay for at least a frame every time and due to the update check delay thing i mentioned, you could basically say it trades ingame time for real time/more lag... that is why the ingame timer gets stuck there in aruk's run which makes it look like my run would end on a lower timer in level 2 but only because aruk's timer was basically cheated in this room.
What we're primarily looking for are gameplay improvements that aren't related to lag management.
if you are serious about that, then you are clearly missing the point of this entire game, because the main goal of any speedrun in this game is almost completely based on lag reduction.
that is also why i don't really understand why anyone is still defending the KC version since there is less lag that can be reduced... (meaning he whole objective is just more limited, but whatever).
Also on the topic of the comparison video, at 4:39 aruk seems to despawn an eyeball while this movie does not. Is there a particular reason why it wasn't despawned?
and you should rather ask the other way around, why did aruk despawn that eyeball?
but i can give you two possible answers:
either it is faster in the KC version (which i highly doubt but you never know...)
or it's just another one of numerous sections in the game that i performed way cleaner in my new run now, even in my old. (because this strategy would be a hell lot slower in the original version)
yo chill out man, i didn't mean to bash on you or anything.
but nice to see that people still have a physical copy of the game even after all those years.
i too have a cartridge of the EU and the JP version of CVA, which is still unrelated to my playtime though.
anyway, i was just trying to point out that the time and effort of the development does have little to do with this submission here.
if there is still anything missing on my end, let me know about it.
i was reading through samsara's post again and i'm not too sure what i can exactly respond to that.
like what can i really break down that isn't obvious already?
if you watch that comparison video of this other guy, you can already see what's different, at least in terms of strategies and gameplay polishment.
what else can i add to that...?
or to ask in a different way, what do you guys need that i didn't deliver yet?
how dare you call this game rushed!
the more time you invest in it the more you will come to appreciate how deeply the developpers cared for this great game and yes i call it great, because i obviously like it and you can probably tell from my list of publications and achievements that i also care alot for this game on my end.
honestly, there is a good chance that i might have the most understanding of this game on this entire planet (aside from the developpers of course).
and as far as i can tell castlevania the adventure is by no means a rushed game.
everything works very well in terms of the game's intentions and error handling.
at this point i might add that during the making of this submission here, when i reconsidered every strategy and evaluated theories of possible improvements, i found a lot of new evidence that there was a huge amount of thought put into making this game. maybe the next game they made was constructed differently, but for the intentions they had with the first game, they definitely delivered a solid product!
if anything then the konami collection was rushed, because why would they patch out hit detection and butcher the audio which is one of the trademark aspects of the series... you better not argue with assumptions that non of us can confirm.
and what was the result of those version discussions?
i mean what am i even explaining here when there is already a defined conclusion on version differences?
this is so unfair omg...
why the f#%$ do i have to prove and argue about every detail but this other guy can just do whatever he wants without explaining s#%&?! this is biased af...
ok whatever... seems the good cause always has to go the longer way, sigh...
you called for it, so better prepare for a whole assay about basically nothing.
well, where do i start...?
basically there are 3 versions of the game that can be compared to each other and those are the JP, US and EU versions of the original gameboy release of the game.
im not sure about any in depth details on those but to my knowledge only the title screens and maybe the lag pattern can be/are different, but nothing significant for the run itself (to current knowledge).
now listen closely!
EVERY other version, no matter on which platform and appearence, is NOT COMPARABLE to any of the original versions stated above!
anything other than the 3 GB versions are different games and behave differently!
i have no freakin idea how this issue is still a thing but for the sake of avoiding another discussion going downhill for no reason, i will try to even explain this nonsense again for you:
the konami collection on gameboy color, that includes a similar game to castlevania the adventure, loads objects and screen backgrounds in a different way than the original game. for the unknowing eye this doesn't look like too much of an issue, but there is a huge ass catch for this specific game!
the thing is, castlevania the adventure is a very special game that seems to be developped in a way that purposely uses loading times aka lag frames for it's processing to asure a fluid gameplay, that is btw the reason why it is so damn slow... this is all done on purpose and it affects basically all aspects of the game's tasks, like character movement, object loading times, button checks, graphic updates on screen and even the audio.
now back to the issue, let's take the konami collection on GBC as an example.
that version seems to work different than the GB version and actually takes about 80% LONGER to load every screen initially, but therfore there are less loading times during the screen (or room) itself.
this would not be any problem if the game itself wasn't based on this almost specific amount of lag, because by basically squeezing all loading times into the initial loading of a game section, the absence of loading time during the characters traversion through the section/screen/room the gameplay itself is sped up and does run faster obviously.
while on paper this might sound like something positive for you now, in the case of this game it is definitely not.
due to the speed up in the gameplay (which is not actually a speed up, only a lack of lag) the games whole behaviour changes since as i told you, every object and action and audio now loads faster and thus has a different timing (which was not intended in the original game!).
it is hard to notice the gameplay differences just by watching the video, if you have no experience in this game, but if you listen closely, you will notice how the sound is heavily stretched and squeezed when you watch the konami collection.
this is just the audio cue though, but the gameplay itself is also greatly affected by this lack of lag during every sceen, mainly because candles and other objects now don't have to load or update as much as in the original game and this make the whole strategy during each section different, because the actions of the player and the enemies on screen do also change up the amount of loading.
so let's say in the original game the optimal way to process through a room would be to whip a candle to make it unload and then walk on...
in the konami collection it might be optimal to just walk through the screen while ignoring the candle alltogether, since most of the loading was already done before the room was shown (in the added 80% of initial loading time).
now you could say the total time with the added 80% in the beginning might still be shorter than the quicker loading of the original game but added gameplay time during the screen...
the thing is, in the original game you can greatly reduce your lag DURING a screen but in the konami collection there is almost no way to readuce any more of the already reduced lag, which just means the gameplay itself is unable to save loading time, because the game already preloads everything in every screen initially.
TL:DR
so the conclusion is that the konami collection is set to a specific amount of lag (yet still very variable) per played section of a TAS speedrun, while the original game also has the option to reduce it's lag by moving in certain ways within the gameplay's possibilities.
meaning: the konami collection is linear af
and the original game can be optimized greatly (which adds a whole lot more of interesting gameplay to the run, just saying...)
furthermore there are even some minor things in the konami collection that make the watchability worse than in the original.
i already adressed the very unnatural sound speeds of the konami collection, but there is also no visual hit detection on any enemy in the game, which in my opinion makes it looks cheaper than it should.
and before anyone argues with colors... guess why i used the gameboy color emulation to make this TAS... it's not as detailed but it looks way better than in just black and white. and also the boot up of the console takes less time than the regular gameboy boot up.
and to close the circle on this insult of a request, how about you tell me why the konami collection even has a right to exist as a showcase on this site in the first place? i wanna hear any reason that makes it superior to the original game?
your turn...
(not sure if i managed to bring my points across here, as i said im still pissed from the last two times i had to deal with this shit here. plus it is so late here that it's early again, meaning i have been up all night, so maybe my reasoning isn't on point at the moment.)
uhm... didn't i explain it enough in the submission text?
what do you wanna hear exactly?
should i break down every single trick used in the run?!
i mean i can, but is that really necessary?
do you need every saved frame in detail or what?
(i seriously struggle to keep my calm at this point, please forgive me if anything sounds a bit harsh or whatever, i will try to stay reasonable)
hahaha actually...
you're right this time! i just dug out my own cartridge of the game and played it on my GBA SP and in fact i also saw those weird lines comming up even on the actual hardware.
so i appologise for my almost rant before! i was indeed wrong all the time.
i will check this again with my GBC later this week, but maybe i just never noticed it before since the screen is so damn small and only on emulator you have a big screen to look at...
also, i would appreciate if anyone could check if those lines also show up on the original non-color gameboy? because i cannot test this since i have only a color...
to my excuse i could say that i mostly played this game in the castlevania collection on my switch and there were no lines at all, but that is obviously another game basically and therfore doesn't have to work the exact same as the original GB version. so again, sorry for my stupidity, i just didn't play the original game after i discovered this graphics bug, but im actually glad that it comes from the game itself and not the emulator, because now i don't even feel bad for my WR run's quality anymore since it is already shown in the best possible way, nice.
thanks to everyone who tried to help and wasted their time for me.
at least i know better now :D
btw last question that still remains to me is which exact emulator build and stuff would i need to submit a new TAS run of this here?
so far i used bizhawk v2.4.2 on gambatte core and GBC setting to run my GB ROM. is that still fine or do i have to use something else like the build you sent me earlier? as i said there is no intro screen somehow and that makes no sense for a serious run... and i also need that intro screen to change the color palette of the game, so i can't skip that!
i just checked all the stuff i got sent and first thing i noticed is that there is no intro sequence of the gameboy anymore when i run that other bizhawk built (idk if it's the core or whatever i set up there now, im rather confused tbh), but that is besides my point.
the main thing is about the graphics bug thing, that is still present as i mentioned.
and since there are still people who think im freakin retarded, i made some screenshots that show what i mean, but i can't seem to post them here... maybe i am retarded after all :/
also, i made a movie file too, but i don't even think that matters, because you could basically record anything in the game that involves scrolling to the left and it will show the issue. i don't really understand why i should make a specific movie... the bug occurs consistently, just walk left in the game and it will happen on bizhawk (yes i use gambatte core but that makes no difference though)
lol, that was my stream title, i tried to make a bad pun but kinda failed^^
well, basically the main issue is that the bizhawk emulator generates those weird lines when the screen scrolls leftwards specifically! and im sure those are not present in either hardware version of the game, only in bizhawk so far.
the timestamp in the stream video was just to demonstrate this specific issue, nothing else. to make a real hardware verification you only need any movie file that plays far enough in the game to reach a place that has the screen scrolling to the left side.
(funny enough, scrolling to the right like in super mario bros seems to work fine, only leftwards generates those weird texture lines while building the image...).
also, aside from those lines, every other graphical "problems" that you see are from the game itself, those are completely normal and fine. this issue here is only about the small lines of textures that appear while the image is generated.
ps: if anyone knows anyone who is in charge of bizhawk stuff, please tell them this left scrolling problem, because it is rather ugly and maybe it's easy to fix when scrolling right has no problems at all...
thank you for the help!
ok guys, i used the latest video from aruk here as a reference and made a list of what i see there. (no proof as always but feel free to call me a lier if i got something wrong).
no mark before a text means this is very likely true and will save time!
* means a smaller improvement of very few frames but they add up throughout the run.
? means it is more on the theory side and could be ignored if nobody wants to try it.
() are basically just comments, no real suggestions in that sense.
so this is what i got for level 1.
here are the video timestamps:
0:14 - *grab items asap
0:22 - *same but more significant
0:35 - definitely destroy this candle, because rope climb on same screen
0:40 - poorly executed rope jump through screen transition, avoid climbing frames at all cost!
0:41 - ? since you jump already and the landing floor provides a lagless fall, why not whip this candle?
0:43 - another rope dependend candle, begone!
0:47 - *good stuff on the eyeball cycle but it looks like you didn't completely max out this jump attack here, you wanna both jump and later here
0:52 - *in a way it's a tradeoff but it looks worth: just delay the whip a few frames to also get the candle
(you increase the enemy lag by that amount of frames but this lag combines with the candle lag so it's less of a problem than it would be seperately and you easily get those frames back from the candle later, remember there are 2 steps in your screen)
0:56 - dropping down platforms is often faster when you jump down instread of just walking (maybe this little height here is too low but it's always worth a try at least)
0:56 - ? and while you jump already, it's also worth to consider whipping that candle here for free...
1:00 - ? another inevitable jump and 2 candles align before your eyes, at least think about it, buy 1 get 1 free...
1:05 - ? im not too sure how your new walking trick works but this part here looks a little wacky to me, might be possible to jump before the needed frame to get more horizontal momentum on the peak to need less frames of pausing
1:13 - (just a correction on the reasoning here: whipping this candle is good because it is a rope dependent one and NOT for the reason you did it! im pretty sure the invincibility cross doesn't make you lag less, if anything then it might even increase processing power for the blinking)
1:20 - in theory this room should be identical to the one before but i know that the eyeball cycle is very though here. anyway, here is another rope dependent candle plus the location is very convenient for lag reduction. (on a side note, you don't have to worry about the jump lag here, because you can just chain 2 jumps after another and it is likely to avoid the inbetween landing lag)
1:32 - this time i even have proof in form of my submission file but i think i didn't even do it optimally. you should understand it by now, this candle is rope dependent and the longer the rope the more impact it does to your lag.
(oh and in case you somehow didn't figure this out yet: the fireball of the whip is obviously an obcest too, so hit the candle with the first frame of your fireball with an attack at ~10 frames after a jump! NEVER let the fireball show up if you can avoid it!)
1:43 - on a long fall like this it is definitely worth to jump down instead of walking, regardless of langing lag! and here is another 2 for 1 case so just make sure it is not worth to attack those two candles here, it's a free whiplash...
1:48 - ? (there are numerous candle tuples like this throughout the entire run and for those im actually not sure if they are worth to use an extra jump for that you could avoid, it probably depends on the distance the fireball has to travel, but i can imagine it COULD be worth. but i wouldn't make anyone test this, because how hard and tedious it is, so im fine if we ignore those cases.)
2:06 - *i could be wrong but this is how i feel about those falling platforms:
i believe that walking on them while they fall down is less optimal because the platforms lag a lot and moving in lag is always slow obviously. but instead of walking i would suggest to first, jump on them asap to make them fall down asap and second, plan it ahead so that you only do chain jumps on all of them where you only touch them for 1 frame to make them fall and you keep all your momentum without landing lag until the last chain jump ends. that way you can cross them faster than when you always get new landing lag on every jump and have to deal with additional processing when the positions of the objects update with a dependency. (very confusing, sorry. basically just do jump chains over falling platforms!)
2:14 - here is another drop off with a good height, so it is likely that a jump down will end up saving some frames because you gain downward momentum from jumps but not from walk offs.
2:31 - falling platforms again, try to plan ahead and do chain jumps so that the last one lands on the last platform asap to make it fall quickly.
2:35 - another drop off, here you wanna chain the jump down with the jump on the next platform so plan ahead where to jump off from above and also chain that second jump into a third to fall down the platform later and do the fireball shot on the candle.
(i just realized, this might be just a workaround for what you already did with finding the least lagging landing frames... idk)
2:45 - (not much more to say about level 1 only that i believe ending on an even timer has less lag than an odd number, but if you apply all the improvements it is very likely the timer will already be on 6 here, but just in case you will barely end on an odd number, try to test if losing some frames to make it even again is worth the loading time into level 2.)
sorry for typos and stuff, im halfway asleep while i post this...
ok, you won!
i take back what i said, you actually outsmarted me with this discovery...
if this works the way i think it does then you just hit me in my blind spot, making me look like a dumbfuck after all my "use your brain"-talk, because i litterally had all the clues in my hand but didn't put them together... (provided it does work as i think!).
could you explain how it works please, because i wanna make sure that im dumb.
ps: should i refer to my own video to make my improvements list?
because im not going to screenshot every single screen in the game, that's even longer than just watching a full video...
the thing is, my video is clearly outdated now and im not sure how helpful it is if i basically explain improvements that aren't even shown in the video that i refer to...
so again, what video should i refer to when i make a full game improvement list?
do i understand the comparison picture correctly that the one without the candle is further into the screen than the other?
if so, then that is a tremendous timesave since you even have half of a screen to pass when the candle is in the center like the picture shows.
but i somehow feel like im missing something, because why is the pic without candle further when you spent 4 more frames on the other candle as you said...?
it's either very unintuitive, like many things in this game, or i don't understand the comparison correctly.
tell me which is the case here? thanks.
btw i will try to make a list of things i think can be improved, but it would be halpful for me to have any video or so that i can refer to.
otherwise i would have to explain the exact locations and all that stuff too when i have no reference... so anyone tell me what source i can refer to.
(oh and it won't be that quickly though, i am pretty busy this week especially today and tomorrow, so i will only be able to start making the list on friday at best. just saying.)
Relax man. If you have a decent submission that cannot be fixed without a lot of work then it will probably be accepted. Freaking out, acting like a jerk, or over reacting will only make you look bad. Just answer the questions.
ok fuck this...
this is the exact same story from last time all over again.
what the fuck do you guys even want??? i just don't get it!
if you say my run here is not good enough because it could be better, then why on earth is the currently latest published run even accepted in the first place, when it's far worse than my run now while the person who made it had even more knowledge and tools than i used?? this makes no fucking sense to me, im sorry.
there will never be a perfect run so what is even the deal?
im not satisfied with my own run anymore since i know i screwed up in a few places, so im ok if my submission is being talked down, it's only fair since i didn't try hard enough i guess.
but if another movie gets accepted in the future and i can find even a single frame in it that is less optimal than my current run, i can never take this whole site seriously anymore...
i only made this run because of the same reason im whining here about other bs:
i care for improvement and not for who the fuck made what, that is not how i grew to TASing, im used to everyone supporting each other and improving together.
i did even credit arukAdo in my video description because he was the one who found it in the first place. i only wanted to make something more optimal than his latest run (which i did).
so if he or anyone else will make an improvement over my improvement, that is completely fine for me, i even encourage everyone to try that.
but when out of selfishness the next creator just does his own thing and basically invents the same stuff again instead of using my work as a basis to build upon, that is where i start to get angry.
because what is the point of wasting time and progress only to avoid cooperation...
it's like a lose lose on both ends.
i already had a better run than aruk, even before this stuff here, but it never was about being better for me, only about pushing the game. and then i saw his pause trick in his run and that led me to improve further with his knowledge as a basis. and now that i showed what i did with it, he seems to push it even further by cleaning up my oversights with it. but that's rather easy as he stated himself, because the trick isn't as hard when you have the right tools and knowledge of it.
i just hope it also works with the more difficult improvements as well, that involve more intensive planning and thinking. that is everything that scares me.
because seeing how he recieved my stuff so far, it doesn't seem as he is acknowledging much of my effort and instead tries to "proof" that he can do everything on his own... (which would be fine if he wouldn't show those mistakes to me everywhere)
so all i want is to save him time by providing what i already found out so he doesn't have to find it again!
to me TASing is a team sport, so i don't really get why it always sounds like i have to "proof" that i belong to the team...
why would you even mistrust anyone that works for the same goal as you??
why would i talk shit to you? what is the point? how would i benefit from that?
how would the goal benefit from that??
this is basically my whole confusion here.
screw my submission, i don't care, im more about the moral thing of the whole subject.
(btw sorry for the slight outrage)
THANK YOU!!!
finally we are somewhat on the same page @aruk!
but i think you misunderstood what i said here:
arukAdo wrote:
Ive followed instructions (stage4) and the candle is still there, it makes no difference. (except I lost frames...)
as i mentioned, im not entirely sure if that specific case is worth the tradeoff but that the candle spawns is a fact! it's impossible that it didn't work for you without hacking^^
in level 4 you have to destroy the last candle of the line of candles that starts at at the beginning of the level. or in other words, it's the last candle before the boomerang guy. and to verify that you got the right one, i also mentioned the content of that candle which in this case is a heart.
and while we're at it:
The whole idea that a candle content is lagging the game is just wrong, im not gonna bother to explain in details how ive figured this
first of all, i wonder why i always have to proove stuff to everyone in form of a file or video, but you are just allowed to claim stuff like that without evidence?
i just find that unfair and it makes me feel like my words hold less value than that of other people here...
(but whatever, this is not what i was going to get into for now.)
back to what you said. i have no file or anything prepared to proove it, but i kinda wanna appeal to everyones rational thought process here.
(in addition, ONLY seeing something doesn't necessarily mean you understand it, so i think those calls for proof often sound more like a cheap excuse to ignore an extra step of thinking to me, but im drifting away again.)
ok, so lets say the content of a candle would not affect the game at all as you claim. (which is not true, if you believe me or not. because it even matters how fast you grab an item that falls out of a candle in the GB version).
this would still not change anything i said because i only really talked about the candles themselves and how they cause lag.
and if we only look at an isolated case of only those 2 candles that depend on each other to trigger the spawn, we can just elaborate what in theorie would be less laggy, so hear me out:
think of 2 empty rooms with nothing in it.
now think of how christopher walks straight through both rooms with no interaction. this will cause a certain amount of lag or emulated frames overall, so just think of this time as a constant measurement.
ok now we let christopher walk through both rooms with only jumping once and whipping in the air. this will obviously take more time than just walking.
ok so far so obvious.
but now we have christopher walk and whip through a room with a candle in it.
what will happen? obviously the candle would just add a bit more time to what was previously just an empty room and we did the same actions. easy right?
and now we finally come to the actual scenario, we do not whip the candle and walk through the room. will that be faster than whipping it? we don't know yet.
so we now only have to compare the amount of time the candle itself costs and what the jump and whip action costs. and if we would know those times, we could already tell what's faster... right? but even without a prooving file, we can still come very close to a likely answer. (and that is what i kinda miss in this whole submission debate thing here, i feel like nobody is used to thinking anymore...)
you can even figure that comparison out by that simulation.
the first case would be to walk 1/4 of the 2 rooms (the first half of room 1), then whip the candle and walk 3/4 of the whole way (the second half of room 1 and all of room 2) with no interactions.
in the second case we would just walk straight through room 1 without jumping and whipping, so the candle will be there for not only half of the time it takes to walk through a room but it will stay for the whole room.
but not only that, because we now have not whipped the candle in room 1, we now also have to walk through room 2 with a candle in it.
so if we put it all together we have 4/4 of candle lag throughout both rooms in this case but instead we save the jump/whip lag.
and now we have both cases and can put them side by side:
1/4 of candle lag + whip action lag.
vs 4/4 of candle lag but without whip lag.
(we could even cut off the first 1/4 of the walking, since it is the same for both cases, but whatever.)
and now tell me, who in this forum thinks that 1 single whip action produces more lag than the combined candle lag for the screen time of 1.5 entire rooms?
i for myself, do not believe in this!
of course i could still be wrong, but when you played the GB version you will develop a sense of scale for the different lag cases and i even tried to come up with a rule of thumb in the subtitles of my video, where i said that you could roughly say that 2 candles are the maximum of what you can leave on screen when there is no interruption of walking through the room, meaning a flat ground without steps and no ropes or other stuff that makes you spend more time than just walking.
oh wait!! if your rule even says that 2 candles is ok, what were you even talking above then??
you still don't use your brains right? ok, let me explain this too... -_-
imagine 2 candles above each other in a room.
to walk in and out of sight you need 1 full room/screen of movement.
but if you have the candles alined horizontally and you only see 1 first and then the next comes into sight after a while, that is more than 1 room/screen of movement but in exchange you will only have a smaller time where both candles are on screen at the same time, so the start and end of that scene only shows 1 candle on both sides. so it is even less laggy than the case with 2 candles on the same vertical position. (sadly i have no comparison for those cases but i assume the longer the screen time takes the worse, meaning the 2 candles in same spot case might be faster than 2 candles in line)
so what am i saying?
my "rule" says that 2 candles in a line, which i consider the worse case, is not worth destroying any of the candles, but the catch is that the screen time on that case is only slightly more than 1 room/screen that christopher has to walk.
so the worst of cases here would be basically what our level 4 secret candle spawn is. a room where you walk by a candle and right when you lose sight of it, the next candle comes into the screen and the same time has to be spend twice until both are gone.
and that case is more of an extreme one in my rule of thumb idea, so it kinda doesn't really cover that. and obviously it is not a real rule at all, i just thought that would be somewhat helpful to keep in mind for a very rough estimation of the gameplay choices. there could easily be a case where this doesn't fit, but it also often cannot even be applied because of the circumstances and obstacles on your way, so take this all with a big grain of salt ok, i never meant to claim this to be true or so, it's just like a tool to help watching the gameplay basically.
damn, this text is huge... fuck. im sorry guys.
well, there is one more thing i wanted to adress:
i tried to do that trick in level 3 with the floor and ceiling and i actually got it to work. and the same applies for the knights in level 4, it seems i was just too dumb to pull it off when i made my run. but i also didn't have a RAM watch to help me...
im kind ashamed that i didn't manage to find/pull off those parts and seeing them work now makes my run feel very unoptimized in that regard.
so i wouldn't mind if you reject my submission since it now actually is not optimal at all in some places. (here is a good example where video proof works, because it is rather obvious that a major exploit saves a shitload of time, but small actions that only save a few frames here and there aren't really obvious, so i don't think that is worth the struggle of making a file or video for them. just my opinion)
ok, sorry again for the big text, but i will end this post now before i can sell it as a book...
i can help you with some theory stuff if anyone has questions, but i can only tell you what i know and not provide stupid file comparisons and such... in the end the hardcore RAM details and everything have to be tested by whoever wants to publish the next run, i already wasted about 500h on this game, so i think i deserve some rest, thanks for your understanding.
(i hope i didn't start a war here by accident now)
"...research existing runs to avoid rather obvious mistakes..."
isn't that an input file already?? i was actually refering to my submission file with this.
in a way i always feel insulted whenever someone posts a "new" file that even includes mistakes that i tried to avoid in my run and i even tried to explain some of them in my maxed out videos description and my subtitles...
so don't tell me i play the big mystery card here, i basically showed most of my knowledge already, but im also glad if someone wants to hear more.
and to name some examples, this is what bugs me the most and it's not even a TAS related fact, only basic game knowledge that could easily be obtained just by casually playing the game, it's no secret at all, which makes it even more outrageous for me:
the very first candle of the game has to be destroyed!
because this candle triggers another additional candle to spawn in the next screen, the second last that contains a 1UP.
and while it technically doesn't matter too much to leave the first candle (it's even better because of less lag caused by the whip action), if you leave it you will have to deal with additional lag from the the other candle that you purposedly spawn.
and the 1UP candle that spawns isn't just a similar case than the one at the start, because the location is right next to a rope that you will have to climb up.
and since climbing is super slow you will spend a lot of time on that screen before you can get out of sight from the candle, which in return means it does NOT save time to leave this candle there while you climb the rope.
so by leaving the first candle alone and seemingly saving the action of whipping it, only causes you to be forced to still use a whip move on the second candle, so even if both actions were equally time consuming (which they aren't, since the location favours the first candle over the 1UP candle), you would still have half of a screen movement more lag before you can even reach the 1UP candle to destroy it... so no matter what, spawning that candle only costs time in my eyes.
unless someone can proof me wrong with a detailed explanation that convinces me that i missed something...
the same problem goes for level 3 and level 4 as well. (at least level 2 was somewhat fixed in the last video here).
in level 3 you have to destroy the 3rd candle of the level to not spawn the first one in the horizontal scrolling section with a crystal ball in it (which i explicitly explained in my video, so idk why people point at me from all sides, but whatever...).
and in level 4 im not super duper sure if it could somehow be worth to ignore it, but i think unspawning any object sounds like a good idea...
at the start of level 4 you have to destroy the last candle of the low ground section where the knights are, that candle contains a heart and if you leave it, it spawns the second candle in the next screen which contains a blinking heart.
(but guess what, i explained that candle spawn too in my video...)
why is every time such a drama here, i just wanted to enrich your lifes with my work, but instead i kinda feel like i did something wrong when i made my run :(
sorry if i did, my only intention was to push the game further than ever before...
i appologise if i accidentally insulted anyone or whatever, i just feel disrespected if my improvements get ignored on purpose even when i showed and even explained how they could be achieved...
there are some other things i dislike too regarding gameplay optimization but they are only a matter of a few frames which im not sure if it even matters in the color version, but it certainly does in the regular version.