It is, marginally. The advantage of the shinespark was that it could drain energy to help for the Crystal Flash preparation, which made it a faster strategy in the end, but now that the Crystal Flash can be skipped, it lost that advantage.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Joined: 11/14/2005
Posts: 1057
Location: United States
For those who are curious, I have made 2 different smv files showing the difference between the walljump and the shinespark. The walljump was 1 frame faster than the shinespark. shinespark vs. walljump. What people don't realize is that the shinespark requires a 3 second cooldown period and the charge up period waste about 2 seconds also.
Also, I don't think any author should be compared to another. Everyone works at a different pace. I for one would rather pump a run out quickly then make a revised version later on, because there will be inevitable discoveries that would make the run obsolete anyway. JXQ's 100% run is a prime example. Even though it is very tightly optimized, there have been so many recent discoveries that it could be obsoleted by a much sloppier run. This however is not the fault of the author, because he or she does not know at the time of such discoveries. Super Metroid is the type of game that will never be perfect anyway (and I really mean never).
Joined: 11/14/2005
Posts: 1057
Location: United States
Earlier in this thread I made the claim that phantoon could be 1 rounded with just 3 packs. I shouldn't have said that without proof, but now I have finished testing it and it is easily possible, check it out here.
As people may or may not know, I am going to improve my current run. I estimate about 15-20 seconds of improvements (or ideas that may or may not work), but there are some issues that need addresing before I even begin.
My biggest issue by far is whether or not I am going to use this 1 round phanton tech that I just posted. If you have seen my published run, you know that I only have 2 missle packs when I face phantoon. The closest pack is the one by the charge beam which requires almost no time at all to collect. The problem is that I only need 3 packs to finish the game, and this means I will have to literally jump over the missle pack on the way to the wave beam. Jumping over the missle pack is probably one of the absolute ugliest and stupidest looking things ever. I do not want the run to look stupid at that part, so the alternative is to simply collect the extra missle pack. The problem with collecting the extra missle pack is that it is not required to collect to finish the game, and it wastes 7 seconds in real time. Though I aim strictly for real time, I am tempted to make an exception here simply because it looks like shit jumping over it. Any feed back would be cool.
I plan on starting the revised run in the summer, so this leaves me time to find all my mistakes. Right now I am meticulously combing through the published run looking for even the tiniest improvement. If you noticed anything at all that looked improvable, please post it.
btw is anyone interested in a new low%tas?
is this maybe usefull for the speedbooster variant? i remembered this trick was only possible with the blue suit..
but now with the supershortcharge there is no need to keep the blue suit all the way..
http://dehacked.2y.net/microstorage.php/info/4066/speedboosterlow%25.smv
Yes, I'm currently preparing to do this run. But first I need to figure out how to cross Craterian missile lake without taking the missile (I want to take Speed Booster instead of Ice, and 15 supers + 5 missiels instead of 10+10) and losing much time doing that.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
Joined: 11/14/2005
Posts: 1057
Location: United States
evilchen, he will not have the speed booster at that point. If he is using the same route as in my run (I think he is). He must bomb jump over the missle pack.
Both simple and extended diagonal jumps won't work. The only thing that seems to work is very slow hovering while adjusting my position by a couple of pixels with every bomb. I have yet to determine the fastest possible way of doing it and compare it to the possible saving an additional super missile would give.
After deciding that, I can start rolling. :)
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
I agree that jumping over a missile would be cool. Getting it would be terribly inconsistent, as you could ask why other nearby items weren't grabbed.
You could probably jump over the missile all the way to the opposite wall and walljump from there, which would look less like "I'm jumping purely to avoid this item".
I will watch through your existing run in detail at some point before your start your improved version and let you know if I see anything that could be improved, while Super Metroid is still fresh in my mind.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
I already mentioned it somewhere, the Shinespark is exactly 20 frames faster compared to Walljumping IF taking the last Power Bomb refills from the Drainer enemies in both cases. You can start shortcharging immediately after landing on the final platform and still get the refills in time that way. If you make Walljumps you need to wait a bit more before starting to run into the next room. But even without this refills the shinespark is at least 10 frames faster if looking into my test results. Pretty minor but still useful time saver. :-)
hero of the day wrote:
Super Metroid is the type of game that will never be perfect anyway (and I really mean never).
Wrong. Except of the item refills and Boss refill drops that affect the next parts of the run slightly, I think it's very well possible to make a perfect run in this game. It just requires alot of patience, very good knowledge of the game mechanics and time (speaking of in-game time obviously, realtime is a harder task of course).
In order to pass the WS lake room without Grapple, Speedbooster, and without touching the Missile pack, moozooh needs to use horizontal bomb jump (not diagonal because it gains height as well, which wouldn't work because Samus would hit the roof too early and therefore forced to cancel the HBJ). He is testing this room for 1-2 weeks already (I think). If you were successful with that moozooh, I would be very interested to hear what you have to say to it, and if it's worth the big time delay HBJ requires to use it at all there.
hero of the day wrote:
Earlier in this thread I made the claim that phantoon could be 1 rounded with just 3 packs. I shouldn't have said that without proof, but now I have finished testing it and it is easily possible, check it out here.
As people may or may not know, I am going to improve my current run. I estimate about 15-20 seconds of improvements (or ideas that may or may not work), but there are some issues that need addresing before I even begin.
My biggest issue by far is whether or not I am going to use this 1 round phanton tech that I just posted. If you have seen my published run, you know that I only have 2 missle packs when I face phantoon. The closest pack is the one by the charge beam which requires almost no time at all to collect. The problem is that I only need 3 packs to finish the game, and this means I will have to literally jump over the missle pack on the way to the wave beam. Jumping over the missle pack is probably one of the absolute ugliest and stupidest looking things ever. I do not want the run to look stupid at that part, so the alternative is to simply collect the extra missle pack. The problem with collecting the extra missle pack is that it is not required to collect to finish the game, and it wastes 7 seconds in real time. Though I aim strictly for real time, I am tempted to make an exception here simply because it looks like shit jumping over it. Any feed back would be cool.
I plan on starting the revised run in the summer, so this leaves me time to find all my mistakes. Right now I am meticulously combing through the published run looking for even the tiniest improvement. If you noticed anything at all that looked improvable, please post it.
Very nice smv. Though you are forced to wait a couple more frames at the Phantoon fight to firstly shot down the blue lights for refills and only then start the doppler-attack. Not sure if it would save more time compared to a 4 Missile-pack variant, because this additional Missiles not only save some time in the fight itself by starting the Phantoon killing immediately, but also some more time in the rest of the run. In your case I would still pick 3-packs and one round him, then just jump over the one right before Wave Beam. It wouldn't look that bad I think, and the one-round kill should make up for the 400 frames long item-screen delay of the additional Missile-pack IMO.
Ok, I just checked, heros smvs and my Walljumping test there was indeed not 100% optimal. Heros one with a 13-frame interval seems to be 100% perfect, and he only got 1 frame faster with it because he didn't optimize the Shinespark. I redid the shinespark part myself and ended 3 frames faster than his perfect walljump test. But keep in mind this is without refills. In the 100% run you will badly need PBs for Maridia, and since you get whole 4 from the drainer in one there, it should be worth it IMO. So the delay would definitely be more than 10 frames with them, still.
Hero's walljumping is improvable by at least one frame.
In actual play, the amount of enemies that are attached to you affect the length of the door transition. By running immediately instead of charging, only one enemy is attached to me, and as a result, I saved five additional frames at that door transition.
However, when I did my tests, they were faster for both times, so I'm not sure why.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Joined: 11/14/2005
Posts: 1057
Location: United States
Well, by perfection of Super Metroid I meant absolute frame perfection (like DK or SMB). I highly doubt this to be possible, mainly because new discoveries are likely to be found. Also the game has a lot of sub pixel positioning that can have an affect on many doors and elevators, these things can be overlooked easily (is ceres station even perfect yet?). The boss and enemy refills can never be perfect, you will have to either delay the fight to ensure better pickups or spend more time collecting the pickups because they fell into a worse position in the room. Finally, Snes9x has some lame lag issues which cause the game to lose time anyway (maybe this can be fixed eventually). Though it would be awesome to see an author with Phil's confidence submit a Super Metroid run that said "this is perfect, it cannot be beaten", that would make my day ;)
JXQ:
The solution is so simple: If you say the more drainer are attached to you the longer the door transition, then just kill them and get PB refills (the more the better) for a faster Maridia. Killing them costs exactly 1 frame delay to lift up the gun right after landing, and after that you can start the shinespark charge immediately and still get the refills in time, as said.
Result:
- Fastest possible door transition
- max possible PB refills (faster quests in the later progress)
- fastest possible reaching of the top morphball hole with the PB pack by using the spark
Not sure if there can be a reason to not use this strategy. :-)
Hero:
Yes, I agree then. It's just because of the refills, nothing else though. Sub-pixel managment is hard, but very well controlable nevertheless. Enemy knockbacks and enemy positions are always the same if you use the same moves in a room. Ok, the elevator randomness is probably also not manipulable, but this is a 1 frame issue. In the end of a any% run, I'm dare to say it can be optimized within a 100 frame margin towards the absolute perfection (ingame time).
And regarding Ceres, I think my escape in the any% TAS with a time of 49'15 is 100% perfect by now. I spent more than a week on it alone, and even though it's more than a half year old, I still can't see any possible improvements there.
And regarding Ceres, I think my escape in the any% TAS with a time of 49'15 is 100% perfect by now. I spent more than a week on it alone, and even though it's more than a half year old, I still can't see any possible improvements there.
What I'm more curious about: is it possible to reproduce such escape time while saving more frames in the pre-Ridley part? It seems that the doors operate on a rather strict frame rule, especially in the escape part. As in, they won't start opening if the door sprite is "disconnected" from the wall, and it seems that there are four different frames of door shaking animation, of which only one is good, and they're running on a global timer. I'll test it later.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
