Journey to silius

Edit history:

ktwo: 2009-11-18 11:39:07 am

2009-07-31 02:45:49 pm

It's a shame that this awesome game (well, at least the music is totally awesome) doesn't have a speedrun yet. I'll see if I can change that. There are a couple of areas that I don't feel confident in, but today I managed to get my first no-death run (always a nice mental barrier to break). This makes the project much more likely to fly, so I thought I'd make a post about it. I clocked the run at around 14.40 (pal-version). I used (read: tried to use) most speedrun tricks that I have developed so far, but the execution was a bit off in many places. I hope to be able to shave off around 30 seconds, maybe even get it under 14 min (but that's pure speculation at this point).

Troublespots:
Preboss 1 - I hate those robo-grasshoppers (or whatever), if anyone has a good strategy to wipe them out quickly, feel free to share it with me...
Level 2 - the first shaft down is tricky, I don't have a good strategy to get past the turrets shooting in all directions. The last turret of that kind (right before the boss) also gives me serious headache.
Level 3 - everything about it, getting lucky, keeping the health and surviving the last corridor while keeping the speed up (as much as possible at least)
Boss 3 - there are lots of ways to take it out, but all speedy ones that I've found involve taking a fair amount of hits

Level 4 actually isn't that bad. It's hard, but the luck factor is insignificant and the health isn't an as critical issue as in level 3, so once the layout is memorized, it's all about a tight execution. Level 5 is just a formality.

Weapon strategies:
Preboss 1 - normal gun
Boss 1 - normal gun
Preboss 2 - normal gun
Boss 2 - missiles
Preboss 3 - missiles
Boss 3 - missiles
Preboss 4 - laser
Boss 4 - grenades
Preboss 5 (the ship) - 5 grenades on the left side and then normal gun. Normal gun on the right side as well.
Boss 5 (alien) - remaining grenades and then normal gun
I use the normal gun on all levels (switch to it immediately after the boss fights). I haven't found an area where using the special weapons would save more time than it takes to switch in the menu.

Well, if there is anyone out there with knowledge about this title, I would love hear his/hers thoughts about a run. Should anyone be interested, I can also post the run I made on youtube, even though it looks pretty rough...

Edit (181109):
--------------
Modified the Preboss 5 strategy.

It's a shame that this awesome game (well, at least the music is totally awesome) doesn't have a speedrun yet. I'll see if I can change that. There are a couple of areas that I don't feel confident in, but today I managed to get my first no-death run (always a nice mental barrier to break). This makes the project much more likely to fly, so I thought I'd make a post about it. I clocked the run at around 14.40 (pal-version). I used (read: tried to use) most speedrun tricks that I have developed so far, but the execution was a bit off in many places. I hope to be able to shave off around 30 seconds, maybe even get it under 14 min (but that's pure speculation at this point).

Troublespots:
Preboss 1 - I hate those robo-grasshoppers (or whatever), if anyone has a good strategy to wipe them out quickly, feel free to share it with me...
Level 2 - the first shaft down is tricky, I don't have a good strategy to get past the turrets shooting in all directions. The last turret of that kind (right before the boss) also gives me serious headache.
Level 3 - everything about it, getting lucky, keeping the health and surviving the last corridor while keeping the speed up (as much as possible at least)
Boss 3 - there are lots of ways to take it out, but all speedy ones that I've found involve taking a fair amount of hits

Level 4 actually isn't that bad. It's hard, but the luck factor is insignificant and the health isn't an as critical issue as in level 3, so once the layout is memorized, it's all about a tight execution. Level 5 is just a formality.

Weapon strategies:
Preboss 1 - normal gun
Boss 1 - normal gun
Preboss 2 - normal gun
Boss 2 - missiles
Preboss 3 - missiles
Boss 3 - missiles
Preboss 4 - laser
Boss 4 - grenades
Preboss 5 (the ship) - 5 grenades on the left side and then normal gun. Normal gun on the right side as well.
Boss 5 (alien) - remaining grenades and then normal gun
I use the normal gun on all levels (switch to it immediately after the boss fights). I haven't found an area where using the special weapons would save more time than it takes to switch in the menu.

Well, if there is anyone out there with knowledge about this title, I would love hear his/hers thoughts about a run. Should anyone be interested, I can also post the run I made on youtube, even though it looks pretty rough...

Edit (181109):
--------------
Modified the Preboss 5 strategy.

Edit history:

ktwo: 2009-11-18 11:41:27 am

2009-08-11 04:22:00 pm

Sorry for double posting, but I thought I'd share a wip from this w-e. Now it's at least possible to see what I'm trying to achieve.
Overall, this turns out to be pretty tricky for me despite the humble length of the game. I guess I'm better at planning and not that good at giving a lot of input quickly. Anyways, I continue playing for now and am down to around 14.15. I lost 10 seconds for no obvious reason on the last alien boss and 15 seconds (very approximately) in other spots. Next goal is to get under 14 min, which I now realize should be a strict minimum requirement before considering submission (maybe even lower, we'll see).
(part 1)
(part 2)

For your reference, all hits I take in the back are intentional and all hits except one (against the big canon in level 2) that I take upfront like a man are execution mistakes.
If there are any potential viewers out there, I would very much appreciate frank feedback (either here or at youtube). In many places I have pretty cautious strategies, so more time than the 15 seconds above can theoretically be cut out. I'm not overly motivated to go for gutsier moves, but I prefer to stop this project now should that be the case instead of ending up with an insufficient finished product.

Edit (280809) :
---------------
New pb around 14.10. It still sucks.

Edit (290809) :
---------------
14.02. Pretty good, but still some rough corners. If I manage to improve this, I will likely ask it to be verified. I'll continue working on it this afternoon (provided the Scandinavian weather will stay shitty).

Edit (310809) :
---------------
After a marathon w-e with an insane amount of tries (well, actually from Friday to Monday), I finally got a decent run. Still plenty of things to improve, but I'm not too embarrased over it either. I've requested to get it verifyed.

Edit (181109) :
---------------
Removed the youtube clips since the final version is now available on sda.

Sorry for double posting, but I thought I'd share a wip from this w-e. Now it's at least possible to see what I'm trying to achieve.
Overall, this turns out to be pretty tricky for me despite the humble length of the game. I guess I'm better at planning and not that good at giving a lot of input quickly. Anyways, I continue playing for now and am down to around 14.15. I lost 10 seconds for no obvious reason on the last alien boss and 15 seconds (very approximately) in other spots. Next goal is to get under 14 min, which I now realize should be a strict minimum requirement before considering submission (maybe even lower, we'll see).
[s]http://www.youtube.com/watch?v=iYZx4mC2tgo[/s] (part 1)
[s]http://www.youtube.com/watch?v=ahqWYTa5bY0[/s] (part 2)

For your reference, all hits I take in the back are intentional and all hits except one (against the big canon in level 2) that I take upfront like a man are execution mistakes.
If there are any potential viewers out there, I would very much appreciate frank feedback (either here or at youtube). In many places I have pretty cautious strategies, so more time than the 15 seconds above can theoretically be cut out. I'm not overly motivated to go for gutsier moves, but I prefer to stop this project now should that be the case instead of ending up with an insufficient finished product.

Edit (280809) :
---------------
New pb around 14.10. It still sucks.

Edit (290809) :
---------------
14.02. Pretty good, but still some rough corners. If I manage to improve this, I will likely ask it to be verified. I'll continue working on it this afternoon (provided the Scandinavian weather will stay shitty).

Edit (310809) :
---------------
After a marathon w-e with an insane amount of tries (well, actually from Friday to Monday), I finally got a decent run. Still plenty of things to improve, but I'm not too embarrased over it either. I've requested to get it verifyed.

Edit (181109) :
---------------
Removed the youtube clips since the final version is now available on sda.

ConHuevos

2012-05-05 06:45:10 pm

Ciento Dos Huevos

Sorry for bumping an old thread, but I've begun attempting to speed run this game. Been studying the TAS and put in some practice while I was streaming at twitch. Haven't made a full run yet, just practiced the 1st and 2nd stage a bit until finals are over, then I'll have more time to dedicate. I have the actual cart, but don't have a capture card (yet), so I'm just practicing on emu until I'm ready to do a full SS.

http://www.twitch.tv/ilikeyourbootybutlmnotgay/b/317222519

Edit history:

Reaif: 2012-10-20 03:49:06 am

2012-10-20 03:48:23 am

Yay! Bumping a thread that has isn't as dead as it was the last time it was bumped!

Anyhow, I've been making attempts at this lately and I got a pretty good one that I feel quite good about.

The final time for this one is 12:20.4, going by first frame input and last frame hit on the boss. Overall, I like how I did on this, but I'll try a few more sessions of attempts to get a better run. I'm not sure a better time will come around, but we'll see how it goes.

BTW, this is the NTSC US version.

i wish i was ktwo

2012-10-20 04:39:29 am

hi reaif

wow, it's really cool to see renewed interest in this game =) thanks for the cool run! There were some really nice moments in there, like all boss fights and the level 3 clutch =D I'd love to see an even better run from you, especially of stages 3 and 4!

Here's hoping you keep at it! Interest hasn't died, even though this thread pretty much has =P

Svenne

Edit history:

Svenne: 2012-10-20 06:18:12 am

Svenne: 2012-10-20 06:17:45 am

2012-10-20 06:17:34 am

Smörgåsbord

No, interest is strong in this one!

2012-10-20 09:54:14 am

Ha ha thanks guys!

I know what you mean by stages 3 and 4. Those were a little rough. That being said, I think 12:00 is possible, though it would really take a miracle to pull it off.

Boxfat

2012-10-21 01:25:56 pm

Arbitrary forum emu

Nice! I've learned some things from this run and might see if I can challenge this run some day. Hope you can get a better level 3/4!

2012-10-21 08:35:37 pm

Sounds good. This game could use some more speedrunnin' love.

Countneko

2013-01-16 06:11:19 pm

Does anyone have strategies for killing the three hopping things that come out of the helicopter boss in level 1? I don't know how to exploit their AI but it seems they always react to your shots and the run I saw just destroys them before they even hit the player. Is it just luck what direction they go in when they first come out or is there another element to it?

Edit history:

Reaif: 2013-01-16 08:48:54 pm

2013-01-16 08:46:29 pm

You can see my strategy in this video. It's very consistent and, I think, fully justifies switching weapons for the fight.

Pretty much I just jump right before they hit the ground and blast them. They always do a high jump when they first hit the ground. That might have a qualifier or two, but when I stand about that far away, I can hit that strat about 95% of the time.

Countneko

2013-01-19 11:24:15 pm

That should make it a lot easier, at least for that bit. Though I'm having trouble getting 2 missiles in the air on the second boss to hit it's core, they're frequently out of alignment. I'm probably just not mashing the button hard enough or mistiming it.

Rodiath

Edit history:

Rodiath: 2013-03-04 08:21:01 pm

2013-03-03 09:18:50 pm

So how do you make that turret at the end of level 1 not shoot as you approach it? Does timing your shots just right make the first shots not come out for some reason`? Whenever I try to approach it the way you do I get shot on the way :X

Edit: Maybe it has to do with the fact that I'm playing on emu? Not as much sprite lag and such? if so that would kind of suck. Also wondering about the part in level 1 on the brick tiles, where you lay down, shoot the mine across the gap, jump over and fire three shots at the missile launcher up on the ledge. I know the missile launcher has 3 hp, but whenever I do that strat it seems one of the hits never registers, which means often I'll get hit once as I get up to shoot it the last time.

2013-03-18 05:43:40 pm

Quote from Rodiath:

So how do you make that turret at the end of level 1 not shoot as you approach it? Does timing your shots just right make the first shots not come out for some reason`? Whenever I try to approach it the way you do I get shot on the way :X

Edit: Maybe it has to do with the fact that I'm playing on emu? Not as much sprite lag and such? if so that would kind of suck. Also wondering about the part in level 1 on the brick tiles, where you lay down, shoot the mine across the gap, jump over and fire three shots at the missile launcher up on the ledge. I know the missile launcher has 3 hp, but whenever I do that strat it seems one of the hits never registers, which means often I'll get hit once as I get up to shoot it the last time.

Hmm, maybe it's a bit late to answer but here it is anyway. :p

The cannon shoots, I'm pretty sure it's just bad quality on the video.

Don't know why you have problem with the missile launcher, its hitbox is pretty big but try to shoot as early as possible. Remember also that the bullet also can be fired in a way that it's almost inside the ground giving you a few more frames to perform this.

[quote="Rodiath"]So how do you make that turret at the end of level 1 not shoot as you approach it? Does timing your shots just right make the first shots not come out for some reason`? Whenever I try to approach it the way you do I get shot on the way :X

Edit: Maybe it has to do with the fact that I'm playing on emu? Not as much sprite lag and such? if so that would kind of suck. Also wondering about the part in level 1 on the brick tiles, where you lay down, shoot the mine across the gap, jump over and fire three shots at the missile launcher up on the ledge. I know the missile launcher has 3 hp, but whenever I do that strat it seems one of the hits never registers, which means often I'll get hit once as I get up to shoot it the last time.[/quote]
Hmm, maybe it's a bit late to answer but here it is anyway. :p

The cannon shoots, I'm pretty sure it's just bad quality on the video.

Don't know why you have problem with the missile launcher, its hitbox is pretty big but try to shoot as early as possible. Remember also that the bullet also can be fired in a way that it's almost inside the ground giving you a few more frames to perform this.

2013-03-23 12:48:58 pm

Well there doesn't seem to be much going on with this game so I might just as well post my 12:12 run here. Maybe someone will find it interesting. Smiley

There is one thing that I would like to know more about though. Once when I practised (on emu) on the first boss, after I killed it, the health regen started during its death animation instead of after. Just as the victory fanfare started it got cut off and the next level started. I would like to know more about this since it's a huge time saver but I have no idea what triggered it and it only happened once. It's not in the TAS either. So I'm wondering if this has happened to anyone here?

2013-03-24 03:28:39 am

The quick health regen you mentioned never happened to me (and I'm pretty sure I would have noticed if it did). If it can be controlled somehow, it sounds like a big time saver. The best reported time for this game that I know of is 11:17 (rta, so 11:12 SDA-timing) by Hotarubi. Trying to get close to that time would be pretty intimidating, so finding new tricks like this would be a big help.

If you don't get a reply here, you can also try to contact one of the TASers and see if they are willing to look into it (klmz answered a few of my questions when I did my run).

2013-03-24 07:26:33 am

Damn that's pretty fast, you know where I can watch that video? I looked but couldn't find any.

2013-03-24 09:20:42 am

Unfortunately no. It's listed on a Japanese site that does not host videos and there are no external links either: http://www27.atwiki.jp/timeattack/pages/50.html

I know that the legitness of just a number without a video on a website can be questioned (and it's kind of a pity for the community too). However, considering how often Hotarubi's name has shown up next to reports of amazing speedruns (sometimes with videos), I think it's likely that the time is valid.

Even without a video, there are a few interesting things to say about that time. First of all, if the time of my run (13:48) were converted to a comparable time for ntsc (with the 5/6-rule, which I believe to be correct for this game, and offset for the difference between SDA- and rta-timing), Hotarubi's run would correspond to ~18 seconds of better play quality. I estimate that if I had made no mistakes, had gotten perfect luck and implemented the riskier strats that I knew of at the time, that would still have fallen just short... There are of course a few uncertainties in here, but a time of 11:17 (rta) must have been extremely good and/or used some additional time savers that I didn't (and still don't) know of. The second interesting thing to note is that there are a few more times listed from the same date, indicating some kind of race or competition having taken place. It'd be surprising if a "perfectly" executed run was achieved under such conditions (but it's of course possible).

Unfortunately no. It's listed on a Japanese site that does not host videos and there are no external links either: http://www27.atwiki.jp/timeattack/pages/50.html

I know that the legitness of just a number without a video on a website can be questioned (and it's kind of a pity for the community too). However, considering how often Hotarubi's name has shown up next to reports of amazing speedruns (sometimes with videos), I think it's likely that the time is valid.

Even without a video, there are a few interesting things to say about that time. First of all, if the time of my run (13:48) were converted to a comparable time for ntsc (with the 5/6-rule, which I believe to be correct for this game, and offset for the difference between SDA- and rta-timing), Hotarubi's run would correspond to ~18 seconds of better play quality. I estimate that if I had made no mistakes, had gotten perfect luck and implemented the riskier strats that I knew of at the time, that would still have fallen just short... There are of course a few uncertainties in here, but a time of 11:17 (rta) must have been extremely good and/or used some additional time savers that I didn't (and still don't) know of. The second interesting thing to note is that there are a few more times listed from the same date, indicating some kind of race or competition having taken place. It'd be surprising if a "perfectly" executed run was achieved under such conditions (but it's of course possible).

2013-03-24 09:56:37 am

That's too bad really. So what are you saying here, that we shouldn't take this time as 100% true? I wouldn't take an unrecorded run as a wr but are these guys know to "exaggerate" their times? Personally I feel that this time is something to try to strive for.

Also since I'm talking to "the" ktwo ;D, would you mind giving me a few pointers? Just like you I tried to convert the time to NTSC and that would make it about 40sec faster than mine. I seriously don't know where that time would come from. I know the mistakes are pretty obvious and could have some tighter 1,2 and 5 boss fights, but 40 sec...
I would also like to know how you did manipulate the first boss little jumpers. I've tried over and over and they simple won't behave like in your run. So any help would be appreciated. Smiley

2013-03-25 10:08:11 am

I personally think it's very likely that a time reported by Hotarubi is legit, but without a verifiable video it's obviously impossible to know for sure. Sadly, the Japanese speedrun community doesn't have the same tradition as the western counterpart of uploading videos of their speedruns. They don't seem to have particularly good platforms for hosting their videos though, so that might be an explanation and it's not necessarily a sign of cheating/bad will. It looks like there are a few ways to contact Hotarubi (http://com.nicovideo.jp/community/co1254961), so you can always try and see if he responds if you reach out to him. I'd be very interested in watching his run as well.

I watched your run.
Stage 1:
* I don't think the hit at 0:23 saved time compared to my strat.
* It looks like you stopped for a split second at 0:49. If you jump at the correct spot, you don't have to stop.
* Other than those two things, I noticed you had two minor strats that I chose to exclude in my run for the sake of consistency.
Boss: I don't know how to manipulate the little jumpers at the boss. To be honest, it was so luck-dependent that I didn't even bother resetting if the jumpers bounced around a bit. I was just a bit lucky to get a good pattern in that attempt and then managed to keep it together for the rest of the run. A few posts above, Reaif describes a different way to approach the boss battle. Maybe that can help you save time consistently? By the way, it took you 8 jumps to take out the helicopter (5 for me). You probably lost 5-6 seconds here in total.

Stage 2:
* You got stuck twice on the first set of stairs. I got stuck once. It's possible to clear it without getting stuck at all (a bit tricky though).
* You stopped the forward momentum waiting for the laser to pass by. I don't remember the details anymore, but I think there are two patterns for that laser cannon and consistent ways to get through without stopping for both patterns.
* A bunch of hits in first shaft down. That section is hard though, but it probably cost 2-3 seconds.
* You can duck under the first shot from the "big cannon" (before the first elevator). If you jump, time that could have been spent filling it with lead was instead spent in the air.
* You stopped to dodge the bullets from the "turrent that shoots in all directions" between the first and the second elevator. If you set up the screen position differently (Jay slightly more advanced on the screen), you can get through without stopping.
* A missed jump right before the pre-boss.
Pre-boss: Missed a few shots and had to jump an extra time or two.
Boss: It looks like you started jumping a bit late. I tried to start jumping as soon as the boss' "backpack" was fully visible and then immediately jumped again each time I landed. I remember the timing was a bit tight for this, but it wasn't too hard to be consistent at not getting hit.

Without having done any testing at all, I'd still say that pretty much all of the things I mentioned for stage 2 look like around one second mistakes or more. For the rest of the run, just ask yourself every time you got hit or had to stop if it was necessary. I think you will see how it all adds up to quite a bit of time that could have been saved. I should also say that in stage 3 (and 4), you need to think about Jay's position on the screen. For example, if you take the hits in the back between elevator one and two, Jay's sprite will advance slightly on the screen and you can then get through the laser barriers without stopping.

As for some ideas for improving upon the strats I used, it would be good with a fresh pair of eyes looking at the last section of stage 3 and also to look at the last boss. My method for the last boss was quite different from the tas, so more testing could possibly result in finding a faster way than mine.

I personally think it's very likely that a time reported by Hotarubi is legit, but without a verifiable video it's obviously impossible to know for sure. Sadly, the Japanese speedrun community doesn't have the same tradition as the western counterpart of uploading videos of their speedruns. They don't seem to have particularly good platforms for hosting their videos though, so that might be an explanation and it's not necessarily a sign of cheating/bad will. It looks like there are a few ways to contact Hotarubi (http://com.nicovideo.jp/community/co1254961), so you can always try and see if he responds if you reach out to him. I'd be very interested in watching his run as well.

I watched your run.
Stage 1:
* I don't think the hit at 0:23 saved time compared to my strat. 
* It looks like you stopped for a split second at 0:49. If you jump at the correct spot, you don't have to stop. 
* Other than those two things, I noticed you had two minor strats that I chose to exclude in my run for the sake of consistency. 
Boss: I don't know how to manipulate the little jumpers at the boss. To be honest, it was so luck-dependent that I didn't even bother resetting if the jumpers bounced around a bit. I was just a bit lucky to get a good pattern in that attempt and then managed to keep it together for the rest of the run. A few posts above, Reaif describes a different way to approach the boss battle. Maybe that can help you save time consistently? By the way, it took you 8 jumps to take out the helicopter (5 for me). You probably lost 5-6 seconds here in total.

Stage 2:
* You got stuck twice on the first set of stairs. I got stuck once. It's possible to clear it without getting stuck at all (a bit tricky though).
* You stopped the forward momentum waiting for the laser to pass by. I don't remember the details anymore, but I think there are two patterns for that laser cannon and consistent ways to get through without stopping for both patterns.
* A bunch of hits in first shaft down. That section is hard though, but it probably cost 2-3 seconds. 
* You can duck under the first shot from the "big cannon" (before the first elevator). If you jump, time that could have been spent filling it with lead was instead spent in the air. 
* You stopped to dodge the bullets from the "turrent that shoots in all directions" between the first and the second elevator. If you set up the screen position differently (Jay slightly more advanced on the screen), you can get through without stopping. 
* A missed jump right before the pre-boss.
Pre-boss: Missed a few shots and had to jump an extra time or two.
Boss: It looks like you started jumping a bit late. I tried to start jumping as soon as the boss' "backpack" was fully visible and then immediately jumped again each time I landed. I remember the timing was a bit tight for this, but it wasn't too hard to be consistent at not getting hit.

Without having done any testing at all, I'd still say that pretty much all of the things I mentioned for stage 2 look like around one second mistakes or more. For the rest of the run, just ask yourself every time you got hit or had to stop if it was necessary. I think you will see how it all adds up to quite a bit of time that could have been saved. I should also say that in stage 3 (and 4), you need to think about Jay's position on the screen. For example, if you take the hits in the back between elevator one and two, Jay's sprite will advance slightly on the screen and you can then get through the laser barriers without stopping.

As for some ideas for improving upon the strats I used, it would be good with a fresh pair of eyes looking at the last section of stage 3 and also to look at the last boss. My method for the last boss was quite different from the tas, so more testing could possibly result in finding a faster way than mine.