I am Raglan, Lord of Darkness king of Destruction. Well, you will not be for too long now that Sky, the son of the Sky-lord, decided to do something about it.

Game Objectives

  • Emulator Used: Snes9x 1.43 v17 svn146
  • Takes Damage to Save Time
  • Manipulates Luck
  • Heavy Gltch Abuse

Dooty's Comments

This run uses the good rom of Sky Blazer in an attempt to improve the currently published run that uses a beta 2 rom. There's slightly differences in the game engine between the two versions and I'll try to describe them now;
  • The loading times are longer.
  • Sky lands one frame later after a jump.
  • The conversations can be skipped faster.
  • There's more lag during some stages.

Stage by Stage

The first difference, the longer loading times can be noticed right at the start of the game, my first input can only be given 91 frames later than it's possible on the beta rom.

Intro Stage

Taking damage on the first stage is not meant to save time, but to avoid wasting more time on the cutscene with Ashura. 89 frames slower.

Faltine's Woods

Eventhough this stage was redone 3 times, it's still slower than the current run, I'd say it's another difference on the engine, but it may sound like a lame excuse. 16 frames slower.

Temple Infernus

Finally a faster stage, thanks to a better positioning at the end of the boss fight. 22 frames faster.

Cliffs of Peril

Okay, there's no excuse to be slower on an auto scroller stage, but I couldn't avoid that extra frame of lag. 1 frame slower.

Tower of the Tarolisk

The different height of the last two spiked platforms made me lose a lot of time on this stage. 225 frames slower.

Gateway to the Eastern Plain

I think I can't brag about those frames saved on this auto scroller because I don't know how it was faster. 10 frames faster.

Petrolith Castle

It's not possible to jump while Sky is crouched under those crushing ceilings on the final version of the game, so, another slower stage. 14 frames slower.

The Falls of Torment

Better luck with those green slime balls to get what I wanted and this little stage is now a little faster. 38 frames faster.

Lair of Kharyon

Unlike on the current run, I don't get a magic replenishment after the previous stage, so I have to deal with what I have left, also one of the eels are on a different location. 85 frames slower.

The Sand Rivers of Shirol

Another stage that will not give me any magic refill at the end, so I must grab what I can before the next stage. 48 frames slower.

Gateway of Eternal Storms

I don't think it's fair to compare time on this stage as it's a completely different stage, and again, I'll not have any free magic refills at the end. 187 frames slower.

Storm Fortress of Kh'lar

Another stage that have a different layout, I don't know why they excluded the center platform on the boss fight, but it really pissed me off. 418 frames slower.

Fortress Shirol

Auto scrollers are supposed to last the same amount of time no matter what, and this one fortunately followed the rule. 0 frames faster/slower.

Caverns of Shirol

This stage have just one name on the final version, but they also eliminated a really useful glitch; the Zipping Left Through the Roof. 228 frames slower.

Dragonhil Forest

Finally a stage that's untouched on the final version, and since the same strategy was used, I got the same result. 0 frames faster/slower.

The Great Tower

Only the first part of this stage is different now, but I still don't get why they insist in making my life a lot more difficult. 286 frames slower.

Shrine of Destiny

The exact same input from Gateway to the Eastern Plain was used here, so what I said there still stands, it's just one frame slower than it was there. 9 frames faster.

Raglan's Citadel

The programmers reserved a really good surprise for me after so many difficult and slower stages; it's not needed to skip the conversation of the Old Man with the Sorceress anymore and the input can be ended a lot sooner. 2326 frames faster.

More Differences

I've spent quite some time to optimize this run, but overall it's slower than the previous run for a myriad of reasons, so only comparing time stage by stage doesn't make much sense. Here's a chart with the frames spent on loadings and talkings, it shows how different the two versions are;
LoadingTalkingTotalVersion
8732987818610Beta 2
10308657416882Final

Possible Improvements

What kind of improvement is that when there's still possible improvements left? well, I only found the speed address of Sky when the run was completed, so it may be possible to save time watching this address, and of course the use of a different rom made it quite difficult to keep the previous run's pace, but now with the same version it will be easier.

Useful Tools

TAS Movie Editor helped me immensely in the creation of this run.

Memory Addresses

  • 7EF340, 2u = Sky's Speed
  • 7EF830, 1u = Ifrit's Health
  • 7EF821, 1u = Beholder's Health
  • 7EF97x, 1u = Evil Wall's Health (x = E,D,F)
  • 7EF83x, 1u = Kharyon's Health (x = C,D,E,F)
  • 7EF824, 1u = Harpy's Health
  • 7EF821, 1u = Kappa's Health
  • 7EF821, 1u = Great Dragon's Health
  • 7EF820, 1u = Ashura's Health
  • 7EF83E, 1u = Raglan's Health

Suggested Screenshot

suggestions?

Nach: Since the old run was made with a beta version, and a final version exists, I'm disqualifying the old run. Therefore I am accepting this run as an improvement to an existing run.

Dacicus: Publishing


TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 15594
Location: 127.0.0.1
This topic is for the purpose of discussing #3458: Dooty's SNES Sky Blazer in 24:27.02
Guga
He/Him
Joined: 1/17/2012
Posts: 838
Location: Chile
Can somebody be so nice to do a temp encode?
Post subject: sky blazer: youtube
Joined: 3/18/2006
Posts: 971
Location: Great Britain
mklip2001
He/Him
Editor
Joined: 6/23/2009
Posts: 2227
Location: Georgia, USA
I haven't had a chance to watch this run yet, but from reading the submission comments, it seems that this run loses time overall except for the ending input! Because of this, I ask: how can we tell if the gameplay is actually better? Should this be treated like the Eggerland movie, where the slower time is due to better emulation / ROM choice?
Used to be a frequent submissions commenter. My new computer has had some issues running emulators, so I've been here more sporadically. Still haven't gotten around to actually TASing yet... I was going to improve Kid Dracula for GB. It seems I was beaten to it, though, with a recent awesome run by Hetfield90 and StarvinStruthers. (http://tasvideos.org/2928M.html.) Thanks to goofydylan8 for running Gargoyle's Quest 2 because I mentioned the game! (http://tasvideos.org/2001M.html) Thanks to feos and MESHUGGAH for taking up runs of Duck Tales 2 because of my old signature! Thanks also to Samsara for finishing a Treasure Master run. From the submission comments:
Shoutouts and thanks to mklip2001 for arguably being the nicest and most supportive person on the forums.
Skilled player (1326)
Joined: 9/7/2007
Posts: 1354
Location: U.S.
mklip: The reason its slower is because of version differences. Bablo used a beta dump which is actually against the rules (but, whatever). Dooty's run uses the proper dump, which sadly has some glitches removed from the beta run. So some glitches couldnt be done in this run sadly. On the subject of the movie itself. That was an amazing improvement to the previous run. There seems to also be more stylistic moves, which I enjoy a lot. Yes Vote! What will you do next Dooty?
Editor, Player (69)
Joined: 1/18/2008
Posts: 663
Enemies must explode, organic or not, in any real SNES platformer. It's a rule. This game doesn't break that rule. From the encode, it looks like there is some unnecessary slowing at parts. Hopefully I'm wrong.
true on twitch - lsnes windows builds 20230425 - the date this site is buried
Joined: 7/2/2007
Posts: 3960
Excellent work! The only thing that bothered me was that you used the same input for both bonus stages; it would have been nice if you'd changed things up for the second one. But that's hardly a dealbreaker. To my eyes this looked like an excellently-optimized run. The frequent use of Comet with minimal stops to recharge magic were quite nice, and there were noticeable route optimizations over the previous run. The only places where I couldn't say for sure if the run was as fast as possible were the horizontal sections with scrolling walls, but they looked pretty dang fast to me. Great job, Dooty! Easy yes vote.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
mklip2001
He/Him
Editor
Joined: 6/23/2009
Posts: 2227
Location: Georgia, USA
Alright, I managed to watch this. It's a shame that many stages ended up slower, but things look pretty optimized, especially the powerup manipulation. The only collections that looked suspicious were the ones in Lair of Kharyon (especially just before the boss fight). It's also unfortunate that you don't have the really quick boss kill anymore in the Storm Fortress of Kh'lar. Yes vote.
Used to be a frequent submissions commenter. My new computer has had some issues running emulators, so I've been here more sporadically. Still haven't gotten around to actually TASing yet... I was going to improve Kid Dracula for GB. It seems I was beaten to it, though, with a recent awesome run by Hetfield90 and StarvinStruthers. (http://tasvideos.org/2928M.html.) Thanks to goofydylan8 for running Gargoyle's Quest 2 because I mentioned the game! (http://tasvideos.org/2001M.html) Thanks to feos and MESHUGGAH for taking up runs of Duck Tales 2 because of my old signature! Thanks also to Samsara for finishing a Treasure Master run. From the submission comments:
Shoutouts and thanks to mklip2001 for arguably being the nicest and most supportive person on the forums.
Expert player (2468)
Joined: 6/2/2009
Posts: 1182
Location: Teresópolis - Rio de Janeiro - Brazil
Thank you all guys!
Sonikkustar wrote:
What will you do next Dooty?
I have a DKC2 project on hold for a very long time on the forum, I may resume or probably restart it soon.
mklip2001 wrote:
The only collections that looked suspicious were the ones in Lair of Kharyon (especially just before the boss fight)
They shouldn't have changed the location of that last eel, the stream forced me to use another Comet Flash, but I also started this stage with less magic, so I have to collect more refills than Bablo.
Derakon wrote:
The only places where I couldn't say for sure if the run was as fast as possible were the horizontal sections with scrolling walls
It's not possible to jump while crouched under those walls anymore, but it's still possible to reproduce Bablo's glitch in those stages with a little trick, so, those stages are indeed slower, but as fast as possible in this version.
I am old enough to know better, but not enough to do it.
Expert player (3644)
Joined: 11/9/2007
Posts: 375
Location: Varberg, Sweden
It'd be nice if you, for each stage, not just mentioned the total amount of frames gained/lost but also gave the exact (or at least estimated) # of frames gainedlost/lost due the version differences and the # of frames gained/lost in the areas in both versions that are directly comparable. As it's presented now, to someone who hasn't followed your progress during the making of the run it's not obvious at all that it's a technical improvement to the previous run, even though I'm almost certain it is.
feos wrote:
Only Aglar can improve this now.
Expert player (2468)
Joined: 6/2/2009
Posts: 1182
Location: Teresópolis - Rio de Janeiro - Brazil
Okay, I'll try to do a comparison chart including the conversations and loading times, but almost all stages are different in a way or another, so I'm not sure how it will work. In the submission text I'm only comparing the stages, from the first possible input, to the ending of the stage.
I am old enough to know better, but not enough to do it.
10RingBox
He/Him
Joined: 10/4/2010
Posts: 19
Was having 4 lives and 20 gems (4:20) intentional on the autoscrolling cliff stage? :P
Expert player (2468)
Joined: 6/2/2009
Posts: 1182
Location: Teresópolis - Rio de Janeiro - Brazil
Thank you for paying that much attention to my run :) but no, it was not, I was trying to collect all items and kill all enemies, but when I couldn't reach one gem I decided to avoid everything after that.
I am old enough to know better, but not enough to do it.
Player (37)
Joined: 2/16/2012
Posts: 282
Great job! I watched through it this afternoon, but I will take a look again soon to pick up on anything I missed. A couple things I remember off the top of my head: -In the beginning stage, I'm not sure if it's faster or slower to jump before that second hit or not. I'm not entirely sure, but I feel like the stun frames end as soon as you land so you can get back to ground-based quick zips sooner. I think there's still some momentum effects from doing a turn while running as well, but I'm not sure how it all plays out framewise. -I feel like there could be a slightly faster route using the middle path on the Sand River, but again I'm not certain in terms of frame-for-frame tradeoffs. -For both of the flying mini-games I think you can position yourself lower and still get hit at the same time by the mines. That should save a few frames, I think. I may steal some of those magic pickup points for use in my own regular runs ;)
Expert player (2468)
Joined: 6/2/2009
Posts: 1182
Location: Teresópolis - Rio de Janeiro - Brazil
Hi Omnigamer, I'll test the points you mentioned, thank you and welcome to TASvideos!
I am old enough to know better, but not enough to do it.
Player (224)
Joined: 10/17/2005
Posts: 399
You've got my yes vote. Also I can confirm that indeed this movie is more optimized than mine.
<adelikat> I've been quoted with worse
Expert player (2468)
Joined: 6/2/2009
Posts: 1182
Location: Teresópolis - Rio de Janeiro - Brazil
Thanks Bablo, I also must thank you for all the tricks used on my run. I did the tests;
Omnigamer wrote:
I'm not sure if it's faster or slower to jump before that second hit
That part is kinda tricky, one frame faster and I lose the chance to take damage on that last enemy, also, it's not possible to zip-jump at the left edge of ledges or on some platforms.
Omnigamer wrote:
For both of the flying mini-games I think you can position yourself lower
Hey, you just found out why I'm ten/nine frames faster than the previous run; One pixel lower and I would take damage ten/nine frames later, weird!
Omnigamer wrote:
the middle path on the Sand River (could be slightly faster)
There's no need to grab any walls on the middle path, but the luck manipulation there is a lot harder than it is on the upper path, so it will depend on how lucky you are.
I am old enough to know better, but not enough to do it.
Player (37)
Joined: 2/16/2012
Posts: 282
Thanks for checking those out then. I tried the flying stuff a bit for myself after I said it (in real-time) but the hitboxes make more sense to me now. Either Sky himself has a conical hitbox or the hitbox on gems/mines grow as they get closer to the screen (more likely). In any case, good work, I wasn't able to find any other parts that looked curious :)
Joined: 5/29/2004
Posts: 757
I am a bit concerned by the fact that the author openly admits there's possibly a lot of improvements due to only finding out about stuff at the end of the run. That alone makes a Yes vote falter hard. It looks great, but knowing there's serious potential left over makes it hard to just let that pass. I know how hard it is to make a TAS and how frustrating it is to find out at the end of something big, but in this case, it seems like serious potential through out and not like "We did an SM64 120 star run and 5 minutes before we found a 3 second improvement in the first 2 stars." where that'd be a hell of a lot more acceptable. I also understand the excitement to get this out when it's fresh and anticipated, but knowing there's a lot to go... it sets a dangerous precedent on such a short run. If it was over an hour that's a bit of a better tasting pill to swallow.. but still. I know this isn't the most popular thought considering how much positive there is on this run, and it was done very well and very entertaining.. but in this current moment.. that's not enough. [And this is also not like Super Metroid where there's countless new things being found out after the fact..] Mr. Kelly R. Flewin
Mr. Kelly R. Flewin Just another random gamer ---- <OmnipotentEntity> How do you people get bored in the span of 10 seconds? Worst ADD ever.
Player (224)
Joined: 10/17/2005
Posts: 399
Nothing new was found during the run, just the way to monitor the character's speed. So only very minor tweaks can be made based on this.
<adelikat> I've been quoted with worse
Editor, Player (69)
Joined: 6/22/2005
Posts: 1050
Archive.org collection Archive.org download Archive.org 10-bit YUV444 download Here are some potential screenshots for your voting pleasure while I work on the encodes. As always, feel free to suggest your own. EDIT: Expired image link removed.
Current Projects: TAS: Wizards & Warriors III.
Expert player (2468)
Joined: 6/2/2009
Posts: 1182
Location: Teresópolis - Rio de Janeiro - Brazil
I like the second one, with Sky stuck on the floor.
I am old enough to know better, but not enough to do it.
Post subject: Movie published
TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 15594
Location: 127.0.0.1
This movie has been published. The posts before this message apply to the submission, and posts after this message apply to the published movie. ---- [1983] SNES Skyblazer by Dooty in 24:27.02