Emulator used: Gens 9f & Gens 9f-modified (plays back correctly with either)
Aims for fastest time
Takes damage to save time
Abuses programming errors
Manipulates luck
My goal priority was as follows:
Fastest level times possible
Fastest overall time possible
Interesting to watch during waiting sequences
With these goals, I strive for the fastest in-game clock time for each act. This means I will not be slowing down near the end in order to collect a smaller bonus (which could be shorter in overall time). By the same token, I will not optimize only to the game clock, but to the frame counter as well. Frames saved are just as important whether or not it makes a difference on the game clock of a particular level. I occasionally bent this rule (for certain bonus tags, or at the end of the run) on a case-by-case basis.
Level
Time
Frames Improved
TSC Record
Green Hill 1
0:24
1 vs. Xebra-v3
0:25
Green Hill 2
0:17
35 vs. Xebra-v3
0:17
Green Hill 3
0:31
11 vs. Xebra-v3
0:34
Marble 1
0:46
113 vs. Xebra-v2
0:54
Marble 2
0:53
21 vs. Xebra-v2
1:02
Marble 3
1:13
73 vs. Xebra-v2
1:19
Spring Yard 1
0:22
43 vs. Xebra-v2
0:23
Spring Yard 2
0:28
37 vs. Xebra-v2
0:31
Spring Yard 3
0:59
156 vs. Xebra-v2
1:05
Labyrinth 1
0:40
58 vs. Xebra-v2
0:47
Labyrinth 2
0:52
N/A
0:58
Labyrinth 3
1:08
N/A
1:28
Star Light 1
0:22
160 vs. HCc-SL1
0:26
Star Light 2
0:17
18 vs. Ouzo-SL2
0:22
Star Light 3
0:44
132 vs. Ouzo-SL3
0:59
Scrap Brain 1
0:32
N/A
0:39
Scrap Brain 2
0:43
N/A
0:46
Scrap Brain 3
0:18
31 vs. HCc-SB3
0:20
Final Zone
1:13
N/A
1:13
TSC Records are accurate as of date of submission.
Points of Interest:
Green Hill 3 - Robotnik cannot be hit 8 consecutive times. This is because during the first hits, he is moving downward, making it appear as though Sonic is bouncing high enough on his own to land another hit. Once Robotnik moves left instead of down, Sonic doesn't have the necessary height, and must land once to deliver the final blow.
Marble 1 - There are two spots where Sonic jumps late, although it's difficult to spot at 100% speed. These pauses are done to manipulate the frame that the moving platform will emerge from the wall later - if it comes out too late, Sonic can't jump directly onto the spiked chandelier above, and must wait two seconds to proceed. Additionally, the number of rings Sonic has when jumping onto the lava can affect which way he bounces, which is why rings are avoided near the pistons earlier.
Marble 3 - A block is pushed with extra speed. This happens when Sonic is a certain height above the block and pushing towards it. In normal circumstances, this technique is slower, because Sonic can't jump this low, but here, the block is both at a perfect height and a short enough distance to push that it saves time. At the boss fight, falling into the lava allows the Boss to appear two frames sooner than if the lava was jumped.
Spring Yard 3 - Robotnik cannot be hit when at the height of his flying pattern, so there is a slight delay between the 6th and 7th hit.
Labyrinth 2 - The extra bonus tags are located about a screen's length before the final sign. Jumping to get them costs 5 frames, and they are only worth 300 points together (as opposed to the 11000 points I do get), so they were skipped. The jumping at the end of the level is just to show that they aren't there.
Labyrinth 3 - This was the most frustrating of all the levels for me, but it turned out great. I first did a slightly-tool-assisted preliminary run (to have something to run against; this seems to help me at this game), where I found out I didn't need to stop for air bubbles. When I actually did the fully optimized run, I was running out of air. Turns out that in my preliminary run, I jumped an extra time at the beginning of the water, extending my air, but this wasted about 20 frames. Thus began the quest to optimize this level to death. The names of my backups tell the story best:
L3 drown.gmv
L3 even closer but still drowning.gmv
L3 disgusting how close I am.gmv
L3 stupid (bleep)ing water.gmv
L3 SWEET GLORIOUS WATER.gmv
Sonic jumps out of the water at the absolute last available frame before he would drown. This makes for an exciting watch!
Scrap Brain 1 - The basics of this glitch involve ducking on the platform to enter the wall. Once inside, pressing left will blast sonic to the right at a ridiculous velocity the screen can't hope to match. Because Sonic is going so fast, he can pass through certain walls / obstacles. By the time the screen finds Sonic, he has already ran through an area that normally would require a lot of pausing, due to fire and crushers.
Scrap Brain 2 - I was unable to get to the section with the sawblades fast enough to make it under all of them, so taking the hit on the second one was the fastest choice (as I can still take advantage of the third conveyor belt moving to the right).
Final Zone - There are two different ways to "finish" the game, each of them getting a different input time, but both fading out the screen at the same time:
18.26.05 - Sonic will run far enough to fade the screen at the earliest time
18.28.15 - Input continues long enough for Sonic to hit Robotnik one last time (not necessary, but sort of anti-climatic if it's left out).
Thus, the input could be terminated two seconds earlier and still make it to the ending.
Suggested description: Somewhat fast bouncing around. The lack of a spin-dash makes the strategy a bit different than the other Sonic runs. Completed without collecting any emeralds.
Joined: 5/17/2004
Posts: 106
Location: Göteborg, Sweden
Very good run, on a long-cursed and delayed game. Heartily agree about the end-of-movie call: the important thing is the speed & quality of the actual run, not the size of the movie file.
Also, I'm not quite sure interpersonal conflict really belongs in a run submission text, but meh.
Joined: 4/20/2005
Posts: 2161
Location: Norrköping, Sweden
Allow me to throw yet another YES vote on this. :) Very nice work, my favourite part was on the third stage, where you practically flew above the entire level. Pure awesomeness. Three, umm I mean, two thumbs up! :)
Joined: 6/21/2004
Posts: 247
Location: New Hampshire
Easy yes. Great work, JXQ!
As much of a jerk as xebra was in the planning thread, I'd have to agree. IMO it'd be best to just let it go.... grudges aren't healthy :)
I don't intend to hold a grudge. I included that section in the submission because it temporarily affected the quality of the run and how much motivation I had to work on it - the submission guidelines say to "describe the making of the movie." I also thought my message of not being bullied was important to voice.
Thanks for the comments guys!
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Joined: 4/21/2004
Posts: 3517
Location: Stockholm, Sweden
Hurry! ^_^
JXQ my man, what can I say, I just watched and enjoyed every frame you played and this is truly a shhweet Christmas present for us. Im MIGHTY impressed how fast you completed this magnificent run, a true role model for us kids who never gave up despite what happened in the Sonic thread :) JXQ rocks! Of course voting Yesh.
Very very good. I have no complaints. You have done an excellent job. There is no doubt this deserves a star.
Excellent job especially on the Labyrinth and Scrap Brain Zones. Those were always the hardest to get through and it's awesome to finally see them done perfectly.
Holy TAS Batman! This Sonic run is amazing!! Very well done, I've never seen the first three stages done that fast. This is truly a work of art, and done in such short time too. Definate YES.
Guybrush: "I'm selling these fine leather jackets."
Wally: "Really?"
Guybrush: "No. I 'm lying."
Wally: "In that case, I don't want one!"
Currently working on:
Nothing at the moment.
Joined: 11/29/2005
Posts: 317
Location: Sao Paulo - Brazil
If I could vote, I would certainly vote YES..
But I can't vote =/ (am I a lurker?)..
So.. Impressive ones: When you fly over the stage, when you go up in Labirynth, last stage..
Very nice one!
As always, your personal attacks warm my heart. You no doubt still have negative feelings towards me, so I encourage you to further explore your grievances in this thread, or whichever public forum you deem appropriate.
On to the run:
No surprises here, I still think your playing style is spastic to the extreme. Considering you lose time against existing WIPs in a number of places, I'm surprised you weren't willing to make greater allowances in order to make your play look smoother and more effortless. Just as an example, if you have to tap back and forth in quick succession, it's very difficult to see you doing it if you are balled up, and thus looks a lot less choppy than otherwise.
You did not discover the optimal path for Marble 1, Marble 2, or Spring Yard 3, at the very least. You also missed a minor collision/boost in Labyrinth 1, though by itself it's not enough to improve the clock. The other levels use paths better than anything I have knowledge of.
I'm curious, why do you single out my Spring Yard 1 as being excellent when you improved comparatively less on some other levels?
Other than the errors I have mentioned, the choppiness of Sonic's movements, and your inability to bury hatchets, the run is good.
Thanks to everyone for the comments!
Questions to answer (for xebra, since no one else asked questions):
There were two levels where your WIPs stayed on the screen after hitting the sign, and I included those in the "frames gained" column. The two levels were Green Hill 2 and Spring Yard 1, in each of these I gained only one frame before the sign; the rest were the result of choosing to not stay on the screen until after the score counting screen started. That's why I singled Spring Yard 1 as good, along with the entire v3 WIP you did.
As far as smoother / effortless play - to me, the run doesn't look more playable by a human if it's smoother, but still perfectly cutting corners, making jumps, and so forth. It also is no secret that my weak point of making runs is the aesthetics. For example, in the Super Demo World, anything that looks remotely cool was probably done by Fabian. I was the frame hunter, and that's how I do these runs the best. (Edit; trying to explain better) I also never cared for the idea of optimizing as much as possible, and then picking and choosing which optimizations to discard in trade off for a smoother watch. I would feel like I was wasting frames in areas that I knew could be done faster.
As far as losing time against existing WIPs, I tried hard to avoid this. Perhaps I don't have these WIPs you speak of? For each level that I had a run to go against, I did the same thing I was doing at the beginning of the run in the forums; making a bunch of checkpoints and comparing times to make sure I wasn't falling behind for part of the level, even if I gained some frames earlier. I'm not saying this couldn't have happened, but it was something I tried to minimize very much.
I'm curious, do you know off hand how much these improvements would add up to?
One more question, do you think you'll ever continue work on your run? Competition breeds perfection, they say.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
I think this should get a star if you didn't get hit by that ball in Scape Brain 2. I know what did was the fastest, but new viewers might be turned off by it. (Since star'd runs are intended to get new viewers started)
And...
Next: Super Mario Bros. 3 (no warps)
Can't wait =)