Posts for HappyLee


Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Mars608 Before 2019, he was known as an SMB TASer. But this year he really pushed Contra TASes to their limits, along with aiqiyou. We would have submitted at least two SMB TASes in 2019 if it weren't for my absence (sorry for that).
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Just finished watching this (sorry for the late comment). A beautiful TAS! I love all your entertainment choices. warmCabin, you're a talented TASer with good taste, and have put lots of efforts into your work. Thank you!
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Post subject: Re: #6514: Soig's NES Contra Force in 10:27.72
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
EZGames69 wrote:
In the case here, 6 seconds in the first stage is a very significant improvement, I do not believe that all 6 of those seconds would be erased simply from RNG changes.
If 6 seconds are considered to be very significant, this TAS improves GlitchMan's published TAS for 2 minutes, and that should be super duper significant, yet nobody ever talks about that.
ThunderAxe31 wrote:
I agree that the text of the rule could be misleading. Yet, we have at least one precedent of a submission being rejected just for being slower than RTA during a single boss fight alone: #5619: Mitjitsu's Saturn Astal in 20:29.92
The precedent is slower than existing RTA runs during a boss fight. That to me just sounds like a pretty big mistake, because TASers should at least consider and compare RTA strategies when making the TAS. But Soig's TAS is slower than a not-previously-shared WIP, and I don't think the two examples can compare.
Mothrayas wrote:
If a movie loses time to a WIP, it does not necessarily mean it loses to existing records. However, it can be an indication that a movie is clearly or provably sub-optimal, so such a movie can be rejected for that reason.
Virtually every TAS ever created have potential improvements in theory, but that doesn't mean that the TAS is sub-optimal, because it can simply mean the author is missing something. We might all make mistakes. Although most of us are pursuing perfection, but most of the times we fail, because we are humans after all. It's important to know that we TASVideos should publish the best or better TASes, not "perfect TASes", because such thing may not exist. And this TAS, as I see it, is without a doubt, the best Contra Force TAS so far, or at least before Xipo completes his movie. If it's not a big mistake (like slower than RTA runs in some parts), I think we should try our best to tolerate it, and recognize its good parts (2-minute improvement should mean something, right?), instead of calling it sub-optimal and shutting it out.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Post subject: Re: #6514: Soig's NES Contra Force in 10:27.72
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
EZGames69 wrote:
Existing records also includes WIPs. While I do think it sucks that Soig wasn’t aware of this wip, and the fact that Xipo never publicly posted this wip (as far as I can tell), that doesn’t discount the fact that they’re still existing records.
If by this standard, a lot of movies could be rejected. For example, someone submits a Super Mario 64 "120 stars" run, and someone else points out that there is a mistake in a stage, and posts his WIP that's maybe 3 frames faster, and so the judge must reject the whole run just by a WIP that shows a possible improvement in a single stage? When I read "existing records", I'm think of full complete runs, not a segment of a run. Since when a 2-minute WIP can be seen as a "record"?
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Post subject: Re: #6514: Soig's NES Contra Force in 10:27.72
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
ThunderAxe31 wrote:
Due to the existence of a movie that beats level 1 faster, this submission has to be rejected.
I find it shocking to see this. This is such an absurd reason for rejection! The movie rules say that the submitted run must "beat all existing records", but it doesn't say that a run must "beat all WIP out there in every stage" to be published. From what I know, Soig's run is the best movie for now that beats every existing records, since Xipo hasn't completed his movie and his WIP is only 2 minutes long. Are we seriously changing the movie rules to "must beat all WIP out there in every stage"?
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
The new improvement is awesome. Thanks for all your efforts!
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
zdoroviy_antony wrote:
Can somebody confirm is it new WR or what? Down press 1 frame earlier in walljump room, but idk what it is. https://www.youtube.com/watch?v=3sJ6oexgfZQ Edit: Looks like cheating and I feel like it already happened before
I'd assume it's simply cheating. He used almost all my inputs, and probably changed Mario's speed after the wall jump.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
That new strat in 8-2 is for RTA only, nothing to do with TAS. GTAce99's finding is not enough, because you need to come up with a whole and stable setup for a possible time save (I'm working on it). Despawning that Piranha Plant in 8-2 (or anything, really) is theoretically possible, but I couldn't imagine anyone crazy enough to use it as an RTA strategy, because it requires 3 Bullet Bills at the same frame. That's right, 3. GTAce99 got it by accident, and it almost couldn't be got on purpose, because it's such a rare opportunity. 4-2 needs no further testing. It's not possible because 4-2 has no Bullet Bills.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
So many great TASes lately. Congratulations to improving the Contra TAS to almost perfection. Comparison video: Link to video
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Shinryuu wrote:
I'm requesting that I can use a fixed version even it doesn't save any time. I want to have a polished product instead. I'm available at Thursday evening when I can verify this little mistake by myself even warmCabin already pointed that out and uploaded a movie file as well.
I truly respect that. So my comparison video can wait until Thursday evening. :)
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
The difference happens right after the "Ready" screen, after suicide. Please help me confirm this so I can keep working on the comparison video (or I get too confused).
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Question about Woodman stage: why is it 4 frames slower during suicide to meeting the boss? I see you pressed Start 4 frames later after landing... Is there a framerule involved? (sorry if it's a dumb question)
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Incredible work! I'm so happy to finally see the finished project. This has everything I've hoped for. My version of comparison video (classic two-screen comparison) coming soon... Edit: Comparison video: Link to video BTW, Crashman has faced a terrible color pattern crash. I hope the publisher doesn't pick that screenshot because it doesn't look too much like Rockman.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Question: Since there have been years without an ideal TAS emulator for Arcade (all current ones have crashes and desync and other issues), how hard would it be for BizHawk to work with Arcade? Is it almost impossible, or just a matter of time? Thanks. :)
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Awesome TAS. Great improvements to an already amazing TAS. Easy Yes vote (who'd vote no on this?).
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Awesome run! I never thought this game could be broken like this. Masterjun never fails to impress us. Easy Yes vote.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Request from Mars608 and aiqiyou: replacing the movie file with this updated version (with double suicide along with the final boss): http://tasvideos.org/userfiles/info/55785055604970913
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Great improvements! Really pushing this category to its limit. The release date & cover have made this run even more significant.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
EZGames69 wrote:
Since you’re here Happylee, can you tell us if the jumps over the pillars before the world 1 bowser can be done faster? It seems like there should be a better way to get around it.
Of course. This movie isn't optimized in many places. For optimized version, please see: [3456] SNES Super Mario All-Stars: The Lost Levels "warpless, Mario" by HappyLee in 34:36.61.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
You copied my inputs for 1-1. I'm not that surprised since I read "1-1: Ground clip for full FPG". But you shouldn't record your run over my movie and pretend you did 1-1 by yourself.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Darkdevil wrote:
I guess Mario ate all of the spaghetti.
Wow. That makes perfect sense! And probably also Luigi.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
warmCabin wrote:
5 second time save in Air! It works for both zipless and any%.
At first I was shocked, and then I realized it was an April Fools joke. :p
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
There goes the NES TAS of 2018... Great job improving this run! Solid YES vote.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
Making a comparison version right now. Edit: Comparison video: Link to video This is great improvement. Considering that Contra is one of the most popular games in China (probably next to Super Mario Bros.), this is really significant. I suggested Mars608 to wait for alex_ik's run gets published, but apparently the authors can't wait. I still have no objection with alex_ik's run being published first, since it's already "publication underway", and alex_ik's run is good that probably deserves a publication.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Experienced Forum User, Published Author, Expert player (2454)
Joined: 12/23/2007
Posts: 822
According to my small scale test (SMB TAS, crf 20), x265 takes about twice the encoding time, but produces about 50% larger file than x264. Also there's many settings we use for x264 encoding that x265 doesn't have. In short, x265 probably sucks for TAS encoding. I haven't tested VP9 because I'm not familiar with the encoder and settings at all.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).