Posts for Zinfidel

Posted: 2/25/2020 5:20 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

SHORT STORY: My TAS is slower than WR RTA due to RTAs using different consoles with better load times (PS2/PSTV). Does this fact disqualify my TAS from the Vault category? LONG STORY: Welp, I've made some heartbreaking discoveries about the TAS that I'm working on, and unfortunately, I made the discoveries really late into the TASing process. My TAS for Armored Core on PSX is nearly 8 minutes slower than the WR RTA in the same category. The reason for this huge difference I've found out is because the Armored Core speedrunning community universally uses late-model PS2s, and more recently, the PSTV service to run the game because those consoles have significantly faster load times. An example, the load time for the first mission in the game on Octoshock/Mednafen/PSX is about 9 seconds long. The load time for the same mission on PSTV is ~3 seconds, so 3x as fast. The category that I've TASed involves a LOT of loading. There are 33 missions skipped in a row, which means that almost the entirety of the middle of this category is loading a mission, then exiting a mission (which also has a load screen). My individual split times for missions on my TAS are way, WAY faster than the RTA times, but I lose so much time to load screens that the it doesn't matter. There is an RTA available to compare against that is run on actual PSX hardware in the same category, and that RTA is about 38 minutes long, so my TAS improves on that by 10 minutes. However, the WR RTA for the same category on the NTSC game is just over 20 minutes long, but is using PSTV for the run. So, what's the situation here? This TAS is definitely a Vault category, there's no way around that. It is slower than the WR RTA though due to technical reasons, but much faster than the most similar RTA I can find. Is this an automatic rejection?

Posted: 2/24/2020 9:17 AM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

I just got done spending a couple days optimizing the second melee in Destroy Floating Mines. The input that I finally came up with is a few seconds faster than JAX's input. I came up with many alternate branches for this melee though, and a few of them were visually interesting. Here's an "outtake" from the TAS. This particular input is nearly 6 seconds slower than JAX's TAS but its got a lot more swag than the solution I settled on. Link to video

Posted: 2/20/2020 7:43 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

I would love to see Poy Poy TASed. My brother and I somehow managed to find a copy of this game when we were young and it literally played the game until the disc was too badly damaged to read anymore. We were never able to replace it because the game is quite rare. It would be super fun to see the kind of tricks and frame-perfect beatdowns a TAS could deliver in the original party brawler.

Posted: 2/20/2020 10:02 AM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

Did you guys run into a "vertical banding" issue when you were working on deinterlacing PSX video? I'm processing some videos of the PSX game Ehrgeiz, which runs in the same mode as Tekken 3, namely 360x480 interlaced (or 400x480 including the overscan). This issue is much easier to describe with screenshots so here's an album of them. I think the problem is due to dithering. Ehrgeiz has dithering applied to nearly the entire game, which is what I think is causing the light-dark bands that are appearing. Deinterlacing is some how picking this dithering up and causing this visual artifact. Just scaling the still-interlaced video does not produce these vertical bands (mostly), it's ony after deinterlacing that they show up. Here's unmodified, but stretched (point) footage:

And here's after a bob filter and stretching (lanczos):

The banding is fairly obvious, and you can see that there is roughly a light-dark-light pattern that makes it stand out. I've used much more advanced filters like QTGMC and yadifmod and gotten the same result. I've included close-up shots of the banding in the album, including some shots of the Tekken 3 clip that was shared in this thread exhibiting the same problem. I watched the HQ downloads for the Tekken 3 encodes and neither seem to have this problem, so I'm wondering if someone who did the encoding for those videos recalls how they solved this problem, or if they even had this problem at all. I'm using AviSynth and VirtualDub for most of my work. I have tried quite a few options for the QTGMC deinterlacer dealing with noise and grain retention, but haven't made much progress there.Here's a link to the clip I used for these examples.

Posted: 2/16/2020 9:53 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

Original question tl;dr: Does Octoshock do any color correction on its own. Answer: Probably not, issue is most likely due to my graphics drivers applying color correction. Haven't figured out where the issue lies exactly but it does appear to be the culprit.

Posted: 2/10/2020 6:33 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

franpa wrote:

updated from v2.3.2 Windows to v2.4 and got the following error:

That first error is the result of the second error. You just gotta make a whole new folder for 2.4 (may not be strictly necessary) and make a new config file.

Posted: 2/5/2020 10:27 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

The AI/luck manipulation is impressive. Having Andy just blow past so many of the shadow creatures is pretty cool. Getting the death animations play just before some of the loading screens was a nice touch too. Yes vote from me.

Posted: 2/5/2020 9:30 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

I love seeing action-adventure games TASed, especially 3D games because there's so much latitude in optimizing the game. The Tomb Raider series is a great fit for TASing. Movement is crisp and focused. Watching Lara tear through these hidden temples at top speed is great fun. All of the glitchy movements and tricks are fun to watch too, especially when they are bewildering (the 'QWOP' movement in particular). I was confused a lot of the time as to how you even completed some of these levels, which I consider a good thing! Breaking the game and finishing levels in obscure ways is something I look forward to in a TAS. Excellent use of dead time on some levels for entertainment value. KS5 was very amusing and the ending of Desert Train was hilarious - I actually laughed out loud. There is only one nitpick I have for the entire run and that is that the game is really dark in a lot of places. That's not a fault of the TAS though, but with the game. That's a definite yes vote for me.

Posted: 2/3/2020 5:39 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

I finally made it through the mission aborting part of the game after weeks of research on mission routing, and now I can finally do the fun parts of the game. Here's the Destroy "Justice" mission that I just got done optimizing. Doing the full 180 degrees turn while also moving backward in a straight line was the main target of the optimization. Being able to perform the maneuver without ever turning the boosters off was a little tricky but doable and yielded nice improvements. This mission is amusing because it is finished using an exploit, so I don't even bother letting the intro radio transmission finish before I've already finished the level. Link to video

Posted: 2/2/2020 11:00 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

A nice, byte-sized amusing TAS. Yes vote!

Posted: 2/2/2020 10:58 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

Great find and improvement. Got a yes before, getting a yes now.

Posted: 2/1/2020 7:50 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

You'll need to use libTAS, and the game will need to run on Linux. http://tasvideos.org/EmulatorResources/LibTAS.html There's libTAS category on the forums as well you can look at for more info.

Posted: 2/1/2020 8:29 AM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

The lead up to the ACE was entertaining, and as others have said it's a pleasure to see this game absolutely demolished like this. The movie is a great technical achievement - yes vote.

Posted: 1/27/2020 11:52 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

I've got a question about official site encodes and horizontal overscan. For PSX movies, recent encodes retain the horizontal overscan pillarboxes in the video. I'm curious as to the reasons for retaining the overscan. At least on Armored Core for PSX, without overscan clipping turned on, the video output resolutions is no longer 320x240, but 370x240, with uneven pillarboxes on either side of the video. It seems to me like this abnormal resolution would cause problems when uploading to YT, which is picky about resolutions at least the last time I checked. Also, wouldn't retaining the overscan cause problems when correction the video's aspect ratio? Older PSX encodes, such as the Castlevania SOTN encodes have no overscan, so it looks like the decision to retain the overscan was made at some point. Or at least, once emulators that supported showing the overscan were starting to be used, the decision was made to keep it. tl;dr: Why do official PSX encodes retain the horizontal overscan pillarboxes?

Posted: 1/27/2020 7:06 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

Here's an update on the progress: I decided some time ago that I would do an any% run for this game, even though others and I agreed it would be a more boring run than a no Abort/OoB run. The reason for this is scale: the goals and route of the any% run are much more definable, and achievable, than a run that would necessary contain huge amounts of entertainment tradeoffs and difficult to answer routing questions. A no abort/OoB run would be an absolutely enormous undertaking, so I want to start with something more manageable. With that said, the route I've chosen (after a LOT of investigation, believe me) is one such that, after the Ravens' Test, 34 missions are skipped and the Chrome mission path is selected. This is because the Destroy "Justice" mission (the Chrome finale) has an exploit that allows it to be completed very quickly. Then the last two required missions will be finished by heavily abusing out-of-bounds glitching. I am past the Ravens' test and am now undertaking the most monotonous and tedious part of this run which is menuing and skipping 34 missions in a row. In the meantime, if anyone wants to take a stab at optomizing the Ravens' Test, please find my movie here: User movie #60310243168691604 I've spent enough time trying to optimize the Ravens' Test and am not getting anywhere, and I have a strong suspicion that it will just take someone thinking outside of the box that my brain is in to find major improvements. I'm happy to do co-authorship of the movie as well if someone can help substantially with improvements. I have a currently-13-page document I can share with anyone interested in my research on the game so far!

Posted: 1/17/2020 11:39 PM

Post subject: Tecmo's Deception Series

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

For the uninitiated: the series gameplay revolves around playing a passive character with the power to create magical traps in their environment, in which they must use to defeat intruders. The traps range from stuff like springboards that launch intruders into the air, to vases that drop on their heads, to electrified vacuum wall panels. The games feature a combo system too, where you get more points, do more damage, and other bonus effects for chaining trap effects together. The game is normally played passively and slowly because you must be very careful around intruders and activate your traps only when the enemies are in just the right position, but a TAS of the game could set up some truly astounding combos in record time. I think it would be incredible to watch if someone wants to make a TAS for the game. I'm thinking specifically of Deception 3: Dark Delusion because it's the one I'm most familiar with, and has all of the above features that I listed. Here's a demo of the gameplay for those interested: Link to video

Posted: 1/17/2020 3:10 AM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

I had some fun reversing the game's RNG, and then also reversed the weather forecast function. I made some scripts for predicting both. You can get them here. Perhaps they will help someone in the future with TASing the game! Note that I only really tested this with the BSNES core. No idea if snes9x will behave correctly with them or not.

Posted: 1/8/2020 1:21 AM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

You should post a video of the first couple of days when you come up with something you like! I'm sure more people than just myself would enjoy that. I'm sure you've taken steps to prevent loss of work like last time, but just in case: a real easy way to use BizHawk and save your work safely is to set the backup directory path to a cloud location, like dropbox. Then set the option to save a backup with every save to your tasproj file. That way, you have a constant, safe stream of backups to fall back on if you need to. If you are using Windows 10, you really have no excuse, because OneDrive comes pre-installed and ready to go.

Posted: 1/2/2020 6:47 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

I am also a fan of puzzle games, so I did watch most of the video, but inevitably skipped around a bit because the pace of the gameplay is pretty slow. If you aren't the one playing the puzzle game and making decisions, then just watching the game requires some sort of incredible or surprising gameplay to keep it interesting. As such, +1 for excellent destruction of the CPU in the puzzle game, but -1 because it's hard to sit through the video all the way due to, well, the game's pace and mechanics. So it's a meh from me.

Posted: 1/2/2020 6:37 PM

Quote

Zinfidel

He/Him

Experienced Forum User, Published Author, Player (201)

Joined: 11/21/2019
Posts: 247
Location: Washington

It's hard to believe it when an even faster TAS comes out for a game that has already been so thoroughly deconstructed. Insane movie, obvious yes vote.