Topic by Zurreco

Posted: 10/29/2006 12:15 AM

Quote

Zurreco

Former player

Joined: 8/1/2004
Posts: 2687
Location: Seattle, WA

kwinse wrote:

But will a DVD player mangle the quality like that, which is what was used to capture this run?

Computer -> S-Video -> TV -> VCR VCR Tape -> DVR -> DVD Instant quality destruction!

hi nitrodon streamline: cyn-chine

Posted: 10/29/2006 12:52 AM

Quote

BoltR

Former player

Joined: 8/12/2004
Posts: 651
Location: Alberta, Canada

Even if you are going TV -> DVD you still are going to get noise in the cables which degrades and image, so i'm not going to go into that. However, even if you forget about that it will still look similar. And I shall prove it! Here are some screen caps which I took. First a cap from the SDA movie beside one of MFlatley's WIPs (First Super Metroid thing I found on my smv folder)

Here they are blown up 2x so you can see some actual detail.

Ok, big difference yes. The next image is a screen cap from me encoding that section MFlatley's WIP with a similar bitrate used in the SDA encoding. This is without any of the further optimizations which are used by encoders at this site, but I'm not sure if Radix/Nate use any advanced encoding options anyways.

Wait a minute... that looks fairly similar to the SDA run! I had also actually taken photoshop and added some noise/blur filters to the emulator screencap to simulate an encode and make it look the same as the SDA encode too, but my power went out and I lost that picture, and didn't bother to remake it. Also, what does that really prove anyways? I can use photoshop I guess. The only thing left is the color of the two are different. Which if you ask anyone who has dealt with PAL/NTSC colors about, they will be able to give you an easy answer for. NTSC color emulation isn't accurate, it may not even be possible (A joke acronym for NTSC is 'Never the Same Color'), so an emulator will NEVER be the same color as on a console. However, if a TAS was played back on a TV to be recorded with a VCR/DVD/Whatever to be submitted to SDA it would get the NTSC color effect and will look different from the emulator.

Posted: 10/29/2006 4:12 AM

Quote

Maur

Joined: 1/31/2005
Posts: 95

I recommend d/ling the HQ version before doing anything similar to that. Also, the video almost certainly passed this test (another example) before getting the ok. The context of the links is that someone (a guy named metalsmasher) tried submitting emulated vids to SDA and got caught. I'd be surprised if nobody went through to look for this sort of stuff before posting.

Posted: 10/29/2006 4:36 AM

Quote

catnap222

Joined: 2/9/2006
Posts: 41
Location: Japan

Even if you prove the possibility of foul play, it doesn't prove this movie is foul play. Also I can't prove this movie is real. But still, I believe it real. Mabye, I'm one of the closest man to hotarubi. We have discussed SM technique and speedrun at NINTENDON! for 1.5 years. In those discussions, I realized that he has monster skills and unique thoughts. He said he could perform extremely hard play which seemed impossible in real play. People including me doubted his big mouth, and gave him many questions in detail. But his words were filled with confidence and there were no paradox. It seems he is not theorist, but naturally has gifts which are skill, idea, and passion. So I recommended him to make pure speedrun. And I told him SDA rules and all of my knowledges about SM including TAS level, but no strategies for speedrun because it seemed to kill his unique ideas. Since He can't speak English so well, I helped with his submittion to SDA. That's all of my supports, and HE DID IT! I'm proud of him. Congratulations, hotarubi! I can answer some questions. - hotarubi recorded this movie with HDD recorder, and burn it into DVD. - It's temporary-blue-suit performed before Draygon fight. Easiest way to do it is to charge with pressing R and wait until charge disappear. - My translation of author's comment is poor, but in Japanese version, he said he realized 31min impossible empirically.

Posted: 10/29/2006 6:45 AM

Quote

Titus_Kwok

Joined: 4/30/2006
Posts: 480
Location: the secret cow level

He does manage to pull off some pretty crazy tricks, but there's also a lot of little mistakes that could've easily been fixed if he was rerecording.

Posted: 10/29/2006 6:50 AM
(Edited: 10/29/2006 6:52 AM)

Quote

BoltR

Former player

Joined: 8/12/2004
Posts: 651
Location: Alberta, Canada

Well, my intention wasn't to to prove that his run was actually a TAS; I haven't even watched it yet. I'm still in the middle of watching the Lufia run that was posted, and hadn't even downloaded the SM run until mikwuyma asked me a question about it earlier today. I was just pointing out that "it looks different than an emulator run" isn't a valid point of comparison. Though from the description I read on SDA I would have to congratulate hotarubi too, that is quite the time. Maur, it doesn't matter what quality of video I were to try and encode MFlatley's against, i'm pretty confident that with a little tweaking I could get mencoder to spit out something that looks very similar to it. Much more similar than the rush job which I have posted in this thread. As for the stretched out text, i'm going to go out on a limb and say that is caused by some sort of post processing done by the emulator. Filtering such as 2xSal tends to mangle text like that. Edit: Titus, hypothetically if one were to and pass out a TAS as a non-assisted run it would be local it have some errors which would be 'easily fixable', or else it might be very obvious that you have used tools.

Posted: 10/29/2006 6:52 AM

Quote

Zurreco

Former player

Joined: 8/1/2004
Posts: 2687
Location: Seattle, WA

Titus Kwok wrote:

He does manage to pull off some pretty crazy tricks, but there's also a lot of little mistakes that could've easily been fixed if he was rerecording.

This isn't really indicative that it is or isn't emulated. If I were to submit an emulated run in such a uselessly extensive method just to get my name on SDA, I would make sure to make tiny mistakes every once in a while to ensure that no one would difinitively know that it was a fake.

hi nitrodon streamline: cyn-chine

Posted: 10/29/2006 9:21 AM

Quote

Mukki

Skilled player (1090)

Joined: 8/26/2006
Posts: 1139
Location: United Kingdom

Exactly, that's just common sense, but this isn't emulated, some players are just really really good :)

http://tasvideos.org/Mukki.html

Posted: 10/29/2006 9:31 AM

Quote

Tub

Joined: 6/25/2005
Posts: 1377

Dragorn's 120 star Mario64-run didn't look any less TASed than this. He's just been around longer, so nobody's accusing him.

m00

Posted: 10/29/2006 11:33 AM

Quote

xoinx

Former player

Joined: 9/20/2006
Posts: 287
Location: Singapore

I just can't imagine how people can stand all the pressure sitting through 2 hours trying to create a 'perfect' run. I would prolly just die of a heart attack if I screw up on the last stars :S I respect them for their persistence and all man... training for a speedrun is tough and time consuming! It's different from TAS, not so much on finding the best glitches to save time, but more on grinding and grinding and grinding to be able to produce consistent results time after time. If hotarubi really pulled it off (which I am starting to be convinced he did), he will definitely be my most respected speed runner...

Truncated wrote:

Truncated is the most fiendish instrument of torture ever devised to bedevil the days of man. -- xoinx

Posted: 10/29/2006 12:35 PM

Quote

supermegavkoy

Joined: 4/30/2005
Posts: 199

Is this guy, hotarubi, a fan of any other games or does he just play Super Metroid? It would be nice to see a run as good as this one of another game!

Posted: 10/29/2006 1:10 PM

Quote

Warp

Banned User, Former player

Joined: 3/10/2004
Posts: 7698
Location: Finland

My first thought was: "What? A speedrun which is 14 minutes faster than the current best TAS? How is that possible?" Than I read that the 32 minutes is not the actual length of the speedrun, but just the value of a counter (which has little to do with real time) in the game. I wish they used standardized timing for all videos, not a per-game one, especially since in Super Metroid that counter doesn't run constantly, and a longer video can actually have a smaller counter at the end, which doesn't make too much sense.

Posted: 10/29/2006 1:57 PM

Quote

Saturn

Player (209)

Joined: 2/18/2005
Posts: 1451

Warp wrote:

My first thought was: "What? A speedrun which is 14 minutes faster than the current best TAS? How is that possible?" Than I read that the 32 minutes is not the actual length of the speedrun, but just the value of a counter (which has little to do with real time) in the game. I wish they used standardized timing for all videos, not a per-game one, especially since in Super Metroid that counter doesn't run constantly, and a longer video can actually have a smaller counter at the end, which doesn't make too much sense.

It actually makes much more sense to use the ingame counter in Super Metroid, because it is very accurate and only shows time the player did truly play. It skips the many random door transitions that vary in lenth depending on which position you enter them, so I think the ingame time in Super Metroid has much more value for a speedrunner and tells the watcher alot better and correctly on what skill-level the run truly was.

See my perfect 100% movie-walkthroughs of the best RPG games on http://www.freewebs.com/saturnsmovies/index.htm Current TAS project (with new videos): Super Metroid Redesign, any% speedrun

Posted: 10/29/2006 3:01 PM

Quote

Bag_of_Magic_Food

He/Him

Active player (277)

Joined: 5/29/2004
Posts: 5712

But this game won't even tell you the number of seconds! What good is that?

put yourself in my rocketpack if that poochie is one outrageous dude

Posted: 10/29/2006 3:06 PM

Quote

JXQ

Experienced player (750)

Joined: 5/6/2005
Posts: 3132

Although I disagree with the relevance of Super Metroid's in-game timer (for TAS purposes, anyway), I do think it's a good idea for SDA to use it, because it's part of the game's hardware, and so can be trusted as accurate (by its rules anyway), similar to using the frame count when we do a run here. Any other timing mechanism would involve timing the recorded video, and it seems like it would be possible to either mess that up, or tamper with it (by speeding up the video perhaps? Eh, maybe that's not as feasible as I first thought).

Warp wrote:

and a longer video can actually have a smaller counter at the end, which doesn't make too much sense.

This is probably what I like least about the in-game timer.

<Swordless> Go hug a tree, you vegetarian (I bet you really are one)

Posted: 10/29/2006 3:22 PM

Quote

Saturn

Player (209)

Joined: 2/18/2005
Posts: 1451

Surely a disadvantage. But still I prefer ingame time. Geting it as low as possible is the goal that counts in speedruns and TAS (for me at least).

See my perfect 100% movie-walkthroughs of the best RPG games on http://www.freewebs.com/saturnsmovies/index.htm Current TAS project (with new videos): Super Metroid Redesign, any% speedrun

Posted: 10/29/2006 4:47 PM

Quote

Tub

Joined: 6/25/2005
Posts: 1377

Warp wrote:

I wish they used standardized timing for all videos

they do. if there's an ingame-timer that's not broken, it's used. if there's none, real-time is measured with standardized end-points. On games where loading times may differ (pc-games, multi-platform games), either loading-times are removed, or times are tracked seperately. that's different from the usual TAS-timing, but actually more standardized, since there's still no agreement on the proper point for ending your TAS ;)

m00

Posted: 10/29/2006 9:42 PM

Quote

Titus_Kwok

Joined: 4/30/2006
Posts: 480
Location: the secret cow level

The emulated system memory displays seconds, doesn't it? So on a TAS you can get an accurate measurement using the in-game timer?

Posted: 10/29/2006 9:46 PM
(Edited: 11/7/2006 11:17 PM)

Quote

moozooh

Senior Moderator

Joined: 8/4/2005
Posts: 5770
Location: Away

Moreover, you can find even the ingame frame counter that goes from 0 to 65536 (and starts anew after that), in addition to the standard two digit hours, minutes, seconds and frames. Super Metroid allows you to monitor pretty much everything you want with great precision.

Warp wrote:

Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.

Posted: 11/7/2006 11:16 PM

Quote

Ersatz

Joined: 8/9/2004
Posts: 139
Location: Washington State

would the concensus be that this run is real? or do you think foul play? I can't quite decide. But i'm leaning toward real.

Posted: 11/7/2006 11:19 PM

Quote

moozooh

Senior Moderator

Joined: 8/4/2005
Posts: 5770
Location: Away

Personally, I have no doubts it is real. Moreover, I think 00:31 is possible single-segmently, and sub-30 is possible in a low-segmented run. Some day we will see such runs, I hope.

Warp wrote:

Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.

Posted: 11/8/2006 12:06 AM

Quote

mikwuyma

Joined: 12/20/2004
Posts: 226

About SDA timing: What tub said, that's the exact standard we use. If a game needs to be manually timed then we start when the player gains control of the character, and end when the player loses control of the character. Whether I think this run is real or not. I'm confident this run is real. A. The arm pumping is nowhere near frame perfect or turbo controller speed. I don't know where people are getting this notion from. B. The small mistakes make a lot of sense once you think about it. The guy is so precise throughout the whole run, that he didn't make very good risk/reward judgments. He tried to be as precise as possible on every single manuever, and that's why you sometimes see him screwing up wall jumps (even in places where he didn't need to wall jump) and barely missing ledges . With that being said, the run kicked ass.

Posted: 11/8/2006 2:09 AM

Quote

Admiral_Sif

Joined: 7/26/2006
Posts: 53

I saw the run when it was posted up and I was in complete awe at the speed he went through the run, but the small mistakes he made, made me believe that in was real IMO. At least I know what he was doing when he was arm pumping, I thought he was just making the movie more interesting :P

Posted: 11/8/2006 6:53 AM

Quote

Kyrsimys

Former player

Joined: 4/16/2004
Posts: 1286
Location: Finland

How can people see the small mistakes he made as proof of the video being real? As someone already said, if someone was to send a TAS to SDA, he would most probably leave in mistakes to make it beliaveable. No one would be stupid enough to just send in a run with absolutely no mistakes. Same goes for the arm pumping.

Posted: 11/8/2006 8:00 AM

Quote

OmnipotentEntity

He/Him

Player (36)

Joined: 9/11/2004
Posts: 2623

I thought the method used to determine that this run was legit was a comparison of the colors. (Emulators don't handle NTSC colors very accurately)

Build a man a fire, warm him for a day, Set a man on fire, warm him for the rest of his life.