Why Mario glides so much higher in the PAL version of SMW?
Was this intentional or some sort of glitch?
didn't know about this, now I'll have to emulate the PAL version
>>3350769
why should anon buy a pal tv, console and game just to see if some glitch works?
Expert troll, or real rumor?
>>3350769
emulation is so perfect these days the only reason to play on actual hardware is if you're a rich autistic collector
all you need is higan and a crt and you get a flawless snes experience
>>3351015
Someone make a comparison webm
>samefag
>>3351024
>emulation is perfect
You know, I thought so, too, but noticed something recently: Mario RPG and Earthbound.
Remember the Yoshi races in MRPG? Where you had to press A and B in alternating sequence according to the rhythm in order to win? Well, I can do it pretty damn well every time on the console. Emulation, however, is off somehow. I tested it on real hardware and I can still pull it off, but not on emulation for some reason.
As for Earthbound, I used to open and close the status window for that clicky-click noise in beat with the tango music in the game's hotels. I don't have a hard copy of the game to test this on, but I can't do it on emulation at all.
Conclusion? I think there's something real to the input lag people talk about. I thought it was total BS, but it ain't.
>>3351050
So wait, you in NTSC could not simply fly over blocking pipes to get keys etc.?
I thought that was what everyone did!
>>3351050
>>3351058
>also thanks for the webms
>>3351289
You totally could, it was just a bit more difficult.
What I mean is that it was easy in (U) but even easier in (E)
that being said, I wonder what the differences between the European revision 1.0 and 1.1 are. Why did they make a second run?
>>3350640
half-A press
>>3351271
>I can do it pretty damn well every time on the console. Emulation, however, is off somehow. I tested it on real hardware and I can still pull it off, but not on emulation for some reason.
its just input lag from your monitor, it has nothing todo with emulator.
>>3351271
Play it on a modern PC you fucking moron.
>>3351319
Tested both versions, Mario still flies higher.
This is the only difference I could notice.
Although maybe not retro, in the GBA version Luigi flew much higher than Mario (or so I remember).
>>3351024
poorfag detected
kill yourself right now
>>3350640
Euro babies need a crutch :^)
>>3351397
Tcrf lists a bunch of differences between the J U and E versions, but the list is incomplete. Also, although the US version has been hacked to oblivion, everything unused has been found and almost every single byte mapped, no one has really done much research on either of the PAL versions.
I'm gonna do a binary diff on E1.0 and E1.1 and cross-reference them with the SMW central ROM map. Maybe there's something interesting.
>>3351595
Godspeed
Isn't it just because PAL has a higher vertical resolution, so it seems like Mario is going higher up when he really isn't?
>>3352491
You can gain height in NTSC version too tho...
>>3351335
>>3351271
Emulation isn't really perfect, but "it works". One pretty known example of input lag ruining gameplay is in Parappa the Rapper (PSX). 9/10 people that played the original game will tell you the emulated version's timing is off.
tl;dr for games with precise timing (rhythm games, games where you have to use some sort of 1-frame parry), there's a difference.
>>3352510
That's objectively not true.
>>3352419
The vertical resolution is 16px bigger but that doesn't really mean anything because it's padded to the bottom of the level. The same level data is there in the NTSC version too but you simply can't see it because of the lower resolution. See pic.
I did the diff and there's a lot of changes between all three (U1.0, E1.0 and E1.1). I couldn't reference the SMWcentral ROM map because they've changed the addresses to SNES instead of PC. Ugh.
It's also possible that there's some kind of historic mislabeling because the E1.0 doesn't even utilise the real PAL resolution and doesn't have the [!] in the ROM name, whereas E1.1 does both. Maybe E1.0 is just a bad dump or something? It aggravates me that there's no information regarding the subject. I don't know which one is used in >>3351058
>>3352419
He is going higher.
>>3352624
It is - you can gain height with both flying methods (the gentle tap backwards and the dive bombing) but it's a bit harder.
>>3353137
Think you could try with the 1.1 [!]?
>>3353142
Same effect.
I'm not trying to trick anyone, I'll even leave the title screen.
>>3353168
Well I'll be damned. Thanks for your effort.
Also, it turns out that there truly are two real European SMW versions according to SNEScentral - SPAL-MW-0 and SPAL-MW-1.
I can't guarantee that these PCB codes correlate to the 1.0 and 1.1 versions available though, but I'm pretty sure they do. It still makes me uneasy that the 1.0 isn't a verified good dump but I digress.
I wonder if the European flight physics carried over to either of the SMAS+W versions.
Lunar Magic doesn't support the European ROMs, but obviously Lunar Compress does. It would make sense to get an allgfx.bin out of both Euro versions and check if there's something we don't know about. I don't think anyone has decompressed the European graphics like ever.
>>3353195
>I wonder if the European flight physics carried over to either of the SMAS+W versions.
Luckly it's possible to load NTSC saves in the PAL Rom.
And the answer is yes, Mario also glides higher in SMAS+W PAL.
>>3353206
>Mario also glides higher in SMAS+W PAL
What about SMAS+W NTSC?
>Luckly it's possible to load NTSC saves in the PAL Rom.
Interesting. Game Genie / Action Replay codes don't work, though. At least when I tried using the cycle through powerups debug cheat I got a crash in the PAL version. But PAR/GG codes just force values to RAM anyway so I guess it makes sense that the addresses can't all be the same.
>>3353212
>What about SMAS+W NTSC?
Same SMW NTSC physics.
And yes, the US version has the same JP physics, if there was any doubt.
>>3353232
I see now.
Apparently, the byte at 0x57BA (US1.0, non-headered) is the byte for "Cape gain (added to y speed)" and that particular byte is $C8 in the US version but $ED in the PAL1.1 version.
However, the US version's $C8 is part of a byte cluster at 0x57B8 that is 40 10 C8 E0 02 - that same byte cluster is in the European version but at 0x576C.
Not 100% certain, but could be that Nintendo of Europe did a mistake -they didn't intend to change the physics- they accidentally padded some other shit there and caused the flying physics related bytes to be misaligned, but they got lucky and the change was so unnoticeable that they shipped it.
Or it could be simply a pointer mistake, hard to say.
>>3351271
>As for Earthbound, I used to open and close the status window for that clicky-click noise in beat with the tango music in the game's hotels.
>I don't have a hard copy of the game to test this on, but I can't do it on emulation at all.
I think we've discovered a new level of autism, gentlemen.
Tried decompressing the PAL graphics with Lunar Compress.
It says the decompression was successful, but the output looks like shit
>>3353371
Did you set correct graphics format and palette in your editor/viewer?
>>3353450
Yes.
I tried both arguments, 1 0 and 0 1 but the output is always just barely there. The first couple of bytes are okay but then it starts going to shit. It's also possible that FuSoYa never looked deeper into the compression scheme and made it work only with the U and J versions.
im not an expert at this shit. but since pal games run at 50hz dont they speed the games up and shit to compensate and it can cause wacky shit to happen?
>>3353470
Yes. For example, "How many pixels Mario moves per frame" byte is higher in the PAL version, but you can't really notice it because it's 50Hz
>>3350640
Mario also runs slightly faster and jumps slightly higher in the PAL version too.
>>3353612
>jumps slightly higher in the PAL version
untrue
>>3353775
Wow can't believe you went to the trouble of that, but okay then. Myth busted.
>>3353798
Thanks
I can't comment on the speed though, but I'd like to know where you've heard such claim
And even so, what does "Mario runs slightly faster" even mean? Faster in real-world seconds? More pixels per frame? More pixels per 50th of a second? More pixels per 60th of a second?
>>3353825
https://www.youtube.com/watch?v=xxvwRDF8pjQ
>>3353848
I need a better video editor.
>>3353825
I'm not too sure to be honest. I always manage to nab far more coins in the yellow switch room (like one extra life sometimes) on the PAL version compared to the NTSC.
>>3353195
>It would make sense to get an allgfx.bin out of both Euro versions and check if there's something we don't know about. I don't think anyone has decompressed the European graphics like ever.
Probably not. The only thing i could think of are the credits, but these are already found.
>>3353896
Maybe the P-Switch timer is slower in PAL?
>>3353914
Yeah, but there's a shitload of differences in the two games, I'm having a hard time believing that all of them are just pointer fixes and new speed bytes.
>>3353884
This isn't necessarily a Mario's speed thing - it could very well be about enemy init like >>3353953 said
PAL SNES runs at 50fps, NTSC at 60. If games aren't changed at all, they'll simply run slowed down. What a select few games like Mario do is increase a few values such as Mario's speed to compensate, but most things in the game like enemies' movement are not changed.
>>3354002
Maybe during development, all time-related variables were written down for easy PAL porting? So you could just run some macro to make them 50hz compliant
This thread is giving me autism.
I tested Mario's speed in stages where I could just run in a straight line. Either Mario is about 10% faster in the PAL version, or the timer is 10% slower.
SMW guru gives insight on Twitter
>>3354013
>>3354002
Turns out Mario does jump higher in PAL but only from running(?)
>>3354026
That's some anons thought.
The gliding height difference is way more notable though, it's kinda hard to believe that a slightly speed difference could be the cause.
>>3354026
When I saw the OP I thought it had to do with framerate right away.
>>3356746
welp I guess this thread is dead now
>>3353354
Is this true? you better not be joking around you nig nog
And I thought retroarch was a fucking container for different emulators, not an emulator.
Am I nuts?
>>3357878
Yes, it is a container for emulator cores ported to the libretro format.
All RetroArch does is pipe audio and sound from its cores into a standardized setup, so you don't have to change settings between a million different programs.
The bonus to this is being able to use settings like hard-gpu-sync in emulators that otherwise don't have it in their standalone states.
>>3351050
Oh, so that's why I couldn't fly very well last time. I thought I got worse than my 4 year old self.
>>3351058
jesus, why does he go so far up on PAL
>>3357878
It is a container, but the deal is that it's responsible for providing functionality that every emulator would otherwise have to implement on its own (reading the controllers, interfacing with whatever graphics API the host system uses, establishing a connection for online play, etc.) while the emulator cores can just focus on the parts particular to the system they emulate. RA's available graphics settings just happen to let you achieve lower input lag than the settings available in most other emulators since it has so many knobs and dials that most others don't
so speedrunners should be using smw pal
>>3364515
No.
Mario may run just slightly faster, but every animation is slower.
