Ending Pic mini-project

[justin3009]

Did NOT know that was an island.  XD

[FaustWolf]

I increased the dark pixel density on the far-off sky island, so hopefully it will be easier to discern from the sky. I know what you mean -- in the pic's penultimate edition, it could have been mistaken for random pixels that resulted from a bad palette reduction.

[Chrono'99]

Wow. Awesome work! I'll insert the patch in the ROM I have and post the final ips patch later today. The release time is near!

[FaustWolf]

I probably don't even need to say this, but be sure to back up your ROM before you do the ending art work patch. I'm not 100% sure it's not going to have undesirable effects, though I took care to negate the effects of loadscreen.flux in the patch.

[Chrono'99]

There's something strange with this patch. If it's applied to the Prophet's Guile ROM, TF will refuse to open it (error message), and it will freeze if run in ZSNES. Oh well, I didn't bother trying to understand; what I did to resolve this is simply applying the patch on a fresh, unmodified CT ROM and then manually import the concerned tilesets/palettes/map assembly from that ROM to the Prophet's Guile ROM (I have lots of backups) using TF and an hex editor. It works flawlessly, so this problem is resolved.

However, there's another problem. The ending pic works fine, but the animated flame tiles in Magus' Castle became glitched (see attached pic)! At first, I thought I might have done something wrong with my manual hex copy-pastes, but in fact these tiles are not just glitched in the PG ROM patched with the ending pic patch, they're also glitched in the fresh CT ROM patched with the ending pic patch. This means that the glitch comes from... the ending pic patch. Curiously, the tiles look fine in TF, so it's only when playing the game that one can notice the issue.

I'm attaching a patch which contains the final version of the game. It's basically the version that I thought we'd release (*sigh* we're so close...), except it has this bug with the flame tiles that I noticed at the last minute.

...So, how do we fix this?

[attachment deleted by admin]

[justin3009]

If you can't fix it by tilesets, you could always replace some NPC sprites with the flames.

[nightmare975]

I can't seem to even start the game, I select new game, and the after I name Magus, the game goes to a black screen. X_X

[FaustWolf]

Damn, sorry about that. I'll have to get into Magus' room in my test ROM -- is there a savestate repository anywhere for this sort of thing? Found it!

Just to be sure, the names of the patched and unpatched ROMs used as references in Lunar IPS don't matter, do they?

EDIT: Nevermind about the need for testing savestates. I see the problem on my end -- in my test ROM, the lower L1/2 subtiles for Magus' castle, Ozzie's fort, etc., have been overwritten with parts of the ending pic. Hell if I know how that happened -- looks like I did a bad recompress at some point along the way in my test ROM without realizing it. The subtiles in Magus' castle and the subtiles in the court room are not automatically linked in some way, are they? If not, Chrono'99, what I'd urge you to do is to lift all the subtile data for Magus' castle from a pristine unheadered ROM and recompress it into your own test ROM after the ending pic patch is applied.

Waitaminute, even better, I'll try it out for ya if someone -- calling Vehek! -- can tell me where the subtiles for Magus' castle are located. It'll take me a bit to sift through Geiger's NA offsets notes to figure it out on my own.

Oh. My. It looks as if the subtile sets *may* be related between the two locations. I'll have to think on this a bit more. Daaaamn... But if someone could help me figure out the offsets for the Magus' castle graphics packets, it would be a major help.

EDIT: Waitaminute, now it seems Flux isn't showing a change in the Magus' Castle tile packets. It appears that, if I look at one subtile collection in Flux and then switch to another, the subtile list still retains the previous graphics packet. Bah, I'll have to see what's up. But obviously there's a problem with the subtiles in Magus' Castle. I'll explore.

[ZeaLitY]

I see three tiles are messed in in the ending art after my playthrough. It's attached. Two are underneath the shaft of Magus's scythe near the grass, and one on the body of the island of Zeal.

Also, in the second attachment, where Magus is standing you can enter and exit the land around the plateau and the plateau itself. Maybe some solidity is needed there.

I guess that's a wrap! There's still the "others" dialogue in Melchior's throne room scene that should be "other" without the s, but that's the only thing I could find.

Anyone have a savestate of the Programmer's Room?

[attachment deleted by admin]

[FaustWolf]

Thanks for the ending pic pic Zeality. That can be handled easily enough, methinks, with some manual reconfiguring in the final patch. It's the Magus throne chamber that bothers me. And -- ooh! I see Chrono'99 implemented the text overlay. Nice!

I'll get on the ending pic glitches and the Magus room flame glitch tonight, and I'll have either a revised patch or a failure report tomorrow morning.

UPDATE: Aha! I now know the culprit for the problem in the patch I created. It appears that multiple locations use the same graphics packet(s) for tile animations. The reason we can't see the problem in Flux is that Flux doesn't show these animated tiles. I would expect other animated tiles to be affected in the hack, and I'll see if I can get a list together of what those might be. In any case, I should be able to find a workaround.

[FaustWolf]

Yeppers, looks like we've got the culprit figured out. Take a look at the attached pic and see what you guys can recognize. Looks like Norstein Bekkler might have been affected along with the candle flames and the court room foreground crowd. This is a graphics packet I overwrote to insert part of the ending pic; figures.

We have a few options. The most efficient is to re-point the sixth trial subtile source such that it draws from an unused graphics packet. As long as things are aligned the same way within whatever graphics packet we decide to use instead of this one, no reconstruction of the ending art should be necessary.

[attachment deleted by admin]

[FaustWolf]

 

Full patch with ending pic glitch and Magus Inner Sanctum flame glitch corrected. Try it and see what you guys think. As long as nobody's planning on making use of the Millennium Fair tile set, we should be good to go.

[attachment deleted by admin]

[Chrono'99]

It glitched something in the Blackbird tileset, but since there are only 3 affected tiles and since they are used only in the Developer's Room, I simply modified the map so that other tiles are used instead. I've attached a pic of the corrected map (see the former version in the Mapping thread). The reactor tiles were neat, but this small modification is probably an easier solution than having to deal with another graphic hacking fix.

I've attached the final(?) patch of the game, as well as a savestate of the ending with access to the Programmer's Room for Zeality.

[attachment deleted by admin]

[FaustWolf]

I guess I just can't win, eh? Let me know if you want me to take a look at restoring those tiles, Chrono'99. Hope it doesn't make things too messed up.

[Chrono'99]

Nah it's definitely fine. Unless someone discovers another bug or something, this version is probably the final one.