Users browsing this thread: 2 Guest(s)
Mass Sprite Sheet Expansion Battle & NPC

#16
Posts: 676
Threads: 44
Thanks Received: 26
Thanks Given: 21
Joined: Jan 2015
Reputation: 11
Status
Zombie
So far I've:
  • Expanded the Battle Graphics Pointer Table at C2/CE43
  • Expanded the Battle Character Palettes at C2/CE2B
  • Expanded another list of pointers to character graphics at CF/F911

All three of those list now cover from sprite sheet 00-63 (Terra - Ramah in Decimal).
All three list have been moved and now start at 41/0000

I've changed the pointers to those list at:
  • C1/3D66
  • C1/3E20
  • C1/3E20
  • C3/6CF7

I didn't move the list at C0/D0F2 and the bank pointers at C0/D23C for Map Character Graphics because it didn't need to be expanded. I have changed those values on a sprite by sprite basis to point toward the graphics sheets I've put in at 41/0A00 and up to see if it helped... didn't help in battle (obviously) but it did load the proper full sprite sheets that I pointed them too, out of battle naturally.

As of now, I've had no success what so ever in loading the following graphics in battle:
  • 27 - Scholar
  • 28 - Draco
  • 30 - Returner
  • 58 - Fish

The following graphics load without error in battle:
  • 35 - Clyde
  • 40 - Bird
  • 41 - Rachel
  • 47 - Cid
  • 48 - Maduin
  • 55 - Esper Fairy
  • All of the normally usable battle sprites (Terra - Imp + Leo-Gesthal)

I've tried mostly everything from 23-34 (Numbers in Decimal btw) Graphics sheets assigned to a character and none of those worked either. I didn't mark them as officially broke because I didn't write down exactly which ones I tried when I did it. I've been using 30 (Returner) and 27 (Scholar) as test subjects while trying to get it to work. Sheet 30 just shows up with no sprite, sheet 27 glitches and tries to crash the game when you enter battle (Sometimes it crashes, sometimes it just glitches horribly). So far, no success on either of them.

I can however, change the pointer for Locke's sprite sheet in the expanded table, to the location where either the Returner or Scholar sheets are located and they will load and work, just not using their own numbers (as in assign graphics 1B to Character 01).

I would have thought using 41 as a bank pointer instead of FF or in this case 101, would have caused issues but other than those certain numbers, the rest work and load fine from the expanded-expanded Rom space. (as in over 4mb).

Using the extra expanded space, I also thought maybe the broken sprite pointers might be in a "dead zone" of memory that didn't want to read properly while the rest of the table was... however when I ran the debugger it showed that it was loading the proper location from the table while trying to load the sprite (the values added up, and there was no other way for it to get a location of 43/6CE0 unless from that table, with that offset.)

I've checked my table numbers to make sure I didn't make a typo or count them out incorrectly, however, if I was off by 500 or 600 bytes it would still load a glitched sprite or half a sprite. On the broke values, it is like it is pointed at nothing (on some of them), yet as I said, if I put the same value in a working value slot, it loads fine from the location it should be in (where I put the sprite sheet to start with).

Unless I've missed a pointer to the table somewhere that I absolutely can't find, it almost has to be something to do with assigning a value like $1B to X to use as an offset for the table. (I think I said that correctly). As in the standard formula just doesn't want to handle THAT value, but a value of $23 or $2F works fine? I don't see a pattern there, or anything that would have a problem with one that would work with the other. And again, the debugger showed the proper destination for the broke sprite sheet while loading it.

So if the code is using a value like $1B to read the expanded table properly, what would stop it from showing the sprite in battle after initial loading, but before the characters appear on screen?

Could it be an issue in the formula for reading palettes? Moreso reading the table with the higher sprite ID (or something to that nature.)

Yeah, its a very long post. Kinda hoped explaining it all fully would give me an idea of what to look at, but so far nothing.


The only true wisdom is knowing you know nothing.
  Find
Quote  



Messages In This Thread
RE: Mass Sprite Sheet Expansion Battle & NPC - by Catone - 04-01-2015, 09:06 PM

Forum Jump:

Users browsing this thread: 2 Guest(s)


Theme by Madsiur2017Custom Graphics by JamesWhite