øAslickproductions.org/forum/index.php?PHPSESSID=5f0fck550j2m4m2fpbtkj2vkm1&action=profile;u=363;area=showposts;start=135e:/My Web Sites/Slick Productions - FFIV Message Board/slickproductions.org/forum/index2bc7.htmlslickproductions.org/forum/index.php?PHPSESSID=5f0fck550j2m4m2fpbtkj2vkm1&action=profile;area=showposts;u=363e:/My Web Sites/Slick Productions - FFIV Message Board/slickproductions.org/forum/index2bc7.html.zx>–h^ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÈ@>ŸºOKtext/htmlISO-8859-1gzip8:ÖŸºÿÿÿÿÿÿÿÿWed, 11 Mar 2020 07:58:15 GMT0ó°° ®0®P®€§²ð®>–h^Ÿº Show Posts - 13375K31C43R

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - 13375K31C43R

Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 »
136
After some further digging, I think I have the definitive solution. The code file in the OP seems to have something to do with running various animation scripts (includes hits from monsters, various weapons, shields, cape dodges etc.). Here is the location of the shield block animation script:

Code: [Select]
D9/9F68: F0 02 92 05 93 02 82 C6 97 78 78 78 78 93 00 FF

The FF at the end tells the script to terminate. I think similar to the animation scripts in FF6, these bytes are assigned to various animation effects (e.g. 93 00 seems to be the "reset" from shield block pose or other poses, 93 02 might be for getting into a pose, etc.). 78 seems to just be "wait". If I remove some of those 78s from the script, I get a shorter animation. That seems to be all there is to it.

At some point, if it doesn't already exist, someone should start a document on FF5 animation script bytes.

EDIT: After even further investigation, I notice a bunch of scripts with quadruplets of 7X bytes near the end, which seems to indicate wait commands of varying length. So an alternative would be to change the 78s to something more appropriate, like 74s.

EDIT 2: Never mind, those are not varying degrees of wait time, there are different sprite IDs. In other words, 78 is a shield. So yeah, just change the number of occurrences.

137
Final Fantasy V Research & Development / Re: Monsters die slowly...
« on: June 04, 2017, 04:49:01 PM »
OK, I decided to slow the animation back down a bit, but it's still much faster than in the vanilla ROM.

A word to the wise: some monsters' dissolve animations will be faster than others'.

138
Final Fantasy V Research & Development / Speeding up shield animation
« on: June 04, 2017, 06:58:36 AM »
I just managed to fix my problem with the death animation, now I'm looking to shorten the shield animation; namely, the length of time spent standing with a shield up. I think I have found the relevant code, so if anyone wants to try fiddling around with that, go ahead.

139
Final Fantasy V Research & Development / Re: Monsters die slowly...
« on: June 04, 2017, 06:56:53 AM »
Thanks! I haven't tested on every monster, so some of them may still be a bit on the slow side, but hopefully not. Also, this version does not affect the already-faster boss dissolve animation.

I do kinda think maybe the animation is a little too fast now, though. I'll experiment some more to see if reining it back in a little helps.

140
Final Fantasy V Research & Development / Re: Monsters die slowly...
« on: June 04, 2017, 01:12:32 AM »
OK, here's the new version! Be sure to apply the old reverse patch before applying this new patch.

141
Final Fantasy V Research & Development / Re: Monsters die slowly...
« on: June 01, 2017, 12:18:21 AM »
I could try it, but I think I'd prefer it as smooth as possible.

142
Final Fantasy V Research & Development / Re: Monsters die slowly...
« on: May 31, 2017, 10:00:44 PM »
That's basically what I'm trying to do. The vanilla ROM basically calls the dissolve function eight times for each step of the dissolve because each call applies a partial dissolve to a different piece of VRAM, and there are roughly eight steps to the dissolve, meaning 64 calls in total. I found a way to reduce the number of calls to one per step by having each call apply to all the relevant parts of VRAM at once, but for some reason the screen is strobing black everywhere during the animation as a side effect.

143
Final Fantasy V Research & Development / Re: Monsters die slowly...
« on: May 31, 2017, 07:34:02 PM »
OK, now this is REALLY odd. Somehow the function at C1/18D2 is being called the same number of times even with the patch applied. Now I get eight interrupt signals for every one call to C1/FD27 within the dissolve animation. What's up with that? :wtf:

144
Final Fantasy V Research & Development / Re: Monsters die slowly...
« on: May 31, 2017, 06:33:58 PM »
OK, I've been debugging with No$SNS and I do notice a lot of flickering, but less of it if I slow down the dissolve animation. Unfortunately, the only way to get rid of the flickering entirely is to bring it back to what it is normally. For all I can fathom, either there's some other variable I can modify that can fix the flickering issue while retaining the sped-up animation, or what this comes down to is simply too much work for the DMA.

What causes the animation to be so slow is that C1/FD27, which is called 64 times, calls the function at C1/02F2 eight times itself, and one of the things C1/02F2 does is wait for an interrupt signal from the GPU thread; this signal is sent from a function located at C1/18D2. That means that over the course of the dissolve animation, we receive a total of 512 interrupt signals! It's no wonder the animation is so slow.

Stepping through C1/18D2, I noticed that one of the functions called near the beginning activates DMA channel 4 to draw VRAM on the screen. I refuse to believe that there isn't a way to speed up the animation without causing problems.

145
Game Modification Station / Re: New patch: Dead in the Air Bug fix
« on: May 21, 2017, 01:33:42 PM »
:bump: See EDIT 10.

146
Game Modification Station / Re: Rapid Save/Reset glitch?
« on: May 14, 2017, 01:59:24 PM »
I guess that kind of makes sense... I didn't think about that, probably because, as the video shows, there are about four iterations before slot 2 shows a party; until then it shows "Empty", and only when it does not show "Empty" is it actually loaded. What I thought was happening was that the party members weren't getting saved until the fourth iteration or so.

Upon further inspection and reading, I now realize the purpose and use of the SRAM checksums to check for corrupted save files. So yes, that seems to be what's happening: shorter and shorter chunks are overwriting each other, and the save file gets corrupted, but after a few iterations a checksum collision occurs and then the save file looks OK.

Unfortunately, that's just an inherent problem with the checksum: collisions are possible. I don't really know what to do about that other than maybe add a "saving in progress" flag, if that would work.

147
Game Modification Station / Rapid Save/Reset glitch?
« on: May 13, 2017, 07:11:58 PM »
Gi Nattak referred me to this glitch a short while back. From what I gather, the reason it works seems to be because somehow the data from two save files are getting mixed together into one. Specifically, what seems to happen is, when the game saves, the X and Y index registers are set to point to the start of the save slot and the start of the current saveable game data, respectively, and iterate through until 2,560 bytes are copied. Despite this, somehow in the alternating reset-interrupted copy iterations of slots 1 and 3 into 2, what's getting saved is a chunk of slot 1, then a chunk of slot 3, then a chunk of slot 1 etc. when these chunks should be overwriting each other.

This doesn't make sense to me because slots 1 and 3 don't seem to be getting affected in any way, and the index registers are always supposed to start at the beginning, therefore the chunks should be overwriting each other in slot 2. Anyone have any idea what's really going on here?

148
There is also an issue of not being able to jump between buildings in Zoso; I don't know causes that, so I just use another rom when I get to that part.

If you're using Zone Doctor to edit maps, that'll probably be your issue. There's a patch that makes characters face the right way while jumping; Zone Doctor overwrites the event code in the ROM, making the game freeze when it comes time to jump between buildings.

149
I could've done that, but I thought it'd be easier to keep it organized within the sprite bank like all the rest of the sprites. I do have to consider what works well with the code that's there, in terms of both space and logic, and yes, maybe there was a way to wrangle it so Banon's sprite was called at the right time from another location. But the route I went, I found it easier to maintain compatibility with the game itself, which is more important than compatibility with FF3usME. Besides, if I was writing FF3usME, I wouldn't hardcode the pointers. I'd get them from the ROM itself. That would make it more compatible with hacks, and would conceivably even allow for editing of Banon's riding sprite within FF3usME.

150
It's gotta be my Banon Riding Patch. It can't be helped, unfortunately.

The thing is, what Banon Riding does is expand Banon's sprite set to add a chocobo-riding sprite, which is needed if the player goes to Figaro's Chocobo Stable during the "Banon et al." scenario. As a consequence of this, a bunch of pointers to the sprite bank in the ROM had to be updated. My hypothesis is that those pointers are hard-coded in FF3usME, hence the scrambled sprites.

Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 »