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 - Akari

Pages: 1 ... 26 27 28 29 30 [31]
751
Scripting and Reverse Engineering / Render error
« on: 2005-11-03 02:38:29 »
I try to render models from xenogears but when I put textures on it I have some funny ladder effect. It always happened just in few places. And then I rotate model it moving.

Could anyone help me just by looking at this. I'm not good when it comes to textures  :z



752
I released source and compiled version of xenogears bg viewer. i don't made any progress since last time, just get code into something readable.
I continue work on this format after two month of rest =)

Source code can be get here http://server.titansoft.ru/akari/xeno_engine_src.rar

Compiled code can be get here http://server.titansoft.ru/akari/xeno_engine.rar

You'll need extracted background data (i still couldn't read CD by sector (anyone knows any cross-platform library for this?))
It's not legal but you can get it at the same place where all other files (xeno_data.rar)

753
Back to the FFVII engine. Why not to start making it like separate moduls and then compound it like it was when the original game was developed.

Right now I trying to get job of game programmer, so it will be great expierience to me.

L. Spiro, I can help you a little bit. Just tell what and where.

754
Here is the next step in progress: textures

0x001C offset from the beginning of the file are always textures block. It

# textures block header
all offsets start from beginning of this block
0x0000 u32 - number of textures and palettes
next are four bytes offsets to textures and palettes

each texture or palette start from header
01 11 00 00 00 00 E0 01 - are palette ID
00 11 00 00 40 01 00 00 - are texture ID
ps: I met only two of them and they are always the same so I think its just identificator and dont have special meaning.

next four bytes are probably VRAM position

and the last four bytes are width and height (2 bytes each)
ps: in texture case - width must be multiply by 2

Palette - (Each pixel - 2bytes);
Texture - (Each pixel - 1byte offset to palette);

Which palette to be used must be set somewhere else.






755
Quote from: halkun
Just cleaning up the forum cruft a bit, don't mind me wiping off this smudge on the window. Continue like I'm not even here... ^_^


Okey. There are some fixes in mesh header

# mesh header
all offsets start from beginning of this block
0x0000 u32 - number of blocks
00 00 00 00 00 00 00 00 00 00 00 00 - unknown (as I see it always filled with zero)

#header to mech block
00 00 - unknown (zero)
84 00 - number of vertex
73 00 - number of mesh
02 00 - number of mesh block
28 01 00 00 - offset to vertex block
28 01 00 00 - offset to vertex block (again?)
48 05 00 00 - offset to first mesh block
80 3B 00 00 - offset to unknown block
00 00 00 00 - unknown (zero)
00 00 00 00 - unknown (zero)
95 01 74 FF 4F 00 53 01 B2 05 F2 FF 1D 03 00 00 00 00 00 00 B8 11 00 00 - unknown (perhaps some flags)

First mesh block starts with 4 bytes header. First two bytes are still unknown. The second one are the number of mesh in this block. This header folowed by mesh data. Each mesh takes 8 bytes. The only difference between quad and triangle is last two bytes (triangle has only 3 vertex indexes so last two bytes are zero (vertex index start from 1)).




ps: I need to sort and name properly all I know already, three different mesh blocks are too much ^_^

756
Mesh of course ^^'' my mistake

Quote
Did you find how the game locates the files or did you trace the disk-access?


No I didn't found any utilites that can trace sector-access. All they work with files  :z
I, simply, went to the first battle in the game and then create dump of memory that my VGS used. Then simply searched correspondence between them. And so I found this data. Then I check my files table (http://server.titansoft.ru/akari/xeno-gears.txt) go to first file and start work.

ps: previous experience with FFVII background helps a lot.

757
This are the first results on Xenogears battle background.

First of all they are at 246419 - 253729 sectors of the disk. There are pairs of two files. First one are the battle background itself. The purpose of second one are still unknown.

Here are description of .bb (battle background) files

0x0000 u32 - number of sectors (always 05 00 00 00)
0x0004 u32 - 6 offsets to the blocks (last one offset to the end of file)
0x001C - till the next offset - tim textures and background (16 bit per piexl without CLUT)

# textures header
all offsets start from beginning of this block
0x0000 u32 - number of tim
next are for bytes offsets to tim
i still need more time to look at the header of each tim, but 0xC offsets from the beginning of each tim are height and width (2 bytes each)

# mech header
all offsets start from beginning of this block
0x0000 u32 - number of blocks
00 00 00 00 00 00 00 00 00 00 00 00 - unknown (as I see it always filled with zero)

#header to mech block
00 00 - unknown (zero)
84 00 - number of vertex
73 00 - number of mech (probably) (in this case each mech are always 12 bytes)
02 00 - unknown
28 01 00 00 - offset to vertex block
28 01 00 00 - offset to vertex block (again?)
48 05 00 00 - offset to next block (or end of vertex block)
(there something between start of next vertex block and beginning of vertex itself)
80 3B 00 00 - offset to mech block (probably)
00 00 00 00 - unknown (zero)
00 00 00 00 - unknown (zero)
95 01 74 FF 4F 00 53 01 B2 05 F2 FF 1D 03 00 00 00 00 00 00 B8 11 00 00 - unknown (perhaps some flags)

next are the second block and so, till the vertex data.
vertex format are the same as in FFVII
8 byte per vertex:
3 u16 x,y,z
1 u16 (unknown, probably textures coordinate or something)


758
Quote from: Micky
First of all, my documentation and viewer is here. Last time I checked it wasn't part of gears, yet.


My version of GEARS already has it (without source code of course). I use your algoritm in my viewer. I try to understand ways to work with 3d field formats before getting in Xenogears field files. I put it aside for time being and return to it after I finish with FFTactics field formats.

Quote from: Micky
I don't think the sections have a special meaning, except that the first one is used for the players to stand on and switched off/moved for some spells. They are probably just clipped as a block if they are outside the viewing volume.


Perhaps you are right, although I don't think Square leaves in the file something that are not used at all. I'll look more closely to it.

Quote from: Micky
By the way, did you make any progress with Xenogears?


I need one or two weeks more before I can return to work on Xenogears (I'm working on new version of my website (http://rpgames.ru) and FFT)

759
I have a little question about battle backgrounds from FFVII.

The first block contains some unknown info, the second block are the platform where characters stand, 6 - number of blocks-1 - are surrounding items, the last block are the textures.

I have just one question. What are the 3-5 blocks. They all are the same. And contains mech that are far away from the others. What is it, did someone knows?

760
I finished check the both disk. They seems identical in every way, except video files
Code: [Select]

DISK 1

000041 str0001.str capsule launch
000825 str0002.str Grahf, Id and Fei's father
004705 str0003.str Grahf, Id and Fei's father 2
008425 str0004.str Weltall transformation
011055 str0005.str Before Lahan destruction
016743 str0006.str Lahan destruction
022911 str0007.str Meet with Elly
025941 str0008.str Desert trip
036637 str0009.str You are coward
045581 str0010.str Grahf appearing
050893 str0011.str Solaris destroy
052165 str0012.str Solaris destroy 2
053149 str0013.str Solaris destroy 3
054133 str0014.str Opening
098917 str0015.str Yagdrassil launch
104697 str0016.str Playing with mom
106105 str0017.str Squaresoft logo

DISK 2

000041 str0018.str ending, with credits
067433 str0019.str destroying Melkaba 1
072505 str0020.str destroying Melkaba 2
074097 str0021.str Deus appering
079569 str0022.str after final battle
117697 str0023.str Deus destruction
120313 str0024.str music video (Two Wings)
146867 str0017.str Squaresoft logo
149313 str0025.str Elly's sacrifice
158377 str0026.str Fei and Elly
160745 str0027.str From Elly to Miang
166297 str0028.str Fei kills everyone

and block with 3d locations and events
Code: [Select]

DISK 1
120750 - 231361 - great amount of data (more than 100mb) as I figure first file is data descriptor. the second byte is always 12, the first one 01 or 00
after this file are lies one or more files with real data.

DISK 2
185064 - 221427 - much smaller amount of data. It seems that almost all scenes from the first disk was cut as well as many locations (It make sense)



As well I found that 03 00 00 00 14 00 00 00 file identificator in real some sort of archive. The first 4 bytes are number of chunk, the second one are first chunk offset. Next 4 bytes are the second chunk offset and so...
There are few type of data that are stored in this archives (or file packes) and in seems determinated by number of chunks. For example 08 00 00 00 - are pack of TIM graphic.

Code: [Select]
120579 8800 ????.tim tim pack (8 tim) contains compass staff, dialog border and so...

Others are still needed to be decrypted. I met 02, 03, 04 and 05 packes (08 was only once)


Right now I try to understand file that I believe is 3d location of Fei's room in Lahan, where game started. (sector 122139 and 122140)

PS: xeno-gears.txt was apdated a bit.

761
Year, i searched a bit. The most data are the same as it the Disk 1, exept video of course. And it seems like some other data absent too. I'll check more intently soon.

762
Okey, thanks, halkun. Your documentation and Gears helped a bit.
I managed to find standart .str and .tim files. Futher on I found .smd, .sec and .wd files like in Final Fantasy Tactics. They are music files. Right now I try to interpritate files with "30 00 00 00" file type identificator.

Here are all info that I get
http://server.titansoft.ru/akari/xeno-gears.txt

763
Okey. I've got some results here. After week working with raw data and iso9660 specification I can say for sure, there is no usual directory/files of any kind. All path tables are has only record about root directory, and file structure itself doesn't contain info other than two files and link to "." and ".."

The SLUS_006.64 and SYSTEM.CNF are pleced in the middle of the disk (strange???). So I've made two more records in the file structure - F_PART.RAW and L_PART.RAW. They uncover raw date before and after two initial files. The hexes are folowing

Code: [Select]
Offset      0  1  2  3  4  5  6  7   8  9  A  B  C  D  E  F

0000B000   30 00 16 00 00 00 00 00  00 16 00 08 00 00 00 00
0000B010   08 00 62 08 1B 11 1E 00  00 02 00 00 01 00 00 01
0000B020   01 00 00 00 00 00 8D 55  58 41 00 00 00 00 00 00
0000B030   30 00 16 00 00 00 00 00  00 16 00 08 00 00 00 00
0000B040   08 00 62 08 1B 11 1E 00  00 02 00 00 01 00 00 01
0000B050   01 01 00 00 00 00 8D 55  58 41 00 00 00 00 00 00
0000B060   3C 00 3E A8 01 00 00 01  A8 3E 00 A0 04 00 00 04
0000B070   A0 00 62 08 1B 11 1E 00  00 00 00 00 01 00 00 01
0000B080   0D 53 4C 55 53 5F 30 30  36 2E 36 34 3B 31 00 00
0000B090   00 00 0D 55 58 41 00 00  00 00 00 00 3C 00 3D A8
0000B0A0   01 00 00 01 A8 3D 43 00  00 00 00 00 00 43 62 08
0000B0B0   1B 11 1E 00 00 00 00 00  01 00 00 01 0C 53 59 53
0000B0C0   54 45 4D 2E 43 4E 46 3B  31 00 00 00 00 00 0D 55
0000B0D0   58 41 00 00 00 00 00 00  3A 00 18 00 00 00 00 00
0000B0E0   00 18 00 28 41 0D 0D 41  28 00 62 08 1B 11 1E 00
0000B0F0   00 00 00 00 01 00 00 01  0C 46 5F 50 41 52 54 2E
0000B100   52 41 57 3B 31 00 00 00  00 0D 55 58 41 00 00 00
0000B110   00 00 3A 00 D2 A8 01 00  00 01 A8 D2 00 00 07 18
0000B120   18 07 00 00 62 08 1B 11  1E 00 00 00 00 00 01 00
0000B130   00 01 0C 4C 5F 50 41 52  54 2E 52 41 57 3B 31 00
0000B140   00 00 00 0D 55 58 41 00  00 00 00 00 00 00 00 00
0000B150   00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00
0000B160   00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00


This must be placed in sector 22 instead of original data (not in raw mode). I used WinHex for it. (PLEASE NOTE THAT THIS IS FOR DISK 1, disk 2 may have other file pointers)

After that point a confront some trouble. First of all I used Clone CD to made image and then edit it. But when I wrote the disk It seems to be the copy of original, without any changes. I tried once more time, but first 23 sectors became blank all of the sudden.

Next I tried to use UltraISO. I made ISO image edit hexes in it and burn them back. Windows was happy, and I too - all files appear and they was exactly the same as raw data of the original, BUT!! - the PS2 and some emulators refused to play that disk. Though I get files to analize, next step is to allocate the other data and perhaps fix incompartible with PSOne. I start from the video i believe. Don't anyone knows where I can get description of PSOne movie formats?

764
I have a little question about Xenogears CD file system. In the root of the disk there are only two files: SLUS_006.64 and SYSTEM.CNF, but nothing more. Although when I've searched the disk with the special tools i found much more info.

Do all the other files exist or are there only raw data on CD? Does a way to recover all the other files exist? How did they hide all of the game files?

765
Ow! great! I just started working with ff7 raw data, because I found this place not so many time ago. But already wanted to remake FFIIV engine. So this is just in time. You can count on me.

And CVS would be great!

*sorry. My English is really baaad*

Pages: 1 ... 26 27 28 29 30 [31]