Bionicle: The Game Models and textures (Research and Tools)

Recommended Posts

Some of you may already know from my posts in the "Game Archive Extracting (Research and Tools)" topic that I was working on scripts that would allow to view and extact the models and textures from Bionicle: the Game. Well, I think developed them to the point where I can make a public release. The model script became quite complicated with all those "Choose the model" things and the way I made the whole choosing process is not the best, but I think it's bearable.

You can find the scripts here: https://mega.co.nz/#F!go8SGQZC!Xk3MTMLc-AlIAUYI3mT3Hg
I've explained how to use them in the Readme file included in the folder above.

Model script is currently v0.1 beta
Texture script is currently v1.0

Here is a research I did on the game's resources:

Spoiler

I've been searching through the models in search of unused content. I didn't really manage to find much, but there's still alot of things that I found or noticed.

Let's start with Toa Mata. The models are all present and this is not surprising as all of them can be seen in the extras. Here is a picture of all 6:

I suppose you can tell even from this picture that the models of Pohatu and Gali are the least finished. It is especially noticable on Gali as the there are a lot of places where the textures look weird. I suppose the plans for their levels were scrapped even before they could polish the models thus they were never finished.

As for the playable Toa, I noticed thet every one of them has two models - one for the gameplay and one for cutscenes. Here is a picture with them all:

*sigh* Takanuva has the stupidest default pose I've ever seen. On a serious note, some model seem to have quite noticable differences. For example Onua, Takanuva, Tahu Nuva (for some reason mask on both of his models looks horrible).

Moving on to the matoran. I think all of you who played the game noticed that there are a lot of inconsistencies with them. For example on the first level matoran were green instead of red (curiously there is also a model called "Le-Wahi Skydome" inside). There are even more oddities inside the files. Almost every level that has matoran has their models under the name "Mesh - TaMatora" (no "n" because it wouldn't fit in 15 characters limit the format has for model names). The only exceptions are Kopaka's and Pohatu's levels. The packages of Kopaka's level has them listed as "KoMatoran" and the only matoran on Pohatu's level is called "KoMatoran" too, presumably because it's just a recolored Ko Matoran. I took a picture of every matoran I could find (excluding Takua), here's a compilation of them:

I signed them with the filename of the package they are from, which corresponds to a certain level. Here is a list of them:

10 - Tahu Mata
15 - Tahu Nuva
20 - Lewa Nuva
21 - Lewa Nuva
22 - ???
30 - Gali Nuva
40 - Onua Nuva
50 - Pohatu Nuva
52 - Pohatu Nuva??
53 - Pohatu Nuva??
60 - Kopaka Mata
75 - main menu and extras
78 - Final Battle

And finally, the bohrok. All of them are present, but the kal are just recolored standart bohrok. Here's a picture:

Krana models are present in levels of Tahu, Kopaka, Gali Nuva amd Lewa Nuva. It's actually the same model with different textures. By the way, while watching some random let's play in order to confirm something I accidently discovered that there is a difference in Krana on Kopaka's level between PC and PS2 ports (not sure about the others). As you can see on the picture below they have completely different colours.

If you have any issues with the scripts - feel free to ask me .
Oh, and also I don't remember if the models of turaga Matau, Onewa and Whenua were in extras, but I can't find them anywhere. The same goes for Onu matoran. If you happen to find them or any other interesting model/texture - please be sure to post it there .

Share on other sites

I've been trying to get the BioBIGB111.py file to work for hours (bear in mind I'm a bit of a noob). Could you explain how to use it to extract files?

Share on other sites

I used a .bat file with something along the lines of this:

python.exe BioBIGB111.py C:GamesBionicle10_vol10-00.PCS

Share on other sites

Excuse the necroposting, but I tried to use these utilities (trying to get into the PS2 version) However, I have no idea on how to open the Game.WAD file in the extracted ISO. There was a python and BMS file in the downloaded folder with "WAD" in their titles, however they are not even mentioned in the readme, so I have no idea on what they are supposed to do. If anyone is there, can I please get help?

Share on other sites

Just by looking at the OP I will assume that the .py file goes in the \plugins\python\ folder of Noesis, and the .bms files is intended to be used with QuickBMS.

BMS is meant for extracting/converting files, and Noesis is meant to view those files.

Share on other sites

Well, that's a lead. but it's a small one that doesn't move me forward; When I tried putting the .py file in the Noesis plugin folder it didn't show up for some reason (even if it's the only one of Maver1k's scripts I put in the folder) and I did download QuickBMS, however it has trouble reading the file.

I can put down what QuickBMS has been telling me:

Error: incomplete input file 1: c:\path\to\file\GAME.WAD

Can't read 16 bytes from offset 505c3a4c.

coverage file 1   0%  0  1162915840 . offset 505c3a4c

last script line before the error:

17 log name startOff size 1

I did find out that Noesis can extract files from an ISO (even though it leaves a ";1" behind the filetype.) there are other formats in Noesis's plugin folder that use the .wad filetype, but none seem to be the BioniclePS2wad.

(also, just in case you are wondering, the WAD file I put through QuickBMS was taken from the ISO using a normal extractor, not Noesis.)

Share on other sites

I think I figured out what is the problem here.

So I tried to extract the WAD file using the WAD_BionicleTheGame_PS2 script myself and initially got the same results. Looking at the code now I can tell that it's because you are supposed to select DIR file instead. That logic may seem a bit confusing, but DIR files actually contain the list of the files in the WAD which is why I used it as the primary one. Then I called the script "WAD unpacker" because that's where the files actually are. Sorry for the inconvenience, I should have left a note about that somewhere. Well, I did leave a comment inside the script now.

Oh and about that other Noesis script - it's for creating WAD archives.

Share on other sites

Thanks. I'm glad to have that cleared up.

(Edit: The files that were extracted using QuickBMS have the extension "PSS" and "PSM" instead of "PCS" and "PCM". Maybe it's because I'm extracting from the Playstation 2 version or something? I absentmindedly put those files through the BioBIGB111 and they did seem to extract without error. I just hope they will register in Noesis)

Edited by Tim5corpion

Share on other sites

I have a theory that PC in PCS and PS in PSS actually stand for PC and PlayStation respectively. From what I remember their structure is very close so the BioBIGB111 script works fine for both platforms. I'm not sure if you'll be able to get anything to load in Noesis though. I still have some of the extracted files from the PS2 version and it only loads textures and they don't really look correct. I don't remember if I ever tried looking in depth into PS2 model format but I'm pretty sure it was different enough for the plugin to be unable to open them. Well, I also had some barely working code that imported parts of level geometry with various degrees of success. It's not on the public folder because it doesn't really work unless I manually set the parameters and even then the results look something like this at best .

Share on other sites

Yikes.

I could give some information though about ripping from the Gamecube version. The popular GCN emulator, Dolphin, has a built-in ripping utility I have used to successfully rip from the Bionicle GCN Iso. I have discarded the results, though, since none of the focus here appears to be on the Gamecube version. (sorry)

Share on other sites

Sorry for doubleposting, but I can tell you that, yes, the PS2 formats were unable to be viewed.

However, I did manage to get ahold of a PC version and I did get the models from that to appear in Noesis! SUCCESS!!!

I tried exporting Tahu as an OBJ to blender, and the model got there, and apparently the UV data did too, but the textures didn't. Looks like I'll have to put the textures on manually. argh.

I did note that the exported textures have a very high alpha value, being very transparent. Interesting.

UPDATE: Didn't take much to put the textures on, since the part materials corresponded to the exported textures. I just had to set "use alpha" to off and flip the textures by the Y axis, and the end result on Blender turned out swell.

Share on other sites

It's good to hear that my plugins are being useful!

I'm not entirely sure but the alpha in the textures was likely used as a specular map or something similar to that. It's probably useless outside of the game. By the way Noesis may export with UV map upside down for some output formats. There is a check box on the export window to prevent that. Well, flipping the textures works fine too, the end result is the same anyway.

Share on other sites

Although I've been busy lately, I would like to make a small contribution to the ripping research.

You know how some models in the Pcs files look like identical giant masses of polygons? out of morbid curiosity, I took the mass that was supposed to be a lavaboard (10 vol 10-00 pcs model 39) and actually exported it to blender. When I displayed it as a wireframe, I noticed something.

It appears that these mesh-messes in the Pcs are, well... everything. it's every single model in the Pcs combined with a single texture. Even the models that can't display normally I saw in there. I saw the life meter, the mask of elemental energy, and the one that I was looking for, Tahu Nuva's surfboard. (Yes. His Nuva form surfboard is in his Mata level for some reason. Don't ask.)

I picked out the connected polygons that should belong to the surfboard, and there were two places where points of the surfboard connected to obvious other objects. There was an edge leading to a series of "walls", and another edge leading to a rod of some sort.

Despite those two errors, not only was the mesh of the surfboard intact, but the UV's were intact and they looked correct with the surfboard's texture (And not upside down because I used the export option you described to me previously.) I see this as a key discovery, since the "messed up" models ARE in there, but I believe the current viewing script can't seem to distinguish where certain models begin or end, and the end result is a mess of everything in the PCS files.

I do have to wonder though... The PCS extractor splits the PCS into two different blocks. I know I'm not part of the people who made the scripts and you can tell me otherwise, but is there missing data in block 1 that got moved to block 0 during the split? Because i'm sure the viewing script is not getting data from both blocks at once.

Create an account

Register a new account

• Recently Browsing   0 members

×

• Activity

×
• Create New...