Hi there! You are currently browsing as a guest. Why not create an account? Then you get less ads, can thank creators, post feedback, keep a list of your favourites, and more!
Quick Reply
Search this Thread
Field Researcher
Original Poster
#101 Old 9th Mar 2010 at 6:17 AM
I loaded the bed into MilkShape, exported it into the milkshape_ascii format which can be imported into Blender, but it failed loading. "Number of Bones invalid". If I import the ms3d format it goes in okay and I can see the bones which shows me the rotation IS wrong. I looked at the files, but there are some differences, but nothing explicit. More checking though
Advertisement
Field Researcher
Original Poster
#102 Old 9th Mar 2010 at 3:23 PM
I think I've now worked out how the rotation work, just need to test my theories.
Field Researcher
Original Poster
#103 Old 9th Mar 2010 at 9:43 PM Last edited by Ceixari : 10th Mar 2010 at 7:11 AM.
I've uploaded the latest version of the test file (WIP). The bones of the bed looks much better now, although in MilkShape they sit a bit lower. Exporting it out is going to be fun.

If you look in the .s3ascg file under the skin section, you see something like:

Code:
skin 109
0 3717ED16 -1.000000 0.000000 0.000000 -0.782125 0.000000 0.000000 1.000000 0.530208 0.000000 1.000000 0.000000 -0.715816


I've been able to work out that you need to look at it this way:

Code:
skin 109
0 3717ED16 

     X        Y        Z     Co-ordinate 
-1.000000 0.000000 0.000000 -0.782125 
 0.000000 0.000000 1.000000  0.530208 
 0.000000 1.000000 0.000000 -0.715816


It now depends on whether a value in the X,Y,Z is not 0.000000 and in this case the bone is built using (-0.782125, -0.715816, 0.530208). Blender is a bit different, so I need to flip the Y and Z, but the principle stays. The -1 might have a relevance in that I need to multiple whatever is in the X,Y,Z columns with the co-ordinate, I've have a look (this thought came as I was typing this)

In the clock for instance, they look like:

Code:
skin 3
0 257E78E0 

1.000000 0.000000 0.000000 0.000000 
0.000000 1.000000 0.000000 -2.330000 
0.000000 0.000000 1.000000 0.397285


In this case, it's just a straighforward top to bottom: (0.000000, -2.330000, 0.000000)
˙uʍop ǝpᴉsdn ǝɹ,noʎ 'oN
#104 Old 10th Mar 2010 at 1:53 AM
So, that's why it worked on the clock and not on the bed? If so, you may have found a clue to why our beds weren't working right. In MS, they seem flipped. With your importer they seem like they are where they need to be(maybe Wes should know about this?). The only problem was that no vertex groups were assigned. The clock still imported fine with correct vertex assignments.

Just curious, but on export wouldn't the bones stay the same and only the vertex group's assignments change? I thought the bones were just a representation of the bones in the rig file.

"Part of being a mesher is being persistent through your own confusedness" - HystericalParoxysm
| (• ◡•)| (❍ᴥ❍ʋ) [◕ ‿ ◕]
Field Researcher
Original Poster
#105 Old 10th Mar 2010 at 8:03 AM
That "no vertext groups were assigned" had me worried, because the code won't actually create the bones if there is no assignment. This looks like another refresh/redisplay type problem.

At the moment to see the bones you have to switch to another layer and switch back. If you go to the Outliner view, it looks as if there is nothing, but if you just resize that view a bit, voilà! magic happens I'll play around with some refreshes.

I'm going to make a little change for the next version, since we need to change it every time we import something. In the editing options of the Armature, I'll switch on X-Ray and change the Display Options from STICK to OCTAHEDRON
Screenshots
˙uʍop ǝpᴉsdn ǝɹ,noʎ 'oN
#106 Old 10th Mar 2010 at 8:35 AM
Quote:
I'll switch on X-Ray and change the Display Options from STICK to OCTAHEDRON

You read my mind...


eta: Your pic illustrates what I mean. The only bone that has something assigned to it is CD68F001.

"Part of being a mesher is being persistent through your own confusedness" - HystericalParoxysm
| (• ◡•)| (❍ᴥ❍ʋ) [◕ ‿ ◕]
Field Researcher
Original Poster
#107 Old 10th Mar 2010 at 9:04 AM
My pic shows only the assignment to group03. If you expand group00_a, you'll see all the rest there. Unfortunately, there was a reason for doing it this way, rather than assigning all the bones to the same group. It was a change necessary after looking at the error you reported earlier where the bed imported, but the armature was throwing an out of index error. In that version of the script, I attempted to assign all the bones to the same group, but it wasn't happy with that method. Removing duplicates and assigning the bones per mesh group worked much better.
Field Researcher
Original Poster
#108 Old 10th Mar 2010 at 9:12 AM
I've just uploaded Version 0.2.0 in two seperate files. Obviously we still need the version without bones as we still need to export back into MilkShape for the time being and .obj does not export bones.. I'm versioning them the same, so we know that one isn't getting left behind.

Please remember to delete any files in your .script directory that start with s3asc.
Field Researcher
Original Poster
#109 Old 12th Mar 2010 at 11:44 AM
In order to do the exports I need to add extra information (in text objects) which will be saved as part of the .blend.

These would include things like a matrix containing group names against file names which means the export will read these and automatically save each group as the correct file and also be able to thow warnings if for instance you imported 3 groups, but are only attempting to export two.

I'll also be storing most of the import text as it contains data we're not actively using. The VRTF section, for instance, has no relevance in Blender, but we still need to export it.
Test Subject
#110 Old 17th Mar 2010 at 12:21 AM
Can somebody help me please. I looked around here and can make no sense of things sofar. If I am in the wrong place let me know, I dont know how to start a new threat.
I am having problems with the plugins from Weshowe. I dont know where to put them in Blender. I put them with the other dll,s but they dont show in Blender. As such I can not import the files from the extracted files from the sims3. Can anybody explain in plain very plain english what to do. I am a total newbie and dont understand all the gooblediegoob yet. Somebody please help.
Test Subject
#111 Old 17th Mar 2010 at 12:23 AM
me again. I dont do sims yet I am only trying to do objects. Thanks
Alchemist
#112 Old 17th Mar 2010 at 1:04 AM
Hi Flowerpower

WesHowe's plugins are for Milkshape. If you are trying to use them for Blender they will not work. They will only work for Milkshape.

The thread that you have just posted in is a thread for an experimental version of Blender plugins. This experimental version is able to import objects, which have been decompiled using WesHowe's ObjTool, into Blender. It is not able to export them yet.

If you want to create new objects right now you will need the ObjTool from WesHowe and you will need Milkshape.

If you are still learning how to make objects, or are willing to wait for the Blender plugins to be finished, then you can wait until the experimental plugins for Blender are completed.

Do you have Milkshape and would like to learn where to put WesHowe's plugins, or would you rather wait for the Blender tool to be completed?

If you would like to start a new thread there is a button that should be located on the upper left hand side of your screen, right over the left-hand side of the area where you have posted, and it is labeled New Thread. If you click that button you will get a box to post in that is like the one you have posted in to post your posts above.

I hope this helps.
Forum Resident
#113 Old 17th Mar 2010 at 11:37 AM
Hay guys, I'd be happy to test ; ) I really am new to all things meshing though so all the help I could give would be from a super newbie perspective........

Disclaimer: These are the personally, personal opinions of me, myself and I. Yours may vary.
Terms & conditions do not apply
Field Researcher
Original Poster
#114 Old 17th Mar 2010 at 8:19 PM
The export script is coming along nicely, but for the moment I would recommend using MilkShape until it's completed as that is the only viable option at this stage. If you're more comfortable with Blender you can of course use the import script, do what ever changes you want to do then export in .obj format for import into MilkShape.
Test Subject
#115 Old 18th Mar 2010 at 9:35 PM
Thank you orangemittens. I have a trial version of milkshape but can not afford to buy it. So I suppose I have to wait for the Blender plugins. Thank you for your help.
Alchemist
#116 Old 18th Mar 2010 at 9:56 PM
NP. And you can use Milkshape with no problem until your trial period runs out.
Alchemist
#117 Old 19th Mar 2010 at 3:27 AM
Quote: Originally posted by Ceixari
The VRTF section, for instance, has no relevance in Blender, but we still need to export it.


Currently, the usage of the VRTF in the MilkShape plugins is to match the bone names (via hash) to the bone index values. The previous version is parsed for the VRTF data, the bone indices are mapped to match the output, the previous file is renamed as a .bakNN, and the data read from the vrtf is just written to the new file.

The compiler carefully matches the vrtf from each of the .s3ascg files with the previous binary version, before making a new binary file.

We have no way at this time to create or really edit the skeleton. You cannot add new joints, but you do not have to use them all. The VRTF section needs to have them all, but there is no requirement to actually assign any of the mesh to them. MilkShape does a primitive scan for unassigned vertices, but that is there as a helper for people that forget, it is a warning message rather than an error message.

If you like to say what you think, be sure you know which to do first.
Field Researcher
Original Poster
#118 Old 19th Mar 2010 at 8:37 AM
Thanks for highlighting that Wes. Something to keep in mind.
Theorist
#119 Old 21st Mar 2010 at 4:00 AM
Yay! This is so great! I can't wait until this is finished! My trial period ran out a while ago and still being in school without a credit card makes me out of luck...

Hi I'm Paul!
Field Researcher
Original Poster
#120 Old 21st Mar 2010 at 2:37 PM
SO far I have been able to get teh export script to write out the VRTF, the bones, the vertex coordinates, the normals, the UV, the bone assignments and the faces. This is basicaly:

VRTF
SKIN
VBUF [0,1,2,3]
IBUF

Thers is still VBUF [4,5,6] to do, skin weight I can get and tangent, but I'll have to play with these for a while.
Field Researcher
Original Poster
#121 Old 22nd Mar 2010 at 1:31 PM
Wes, if you pass by here again...

What do you do in the case of VBUF-6? Do you just reload what it had before or do you pass zero values?

Thanks
Field Researcher
Original Poster
#122 Old 22nd Mar 2010 at 11:01 PM
Except for VBUF-6, it looks like the export script is now testable. I'l run a few imports and exports tomorrow all the way from S3OC through Blender and all the way back into TS3 to make sure the objects still work and then one or two with some modifications. After that I'll update on the possible upload of the scripts.
Field Researcher
Original Poster
#123 Old 23rd Mar 2010 at 3:42 PM
Well, the ObjTool is recompiling the export files, so that looks good. I'm having an error in some cases, but I know why and I'm having to play with the bones in the import script to fix the error.
Field Researcher
Original Poster
#124 Old 26th Mar 2010 at 3:27 PM
I had a problem with exporting the unique bones set back out to the relevant files they came in as, but I think I have that sorted. I'm fixing errors as I'm recompiling them with the ObjTool. I'm hoping to put version 0.2.1 of the files up in the next few days.
Field Researcher
Original Poster
#125 Old 26th Mar 2010 at 4:14 PM
Current error from S3ObjTool is:
Code:
ERR: Invalid item count 3: expected 5 [vbuf tang: line 19969, group 0].

My files looks quite similar to the exact same object imported and exported to/from MilkShape, so it's taking me having to look at the files line by line to see the difference.
Page 5 of 10
Back to top