Replies: 95 (Who?), Viewed: 36357 times.
Page 2 of 4
Warrior Gryphon
site owner
#26 Old 18th Oct 2009 at 4:19 PM
I'd be willing to work with somebody on this to get accessory meshes working properly using CTU. To do that, I'd need to know some things:

1. What is currently limiting CTU in doing this?
2. What could CTU do better to allow these meshes to work?

Story books are full of fairy tales, of Kings and Queens, and the bluest skies.
Test Subject
#27 Old 18th Oct 2009 at 7:18 PM
I hope that in sharing my learnings of the past 2 days, I can help a bit with sorting this out.

I am making smaller rectangular glasses (normal glasses, not sunglasses) starting with the game mesh "amAccessoryGlassesRectangle".

When using the CTU to extract the meshes, it extracts 8 meshes:

amAccessoryGlassesRectangle_lod0_0x000000006CF8849C.simgeom
amAccessoryGlassesRectangle_lod0_1_0x00000000895912A0.simgeom
amAccessoryGlassesRectangle_lod1_0x000000006CF8849D.simgeom
amAccessoryGlassesRectangle_lod1_1_0x00000000C95BB5D9.simgeom
amAccessoryGlassesRectangle_lod2_0x000000006CF8849E.simgeom
amAccessoryGlassesRectangle_lod2_1_0x00000000C95DF44E.simgeom
amAccessoryGlassesRectangle_lod3_0x000000006CF8849F.simgeom
amAccessoryGlassesRectangle_lod3_0x000000006CF8849F.simgeom

Together with a file named
736884F1_00000001_0000000000EB1B5A.vpxy

I edited all 8 these meshes, and then (tried to) import them into the CTU in the following way:

amAccessoryGlassesRectangle_lod1_0x000000006CF8849D.simgeom into LOD1
amAccessoryGlassesRectangle_lod1_1_0x00000000C95BB5D9.simgeom into LOD 1_1
amAccessoryGlassesRectangle_lod2_0x000000006CF8849E.simgeom into LOD2
amAccessoryGlassesRectangle_lod3_0x000000006CF8849F.simgeom into LOD3

After renaming the mesh "amAccessoryGlassesSmallerRectangle" and hitting "commit", when I add a design (whether it's a new blank design I design from scratch or copied from base) the 3D viewer loads the original game mesh(es).

I then "Save As", close the CTU and open it again to Load the saved package. New meshes load, but the glasses of the glasses are AWOL. They don't show up in-game either, although the newly meshed frame does.

I experimented with loading the meshes into different LOD fields, but nothing worked.

Using TSRW (the current public version, not the new one that's still in testing) I exported the 4 "vertices collections" (or whatever they're called in TSRW) wavefront objects and edited all 4 (in exactly the same way I did the other 8). I re-imported them into TSRW and it assembles just fine, and everything shows up correctly in-game.

I compared the files in MilkShape, and found that the following:

(NOTE: I named the files exported from TSRW in the following way: "Mesh1.obj" for the first (topmost) collection of vertices, Mesh2.obj for the collection below that, Mesh3.obj for the collection listed third and Mesh4.obj for the last (bottom) collection of vertices.)

Mesh1 is the same as amAccessoryGlassesRectangle_lod0_0x000000006CF8849C.simgeom
(This looks like a lower-detail frame)
Mesh2 is the same as amAccessoryGlassesRectangle_lod0_1_0x00000000895912A0.simgeom
(This looks like a set of glasses to fit inside a frame)
Mesh3 is the same as amAccessoryGlassesRectangle_lod1_0x000000006CF8849D.simgeom
(This looks like a higher detail frame)
Mesh4 is the same as amAccessoryGlassesRectangle_lod1_1_0x00000000C95BB5D9.simgeom
(More glasses to fit into a frame)

I tried to load the meshes into the LOD fields of the CTU using this information in the following way:
amAccessoryGlassesRectangle_lod0_0x000000006CF8849C.simgeom into LOD1
amAccessoryGlassesRectangle_lod0_1_0x00000000895912A0.simgeom into LOD 1_1
amAccessoryGlassesRectangle_lod1_0x000000006CF8849D.simgeom into LOD2
amAccessoryGlassesRectangle_lod1_1_0x00000000C95BB5D9.simgeom into LOD3

Still no glasses inside the spectacles.

I am more than willing to accept that this is an ID 10 T error stemming from the fact that I am made of fail, but it seemed interesting nonetheless, and any help would be appreciated.
Sockpuppet
#28 Old 18th Oct 2009 at 7:36 PM
Quote:
Originally Posted by EsmeraldaF
Do you mean sunglasses? I can make sunglasses with TSR Workshop with no problem, EXCEPT, they come out very pixellated, which really doesn't look at all good. TS3 desperately needs decent sunglasses, especially for men, and I've made the ones below for both men and women, which would look good if only they weren't all pixellated around the edges:

TSR Workshop is very limited and it only uses .obj files, which maybe doesn't produce as good quality as .simgeom files do? Or maybe TSR Workshop itself just isn't much good at producing good quality graphics? If I can figure out how to package the .simgeoms using Postal, I wonder if the glasses would look better quality. Have you successfully made non-pixellated-looking glasses that way?


Ill try to make a set with CTU.

I also made earlier accesoires with TSR workshop and on the mesh tab i exported all DDS files and then reimported them back again.( i used to do the same with meshes.)
This will give you the Texture files unpackaged(seperated from the GEOM) and are much easier to edit.
TSR workshop however kills the vpxy files(and all morphs) while CTU gives you all the necesarry files wich i think is better to make a complete package.

Ouch, i extracted all meshes from the unisex glasses......this will take me a while...
Mad Poster
#29 Old 18th Oct 2009 at 11:36 PM
Quote:
Originally Posted by Delphy
I'd be willing to work with somebody on this to get accessory meshes working properly using CTU. To do that, I'd need to know some things:

1. What is currently limiting CTU in doing this?
2. What could CTU do better to allow these meshes to work?


I'm unfortunately too much of a newbie at meshing to know or understand all the ins and outs, but what little I've been able to determine is:

Results depend on which version of CTU you're using - the latest one (which lets you import LODs 0_1, 1_1 and 2_1) or the previous one (which only lets you import LOD1_1). I've packaged the same pair of shades with both versions, importing the frames as LOD0 (or LOD1 in the older CTU) etc and the lenses as LOD0_1 (or LOD1_1) etc. I then opened each of them with Sims 3 Package Explorer.

I found that the older CTU apparently ignores the inputting of the LOD1_1, with S3 Package Explorer showing only three .simgeon files instead of the expected four. On opening each of them in Milkshape, they turned out to be the three frames LODs. Opening the package in CTU, the 3D preview shows empty frames as below:


But, looking at the package created with the latest CTU, S3 Package Explorer shows all LODs accounted for. And in CTU, the 3D preview shows a complete set of sunglasses. So, it looks hopeful and like it should work:


However, both versions of the sunglasses cause TS3 to crash in CAS when scrolling down through the Accessories panel in the attempt to get to the new glasses. I have tried deleting all of my cache files each time, but no joy.

Creating these same glasses in TSR Workshop causes no problems in game and they work fine, but they look like cr*p because they come out so pixellated!
Sockpuppet
#30 Old 19th Oct 2009 at 9:04 AM
The latest CTU says that you only need LOD0 when doing hair or accesoires.
Strange....i do not have that particular glasses...:D
Mad Poster
#31 Old 19th Oct 2009 at 9:30 AM
Quote:
Originally Posted by Base1980
The latest CTU says that you only need LOD0 when doing hair or accesoires.
Strange....i do not have that particular glasses...:D


Do we ONLY need a LOD0 for accessories? I assumed you needed LODs 1, 2 and 3 in addition to LOD0. Could that be the problem, should we only be inputting a LOD0 (and LOD0_1) and no other LODs? Also, as far as I know, you can't use CTU for hair - you need to use DABOOBS for that.
Warrior Gryphon
site owner
#32 Old 19th Oct 2009 at 11:43 AM
It says you only need to USE LOD0 when doing hair/accessories. But you do still need to put the other LODs in

Esmeralda can you send me the package file that was done with CTU that crashes the game?

Story books are full of fairy tales, of Kings and Queens, and the bluest skies.
Lab Assistant
#33 Old 19th Oct 2009 at 4:12 PM
and now i'm having an issue with importing into game >< I can see my mesh in CTU but game shows me only my created designs. help?
Warrior Gryphon
site owner
#34 Old 19th Oct 2009 at 4:28 PM
Merkada: Can you attach the file to the post?

Story books are full of fairy tales, of Kings and Queens, and the bluest skies.
Mad Poster
#35 Old 19th Oct 2009 at 4:36 PM
Quote:
Originally Posted by Delphy
Esmeralda can you send me the package file that was done with CTU that crashes the game?


Delphy, my most abject and embarrassed apologies, it was my own dumb fault those packages were crashing in-game - I'd made the .simgeoms for those packages out of the .obj files that I'd used in TSR Workshop rather than from going back to the original .simgeoms of the TS3 sunglasses I cloned them from and starting again, and I think they must either lose something in the translation from obj to simgeom or get screwed up somehow by TSRW. I'd started wondering if that could be the problem, and decided to quickly mock up a rough draft new version of my sunglasses by re-modifying the original simgeoms. The resulting package does NOT crash TS3.

But, the resulting package still shows up in CTU as complete with frames and lenses, but with no lenses in TS3:




Here's a copy of that package - if you look at it with Package Explorer or similar you can see all the meshes are in there, but for some reason the lenses are not showing up.

http://www.modyourpanties.com/hosti...testnew.package
Warrior Gryphon
site owner
#36 Old 19th Oct 2009 at 4:41 PM Last edited by Delphy : 19th Oct 2009 at 5:40 PM.
Esmeralda: Excellent thanks for the upload. I'll take a look now for you.

Edit: Okay all the VPXY links look ok. Going to test it in game now.

Edit: Lenses indeed don't show up - even with swapping the frame and lens links in the VPXY. This leads me to think that perhaps something else is going wonky - maybe in the simgeom.

I still have some unreleased tools I can use to examine the glasses.

Out of curiosity, what was the original mesh you used as a base?

Story books are full of fairy tales, of Kings and Queens, and the bluest skies.
Lab Assistant
#37 Old 19th Oct 2009 at 5:44 PM
this is what it looks like in CTU (a small hat):


this is what my game gives me:


and there is another issue. sometimes if I try to open my package in CTU, it says "Sorry, we couldn't find a mesh!" or the mesh is completely black and nothing can be done with it.
Mad Poster
#38 Old 19th Oct 2009 at 6:28 PM
Quote:
Originally Posted by Delphy
Esmeralda: Excellent thanks for the upload. I'll take a look now for you.

Edit: Okay all the VPXY links look ok. Going to test it in game now.

Edit: Lenses indeed don't show up - even with swapping the frame and lens links in the VPXY. This leads me to think that perhaps something else is going wonky - maybe in the simgeom.

I still have some unreleased tools I can use to examine the glasses.

Out of curiosity, what was the original mesh you used as a base?


Thanks for looking into this! I used afAccessoryGlassesOversized as the base.
Warrior Gryphon
site owner
#39 Old 19th Oct 2009 at 6:52 PM
Esmeralda: Can you provide the TSRW version too? So far I can't see any major inconsistencies in either the VPXY or the GEOM (or the MTNF) parts

Story books are full of fairy tales, of Kings and Queens, and the bluest skies.
Mad Poster
#40 Old 19th Oct 2009 at 7:41 PM
Quote:
Originally Posted by Delphy
Esmeralda: Can you provide the TSRW version too? So far I can't see any major inconsistencies in either the VPXY or the GEOM (or the MTNF) parts


It's version 0.6.3.0.0.6.3.0.
Warrior Gryphon
site owner
#41 Old 19th Oct 2009 at 10:24 PM
I meant the package file.

Anyway I have actually gone through again comparing the GEOM file for the original lod0_0 lenses, and I note this:



Notice how the MTNF has 17 entries (0 to 16)?

Okay now this:



This is the MTNF from the new lod0_0 lens. Notice how it's only got 10 entries (0 to 9)? This MTNF is the material definition and it controls the transparency of the glasses, as far as I can tell. The modified lens GEOM is missing a lot of the data that was in the original lens GEOM, and I suspect this is what is causing it to go screwy.

One experiment to try would be to import into the package the original lenses (yes, I know they wouldn't match up), and see if they appear correctly in the game.

Story books are full of fairy tales, of Kings and Queens, and the bluest skies.
Mad Poster
#42 Old 19th Oct 2009 at 10:35 PM
Quote:
Originally Posted by Delphy
I meant the package file.


Ha ha, sorry, thought you meant which version of TSRW I'd used. Here you go:

http://www.modyourpanties.com/hosti...es_tsrw.package

I don't know what MTNF means or the significance of the differing numbering, but it does seem like either being reshaped and re-exported, or being packaged, or both, could have somehow changed them. I'll try packaging them with the original lenses soon as I get a chance.
Warrior Gryphon
site owner
#43 Old 19th Oct 2009 at 10:44 PM
The MTNF, as explained, is the MATD of the Sims 3 world - it controls the material definitions for the mesh, including Alpha, Specular, Reflective and all that jazz. CTU itself does not modify or change the MTNF in any way - it merely uses it to discover the bump map. Thus, it's likely that the original GEOM coming out of Milkshape is incorrect.

The TSRW packaged Lenses do indeed have the full complement of MTNF values - essentially an exact copy of the original.

Can you tell me the exact process from start to finish (including extraction, milkshape, plugins, etc) that you are using to generate the LOD0_0 lens geom?

Story books are full of fairy tales, of Kings and Queens, and the bluest skies.
Mad Poster
#44 Old 19th Oct 2009 at 11:21 PM
Well, this is weird. I've just packaged them with the original lenses, and the original lenses showed up perfectly in-game. So something is presumably being lost when exporting the re-shaped lenses from Milkshape, but I haven't got a clue what. All I do is reshape them and export them, without doing anything else to them. Maybe there's something else that one is supposed to do. In any case, I guess it's not a CTU problem then, but some sort of meshing issue. Thank you very much for taking the time to look into this - sorry to waste your time. Maybe that explains the disappearing hat a few posts back, too.

But, it's strange that these disappearing items DO show up in CTU, but not in game!
Sockpuppet
#45 Old 20th Oct 2009 at 12:16 AM Last edited by Base1980 : 20th Oct 2009 at 12:35 AM.
i also had this issue once.
There is something with those lenses when opening them in milkshape.
If you select the full lenses in a window(select faces by vertex) and try to assigne a material to it, it fails.
you keep getting the error ''plz select a group and try again''
you must use the tab select from the group tab or select all from the edit tab before you can assigne a material to it.
I have searched for a floating vertex somewhere that i might been missing, but there is not one.
Select inverse or regrouping doesn't bring any info.
Used the nerdy glasses then.

The thing is that is if you create a 3rd lens from scratch it will not show.
And the reason are the morphs i think as you have to also edit them.(and the Bgeo?)
You can not make custom meshes yet without to edit and include the morphs also.
Tsr kills the morphs and vpxy files and that is prolly why they show?
Mad Poster
#46 Old 20th Oct 2009 at 11:26 AM
Ah, that's interesting, Base1980. I hope someone will be able to figure out what to do about that situation! Wish I had more than just a few weeks experience meshing, as I just don't know enough to be able to troubleshoot.
Warrior Gryphon
site owner DELETED POST
20th Oct 2009 at 12:46 PM
This message has been deleted by Delphy. Reason: Incorrect info
Warrior Gryphon
site owner
#47 Old 20th Oct 2009 at 1:58 PM
Okay I was wrong in thinking it was the MTNF chunk - I was comparing the wrong GEOM

It's definately something in the generated GEOM though since, as you say, importing the old lens over the top works fine.

I'll have to dig deeper and see if anything else could be an issue.

Story books are full of fairy tales, of Kings and Queens, and the bluest skies.
Warrior Gryphon
site owner
#48 Old 20th Oct 2009 at 5:22 PM
The GEOM generated by TSRW has two bones in the bone hash array at the end of the file. Only one, however, seems to be used, like so:

Quote:
Vertex #0
XYZ: -0.056697 0.10805 1.726916
Normal: -0.332593 0.858283 -0.39081
UV: 0.073645 0.914578
Bone: 16843008
Weights: 1 0 0 0
Tangent Normal: 0.9550394 0.2837275 -0.08601385
Vertex #1
XYZ: -0.05666 0.111432 1.737979
Normal: -0.355715 0.926329 -0.12402
UV: 0.070773 0.844468
Bone: 16843008
Weights: 1 0 0 0
Tangent Normal: 0.9350259 0.3461726 -0.07675391


The Bone: 16843008 is repeated for every vertex.

However, in the Milkshake generated GEOM files, we see this:

Quote:
Vertex #0
XYZ: -0.056697 0.10805 1.72855
Normal: -0.3448444 0.8607138 -0.3745053
UV: 0.073645 0.9145781
Bone: 1936223488
Weights: 1 0 0 0
Tangent Normal: 0.9363709 0.2875915 -0.2012475
Vertex #1
XYZ: -0.05666 0.111432 1.73965
Normal: -0.3740949 0.9193669 -0.1217272
UV: 0.07077301 0.8444681
Bone: 1902997760
Weights: 1 0 0 0
Tangent Normal: 0.92267 0.3557443 -0.1487482
Vertex #2
XYZ: -0.067396 0.103809 1.73362
Normal: -0.4830155 0.830047 -0.2787796
UV: 0.025249 0.892759
Bone: 1936945920
Weights: 1 0 0 0
Tangent Normal: 0.8704554 0.4206824 -0.2556045
Vertex #3
XYZ: -0.068829 0.105063 1.742761
Normal: -0.4839657 0.8671033 -0.1179362
UV: 0.016315 0.83382
Bone: 1835888896
Weights: 1 0 0 0
Tangent Normal: 0.8742422 0.473167 -0.108691
Vertex #4
XYZ: -0.027967 0.115094 1.735731
Normal: 0.02091248 0.9866844 -0.1612964
UV: 0.197621 0.8515911
Bone: 1835560192
Weights: 1 0 0 0
Tangent Normal: 0.9957883 -0.03496065 -0.08475534
Vertex #5
XYZ: -0.042872 0.113901 1.73776
Normal: -0.1457747 0.9784722 -0.1460887
UV: 0.131808 0.8479319
Bone: 1936222208
Weights: 1 0 0 0
Tangent Normal: 0.9848748 0.12955 -0.1150589


Notice how the bones are different?

If we go in and set the bones of every vertex in the lens to the same as in the TSRW mesh, guess what we get?



So, the problem is in the generated GEOMs - or more specifically, the bone assignments

Story books are full of fairy tales, of Kings and Queens, and the bluest skies.
Alchemist
#49 Old 20th Oct 2009 at 5:58 PM
While you assigned it to the right joint and fixed it, both GEOMs are single weighted and to one joint.

Only the bone indices which have skin weights are valid. The rest of the bytes in the assignment field are garbage. If you are trying to use an index value that has a zero weight, you will exceed your array bounds. Crash.

If you like to say what you think, be sure you know which to do first.
Warrior Gryphon
site owner
#50 Old 20th Oct 2009 at 6:17 PM
Wes, except the game doesn't crash, even when the vertices are assigned to bones that don't exist, even if they are single weighted.

Story books are full of fairy tales, of Kings and Queens, and the bluest skies.
Page 2 of 4
Back to top