A start on the .MESH file format

[alpaca]

Haha that looks really nice already. You should consider working for a mod team (for example, for me )

[Re Berengario I]

P.S. do you know how to get closer to the figures with the camera???

Free the camera in the game options.

[SigniferOne]

Thanks Shredder
...as you probably noticed I'm moving my first steps into M2TW modding....

Now all seems to work in the right way :)
Here the Venetian infantry with my barbute mixed up with the other vanilla helmets:

Why are the other units so darkened? You sure everything went through okay?

P.S. do you know how to get closer to the figures with the camera???

I think you need to be in "replay" mode, then you can even go through units if you want to. In regular mode the camera stays a short distance away from soldiers.

[Luciano B]

Why are the other units so darkened? You sure everything went through okay?

...It appens because there's the shadow of some palm trees over them :)






P.S. Thanks again to all.

[Luciano B]

Haha that looks really nice already. You should consider working for a mod team (for example, for me )

Thanks for the kind offering, Alpaca ...but I'm really busy to complete our current version for RTW/BI of "Paeninsula Italica":
https://forums.totalwar.org/vb/forumdisplay.php?f=216

...we are also exploring to convert the project for M2TW for a future release running with the new engine.

Once finished the RTW/BI version of "Paeninsula Italica" I planned to help the team working on the mod "Condottieri" which covers the wars in Italy during the early renaissance, so I'm currently doing my modelling experiments working on late medieval/early renaissance stuff... if you need something related to this era you're welcome...

[Luciano B]

great work
how did you mapped your barbute helmet? in max or milkshape?

Thanks Herkus

I used the mapping utility provided with milkshape; I mapped provisionally the new helmet on the vanilla texture


My next goal will be to move the whole helmets matter from the main "figure" to separate attachments... I don't know if it will work ...suggestions will be greatly appreciated :)

[Herkus]

Aha, thanks a lot Luciano.

I am also going to include helmets in game. Made them very different from each other (and very low in polies) and will try to make completely new map by deleting previous helmet textures.
so far... without texture
shape:

wires:

[zxiang1983]

@Herkus

All nice looking helmet, looking forward to them. Did you make them in Max or Milkshape?

[GrumpyOldMan]

Hi All

Just so you don't think I've been slacking while the erudite and talented KnightErrant has been releasing his code, here's a pic of the latest project, this is to help the RTW modders who want to move over to M2TW.



I've made a converter that takes RTW CAS figure files (human only at this stage) and converts them to Milkshape ascii files with the vertices already assigned to the correct M2TW bones. There's a little bit of prep work required but nothin goverly hard or complicated to get them m2tw ready.

I'll put a GUI on it tomorrow (all things being equal) and put it up as a separate program.

Cheers

GrumpyOldMan

[Herkus]

@Herkus

All nice looking helmet, looking forward to them. Did you make them in Max or Milkshape?

Thanks, I made them in Max.

@GrumpyOldMan
Excellent work. This will really help some of the RTW big mods who are planning to convert to M2. I guess that model is around 600 polygons, , that can make difference in game's performance.
Did you assigned the same texture file for figure and attachments or made separate texture files?

[GrumpyOldMan]

Hi Herkus

@GrumpyOldMan
Excellent work. This will really help some of the RTW big mods who are planning to convert to M2. I guess that model is around 600 polygons, , that can make difference in game's performance.
Did you assigned the same texture file for figure and attachments or made separate texture files?

Yes it's the same number of tris as a RTW cas, I must admit I did cheat a bit though and use the same model for lod0, lod1 and lod2 It's around the same as a m2tw lod1/2 figure.

I didn't mess with the usual texture set up and used the same texture for both but with one in the model texture folder and one in the attachmentset folder. I'll leave such intrepid adventuring to any volunteering test pilots

Cheers

GrumpyOldMan

[Luciano B]

Aha, thanks a lot Luciano.

I am also going to include helmets in game. Made them very different from each other (and very low in polies) and will try to make completely new map by deleting previous helmet textures.

Well done, Herkus.

I guess that I have to decrease the numbers of polis in my barbute... however it still works smoothly running in my old PC...

[Andromachus Theodoulos]

@All

Good stuff going on here!!!!

@GOM

Grumpy, I figured I would post this out here so others can read this who may not understand.

Could you give an explanation of (in MilkShape) what the comments mean under the Model tab (they look like some kind of coordinates) and under the Group tab.

Thanks,

Oh, Can't wait to try out the CAS/MESH converter!!!!

AT

[Athokas]

I feel like a moron for asking this but how do you import the model into 3Dmax?
I converted the meshes to the only thing possible, MS3D, but Max won't load it. Read the readme carefully, but it is mostly directed at Milkshape users

Sorry for being a retard but please

[Herkus]

I feel like a moron for asking this but how do you import the model into 3Dmax?
I converted the meshes to the only thing possible, MS3D, but Max won't load it. Read the readme carefully, but it is mostly directed at Milkshape users

Sorry for being a retard but please

Open the ms3d model in milkshape, then go to - file->export->Autodesk 3DS->save the file.

After that - open Max and go to file->import-> and browse for the 3ds format which you saved with milkshape.

[KnightErrant]

Hi AT,

The data in the model comment is the bounding sphere data. The order
is x, y, z, and radius. You may need to change the 4th entry, radius,
if you switch out a short weapon and replace it with a longer one. Check
what value the unit has that you take the weapon from and see if its
very different.

The comments under the group tab are stored there because Milkshape's
format only allows 32 characters for the group name. The mesh format
requires a group type as well as a group name so we put them in the
group comments to be able to read them out again and put them with their
corresponding triangle group. First entry is the group type like Head, and
the second is the group name like Head_09.

[Athokas]

OMG I have to own BOTH Milkshape and Max? Damn, please extent the converter soon to make it convert Mesh -> 3DS

[Andromachus Theodoulos]

@KE,

Thank you sir for that explanation...

I knew you and Grumpy had discussed it before but I couldn't find it...

Thanks!!

AT

[GrumpyOldMan]

Hi Athokas

OMG I have to own BOTH Milkshape and Max? Damn, please extent the converter soon to make it convert Mesh -> 3DS

Well, for low poly game models you don't have to own Max at all, Milkshape has got enough features to cope.

If you want to export the ms3d models to 3ds or obj format you can use LithUnwrap, a free program to convert. Just be aware that you lose all bones, vertex assignments and weights (not part of the 3ds or obj format). I've been looking for plugins or formats that could be used but no luck yet.

There is a max to ms3d plugin available from http://legolas.mdh.se/~elt01mcg/ but from what people tell me, it doesn't export rtw cas files successfully. The plugin has source included so maybe somebody that has knowledge of Max can work on it.

Cheers

GrumpyOldMan

[zxiang1983]

The data in the model comment is the bounding sphere data. The order
is x, y, z, and radius.

Hi, KE, thanks for the explanation. But what if the unit has two weapons, one primary and one secondry? Which weapon does the radius mean?

And another question. I had a quick check on one of my self-made unit. Just find that he has five comments! What does this mean? BTW, he's a knight with a cavalry spear and a sword. Several parts of him is taken from other units. And it "seems" work fine in game. So now do I have to delete some of these comments and then modify the rest?


Thanks in advance.

[SigniferOne]

Did you guys ever figure out what the bounding sphere was for?

[KnightErrant]

@zxiang1983

Hmmm, for the first question I think it should apply to the primary
weapon as that should determine what happens on initial contact.
We haven't done enough experiments with bounding sphere data
to know for sure everything that it affects.

Haven't seen the multiple entries before. This may be an artifact
from the merge; I didn't do a merge facility in my Python script so
this will have to wait for GrumpyOldMan. I would guess that it works
for the back conversion because the comment would be read out,
tokenized by carriage return/linefeed codes and the first four float
values would be used for writing back into the mesh format. At least,
that how I parse it in Python. Does mean that only the first four
values would be used and the others ignored.

Did you do four or five merges to get the final model? For debugging
it would probably be good to know.

[zxiang1983]

@KE
Thank you very much. You are quick:P

It's difficult to remember how many time I do the merges... Due to my memory,
1, merge the body with a helmet
2, merge 1 with a feather
3, merge 2 with shield
4, merge 3 with spear & sword

So, yes, it's four time I did the merges. I think it must be the side product of merge. Now I'm going to delete four of the comments and see what will happen.


Edit: Seems all goes fine. I keep the middle one(the one with radius of 2.05549) and delete other 4 entry. The change of radius is obvious. Enemy unit was taken down even when the weapon did not touch him.

[KnightErrant]

Sounds like an excellent experiment to do. I may have spoken in
ignorance, maybe this is what's supposed to happen. If you merge
maybe you're supposed to get a choice of bounding sphere data.
Helmut, feather, then spear, etc. looks like one of the bounding sphere
entries has a larger radius, see where it goes from 1.32 to 2.055. So maybe
this is just a way of giving the user a choice for the appropriate
data for his modifications. Been busy so I haven't kept up to date with all the
great stuff going on here.

Edit: Excellent spot, radius works with weapon kill. This was speculated on a few
pages back in the wiki. Could you do a few extreme checks like bounding radius
of 5.0 or 10.0 to see if a weapon becomes a long distance killer? I'm not advocating
Medieval Age directed energy weapons or anything but this could be interesting.

[zxiang1983]

@KE
The feather is taken from French Lancers. So I guess the radius is Lancers' and then it's not surprise the number is much larger than others.

Edit: yes, I checked the Lancers's comment. It's exactly a match.

0.736253
0.358799
1.12306
2.05549
/

Edit: Did another experiment, changing the radius from 2.05549 to 5. Due to my observation the effect is not obviouse. I should say it's the same as 2.05549. So maybe there's a maximum number?

[KnightErrant]

Very good, so the model comments build up successive merge information
with new bounding sphere data which you can then pick and choose.

@SignifierOne
I think zxiang1983 is answering your question, radius seems to be related
to where weapon effectiveness turns on. We had thought that the collision
mechanism wasn't as detailed as a per poly process and that seems more
likely now. Must go back and read again a few pages back in this thread.

[GrumpyOldMan]

Hi KE & ZX

The change in kill radius could also be due to the 'z' value in the third group being for a mounted figure while the others are for foot figures. Check the values of the vanilla figure that you took the lance from, this should give you a rough idea of what would be the proper collision sphere. Generally a 'z' value around 1 means mounted on a horse, and the longer the weapon, the larger the radius or 4th value will be.

With the comments, because people might be keeping their own notes in the comments, what I've done with the merge is to copy the comments from both files and merge them as well. Only the first four lines of the Model Comments are read by the converter for the re-conversion process, first three line for Group Comments. The order is first file chosen to merge has the first lines of comments written.

@KE I've put a response up on the Animations thread

Cheers

GrumpyOldMan

[KnightErrant]

@GOM,

Took a quick look at the animations thread. Much thanks but
its late for me so I'll scan more carefully tomorrow. Yes, I did
see the pelvis movement thinking that is the overall animation
change to the (center of mass?).

I think when my brain is rested, I'll have better questions re Euler
conventions. (Have to study the code you've posted.) You were
right, plenty of adventures lie in store.

It sounds though apart from calculated roll, calculated pitch, calculated yaw,
being assigned a different name, that Mete must be following a standard
convention: Said badly, I mean he's not doing a 212 when the rest of the
world does 321 (or 123 depending on the literature you read).

Ok, reread your response again before I hit the submit reply button,
the extra bytes of floats have some global stuff with it. I think I would like
to take the language for a shareware ride. Please send what you can,
I'll try to understand it tomorrow.

(I do appreciate this, starting absolutely from scratch on animation stuff.)

Regards,

KE

[zxiang1983]

Hi, all!

I've got a new problem here. I made a unit which has polish_noble_ug1's body and ee_spear_militia's spear and shield. The model looks fine and even works well in game. However there's something a little wierd. The symptom is as below:


Most of them fight well, but a few of them just stand still and do nothing after the fight begins for a moment. And in most cases they even won't follow your orders such as moving or attacking. Is it a problem of the model or a bug of the game? Because neither of vanilla polish_noble_ug1 and ee_spear_militia has this symptom.

I'm guessing it is related to bounding sphere data? Because polish_noble_ug1 is a cavalry unit but now I put him on foot. But I've also tried to use dismounted_polish_noble_ug1 or ee_spear_militia's bounding sphere data. No luck.

I've upload my mesh files and entries here in case anyone wants to research:
http://rapidshare.com/files/25569499/Wlocznicy.zip

Any help is appreciated!


Edit: Maybe I made some mistake in EDU, coz after I re-write the EDU entry I cannot reproduce this bug. So, problem sovled!:)

[GrumpyOldMan]

Hi Zxiang

Edit: Maybe I made some mistake in EDU, coz after I re-write the EDU entry I cannot reproduce this bug. So, problem sovled!:)

That's what I like to see, problems raised and then solving themselves without any input from me .

Cheers

GrumpyOldMan