XnaLara for 3dsmax

[Tosyk]

Originally Posted by XNAaraL

Than use the demonic 2 formats (ASCII/SMD). But you saying on the same time, that it not works for you. It it would be "just fine", then you would not need any change requests!

though formats from daemon (ASCII/SMD) holds everything I need their converting is not any tools can support properly, that's why I'm asking here:
ASCII doesn't have native support of axis and SMD doesn't have support of more than 1 uv channel.

Originally Posted by XNAaraL

Then please send me the direct link (No XenTax link!) for this demonic noesis python ASCII script And I will take a look into it, to fix the issue.

here's the script that daemon fixed: https://drive.google.com/open?id=1AV...PBZ5g02mPE7H-y

Originally Posted by XNAaraL

If you means "bone direction", then yes. It is implemented in NGFF (New Generic File Format) since 2 years.

I could be wrong with terms but this "axis" might be called "directions" I suppose

Originally Posted by XNAaraL

About your request to "convert this ASCII to FBX":
There are 4 ways, which works for myself.

I like noesis because it allows to use it in console mode so I'm creating handy CMD file with drag&drop feature to convert any model(s) with half of a click. I'm still trying to find a way to get blender to work the same or similar way

Originally Posted by XNAaraL

1) Use my own Generic_Item.mesh.ascii noesis phyton script. The result can exported from Noesis to FBX without any trouble.
Direct link https://www.sendspace.com/file/pj3ez6
Always latest version http://************/viewtopic.php?f=17&t=1139

I'm aware of that script though I didn't knew it's from you. Unfortunately this script works the way I call "bad" When I doubleclicking ASCII model from windows noesis shows error message says that model can't be loaded because 'fmt_xps_0_9_2.py' can't be found from noesis 'python' folder — I used latest version (fmt_xps_0_9_4.py) but it still demands for 'fmt_xps_0_9_2.py' for some reason - though 0.9.4 script has mentions about 0.9.2 and 0.9.1 in it so I kept all 3 verions together. It works fine if I open ASCII model through noesis interface.

I illustrated bone axis (directions) bellow:

here's how axis looks if I open jensen model with your (0.9.4) version of script:
[IMG]https://media.*******app.net/attachments/315940320107823104/482175432238825472/unknown.png?width=745&height=588[/IMG]

and here's how axis looks when I open same model with modified script from daemon:
[IMG]https://media.*******app.net/attachments/315940320107823104/482174525912383493/unknown.png?width=745&height=594[/IMG]

as you can see your version ignores axis which I'm actually after.

p.s.: blender is a good tool, but it's hard for me to fully understand it or use it for fix/convert models — my field is a 3ds max

[XNAaraL]

Originally Posted by Tosyk

here's the script that daemon fixed:

He has modified my own script, without asking me. And he has messed everything. He has even removed the whole part to export a model as .mesh.ascii

Originally Posted by Tosyk

though formats from daemon (ASCII/SMD) holds everything I need their converting is not any tools can support properly, that's why I'm asking here:
ASCII doesn't have native support of axis and SMD doesn't have support of more than 1 uv channel.

I will explain this later in great detail. The daemon format holds not everything you need. It only contains garbage data.

Bone axis

Originally Posted by Tosyk

Noesis doesn't visualize this kind of data

Originally Posted by Tosyk

I illustrated bone axis (directions) bellow:

You learn quickly.

Originally Posted by Tosyk

I illustrated bone axis (directions) bellow:
here's how axis looks if I open jensen model with your (0.9.4) version of script:
https://media.*******app.net/attachm...745&height=588

and here's how axis looks when I open same model with modified script from daemon:
https://media.*******app.net/attachm...745&height=594
as you can see your version ignores axis which I'm actually after.

It is a matter of "global" bone pose mode versus "local" bone pose mode.

- My original script set the bone orientaion as global mode.
- The daemon has tried to set it to local mode. But he has messed everything.

Now, after you know how to visualize the orientation, take a closer look at the directions:
- In "Global mode", the blue Z-axis always point upwards. My script makes it correct.
- In "Local Mode", the blue Z-axis always points in the direction of the bone. The demon script does it wrong. Totally random:

GROUND bone, almost right

http://image.noelshack.com/fichiers/...291-ground.jpg

calf bone, 90° false rotated

http://image.noelshack.com/fichiers/...70291-calf.jpg

foot bone, 2 times 90° wrong. Look also at the red arrow.

http://image.noelshack.com/fichiers/...70294-foot.jpg

elbow bone, incorrect

http://image.noelshack.com/fichiers/...0294-elbow.jpg

hand bone, just wrong

http://image.noelshack.com/fichiers/...70294-hand.jpg

finger bone, random direction

http://image.noelshack.com/fichiers/...294-finger.jpg

head bone, just a mess

http://image.noelshack.com/fichiers/...70320-head.jpg

...
...

--> The daemon format holds not everything you need. It only contains garbage data.


Multiple UV channel

More than 1 uv channel are supported by .mesh.ascii since the first version. If I remeber right, the first model was Laras Yacht by Dusan himself http://www.mediafire.com/file/mwakzz...Yacht.zip/file
Shader code.

    VertexShaderOutput output;
    float4x4 WorldViewProjection = mul(mul(World, View), Projection);
    float3x3 World3x3 = (float3x3)World;
    output.Position = mul(mul(input.Position, RootMatrix), WorldViewProjection);
    output.TexCoord1 = input.TexCoord1;
    output.TexCoord2 = input.TexCoord2;
    float3 normal3 = (float3)input.Normal;
    output.WorldNormal = normalize(mul(normal3, World3x3));
    return output;

daemon has messed the uv channel in the noesis script and he has messed it in the ascii file.

			uvList.append(uv)
			uvList2.append(uv2)
			uvList3.append(uv3)
			uvList4.append(uv4)

Now, in noesis, every model has exactly 4 uv channel, mostly filled with garbage (zero values).
He also did not understand what causes this UV layer. A UV layer points to a Material (to a xy coordinate from a texture image). That is why it is called TexCoord2 (Texture coordinate 2) in the shader code.
He pass no informations to noesis about the matching material.

Even worse, he already sets this "Material" information to null in the ascii file!

sm_0
1
0
9179
-0.570395 -0.026418 1.152765
-0.063099 -0.228607 -0.971472
0 0 0 255
0.016275 0.297107
257 5 5 5
1 0 0 0

Ex: Line 938 from the demonic ASCII file.

No wonder noesis throws a NULL pointer exception when you try to export this data as an FBX, for example.

--> The daemon format holds not everything you need. It only contains garbage data.

Originally Posted by Tosyk

I'm aware of that script though I didn't knew it's from you. Unfortunately this script works the way I call "bad" When I doubleclicking ASCII model from windows noesis shows error message

Thanks for this good BUG report. I will fix it in version 0.9.6

[Tosyk]

Originally Posted by XNAaraL

It is a matter of "global" bone pose mode versus "local" bone pose mode

3ds max needs local bone axis, and not global.

Originally Posted by XNAaraL

Now, after you know how to visualize the orientation, take a closer look at the directions:
- In "Global mode", the blue Z-axis always point upwards. My script makes it correct.
- In "Local Mode", the blue Z-axis always points in the direction of the bone. The demon script does it wrong. Totally random

how do you choose what mode to show in noesis?

edit: also we found that broken bone axis caused by noesis bug. everything working fine for me in noesis v4231

[XNAaraL]

This is my last comment on this issue. I am so tired of this demon spread here such a lot of falsehoods; although I provide here verifiable evidence.

Originally Posted by Tosyk

3ds max needs local bone axis, and not global

This statement is false. Wrong information given by you.

True is:
The Reference Coordinate System list lets you specify the coordinate system used for a transformation (Move, Rotate, and Scale). Options include View, Screen, World, Parent, Local, Gimbal, Grid, Working, and Pick.

World (Global)
Uses the world coordinate system. Seen from the front:

• X runs in a positive direction to the right.
• Z runs in a positive direction upward.
• Y runs in a positive direction away from you.

The World coordinate system is always fixed.


Autodesk Reference: https://knowledge.autodesk.com/suppo...B07E3-htm.html

Originally Posted by Tosyk

edit: also we found that broken bone axis caused by noesis bug. everything working fine for me in noesis v4231

This statement is false. Wrong information given by you and given by demon.

True is:
The demon model data are corrupt. The demonic "Noesis plugin" is corrupt.


Correct models looks like

Local uses an individual coordinate system specific to each object.
Z align with the bone axis

Noesis reference http://www.richwhitehouse.com/filemi...oesisv4231.zip
Autodesk reference https://knowledge.autodesk.com/suppo...07164-htm.html

EOT

[Tosyk]

why my message was deleted? I only share the solution

here's it:
I discovered that after Autodesk FBX export plugin for noesis been updated to FBX SDK 2017.1 (from noesis version 4.24) it produces a bug which causes this broken bone axis rotations. Old 2014-based plugin is in "optionalplugins" dir of noesis main folder and it's called autodesk_fbx_2014.dll. So to get bone axis rotations from original game (from any game that daemon made tools for) models use <attached> modified script to open ASCII files and then export to FBX with Old 2014-based plugin.

p.s.: XNAaraL, I'm sure you're good mathematician but to be able to evolve in any kind of things a technology, mind and personality must be flexible.

<attachment>: ask me for the script, XNAaraL doesn't allow to post it here by reports to the moderator.

[Cochrane]

Since I was just dragged into this (via a private message), I'd just like to say:

• I have no idea what any of you are talking about.
• Reading the thread has not helped.
• I'm too lazy to Google it.
• Whatever it is, nobody has said to me "GLLara is broken because of this", possibly because (as far as I can tell) nobody uses GLLara. So it doesn't affect me negatively.
• If you do need input from me on anything, please provide a link to a sample file, explain what it should do and how what you're seeing differs. If language is an issue, I accept german or english, whichever you feel more comfortable with (if it's a wash then I prefer english because all my knowledge about 3D programming is in english).

Also, settling technical disputes (if it even is one, I honestly can't tell) by calling mods and invoking copyright claims? Well, okay, I don't want to offend anyone about how they choose to regulate access to their work. But I will say that GLLara is open source (GPL v2 or later) and will remain open source, and if you make a fork of it and do something differently, maybe even screw it up, that is 100% all right with me. Personally I think this is the most useful way to develop non-commercial tools for fans, and I don't really understand why so many people see things differently, but again, I don't want to criticise anyone.

Just please don't drag me into this.

[Tosyk]

Originally Posted by Cochrane

Since I was just dragged into this (via a private message)

hello, not sure why you need to be dragged in and who you actually are

Originally Posted by Cochrane

I have no idea what any of you are talking about.

basically I asked If anyone could modify maxscript to load ASCII with 'additional data' in it and XNAaraL answer by calling this data "wrong", my talk is BS and both me and my friend are idiots

Originally Posted by Cochrane

If you do need input from me on anything, please provide a link to a sample file, explain what it should do and how what you're seeing differs<...>

beside the maxscript there's a Noesis script for ASCII (made by XNAaraL). It was modified by my friend to load that 'additional data' and I used it to convert ASCII->FBX. Also because of broken Noesis FBX exporter it's produced a corrupted FBX. That's why I asked for fixing maxscript.

Originally Posted by Cochrane

Also, settling technical disputes (if it even is one, I honestly can't tell) by calling mods and invoking copyright claims? Well, okay, I don't want to offend anyone about how they choose to regulate access to their work. But I will say that GLLara is open source (GPL v2 or later) and will remain open source, and if you make a fork of it and do something differently, maybe even screw it up, that is 100% all right with me. Personally I think this is the most useful way to develop non-commercial tools for fans, and I don't really understand why so many people see things differently, but again, I don't want to criticise anyone.

totally agree with your statements though the issue was only when I posted modified Noesis script of XNAaraL — he didn't like it so he reported.


p.s.: a friend of mine - daemon - creating tools to convert models from games. he choosed ASCII format to hold mesh, skeleton and - what's more important - multiple uv channels and local bone axis rotations (as 'additional data'). so after he expanded ASCII format he modified XNAaraL's ASCII noesis script to load that 'additional data'. I shared that script here and my post was deleted.

[German_Hunter]

I've encountered a problem with this script:
Using 3DSMax7 here and when loading up the script I get the error message that the script isn't compatible to my version of 3DSMax. When trying to export the model into XPS (.mesh.ascii), it simply doesn't work.
Now I've read in the opening post from here that the script is comptaible to all Max versions since 3DSMax8.

Question from me is: Does a Max8 installation exist somewhere on the internet for everyone to download?

[XNAaraL]

Originally Posted by German_Hunter

I've encountered a problem with this script:
Using 3DSMax7 here and when loading up the script I get the error message that the script isn't compatible to my version of 3DSMax. When trying to export the model into XPS (.mesh.ascii), it simply doesn't work.
Now I've read in the opening post from here that the script is comptaible to all Max versions since 3DSMax8.

Question from me is: Does a Max8 installation exist somewhere on the internet for everyone to download?

I've read in the opening post from here, that the maxscripts is tested on 3dsmax8-3dsmax2012 should work across all versions of max past v8(2005).
The opening post do not claim that is comptaible to all Max versions since 3DSMax8
Of course, I read also:

Originally Posted by AnnaShepard

scratch my previous statement. It works in 3DS Max 2014. :>

Google is your friend.
Autodesk only supports a few versions back officially.

Is there any way to download 3ds max 2012? : 3dsmax - Reddit

Autodesks Virtual Assistant.

Solved: Download old versions of 3ds Max - Autodesk Community- 3ds ...

You can also try:
3ds-max free trial

BTW:

Dependent on your goal, there are uncomplicated alternatives to the maxscript.
(Viele Wege führen nach Rom)

How to load a GenericItem.mesh.ascii model (XNALara/XPS) in 3ds max 7 ?

1. Download Noesis version 4.274 and Noesis to XNALara/XPS importer/exporter plugin 0.9.6
2. Load the .mesh.ascii file in Noesis version 4.274 (If you not have the .mesh.ascii format, then generate it in XNALara/XPS, using the shortcut F11)
3. Exported it from Noesis as "Autodesk FBX" (or any other format supported by 3ds max 7)
4. Load the FBX file in 3ds max

How to convert a model from 3ds max 7 to XNALara/XPS ?

1. Download Noesis version 4.274 and Noesis to XNALara/XPS importer/exporter plugin 0.9.6
2. Export the model from 3ds max as .FBX
3. Load the FBX model in Noesis
4. Export it from Noesis, using "File export" and choose "Main output type: .mesh.ascii XNALara/XPS 0.9.6"
5. Load the .mesh.ascii file in XPS 11.8. (Drag&Drop or "Modify-->Load GenericItem ..."

Step by step instruction here: https://www.tombraiderforums.com/sho...postcount=6430

[Dragonknight]

Whenever I try to export a model as a mesh.ascii file the program freezes. Do you have any idea what is going on?